Managerial Economics, 5th Edition PDF

Managerial Economics
Now in its fifth edition, Ivan Png’s Managerial Economics has been extensively
revised with
• an introductory chapter emphasizing decision-making and behavioral biases,

• an intensive application to recent business decisions, as well as
• a streamlined presentation focusing on the economics that managers need
to know.
As always, the text presents the key concepts of microeconomics intuitively,

without sophisticated mathematics. Throughout, it emphasizes actual manage-
ment applications, and links to other functions including marketing and finance.
The new fifth edition is updated with fresh up-to-date discussion questions
from all over the world and enhanced with detailed instructor supplements. It
is an ideal text for any course focusing on the practical application of microeco-
nomic principles to management.
Truly useful economics for managers. In the words of one professor, “I can use
your book for serious conversation with adult students.”
Ivan Png is a Distinguished Professor in the School of Business and Department

of Economics at the National University of Singapore. Previously, Dr Png was
a faculty member at the Anderson School, University of California, Los Angeles
(1985–1996) and the Hong Kong University of Science and Technology (1993–
1996), and Visiting Professor at the Tuck School of Business, Dartmouth Col-
lege (2011–12). His book, Managerial Economics, has been published in multiple
editions and adapted into Chinese (traditional and simplified characters), Korean,
and Polish. He received the NUS-UCLA Executive MBA Teaching Excellence
Award in 2008 and 2011. Dr Png was a nominated MP (10th Parliament of Singa-
pore), 2005–2006, a member of the Trustworthy Computing Academic Advisory
Board, Microsoft Corporation, 2006–10, and an independent director of Health-
way Medical Corporation, 2008–2011.
“I often use this book in my MBA-level Managerial Economics classes. It
provides clear and rigorous explanations of major microeconomic theories and
supports them with a large number of relevant examples from around the world.
I find this book well-written, concise, and interesting.”
Annamaria Conti, Assistant Professor, Scheller College of
Business, Georgia Institute of Technology, USA
“Ivan Png’s Managerial Economics is a concise and effective textbook that makes
use of endless real-life examples not only to illustrate a point, but rather to motivate
it, explain it, and convince students of its relevance. The wide range of situations,
industries, and historical events covered make this book quite an enriching read.
In my opinion, Managerial Economics is an excellent tool when used to teach
MBAs and students without much economics background. Economics majors
will benefit from reading it as they will gain valuable practical knowledge of the
use of their tools.”
Eugenio J. Miravete, Rex G. Baker Jr. Professor of Political
Economy, University of Texas at Austin, USA
“Ivan Png’s Managerial Economics has made my life easy as a lecturer! It is a

fantastic integration of theory, real life examples, and case studies, which makes
both teaching and learning a joy. The chapters are well organized and highly
relevant – keeping in mind business and management students’ needs. I have
thoroughly enjoyed using this book as the main text for my postgraduate manage-
ment economics class.”
Abhijit Sengupta, Lecturer, Essex Business School,
University of Essex, UK
“I have been using this text for teaching the EMBA Global Asia program (a joint
program of Columbia Business School, HKU Business School, and London
Business School) in the last six years. Students like the text very much as it offers
possibly the simplest and most concise explanation of economic concepts and
principles that are applicable for business decision making.”
Tao Zhigang, Professor of Economics and Strategy, Faculty of
Business and Economics, University of Hong Kong
Fifth Edition
Ivan Png
First published 2016
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
711 Third Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa
business
© 2016 Ivan Png
The right of Ivan Png to be identified as author of this work has
been asserted by him in accordance with sections 77 and 78 of the
Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or
reproduced or utilised in any form or by any electronic, mechanical,
or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or
retrieval system, without permission in writing from the publishers.
Trademark notice: Product or corporate names may be trademarks
or registered trademarks, and are used only for identification and
explanation without intent to infringe.
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
Png, Ivan, 1957-
Managerial economics / Ivan Png. – 5th edition.
1. Managerial economics. I. Title.
HD30.22.P62 2015
338.5024’658–dc23 2014049719
ISBN: 978-1-138-81025-9 (hbk)

ISBN: 978-1-138-81026-6 (pbk)
ISBN: 978-1-315-74964-8 (ebk)
Typeset in Times
by Sunrise Setting Ltd, Paignton, UK
For my parents and three Cs – CW, CY, CH
This page intentionally left blank
Contents
List of Illustrations ix
Preface xii
Acknowledgments xv
About the Author xvi
1 Introduction to Managerial Economics 1
Part I Competitive Markets 19
2 Demand 21
3 Elasticity 43
4 Supply 65
5 Market Equilibrium 96
6 Economic Efficiency 118
Part II Market Power 137
7 Costs 139
8 Monopoly 167
9 Pricing 193
10 Strategic Thinking 218
11 Oligopoly 249
Part III Imperfect Markets 277
12 Externalities 279
13 Asymmetric Information 305
viii Contents
14 Incentives and Organization 329

15 Regulation 357
16 Answers to Selected Progress Check and
Review Questions 378
Index 400
Illustrations
Figures
1.1 Value added 4

2.1 Individual demand curve 23
2.2 Individual demand curve with lower income 26
2.3 Individual demand curve with more expensive
complement 29
2.4 Individual buyer surplus 33
2.5 Package deal 34
2.6 Demand curve for mobile telephone service 36
2.7 Market demand curve 42
3.1 Calculating own-price elasticity 46
3.2 Short- and long-run demand for a non-durable item 57
4.1 Short-run total cost 69
4.2 Short-run marginal, average variable, and average costs 72
4.3 Short-run profit 74
4.4 Short-run production rate 76
4.5 Lower input price 79
4.6 Long-run production rate 82
4.7 Individual seller surplus 85
4.8 Price elasticity of supply 87
4.9 Market supply 95
5.1 Market equilibrium 101
5.2 Supply shift 103
5.3 Price elasticities of demand and supply 104
5.4 Demand shift 107
5.5 Short-run market equilibrium 108
5.6 Long-run market equilibrium 109
5.7 Demand increase: short and long run 111
5.8 Demand reduction: short and long run 111
6.1 Air travel market 123
6.2 Pricing and freight cost 128
6.3 Air travel tax 131
7.1 Costs in a single period of production 140
7.2 Transfer price 145
x Illustrations
7.3 Economies of scale 152

7.4 New tanker prices: January 2006 155
7.5 Experience curve 159
8.1 Monopoly production scale 175
8.2 Demand increase 177
8.3 Reduction in marginal cost 178
8.4 Market structure 183
8.5 Monopsony purchasing 186
9.1 Uniform pricing 195
9.2 Complete price discrimination 199
9.3 Direct segment discrimination 204
10.1 Scheduling the evening news: extensive form 229
10.2 Scheduling the evening news: uncertain consequences 231
10.3 Narrow-body jets: new entry 232
10.4 Scheduling the evening news: TV Delta outsources
production facilities 233
10.5 Bank deposit – without deposit insurance 236
10.6 Bank deposit – with deposit insurance 236
10.7 Strike 237
10.8 Poison pill 239
10.9 Solving Nash equilibrium in randomized strategies 246
11.1 Monopoly 251
11.2 Price competition with differentiated products:
residual demand 253
11.3 Price competition with differentiated products:
best response functions 254
11.4 Strategic move 258
11.5 Limit pricing 258
11.6 Capacity competition: residual demand 260
11.7 Capacity competition: best response functions 261
11.8 Capacity leadership 264
12.1 Positive externality 282
12.2 Negative externality 284
12.3 Rivalness 294
12.4 Economically efficient provision of public good 295
13.1 Market with symmetric information 310
13.2 Market with adverse selection 310
13.3 Market failure 313
14.1 Economically efficient effort 332
14.2 Performance pay 336
14.3 Performance quota 337
14.4 Vertical integration 349
15.1 Price regulation 360
15.2 Moral hazard in medical services 367
Illustrations xi
15.3 Economic efficiency in emissions 370

15.4 User fee 370
15.5 Accidents 373
Tables
2.1 Individual demand 23

2.2 Individual demand with lower income 26
2.3 Market demand 42
4.1 Short-run weekly expenses 68
4.2 Analysis of short-run costs 68
4.3 Short-run profit 74
4.4 Long-run weekly expenses 81
4.5 Analysis of long-run costs 81
4.6 Long-run profit 82
4.7 Barrick Gold 92
7.1 Conventional income statement ($ million) 141
7.2 Income statement showing alternatives ($ million) 142
7.3 Income statement showing opportunity cost ($ million) 142
7.4 Conventional income statement ($ million) 147
7.5 Income statement showing alternatives ($ million) 147
7.6 Income statement omitting sunk costs ($ million) 147
7.7 Daily expenses for newspaper production 150
7.8 Analysis of fixed and variable costs 151
7.9 Expenses for two products 157
7.10 Chrysler LLC liquidation ($ million) 164
7.11 Punch Taverns (£ million) 165
8.1 Monopoly revenue, cost, and profit 173
8.2 Advertising and R&D, 2013 182
8.3 Google profit and loss, 2010–2013 190
9.1 Indirect segment discrimination in air services 209
9.2 Pricing policies: profitability and information 211
9.3 DEWA: electricity rates 215
10.1 Gasoline stations: price war 221
10.2 Battle of the Bismarck Sea 222
10.3 Narrow-body jets: new entry 224
10.4 Gasoline stations: price war (modified) 225
10.5 Scheduling evening news 228
11.1 Memory industry: share of production 250
12.1 Customer traffic and profit 281
12.2 Customer traffic and externality 283
13.1 Southwest Airlines 325
15.1 Competition laws 363
Preface
Managerial economics is the science of directing scarce resources in the manage-

ment of a business or other organization. This book presents tools and concepts
from managerial economics that practicing managers can and do use. It
• emphasizes simple, practical ideas,

• focuses on application to business decision-making,
• integrates global business issues and practice,
• provides conceptual rigor without mathematical complexity.
This book is aimed at business students as well as practitioners. Accordingly, it

is deliberately written in a simple and accessible style. It presents a minimum of
technical jargon, complicated figures, and highbrow mathematics. It starts with
the very basics and does not presume any prior knowledge of economics. While
the mathematics is minimal, the economics is rigorous. The application of eco-
nomic concepts to business practice will challenge even readers with some back-
ground in economics.
Managerial economics is unique in integrating the various functions of man-
agement. In addition to presenting the essentials of managerial economics, this
book includes many links to other management functions. Some examples are
accounting (transfer pricing), finance (opportunity cost of capital and takeover
strategies), human resource management (incentives and organization), and mar-
keting (advertising and pricing).
In addition to the managerial focus, two features are worth emphasizing. First,
the same principles of managerial economics apply globally. Reflecting this unity,
the book includes examples and cases from throughout the world. Second, the
book uses examples from both consumer and industrial markets. The reasons
are simple: a customer is as likely to be another business as a human being, and
likewise for suppliers.
For most readers, this may be their only formal book on economics. Accord-
ingly, the book eschews sophisticated theories and models, such as indifference
curves and production functions, which are more useful in advanced economics
courses. Further, the book recognizes that many topics traditionally covered by
managerial economics textbooks are now the domain of other basic management
courses. Accordingly, the book omits linear programming and capital budgeting.
Preface xiii
Regarding language, this book refers to businesses rather than firms. Realisti-
cally, many firms are involved in a wide range of businesses. In economics, the usual
unit of analysis is a business, industry, or market rather than a firm. Also, the
book refers to buyers and sellers rather than consumers and firms, since in most
real markets, demand and supply do not neatly divide among households and
businesses. To cite just two examples, in the market for telecommunications, the
demand side consists of businesses and households, while in the market for human
resources the supply side comprises households and businesses. Outsourcing has
reinforced this diversity of suppliers.
Managerial economics is a practical science. Just as no one learns swimming
or tennis simply by watching a professional, so no one can learn managerial eco-
nomics merely by reading this book. Every chapter of this book includes prog-
ress checks, and review and discussion questions. The progress checks and review
questions are to help the reader check and reinforce the chapter material.
Readers must practice their new-found skills on these checks and questions. The
discussion questions are intended to challenge, provoke, and stretch. They will be
useful for class and group discussions.
Key Features
• Simple, practical ideas for business decision-making

• Integrates managerial economics into finance, accounting, human resources,
and other management functions
• Mini-cases and examples from around the world
• Every chapter is reinforced with progress checks, review questions, and discussion
questions
• Easy to read, with minimal technical jargon, figures, and mathematics
• Complete instructor’s supplements – transparency masters, answers to discussion
questions, casebank, and testbank.
Organization
This book is organized into three parts. Following the Introduction, Part I pres-
ents the framework of perfectly competitive markets. Chapters 2–6 are the basic
starting point of managerial economics. These are presented at a very gradual
pace, accessible to readers with no prior background in economics.
The book gathers pace in Parts II and III. These are relatively self-contained,
so the reader may skip Part II and go directly to Part III. Part II broadens the
perspective to situations of market power, while Part III focuses on the issues of
management in imperfect markets. Chapter 15 on regulation is the only chapter in
Part III that depends on understanding Part II.
xiv Preface
A complete course in managerial economics would cover the entire book. For
shorter courses, there are three alternatives. One is a course focusing on the man-
agerial economics of strategy, which would comprise Chapters 1–11. Another
alternative focuses on the managerial economics of organization, comprising
Chapters 1–7 and 12–14. The third alternative focuses on modern managerial
economics – strategy and organization – and would comprise Chapters 1–4 and
7–14.
Website
Online support for this book can be found at https://sites.google.com/site/pngecon/.

The site contains additional cases and applications, as well as updates and correc-
tions to the book. The site also contains a link to resources for instructors, includ-
ing transparencies, answers to discussion questions, a testbank, and a casebank.
Acknowledgments
In preparing the fifth edition, I gratefully acknowledge advice and suggestions from
Nicholas Snowden, Richard Leigh, and Joy Cheng. Finally, I thank generations
of students at NUS, HKUST, and UCLA for their enthusiastic support and
encouragement.
About the Author
Ivan Png is a Distinguished Professor in the School of Business and Department

of Economics at the National University of Singapore. He was a Visiting Profes-
sor at the Tuck School of Business, Dartmouth College (2011–2012). Previously,
he was a faculty member at the Anderson School, University of California, Los
Angeles (1985–1996) and the Hong Kong University of Science and Technology
(1993–1996).
Dr Png attended the Anglo-Chinese School, Singapore, and graduated with
first class honors in economics from the University of Cambridge (1978) and a
PhD from the Stanford Graduate School of Business (1985).
His research has been published in leading management and economics journals
including Management Science, American Economic Review, and Journal of Politi-
cal Economy. He received the NUS-UCLA Executive MBA Teaching Excellence
Award in 2008 and again in 2011.
Dr Png was a nominated MP in the 10th Parliament of Singapore (2005–2006),
member of the Trustworthy Computing Academic Advisory Board, Microsoft Cor-
poration (2006–2010), and an independent director of Hyflux Water Trust Man-
agement Pte Ltd (2007–2011) and Healthway Medical Corporation (2008–2011).
Dr Png speaks English and Chinese (Mandarin). He is married to Ms Joy Cheng.
They have two sons, Max and Lucas.
C H A P T E R
1
Introduction to
LEARNING OBJECTIVES
• Appreciate the objective of managerial economics.
• Understand value added and economic profit.
• Apply total benefit and total cost to decide participation.
• Apply marginal benefit and marginal cost to decide extent.
• Appreciate the effect of bounded rationality on decision-making.
• Apply net present value to evaluate benefits and costs that flow
over time.
• Understand the vertical and horizontal boundaries of an
organization.
• Distinguish competitive markets, market power, and imperfect
markets.
1. What Is Managerial Economics?
Airbus and Boeing dominate the manufacturing of large commercial jet aircraft
(with 150 or more seats).1 Boeing’s most successful and profitable plane is the
Boeing 737, a twin-engine, single-aisle, medium-range jet. First flown in 1967, the
Boeing 737 has been developed into nine models. As of December 2010, Boeing
had delivered 6,687 units of the 737, with a further 2,186 on order. The Boe-
ing 737 competes with Airbus’s A320 family, comprising five models – the A318,
A319, A320, A321, and ACJ business jet. According to Boeing forecasts, airlines
would buy 23,370 new single-aisle planes in the next 20 years.
However, at the Paris Air Show in June 2011, Jim Albaugh, CEO of Boeing
Commercial Airplanes, conceded: “The days of the duopoly with Airbus are
2 1 Introduction to Managerial Economics
over.” Manufacturers, from China, Canada, Russia, and Brazil, have developed,
launched, or are poised to launch new aircraft to compete with Boeing’s 737 and
Airbus’s A320 family.
In November 2010, the Commercial Aircraft Corporation of China (COMAC)
announced that it had booked 100 orders for the C919, a new 150-seat single-aisle
plane, the leading customers being Air China, China Southern Airlines, and
China Eastern Airlines. By September 2014, COMAC reported an increase to
400 orders from 16 customers. The maiden flight of the C919 is scheduled for
late 2015.
Bombardier, headquartered in Canada, had long manufactured regional jets,
which are smaller short-range planes with up to 100 seats. It aspired to expand
into larger aircraft, but only began development in 2008, upon securing a letter
of interest for 60 planes from Deutsche Lufthansa. The new CSeries, a family of
100- to 149-seat aircraft, is scheduled to enter service in 2015. The CSeries will
reduce fuel consumption by 20% through use of advanced materials and a more
fuel-efficient engine, the PW1000G from Pratt & Whitney. As of September 2014,
Bombardier had 203 orders.
The Russian manufacturer of military aircraft, Irkut, is diversifying into
commercial jets and launched the MC-21 in 2007. By February 2014, Irkut had
secured 170 orders, all from Russian airlines. The MC-21 will reduce fuel con-
sumption through lower weight, better aerodynamics, and more efficient engines.
It is scheduled to enter service around 2016.
Like Bombardier, the Brazilian manufacturer, Embraer, is an established manu-
facturer of regional jets. However, as of June 2011, it had not secured any orders
and had deferred a decision on whether to commence development. CEO Fred-
erico Curado remarked: “Going up against Boeing and Airbus in head-to-head
competition is really tough, not only because of their size, but because of their
existing product line and industrial capacity . . . . They can have a very quick
response and literally flood the market.” Moreover, Tom Enders, CEO of Airbus,
cautioned that there might not be sufficient room for six manufacturers.
In December 2010, Airbus announced that it would develop a new version of
the A320 – the A320neo (“new engine option”). The A320neo would be powered
by either CFM International’s LEAP-X engine or the engine in the Bombardier’s
CSeries, Pratt & Whitney’s PW1000G. In March 2011, Airbus announced that it
would raise production of the A320 family from 34 to 36 units per month.
Boeing’s Jim Albaugh acknowledged that the CSeries 300 appeared to target
customers of the Boeing 737. He insisted that Boeing would respond: “I look at
the 737 business that we have and it is one of the cornerstones of Boeing Com-
mercial and it is a marketplace we are going to defend.” Just before the Paris
Air Show, Boeing announced an increase in production of the Boeing 737 from
31.5 units to 42 units per month.
Why did Bombardier wait until securing orders from Lufthansa before launch-
ing the CSeries? Why does Bombardier emphasize the fuel efficiency of its CSeries?
Who among the new entrants that have commenced development – Bombardier,
Introduction to Managerial Economics 3
COMAC, and Irkut – has the best chance of succeeding in competition with Boeing
and Airbus? Should Embraer stay out of the market?
What about Boeing? How should it respond to the new entry? Should it
launch a new product or, like Airbus, modify its existing plane? Did it make
sense to expand production of the 737? Why did Airbus respond to the entry of
Bombardier and COMAC with a new version of the A320 rather than a totally
new plane?
All of these are questions of managerial economics. Managerial economics is
the science of cost-effective management of scarce resources. Wherever resources
are scarce, managers can make more cost-effective decisions by applying the
discipline of managerial economics. The decisions may
regard customers, suppliers, competitors, or the internal Managerial economics:
workings of the organization. Whether the organization is a The science of cost-
effective management of
profit-oriented business, non-profit organization, or house- scarce resources.
hold, managers must make the best use of scarce resources.
Boeing has limited financial, human, and physical resources.
Boeing managers seek to maximize the financial return from these limited
resources. The same is true of Airbus. While Boeing is a publicly traded company
with diversified shareholders, Airbus is controlled by French, German, and Span-
ish corporate shareholders. Despite the differences in organization, the principles
of managerial economics apply to both Airbus and Boeing. Each must compete
effectively against the other and against Bombardier, COMAC, and Irkut, each
must allocate resources to research and development (R&D), each must manage
demand and costs, and each must set prices.
Managerial economics consists of three branches: competitive markets, mar-
ket power, and imperfect markets. Accordingly, this book is organized into three
parts. Before introducing the three branches of managerial economics, let us first
develop some background.
2. Value Added
For the most part, this book takes the viewpoint of a profit-oriented business,
while also considering the management of non-profit organizations and house-
holds. The primary goal of a profit-oriented business is to maximize profit.
Indeed, the aim of competitive strategy is to deliver sustained profit above the
competitive level.
Accordingly, an essential concept for managerial decision-making is economic
profit. To appreciate the concept of economic profit, consider the basic equation
of managerial economics:
Value added = Buyer benefit − Seller cost

= Buyer surplus + Seller economic profit. (1.1)
Buyer
surplus
Value added
Seller
Buyer benefit economic
profit
Buyer’s expenditure =
Seller’s revenue
Seller
cost
FIGURE 1.1 Value added.

Note: Value added=Buyer benefit−Seller cost=Buyer surplus+Seller economic profit.
Source: adapted from Luke M. Froeb and Brian T. McCann, Managerial Economics: A Problem-
Solving Approach, Mason, OH: South-Western, 2010, p. 127.
This equation states that value added is the difference between buyer benefit
and seller cost. It is only to the extent that businesses deliver
Value added: Buyer benefit to buyers that exceeds the cost of production that they
benefit less seller cost. create value. Equation (1.1) is basic to all organizations –
Comprises buyer surplus
whether profit-oriented, non-profits, or households. To cre-
and economic profit.
ate value, they must deliver benefit that exceeds cost. Anyone
who delivers benefit which is less than the cost of production
is destroying value.
Referring to Figure 1.1, value added is shared by buyer and seller. The buyer
gets some part of the value added in buyer surplus, which is the difference
between the buyer’s benefit and their expenditure. The seller gets the other part
of the value added in economic profit, which is the difference between the rev-
enue that the seller receives (equal to the buyer’s expenditure) and the cost of
production.
The larger is the value added, the larger is the amount to be shared by buyer
and seller. For profit-oriented businesses, that means the potential for economic
profit is greater!
The concept of value added applies to governments and non-profits as well.
Suppose that the government provides free healthcare. Since the healthcare is
free, the government receives no revenue. While the healthcare service makes a
financial “loss,” that does not mean that the service is a mistake. The healthcare
provides a benefit. Referring to Figure 1.1, the value added is the buyer benefit
minus the cost of provision. So long as the benefit exceeds the cost, the service
adds value.
PROGRESS CHECK 1A
Explain the relation among the following: buyer benefit, seller cost, value
added, buyer surplus, and economic profit.
3. Decision-Making
The two fundamental decisions in business can be stated simply as participa-

tion (“which”) and extent (“how much”). Which market to enter? How much
to invest? Which products to sell? How much to produce? Which R&D strategy
to follow? How much to spend on R&D? Which job to take? How many hours
to work?
Here, we present fundamental techniques to decide on participation and extent.
Then we discuss psychological limitations in individual decision-making. All
organizations – whether businesses, non-profits, or households – are managed by
individual human beings. To the extent that those individuals are subject to biases,
the biases affect the organizations.
Which and How Much?

The decisions on participation (which) and extent (how much) resolve into
analyzing the total and marginal benefits and costs. Let us use the following exam-
ple to introduce the concepts of total and marginal, and then relate them to the
decisions of which and how much.
Annabel must decide how to invest $10,000. Her bank pays 2% interest on sav-
ings accounts of any amount. The bank also offers a fund in units of $10,000 with
an interest rate of 3%.
If Annabel were to deposit the money in a savings account, her interest income
would be 2% × $10,000 = $200. If Annabel were to deposit the money in the fund,
her interest income would be 3% × $10,000 = $300. So, to maximize her interest
income, Annabel should invest in the fund.
In deciding which investment to make, Annabel should choose according to the
total interest income. She should invest in the fund.
Closely related to the total is the concept of average. Generally, the average
value of a variable with respect to some measure is the total
value of the variable divided by the total quantity of the mea- Average value: The
total value of the variable
sure. Annabel could also choose the investment according to divided by the total
the average interest income. Her average interest income quantity of the measure.
would be 2% from the savings account and 3% from the fund.
Now, suppose that Annabel’s uncle gives her $1,000. How
should she invest the gift? For incremental decisions such as how to invest the
additional $1,000, or generally, how much to invest, the decision-maker should
consider the marginal benefits and costs.
The marginal value of a variable with respect to some measure is the change
in the variable associated with a unit increase in the measure.
If Annabel were to deposit the gift in the savings account, Marginal value: The
change in the variable
her interest income would be 2% × $1,000 = $20. Hence, her associated with a unit
marginal interest income from a deposit of $1,000 in the sav- increase in a measure.
ings account would be $20.
Annabel cannot deposit the gift in the fund because the fund is sold in units of
$10,000. So, the marginal interest income from a deposit of $1,000 in the money
market fund is zero. Thus, to maximize her interest income, Annabel should
deposit the gift in the savings account.
Generally, the marginal value of a variable may be less than, equal to, or greater
than the average value. The relation between the marginal and average values with
respect to some measure depends on whether the average value is decreasing, con-
stant, or increasing with respect to the measure.
PROGRESS CHECK 1B
What would be Annabel’s marginal interest income from deposits of $1,000,
$2,000, . . ., $9,000, $10,000 in the fund?
Here is another example of using total and marginal benefits and costs to
decide on participation and extent. Max is working as an associate in a manage-
ment consulting firm for after-tax earnings of $30,000 per year. Should he get
an MBA, forgoing two years of earnings and possibly securing a higher-paid
job after graduation? After the MBA, having secured a job, how many hours
should he work?
In deciding which career path to follow – whether to continue with his current
job or get an MBA – Max should consider the total earnings (and total costs)
of each alternative. To be precise, since these earnings and costs flow over time,
Max should consider the net present value of the total earnings and total costs
of each alternative. (The next section explains the concept of net present value.)
He should choose the alternative that maximizes the net present value of the total
earnings and total costs.
Having secured a job, Max must decide how many hours to work. In deciding
how much to work, Max should consider the marginal earnings and marginal cost
of each additional hour.
He should work up to the point that the marginal earnings per hour equal
the marginal cost per hour. If the marginal earnings exceed the marginal cost,
then he should work more. The additional earnings would exceed the additional
cost. By contrast, if the marginal earnings are less than the marginal cost, then
he should work less. The reduction in earnings would be less than the reduction
in cost.
Generally,
• in decisions on participation – which market, which product, which job – the

manager should compare the total benefit with the total cost;
• in decisions on extent – how much to invest, how much to produce, how
many hours to work – the manager should compare the marginal benefit
with the marginal cost.
Bounded Rationality
Managers are human and as such are subject to bounded rationality. Typically,
managerial economics models assume that people make decisions rationally, in
the sense that individuals always choose the alternative that maximizes the differ-
ence between benefit and cost.
However, people do not always behave rationally, and indeed, they make sys-
tematic errors in decisions. Human beings behave with bounded rationality (less
than full rationality) because they have limited cognitive abilities and cannot fully
exercise self-control.
Individuals tend to adopt simplified rules in making decisions, especially under
conditions of uncertainty. These simplified rules result in systematic biases,
including the following:
• Sunk-cost fallacy. The psychologists Hal Arkes and Catherine Blumer

gave discounts at random to people buying season theater subscriptions.
Consumers who paid the regular price attended more plays than those
who received the unanticipated discounts. But the subscription was a
sunk cost and did not affect the (forward-looking) benefit from attending
any particular play. However, the experiment showed that consumers
who had incurred a larger sunk cost tended to consume more. The
sunk-cost fallacy leads managers responsible for failing projects to
continue investing, throwing good money after bad, to rationalize sunk
investments.
• Status quo bias. The economists Jack Knetsch and Jack Sinden randomly
gave students either $3 cash or a lottery ticket. The students with $3 cash
were offered the opportunity to buy lottery tickets for $3, while those
with lottery tickets were offered the opportunity to sell the tickets for $3
in cash. Systematically, those with cash preferred to keep the cash, while
those with lottery tickets preferred to keep the tickets. However, if the
status quo had no effect, the proportions of students buying/keeping the
tickets should be the same. Individuals tend (perhaps out of sheer inertia)
to prefer the status quo. Policy-makers in charge of retirement planning
and organ donations exploit status quo bias by designing defaults in favor
of better choices.
• Anchoring. The behavioral economists Amos Tversky and Daniel Kahneman
posed the multiplication problem
8×7×6×5×4×3×2×1=?
to one group of students and the problem
1×2×3×4×5×6×7×8=?
to another group. The students were not given enough time to complete the
calculation. While the correct answer is 40,320, the median estimates were
2,250 in the former group of students and 512 in the latter. The students
anchored on the first few numbers that they could calculate. In online
retailing, sellers exploit consumer anchoring. Some consumers anchor on
the price of the item, ignoring the cost of shipping. So, online retailers set
low prices for products and earn profits on shipping.
To the extent that individuals are subject to bounded rationality, the role for
managerial economics is even larger than when individuals are fully rational. The
techniques of managerial economics help to correct systematic biases in individ-
ual decision-making as well as show how to make better overall decisions.
4. Timing
Managerial economics analysis includes two types of models. Static models

describe behavior at a single point in time, or equivalently, disregard differences
in the sequence of actions and payments. Examples include the model of com-
petitive markets (Chapters 2–5) and the analysis of organizational architecture
(Chapter 14).
By contrast, dynamic models explicitly focus on the timing and sequence of
actions and payments. Examples include games in extensive form (Chapter 10),
one seller’s commitment to increase production and so persuade competitors to
produce less (Chapter 11), and the effect of critical mass in a market with network
effects (Chapter 12).
In dynamic settings, receipts and expenditures often occur at different times.
In principle, one dollar now is different from one dollar in the future. To put the
two amounts on the same basis, the dollar in the future must be discounted to
its present value. While a comprehensive analysis of present value is the subject
of financial theory and outside the scope of this book, a brief introduction here
would be helpful.
Discounting
Investments necessarily involve using resources at certain times in order to receive
benefits at other times. In order to account correctly for the
Discounting: A procedure importance of time for managerial decisions, it is necessary
to transform future dollars
into an equivalent number
to discount future values so that they can be compared with
of present dollars. the present.
If you put $1 in the bank today and it earns 10% interest,
it will grow to $1.10 next year. Hence, this year’s dollars and
next year’s dollars should be treated as if they are measured in different units ( just
as if they were different currencies).
Specifically, divide next year’s dollars by 1.10 to show that $1 next year is equiv-
alent to $1/1.10 this year, or approximately $0.91. The reason why $1 next year
is only worth 91 cents now is that the 91 cents, after growing for one year at
10% interest will become $1 next year. For similar reasons, $1 two years from now
is worth around $0.83 now (since $0.83 now, after growing at 10% interest for two
years, will become $1 in two years’ time).
Present value can be calculated over any period of time – years, months, and
even weeks. Whatever the period, the key is to apply an appropriate discount rate
for that period.
Net Present Value

Evaluating flows of revenues and costs over time requires repeated application of
the principle of discounting. Every dollar amount should be
discounted according to how far in the future it occurs, to Net present value: The
sum of discounted values
evaluate its present value. The net present value (NPV) is of inflows and outflows
the sum of the discounted values of inflows and outflows over time.
over time. Intuitively, the NPV represents the current valua-
tion of various flows of dollars over time.
Consider, for instance, Max’s decision whether to get an MBA. Suppose that he
is limiting his planning to the next five years and that his discount rate is 8% per
year. Assuming no increase in earnings, the NPV of continuing in the current job is
$30,000
, $30,000
, $30,000
, $30,000
$30,000 + + + + = $129,364.
1.08 1.082 083
11.08 1.084
Suppose that the tuition and other costs of the MBA are $50,000 for each of
two years, and that Max expects after-tax earnings of $95,000 after graduation.
Then the NPV of getting an MBA is
$50,000 $95,000 $95,000 $95,000

−$50,000 − + + + = $130,393.
1.08 1.082 083
11.08 1.084
So, Max would get a higher NPV from the MBA.

Using NPV, a manager can evaluate a series of inflows and outflows that occur
at different times from the vantage point of the present. If the NPV is positive,
then the inflows exceed the outflows after accounting for the timing of the inflows
and outflows. Conversely, if the NPV is negative, then the outflows exceed the
inflows. In Max’s case, the NPV of the MBA exceeds the NPV of continuing in
the current job, so he should get the MBA.
The key to calculating the NPV is the discount rate. When borrowing money to
purchase a car, the discount rate should be the interest rate on the loan. When using
money from your bank account to invest in real estate, the discount rate should be
the bank’s interest rate (since you forgo the opportunity to earn the interest).
PROGRESS CHECK 1C
If Max’s discount rate is 10%, should he get the MBA?
5. Organization
Throughout this book, we will take the viewpoint of an organization, which may
be a business, non-profit, or household. Managers of all such organizations face
the same issue of how to effectively manage scarce resources. Since our analysis
focuses on the organization, we first must identify its boundaries. We briefly dis-
cuss this issue here, while leaving the detailed analysis to Chapters 7 and 14.
Organizational Boundaries
The activities of an organization are subject to vertical and horizontal boundaries.
The vertical boundaries of an organization delineate activi-
Vertical boundaries: ties closer to or further from the end user. By contrast, the
Delineate activities closer horizontal boundaries of an organization are defined by the
to or further from the end organization’s scale and scope of operations. Scale refers to
user.
the rate of production or delivery of a good or service, while
scope refers to the range of different items produced or
Horizontal boundaries: delivered.
Defined by the scale In the aircraft manufacturing industry, the vertical chain
and scope of the of production runs from aluminum and other materials,
organization’s operations. to wings, tails, landing gear, engines, and other parts, and,
finally, to assembly of the aircraft. The end users of jet air-
craft are passengers and shippers of freight.
Consider two aircraft manufacturers. Suppose that one produces wings and
landing gear, while the other does not. With regard to vertical boundaries, the
aircraft manufacturer that produces wings and landing gear is more vertically
integrated than the aircraft manufacturer that does not produce wings and land-
ing gear.
With regard to horizontal boundaries, an aircraft manufacturer that produces
planes at the rate of 40 units per month is producing on a larger scale than one
producing at the rate of 30 units per month. An aircraft manufacturer that pro-
duces both commercial and military aircraft is producing with a larger scope than
one that specializes in commercial aircraft.
In the cable TV industry, the vertical chain runs from content, including mov-
ies, sports, and financial information, to programming and, finally, to distribu-
tion. The end users include households and commercial customers such as hotels
and bars.
With regard to vertical boundaries, a cable TV provider that produces movies is

more vertically integrated than a cable TV provider that buys movies from others.
With regard to horizontal boundaries, a cable TV provider that also provides
broadband service is operating with a larger scope than one that specializes in just
cable TV.
Outsourcing
Outsourcing is the purchase of services or supplies from external sources. It is
the opposite of vertical integration, and affects the vertical
boundaries of the organization. If an aircraft manufacturer Outsourcing: The
outsources the production of wings and landing gear, then it purchase of services or
supplies from external
is shrinking its vertical boundaries. Similarly, if a cable TV
sources.
provider outsources the production of movies, it is shrinking
its vertical boundaries.
Owing to declining costs of transport and communications, and falling barriers
to trade and investment, international outsourcing has grown rapidly. Chapter 14
on incentives and organizations discusses outsourcing in detail.
PROGRESS CHECK 1D
Explain the difference between the vertical and horizontal boundaries of an
organization.
NEW BUSINESS ORGANIZATION: PEER-TO-PEER
Some of the fastest-growing businesses challenge conventional thinking

about business organization. Janus Friis and Niklas Zennstrom applied
peer-to-peer technology to develop Skype, software for voice calls over the
Internet. They remarked: “The telephony market is characterized both by what
we think is rip-off pricing and a reliance on heavily centralized infrastructure.
We just couldn’t resist the opportunity to help shake this up a bit.”
Unlike conventional businesses, Skype is located nowhere and everywhere.
It operates from the computers of over 300 million worldwide users and
through the Internet. As for organization, its vertical chain is short, while its
horizontal boundaries are large in terms of scale but, with just one product,
narrow in terms of scope. In May 2011, software publisher Microsoft acquired
Skype for $8.5 billion.
Sources: CNET, “Kazaa founders tout PTP VoIP,” October 19, 2004; Wired, “Microsoft Buys
Skype for $8.5 billion. Why, Exactly?” May 10, 2011.
6. Markets
One concept of managerial economics – the market – is so fundamental that it

appears in the names of each branch of the discipline.
Market: Buyers and A market consists of buyers and sellers who communicate
sellers who communicate
with one another for voluntary exchange. In this sense, a
with one another for
voluntary exchange. market is not limited to any physical structure or particular
location. The market extends as far as there are buyers or
sellers who can communicate and trade at relatively low cost.
Consider, for instance, the market for cotton. This extends beyond the Intercon-
tinental Exchange in New York to growers in Texas and textile manufacturers in
East Asia. If the price on the Intercontinental Exchange increases, then that price
increase will affect Texas growers and Asian textile manufacturers. Likewise, if
the demand for cotton in Asia increases, this will be reflected in the price on the
Exchange.
In markets for consumer products, the buyers are households and sellers are
businesses. In markets for industrial products, both buyers and sellers are busi-
nesses. Finally, in markets for human resources, the buyers are businesses and
sellers are households.
By contrast with a market, an industry consists of businesses engaged in the
production or delivery of the same or similar items. For
Industry: Businesses instance, the clothing industry consists of all clothing manu-
engaged in production or
delivery of the same or
facturers, and the textile industry consists of all textile manufac-
similar items. turers. Members of an industry can be buyers in one market
and sellers in another. The clothing industry is a buyer in the
textile market and a seller in the clothing market.
Competitive Markets
The global cotton market includes many competing producers and buyers.
How should a producer respond to an increase in the price of water, a drop in
the price of cotton, or a change in labor laws? How will these changes affect
buyers?
The basic starting point of managerial economics is the model of competitive
markets. This applies to markets with many buyers and many sellers. The market
for cotton is an example of a competitive market. In a competitive market, buyers
provide the demand and sellers provide the supply. Accordingly, the model is also
called the demand–supply model.
The model describes the systematic effect of changes in prices and other eco-
nomic variables on buyers and sellers. Further, the model describes the interaction
of these changes. In the cotton example, the model can describe how the cotton
producer should adjust prices when the price of water increases, the price of cot-
ton drops, and labor laws change. These changes affect all cotton producers. The
model also describes the interaction among the adjustments of the various cotton
producers and how these affect buyers.
Part I of this book presents the model of competitive markets. We begin with
the demand side, considering how buyers respond to changes in prices and income
(Chapter 2). Next, we develop a set of quantitative methods that support precise
estimates of changes in economic behavior (Chapter 3). Then we look at the sup-
ply side of the market, considering how sellers respond to changes in the prices
of products and inputs (Chapter 4). We bring demand and supply together and
analyze their interaction in Chapter 5, then show that the outcome of market
competition is efficient (Chapter 6).
THE EXTENT OF E-COMMERCE MARKETS
A conventional bricks-and-mortar retail store serves a geographical area

defined by a reasonable traveling time. By contrast, an Internet retailer serves
a much larger market – defined by the reach of telecommunications and the
cost of shipping.
Founded in 1994, Amazon.com began by retailing books. Twenty years
later, by November 2014, its market capitalization of US$141.27 billion was
over 100 times greater than that of the leading conventional bookstore, Barnes
and Noble. The vast disparity reflects the stock market’s assessment of the
difference in the long-term profitability of the two companies.
Besides serving a much larger geographical market, an e-commerce
business can more readily expand into other product lines. Not having to
maintain physical stores, the e-commerce business may achieve lower costs.
Market Power
In a competitive market, an individual manager may have little freedom of action.
Key variables such as prices, scale of operations, and input mix are determined by
market forces. The role of a manager is simply to follow the market and survive.
Not all markets, however, have so many buyers and sellers to be competitive.
Market power is the ability of a buyer or seller to influ-
Market power: The
ence market conditions. A seller with market power will have ability of a buyer or seller
relatively more freedom to choose suppliers, set prices, and use to influence market
advertising to influence demand. A buyer with market power conditions.
will be able to influence the supply of products that it purchases.
A business with market power must determine its horizontal boundaries. These
depend on how its costs vary with the scale and scope of operations. Accordingly,
businesses with market power – whether buyers or sellers – need to understand
and manage their costs.
In addition to managing costs, sellers with market power need to manage their
demand. Three key tools in managing demand are price, advertising, and pol-
icy toward competitors. What price maximizes profit? A lower price boosts sales,
while a higher price brings in higher margins. What is the best way to compete
with other businesses?
Part II of this book addresses all of these issues. We begin by analyzing costs
(Chapter 7), then consider management in the extreme case of market power,
where there is only one seller or only one buyer (Chapter 8). Next, we discuss pric-
ing policy (Chapter 9), and strategic thinking in general (Chapter 10) and in the
context of competition among several sellers or buyers in particular (Chapter 11).
Imperfect Markets
Businesses with market power have relatively more freedom of action than those
in competitive markets. Managers will also have relatively
Imperfect market: One more freedom of action in markets that are subject to imper-
party directly conveys a
fections. A market may be imperfect in two ways: when one
benefit or cost to others,
or one party has better party directly conveys a benefit or cost to others, or when
information than others. one party has better information than others. The challenge
for managers operating in imperfect markets is to resolve the
imperfection and so enable the cost-effective provision of their products.
Consider the market for residential mortgages. Applicants for mortgages have
better knowledge of their ability and willingness to repay than potential lenders.
In this case, the market is imperfect owing to differences in information. The chal-
lenge for lenders is how to resolve the informational differences so that they can
provide loans in a cost-effective way.
Managers of businesses in imperfect markets need to think strategically. For
instance, a residential mortgage lender may require all loan applicants to pay for a
credit evaluation, with the lender refunding the cost if the credit evaluation is
favorable. The lender might reason that bad borrowers would not be willing to pay
for a credit evaluation because they would fail the check. Good borrowers, how-
ever, would pay for the evaluation because they would get their money back from
the lender. Hence, the credit evaluation requirement will screen out the bad bor-
rowers. This is an example of strategic thinking in an imperfect market.
Differences in information and conflicts of interest can cause a market to be
imperfect. The same imperfection can arise within an organization, where some
members have better information than others and interests diverge. Accordingly,
another issue is how to structure incentives and organization.
Part III of this book addresses all of these issues. We begin by considering
the sources of market imperfections – where one party directly conveys a benefit
or cost to others (Chapter 12) and where one party has better information than
others (Chapter 13). Then we study the appropriate structure of incentives and
organization (Chapter 14). Finally, we discuss how government regulation can
resolve market imperfections (Chapter 15).
PROGRESS CHECK 1E
Distinguish the three branches of managerial economics.
KEY TAKEAWAYS
• Managerial economics is the science of cost-effective management of scarce

resources.
• Value added is the difference between buyer benefit and seller cost, and
comprises buyer surplus and economic profit.
• In decisions on participation, compare the total benefit and total cost.
• In decisions on extent, compare the marginal benefit and marginal cost.
• In decision-making, take care to avoid systematic biases, including the sunk-
cost fallacy, status quo bias, and anchoring.
• When evaluating benefits and costs that flow over time, use net present value
with the appropriate discount rate.
• The vertical boundaries of an organization delineate activities closer to or
further from the end user.
• The horizontal boundaries of an organization are defined by the scale and
scope of operations.
• Businesses with market power must manage their costs, pricing, advertising,
and relations with competitors.
• Businesses in imperfect markets should act strategically to resolve the
imperfection.
REVIEW QUESTIONS
1. Explain the difference between value added and economic profit.

2. Consider a charity that gives free mosquito nets to poor people. Since the
charity receives no revenue while mosquito nets are costly, does the free
distribution mean that the charity is destroying value?
3 Give an example in which the marginal exceeds the average value.
4. Give an example in which the marginal is less than the average value.
5. Why do individuals act with bounded rationality?
6. Explain why an employer expecting $1 million of future pension costs need not
provide $1 million today in order to meet the pension fund’s future obligations.
7. Referring to the net present value example in Section 4 above on timing, under
what circumstances, if any, could the NPV be positive?
8. Describe the vertical boundaries of your local cable television provider. In what
ways could the vertical boundaries be expanded or reduced?
9. Describe the horizontal boundaries of your university. In what ways could the
horizontal boundaries be expanded or reduced?
10. In the context of manufacturing Apple iPhones, explain the difference between
outsourcing and vertical integration.
11. Explain the difference between: (a) the market for electricity; and (b) the electricity
industry.
12. True or false?
(a) In every market, all buyers are consumers.
(b) In every market, all sellers are businesses.
13. What is another name for the model of competitive markets?
14. What distinguishes a manufacturer with market power from one without market
power?
15. Should managers operating in an imperfect market: (a) set high prices to make
up for the imperfection; or (b) act strategically to resolve the imperfection?
DISCUSSION QUESTIONS
1. Mercy Hospital provides healthcare at subsidized prices and is so popular

that patients wait in long queues. Revenue is $75 million a year, while the cost
of providing service is $100 million a year. A government subsidy covers the
difference. Some critics argue that, since Mercy Hospital is losing money, it
should be shut down. The management of Mercy Hospital argues that the long
waiting times justify a larger government subsidy to expand staff and facilities.
(a) What is Mercy Hospital’s economic profit?
(b) What is the minimum benefit that Mercy Hospital must provide to add
value?
(c) Do you agree that Mercy Hospital should be shut down?
(d) What information do you need to decide whether Mercy Hospital
should expand or reduce service?
2. Alan and Hilda are clerks at a department store. The store pays each clerk
$10 per hour for a basic eight-hour day, $15 per hour for overtime of up to four
hours, and $20 for overtime exceeding four hours a day.
(a) Alan works 10 hours a day. What are his: (i) marginal pay; and (ii) average
pay?
(b) Hilda works 14 hours a day. What are her: (i) marginal pay; and
(ii) average pay?
(c) Under what pay structure would the marginal pay be less than the
average pay?
3. Ford offers a three-year or 36,0000-mile warranty on the Explorer car. Consumers
must pay for extended warranties beyond the manufacturer’s warranty period.
The Auto Club offers an extended warranty for a Ford Explorer in Hanover, New
Hampshire, at a price of $1,259. This would cover years 4 and 5, after the expiry
of manufacturer’s warranty.
(a) Explain the role of discounting in your decision whether or not to
purchase the extended warranty.
(b) Suppose that the expected cost of repair is $800 in each of years 4
and 5. If your discount rate is 6% per year, should you purchase the
extended warranty?
(c) Would your decision be different if your discount rate were 1% per year?
4. In each of the following instances, discuss whether horizontal or vertical

boundaries have been changed, and whether they were extended or shrunk.
(a) The Canadian manufacturer of regional jets, Bombardier, launched the
CSeries, a family of 100- to 149-seat, long-range jets.
(b) Software publisher Microsoft acquired Skype, a provider of Internet
telephony services.
(c) Conglomerate General Electric divested its subsidiary, NBC Universal,
which merged with cable TV provider, Comcast.
(d) Dutch financial services group ING divested its insurance and
investment management businesses as NN Group.
5. Referring to the definition of a market, answer the following questions:
(a) Opponents of the Iraqi government periodically sabotage the pipelines
through which Iraq exports oil to the rest of the world. Is Iraq part of
the world market for oil?
(b) Prisoners cannot freely work outside jail. How would changes in
prisoners’ wages affect the national labor market?
(c) The Australian national electricity transmission grid links eastern
states including New South Wales, South Australia, and Victoria, but
not Western Australia. How would price changes in Western Australia
affect the other states?
(d) By executive order, President Barack Obama will allow illegal
immigrants to work legally in the United States. How would that affect
US wages?
You are the consultant!

In your organization or personal experience, identify and explain any decisions
that have been systematically biased by: (a) the sunk-cost fallacy, (b) status
quo bias, or (c) anchoring. Explain how the organization or you could have
achieved a better outcome by controlling the bias.
Notes
1 This discussion is based, in part, on Richard Tortoriello, “Aerospace & defense,” Standard &
Poor’s Industry Surveys, February 10, 2011; “Boeing likely to boost 737, 777 production rates,”
ATWOnline, March 18, 2010; “Airbus and Boeing call end to ‘duopoly’,” Financial Times,
June 21, 2011; “Airbus-Boeing duopoly holds narrow-body startups at bay at Paris Air Show,”
Bloomberg, June 23, 2011; “Comac’s C919 jet to complete assembly by September 2015,” South
China Morning Post, September 25, 2014; “Bombardier C Series said to be favourite for Austrian
fleet revamp,” Bloomberg, October 15, 2014.
PA R T
I
Competitive Markets
C H A P T E R
2
Demand
LEARNING OBJECTIVES
• Appreciate why consumers and business buyers buy more at lower
prices.
• Distinguish consumer demand for normal products and inferior
products.
• Appreciate the impact on demand of changes in the prices of sub-
stitutes and complements.
• Understand differences between consumer and business demand.
• Appreciate the concept of buyer surplus.
• Apply package deals and two-part pricing to extract buyer surplus.
1. Introduction
China Mobile is the world’s largest mobile telephone service provider. As of Sep-
tember 2014, China Mobile served 799 million customers, of which 40.95 million
and 244.5 million subscribed to 4G and 3G services, respectively. The company
employed 170,030 persons at the end of 2013.
In his letter to shareholders for the year 2010, Chairman Wang Jianzhou
reported that the company had added 62 million new customers, “a large part of
whom continued to come from rural and migrant markets.” Analysis of the com-
pany’s operational data reveals that, while the company has continued to grow its
customer base, the rate of growth declined from 24% in 2008 to 14% in 2009, and
to 12% in 2010.
22 2 Demand
Chairman Wang also credited the company with achieving average revenue per
user (ARPU) of 73 yuan per month and “a slowdown in decline.” The company’s
ARPU declined by 7% in 2008 and 2009, and by 5% in 2010.
In discussing the outlook for the company, he stressed the “vast potential for
sustainable development.” Further, he reiterated that “Our commitment is unwav-
ering – we will strive to create value for our shareholders.”
Facing saturation in the demand for mobile service in urban areas, China
Mobile has sought growth among rural and migrant consumers. Yet, amidst the
growth, the company faces the perplexing challenge of declining ARPU. Is the
growth of rural and migrant subscribers somehow related to the decline in ARPU?
Like mobile service providers throughout the world, China Mobile offers sub-
scribers a choice of multiple price plans in each market. Plans with higher monthly
fixed charges offer more “free minutes.” How do consumers choose among these
plans, and how should China Mobile price the plans?
This chapter introduces the concept of a demand curve, which describes the
quantity demanded of an item as a function of its price and other factors. Next,
we consider how demand depends on income, the prices of complementary and
substitute products, and advertising. Businesses can use the model of demand
to influence their sales. China Mobile can use the model to understand how the
growth of rural and migrant consumers is systematically related to the decline
in ARPU.
Using the individual buyer’s demand curve, a seller can determine the maxi-
mum that the buyer is willing to pay for any specified quantity, and thereby extract
the highest possible price. We apply this approach to explain how China Mobile
should set prices for mobile service plans.
Finally, we extend from the individual demand to the demand curve of the
entire market. The principles of consumer demand for final goods and services
apply, with suitable adjustments, to business demand for inputs.
2. Individual Demand
To understand how a price cut will affect sales, we need to know how the cut in
price will affect the purchases of the individual buyers and, generally, how an
individual’s purchases depend on the price of the item. The individual demand
curve provides this information: it is a graph that shows the quantity that the
buyer will purchase at every possible price.
Construction
Let us construct Joy’s demand for movies. We must ask Joy a series of questions
that elicit her responses to changes in price. We first ask: “How many movies
would you attend a month at a price of $20 per movie?” Suppose that Joy’s answer
Demand 23
is: “None.” (Strictly, we pose the question holding “other things equal,” because
Joy’s decision may depend on other factors, such as her income.)
We then pose similar questions to Joy for other possible prices for a movie: $19,
$18, ..., $1, and $0. For each price, Joy says how many movies she would attend a
month. Table 2.1 presents this information and represents Joy’s demand for mov-
ies. (Assuming that the consumer’s demand curve is a straight line, we can draw
the demand without filling all the rows of the table.)
Next we graph the information from Table 2.1 as shown in Figure 2.1. We rep-
resent the price of movies on the vertical axis and the quantity in movies a month
on the horizontal axis. (Note that demand and supply curves do not follow the
scientific convention of representing the independent variable on the horizontal
axis and the dependent variable on the vertical axis.)
At a price of $20, Joy says that she would not go to any movies, so mark the
point with the price equal to $20 and quantity of movies equal to zero. Continuing
with the information from Table 2.1, mark every pair of price and quantity that
Joy reports. Joining these points then yields Joy’s demand curve for movies.
Table 2.1 Individual demand
Price ($ per movie) Quantity (movies per month)
20 0
19 1
18 2
... ...
0 20
20 Demand curve
Price ($ per movie)
19
18
0 1 2 19 20
Quantity (movies a month)
FIGURE 2.1 Individual demand curve.

Notes: The individual demand curve shows, for every possible price, the quantity of movies that Joy
will buy. It also shows how much Joy would be willing to pay for various quantities.
24 2 Demand
Knowing Joy’s demand curve, a movie theater can predict how Joy will respond
to changes in its price. For instance, if presently the theater charges $12 per movie,
Joy will buy eight movies a month. If the theater reduces its price to $11, it knows
that Joy will increase consumption to nine movies a month. By contrast, if the the-
ater raises its price to $13 per movie, Joy would cut back to seven movies a month.
Marginal Benefit
The individual demand curve shows the quantity that the buyer will purchase
at every possible price. Let us now consider the individual demand curve from
another perspective.
Referring to Joy’s demand curve in Figure 2.1, we can use the curve to deter-
mine how much Joy would be willing to pay for various quantities of movies.
Specifically, the curve shows that she is willing to pay $19 per movie for one movie
a month. Further, it shows that Joy is willing to pay $18 per movie for two movies
a month, and so on.
Generally, if the number of movies is larger, the price that Joy is willing to pay is
lower. Equivalently, at a lower price, Joy is willing to buy a larger quantity. These
two related properties of a demand curve reflect the principle of diminishing mar-
ginal benefit.
Any item that a consumer is willing to buy must provide some benefit. We mea-
sure the benefit in monetary terms. The marginal benefit is the benefit provided
by an additional unit of the item. The marginal benefit of
Marginal benefit: The the first movie is the benefit from one movie a month. Simi-
benefit provided by an
larly, the marginal benefit of the second movie is the addi-
additional unit.
tional benefit from seeing a second movie each month.
By the principle of diminishing marginal benefit, each
Diminishing marginal additional unit of consumption provides less benefit than the
benefit: Each additional preceding unit. In Joy’s case, this means that the marginal
unit provides less benefit benefit of the second movie is less than the marginal benefit
than the preceding unit.
of the first movie, the marginal benefit of the third movie is
less than the marginal benefit of the second movie, and so on.
Accordingly, the price that an individual is willing to pay will decrease with
the quantity purchased. Hence, the demand curve will slope downward. This is a
general property of all demand curves: the lower the price, the larger will be the
quantity demanded. The fundamental reason for the downward slope is dimin-
ishing marginal benefit.
PROGRESS CHECK 2A
Suppose that the theater presently charges $11 per movie. By how much must
the theater cut its price for Joy to increase her consumption by three movies a
month?
Demand 25
Preferences
The procedure for constructing a demand curve relies completely on the con-
sumer’s individual preferences. The individual decides how much he or she wants
to buy at each possible price. The demand curve then displays information in a
graphical way.
Consumers may have different preferences and hence different demand curves.
One consumer may like red meat while another is a vegetarian. Further, demand
curves will change with changes in the consumer’s preferences. As a person grows
older, her demand for rock videos and athletic events will decline, while her
demand for healthcare and will increase.
GASOLINE PRICES AND DRIVING
Brittany and Danny Schulz live in Murfreesboro, Tennessee. Early in November

2014, Ms Schulz was pleasantly surprised to fill up her Nissan Altima for only
$45, substantially less than the usual $55.
With the average price of gasoline falling below $3 per gallon, American
households will spend more on electronics, holiday gifts, and other
discretionary items. And buy more gasoline. Chris Christopher, director of
consumer markets at market researcher, IHS, predicts: “People are more
likely to drive – and drive longer distances – this season”. Indeed, Mr and Mrs
Shulz are planning a 300-mile road trip to visit friends in Columbus, Georgia.
Source: “Why Americans don’t trust lower gas prices,” Washington Post, November 4, 2014.
3. Demand and Income
Joy’s demand for movies may vary not only with the price of such movies, but also
with her income. If she gets a raise, how would that affect her demand for movies?
Income Changes
Suppose that Joy’s income is presently $50,000 a year. Table 2.1 and Figure 2.1
represent Joy’s demand for movies with an income of $50,000 a year.
We then ask Joy a series of questions. These questions probe the effect of
changes in income as well as price: “Suppose that your income is $40,000 a year.
How many movies would you buy a month at a price of $20 per movie?” We then
repeat the question with other possible prices and tabulate the information.
Suppose that Table 2.2 represents Joy’s answers. We also represent this infor-
mation in Figure 2.2. Marking the pairs of prices and quantities, and joining the
points, we have Joy’s demand curve with an income of $40,000 a year.
26 2 Demand
Table 2.2 Individual demand with lower income
Price Quantity
($ per movie) (movies per month)
20 0
19 0
... 0
10 0
9 2
8 4
... ...
0 20
20 Demand curve with

Price ($ per movie)
$50,000 income
Demand curve with

$40,000 income
10
8
0 4 12 20
FIGURE 2.2 Individual demand curve with lower income.

Note: As Joy’s income falls from $50,000 to $40,000, her entire demand curve shifts toward the left.
Joy’s demand curve for movies with $40,000 income lies to the left of her demand
curve with $50,000 income. At every price, the quantity demanded with $40,000
income is less than or equal to the quantity demanded with $50,000 income.
Referring to Figure 2.1, if the price of movies drops from $8 to $7 per movie,
while Joy’s income remains unchanged at $50,000 a year, we trace Joy’s response
by moving along her demand curve from the $8 level to the $7 level. By contrast,
referring to Figure 2.2, if her income falls from $50,000 to $40,000, while the price
remains at $8 per movie, we represent Joy’s response by shifting the entire demand
curve to the left. The essential reason for this difference is that the figure, having
just two axes, does not explicitly represent the buyer’s income.
Let us understand the difference in graphical representation between a change
in price and a change in income in another way. On Figure 2.2, at the $8 level, mark
two quantities: a quantity of 12 movies a month when Joy’s income is $50,000,
and another quantity of four movies a month when Joy’s income is $40,000.
Demand 27
Can we join these points to form a demand curve? The answer is definitely “no,”
because each point corresponds to a different income and different demand curve.
A demand curve shows how a buyer’s purchases depend on changes in the price of
some item, holding income and other factors unchanged. Accordingly, for each of
the points, there is a separate demand curve.
In general, we represent a change in the price of the item by a movement along
the demand curve. By contrast, we represent a change in income or any factor
other than the price of the item by a shift in the entire demand curve.
Normal and Inferior Products

When Joy’s income drops from $50,000 to $40,000 a year, her demand for movies
shifts to the left. As Joy’s income falls, her demand for movies also falls. By con-
trast, if her income were to rise, her demand would increase.
Let us compare Joy’s demand for movies in general with her demand for after-
noon matinees. If Joy’s income falls, it is quite possible that she will substitute
cheaper forms of entertainment for more expensive ones. In particular, the drop in
her income may lead to an increase in her demand for afternoon matinees.
By contrast, when Joy’s income increases, we can expect her to switch away
from cheaper forms of entertainment and toward more expensive alternatives. So,
as her income rises, Joy’s demand for afternoon matinees will fall.
Goods and services can be categorized according to the effect of changes in
income on demand. If the demand for an item increases as the buyer’s income
increases, while the demand falls as the buyer’s income falls, then the item is a
normal product. Equivalently, the demand for a normal
product is positively related to the buyer’s income. Normal product:
By contrast, the demand for an inferior product is nega- Demand is positively
tively related to the buyer’s income. This means that the related to buyer’s income.
demand falls as the buyer’s income increases, while the
demand increases as the buyer’s income falls. Inferior product:
For Joy, movies are a normal product, while afternoon Demand is negatively
matinees are an inferior product. Generally, broad catego- related to buyer’s income.
ries of products tend to be normal, while particular products
within the categories may be inferior.
Consider, for instance, consumer electronics. While consumer electronics as a
category are a normal product, particular products such as all-in-one stereos may
be inferior. In passenger transport services, the entire category is probably a nor-
mal product. Within the category, taxis are a normal product while buses may be
an inferior product.
The distinction between normal and inferior products is important for busi-
ness strategy. When the economy is growing and incomes are rising, the demand
for normal products will rise, while the demand for inferior products will fall.
By contrast, when the economy is in recession and incomes are falling, the
demand for normal products will fall, while the demand for inferior products
will rise.
28 2 Demand
The distinction is also useful in international business. The demand for nor-
mal products is relatively higher in richer countries, while the demand for inferior
products is relatively higher in poorer countries. For instance, in developed coun-
tries, relatively more people commute to work by car than bicycle. The reverse is
true in very poor countries.
PROGRESS CHECK 2B
Draw a curve to represent an individual consumer’s demand for all-in-one
stereos. (a) Explain why it slopes downward. (b) How will a drop in the con-
sumer’s income affect the demand curve?
CHINA MOBILE: GROWING SUBSCRIBERS, SHRINKING ARPU
With prepaid mobile service, subscribers pay for specific quantities of air-
time in advance. By contrast, with contract (also called “post-paid”) service,
subscribers enter into an agreement for a minimum period, and may use any
quantity of air-time subject to paying the monthly bill.
Prepaid service and contract service cater to different market segments.
Typical subscribers to prepaid service include consumers with lower income,
such as rural people, migrant workers, and students.
Analysts estimate that, in 2011, over 90% of China Mobile’s subscribers
bought prepaid service. By contrast, 10 years earlier, the prepaid proportion
was 48%. The growth in prepaid customers is consistent with China Mobile’s
push to rural and migrant consumers.
People with lower income typically spend less on consumer goods. China
Mobile would derive lower ARPU from rural and migrant consumers. As a
result, the company’s quest for growth among rural and migrant consumers
is unavoidably associated with lower ARPU.
Sources: Business Monitor International, China Telecommunications Report – Q2 2011;
China Mobile Ltd.
4. Other Factors in Demand
The individual demand may depend on other factors besides the price of the item
and the buyer’s income. The other factors may include the prices of related prod-
ucts, advertising, durability, season, and location. Here, we focus on the prices of
related products and advertising.
Complements and Substitutes

Assume that Joy always eats popcorn when she goes to the movies. How will an
increase in the price of popcorn affect Joy’s demand for movies? A change in the
Demand 29
price of popcorn will affect Joy’s purchases of movies at all movie prices. Hence,
it will cause a shift in the entire demand curve.
Suppose that presently the price of popcorn is $1. Figure 2.3 represents Joy’s
demand curve for movies when the price of popcorn is $1. We next construct Joy’s
demand when the price of popcorn is $2. To do so, we ask Joy how many movies
she would see at various movie prices if the price of popcorn were $2. Figure 2.3
shows the demand curve: when the price of popcorn is higher, the demand curve
for movies is further to the left.
In general, related products can be classified as either com-
plements or substitutes according to the effect of a price
Complements: An
increase in one product on the demand for the other. Two increase in the price of
products are complements if an increase in the price of one causes a fall in the
one causes the demand for the other to fall. By contrast, two demand for the other.
products are substitutes if an increase in the price of one
causes the demand for the other to increase.
For Joy, popcorn and movies are complements. The more
Substitutes: An increase
movies she sees, the more popcorn she will want. Hence, if in the price of one
the price of popcorn is higher, the price of the overall movie causes an increase in the
experience will be higher, and she will go to fewer movies. demand for the other.
How will an increase in the price of online movies affect
Joy’s demand for movies in the theater? Instead of going to
the theater, Joy could watch an online movie. Accordingly, these two products are
substitutes. If there is an increase in the price of online movies, Joy’s demand for
theater movies will increase.
Generally, a demand curve will shift to the left if there is either an increase
in the price of a complement or a fall in the price of a substitute. By contrast,
20
Demand curve
Price ($ per movie)
with $1 popcorn
Demand curve
with $2 popcorn
0 20
FIGURE 2.3 Individual demand curve with more expensive complement.

Note: As the price of popcorn increases, Joy’s entire demand curve shifts toward the left.
30 2 Demand
a demand curve will shift to the right if there is either a fall in the price of a com-
plement or an increase in the price of a substitute.
PROGRESS CHECK 2C
Referring to Figure 2.1, how would a fall in the price of online movies affect
the original demand curve?
GASOLINE PRICES AND THE DEMAND FOR SUVS AND

COMPACT CARS
Sport-utility vehicles (SUVs) like the Cadillac Escalade, Chevrolet Tahoe, and
Lincoln Navigator consume more fuel than compact cars like the Chevrolet
Aveo. Gasoline is a complement to motor vehicles, while compact cars are a
substitute for SUVs.
With gasoline prices falling, the American demand for SUVs and trucks
has increased. J.D. Power and Associates survey consumers on factors that
influence their car buying. Between June 2013 and 2014, fuel economy dropped
from third to seventh in the list, behind reliability, styling, brand preference, ride
and handling, reputation, and price.
Nevertheless, taking a long-term view, Ford redesigned its popular F-series
truck, replacing steel with aluminum to reduce the vehicle weight by 700
pounds. CEO Mark Fields remarked that consumers know that oil prices can
go up as quickly as down and want fuel economy.
Source: “Cheap gas won’t let us off the hook: Ford CEO,” CNBC, November 12, 2014.
Advertising
Advertising expenditure is another factor in demand. For instance, Joy’s demand
for movies may depend on advertising by the theater. Generally, an increase in the
seller’s advertising will increase the buyer’s demand. We represent this by shifting
the buyer’s demand curve to the right.
Advertising may be informative or persuasive. Informative advertising commu-
nicates information to potential buyers and sellers. For instance, movie theaters
list the movies that they are showing and their show times in the daily newspapers.
These listings inform potential customers. Persuasive advertising aims to influ-
ence consumer choice. Manufacturers of cigarettes and cosmetics, for instance,
use advertising to retain the loyalty of existing consumers and attract others to
switch brands.
Demand 31
5. Business Demand
Movies and mobile telephone service are consumer goods. By contrast, TV adver-
tisements and excavators are industrial goods – they are bought by businesses
rather than consumers. In any economy, the majority of economic transactions
are business-to-business sales. Accordingly, managers must understand the prin-
ciples of business demand.
Inputs
Consumers buy goods and services for final consumption or usage. By contrast,
businesses buy goods and services not for their own sake but to use them as inputs
in the production of other goods and services. Chapter 4 will provide a detailed
analysis of business operations. Here, we will review only the essentials necessary
to understand the business demand for inputs.
The inputs purchased by a business can be classified into materials, energy,
labor, and capital. Businesses use these inputs to produce goods and services for
sale to consumers or other businesses. For example, an express delivery service
uses human resources, equipment, and energy to deliver documents and packages
for consumers and businesses.
The inputs may be substitutes or complements. For the express delivery service,
trucks and drivers are complements: a truck without a driver is quite useless, as
is a driver without a truck. Other inputs may be substitutes: the service can use
workers or machines to sort packages and to load shipments.
Demand
A business produces items for sale to consumers or other businesses. The business
earns revenues from sales. By increasing inputs, the business can produce a larger
output and raise revenue. Accordingly, the business can measure its marginal ben-
efit from an input as the increase in revenue arising from an additional unit of the
input.
Using the marginal benefit of an input, we can construct the individual demand
curve of a business. This shows the quantity of the input that the business will
purchase at every possible price. A business should buy an input up to the quan-
tity that its marginal benefit from the input exactly balances the price.
Suppose that, with a larger quantity of an input, each additional unit of the
input generates a smaller increase in revenue. This means that the input provides
a diminishing marginal benefit to the business.
Hence, when the price is lower, the business will buy a larger quantity, and con-
versely, when the price is higher, the business will buy a smaller quantity. Thus, the
demand curve for the input slopes downward.
32 2 Demand
Demand Factors
A change in the price of an input is represented by a movement along the demand
curve. By contrast, changes in other factors will lead to a shift of the entire demand
curve. Business demand does not depend on income but rather on the output of the
item being produced. If the output is larger, then the business will increase its demand
for inputs. If, however, the output is lower, then the demand for inputs will be lower.
The business demand for an input also depends on the prices of complements
and substitutes in the production of the output. For instance, delivery trucks
and drivers are complements, hence an increase in drivers’ wages will reduce the
demand for trucks.
6. Buyer Surplus
The individual demand curve shows the quantity that the buyer will purchase at
every possible price. The demand curve also shows the maximum amount that a
buyer is willing to pay for each unit of the item.
The perspective of “willingness to pay” is important for pricing policy as it
shows the maximum that the buyer can be charged. We will explain two pricing
schemes to extract the maximum that the buyer is willing to pay, and so maximize
the seller’s profit.
Benefit
The demand curve shows the buyer’s marginal benefit from each unit. Using this
information, we can calculate the buyer’s total benefit,
Total benefit: The benefit
which is the benefit provided by all the units that the buyer
provided by all the units
that the buyer consumes. consumes. The total benefit is the marginal benefit from the
first unit plus the marginal benefit from the second unit, and
so on, up to and including marginal benefit from the last
unit that the buyer purchases.
A buyer’s total benefit is the maximum that the buyer is willing to pay. Graph-
ically, the total benefit is represented by the area under the buyer’s demand curve
up to and including the last unit consumed.
Let us calculate Joy’s total benefit from eight movies a month. Her total benefit
is the area under her demand curve up to and including eight movies a month. In
Figure 2.4, this is area 0b d = 12 ($8 8) + $12 × 8 = $128 . So, the maximum that Joy
would be willing to pay for eight movies a month is $128.
Benefit and Expenditure

Suppose that the price of a movie is $12. Then Joy buys eight movies a month.
As already calculated, her total benefit would be $128. Joy, however, needs to pay
only $12 × 8 = $96, which is less than her total benefit.
Demand 33
20 d
Price ($ per movie)
a c
12
10 f
e
b
0 8 10 20
FIGURE 2.4 Individual buyer surplus.

Notes: The individual buyer surplus is the buyer’s total benefit from some quantity of purchases less
the actual expenditure. At a price of $12, Joy’s buyer surplus is the shaded area acd between the
demand curve and the $12 line.
The difference between a buyer’s total benefit from some consumption and her
actual expenditure is the buyer surplus. At the price of $12
per movie, Joy’s buyer surplus is $128 − $96 = $32. Buyer surplus: The
Referring to Figure 2.4, at the price of $12, her expen- difference between the
diture on eight movies a month is represented by the area buyer’s total benefit from
0bcad under the $12 line up to and including eight movies a some consumption and
actual expenditure.
month. Hence, Joy’s buyer surplus is the difference between
the area 0bcad which represents total benefit and the area
0bca which represents expenditure, or the area acd between her demand curve
and the $12 line.
Generally, provided that purchases are voluntary, a buyer must get some sur-
plus. If not, he or she would not buy. The maximum that a seller can charge is
the buyer’s total benefit. If a seller tries to charge more, then the buyer will walk
away.
PROGRESS CHECK 2D
Suppose that the price of movies is $8. On Figure 2.4, mark Joy’s buyer surplus.
Price Changes
Referring to Joy’s demand for movies, at the price of $12, Joy goes to eight movies
a month, and her buyer surplus is area acd. Now suppose that the price drops to
$10. Then Joy will raise her attendance from eight to ten movies a month. Her
34 2 Demand
buyer surplus will increase by area efca: she gets the eight original movies at a
lower price, and she goes to more movies.
Generally, a buyer gains from a price reduction in two ways. First, the buyer
gets a lower price on the quantity that they would have purchased at the original
higher price. Second, she will buy more, gaining buyer surplus on each of the
additional purchases. The extent of the second effect depends on the buyer’s sen-
sitivity to the price reduction. The greater the increase in purchases, the larger will
be the buyer’s gain from the price reduction.
Similarly, a buyer loses from a price increase in two ways – the buyer must pay
a higher price, and will buy less.
Package Deals and Two-Part Pricing

A seller who has complete flexibility over pricing maximizes profit by charging
each buyer just a little less than total benefit. Practically, this can be implemented
in two ways – through package-deal pricing and two-part pricing.
Let us explain these pricing policies in the context of Jania’s demand for mobile
telephone calls, shown in Figure 2.5. If the mobile telephone provider charges a
price of 10 cents per minute, Jania will make 100 minutes of calls a month. The
provider would earn revenue of 100 × 10 cents = $10. Jania would get a buyer
surplus of 12 × 40 × 100 cents = $20 .
Jania’s total benefit from 100 minutes of calls a month is the area under her
demand curve from 0 to 100 minutes, or 100 × 12 × (50 + 10) = $30 $ . Suppose that
the mobile provider offers Jania a package deal of 100 minutes of calls at $29 with
no other alternative. A package deal is a pricing scheme
Package deal: A pricing comprising a fixed payment for a fixed quantity of consump-
scheme comprising a
tion. The package deal will give Jania a buyer surplus of
fixed payment for a fixed
quantity of consumption. $30 − $29 = $1. Since there is no other alternative, Jania will
buy this package.
50
Price (cents per minute)
Demand curve
Total
10
benefit
0 100 125
Quantity (minutes per month)
FIGURE 2.5 Package deal.

Note: Jania will buy a package deal of 100 minutes of calls for any price not exceeding her total
benefit of $30.
Demand 35
The mobile provider would earn $29 in revenue, which is almost three times the
revenue from the 10-cent pricing policy. The package deal enables the service to
soak up almost all of Jania’s buyer surplus. Since Jania’s consumption is the same
under 10-cent and package-deal pricing, the provider’s cost would be the same.
So, its profit would definitely be higher.
The mobile provider could also extract Jania’s buyer surplus through two-part
pricing. A two-part price is a pricing scheme comprising a
Two-part price: A pricing
fixed payment and a charge based on consumption.
scheme comprising a fixed
Suppose that the mobile provider offers a two-part plan payment and a charge
comprising a $19 monthly charge and an airtime charge of based on consumption.
10 cents a minute. Referring to Figure 2.5, under this plan,
Jania would buy 100 minutes of calls a month. Her total benefit
would be the area under her demand curve from 0 to 100 minutes, which we earlier
calculated to be $30. She would pay a $19 monthly charge and 100 × 10 cents = $10
in airtime charges. Hence, her buyer surplus would be $30 − $19 − $10 = $1.
Just like the package deal, the two-part price enables the service provider to
soak up most of the consumer’s buyer surplus. The mobile provider would earn
$29 in revenue and the cost would be the same as under 10-cent pricing.
Providers of many services, including banking, car rentals, telecommunications,
and Internet access, make use of package deals and two-part pricing. They also com-
bine the two pricing techniques so that the monthly charge covers a specified quan-
tity of “free” usage while the user must pay a usage charge beyond the free quantity.
CHINA MOBILE: “WORLDWIDE CONNECT” SERVICE
China Mobile offers nine price plans for its “Worldwide Connect” service in
the city of Nanjing. The cheapest plan provides 350 “free minutes” of calls
for 68 yuan per month. The price of additional calls is 0.29 yuan per minute.
Suppose that Figure 2.6 depicts Lin Jun’s demand curve for mobile telephone
service. The 68 yuan per month plan involves two-part pricing. The plan provides
350 free minutes. At the 351th minute, her marginal benefit exceeds 0.29 yuan
per minute. Accordingly, she would consume beyond 350 minutes. Specifically,
she would consume more minutes up to the quantity where her marginal benefit
is exactly 0.29 yuan per minute. That quantity is 400 minutes per month.
At that level of consumption, Lin Jun’s total benefit would be (0.29 × 400 +
0.5 × 1.00 × 400) = 116 + 200 = 316 yuan per month. Her buyer surplus would
be her total benefit less the monthly fee and less the airtime charge, or 316 −
68 − 0.29 × 50 = 233.50 yuan per month.
Realistically, Lin Jun can choose among the various plans offered by China
Mobile and other service providers. Suppose, however, that China Mobile
is the only provider and offers only the plan with 350 free minutes. Then
China Mobile could raise the monthly charge by 233 yuan. Lin Jun would still
subscribe and get a buyer surplus of 0.50 yuan per month.
36 2 Demand
1.29
Price (yuan per minute)

Demand curve
0.29
0 350 400
Quantity (minutes per month)
FIGURE 2.6 Demand curve for mobile telephone service.

Notes: The 68 yuan per month plan provides 350 free minutes. Lin Jun would consume
beyond the free minutes up to the quantity where her marginal benefit is exactly 0.29 yuan
per minute.
7. Market Demand
For strategy, marketing, and other purposes, businesses may plan on the basis of
the entire market rather than individual customers. They must then understand
the demand of the entire market.
The market demand curve is a graph that shows the quantity that all buyers
will purchase at every possible price. The market demand
Market demand curve: curve is constructed in a similar way to the individual
A graph showing the demand curve. To construct the market demand for an item,
quantity that all buyers ask each potential buyer the quantity that they would buy at
will purchase at every every possible price. Then, at each price, add the reported
possible price.
individual quantities to get the quantity that the market as a
whole will demand. (The appendix to this chapter explains
how to construct the market demand curve by horizontal summation of the indi-
vidual demand curves.)
The properties of the market demand curve are similar to those of the individ-
ual demand curve. As each buyer’s marginal benefit diminishes with the quantity
of consumption, the market demand curve slopes downward. Equivalently, at a
lower price, the market as a whole will buy a larger quantity.
The market demand depends on other factors that affect individual demand.
The market demand for a consumer good depends on buyers’ incomes, the prices
of related products, and advertising. The market demand for a business input
depends on output of the business and the prices of related products.
Demand 37
The market buyer surplus is the difference between the buyers’ total benefit
from consumption and the buyers’ actual expenditure. Graphically, it is the area
between the market demand curve and the price line.
BOMBARDIER CSERIES
In the 1990s, with the price of oil around US$20 per barrel, the cost of fuel
amounted to less than 20% of an airline’s operating costs. By 2010, the price
climbed above US$100 per barrel, and airline fuel costs ballooned to as high
as 40% of operating costs, exceeding the cost of labor. However, with the
expansion of US oil production and weakening growth in demand, the price
of oil has since moderated to around US$50 per barrel.
Depending on their expectations about the future price of oil, airlines
may be willing to pay for new aircraft that provide greater fuel efficiency.
Bombardier’s new CSeries promises to reduce fuel consumption by 20%
through two innovations. One is use of advanced materials – comprising
46% lightweight composites and 24% aluminium lithium – in construction
of the airframe. The other innovation is a more fuel-efficient engine, Pratt &
Whitney’s PW1000G.
Sources: Zubin Jelveh, “Flying on empty,” Portfolio.com, June 4, 2008; Bombardier Incorporated.
KEY TAKEAWAYS
• Owing to diminishing marginal benefit, consumers and business buyers buy

more at lower prices.
• Consumer demand for normal products increases with income, while consumer
demand for inferior products decreases with income.
• The demand for a product increases with the price of a substitute, and
decreases with the price of a complement.
• Business demand increases with the output of the item being produced.
• Buyer surplus is the difference between the buyers’ total benefit from
consumption and the buyers’ actual expenditure.
• A seller can extract the buyers’ surplus and raise profit by selling through
package deals and two-part pricing.
REVIEW QUESTIONS
1. Think of a good or service that you bought recently. Would you have bought
less of the item if the price had been lower? Explain why or why not.
2. Define (a) normal product, and (b) inferior product, and give examples to
illustrate your definition.
38 2 Demand
3. Think of a good or service that you bought recently. Would you have bought
more or less of the item if your income had been lower? Explain why or why not.
4. Define what is meant by (a) a substitute, and (b) a complement, and give
examples to illustrate your definition.
5. A new birth-control device protects women against pregnancy but not sexually
transmitted diseases. How will this new product affect the demand for: (a) male
condoms; (b) birth-control pills?
6. How does Pepsi advertising affect the demand for: (a) Pepsi; and (b) Coca-Cola?
7. A key component of mobile phones is the microprocessor. Explain how
changes in consumer incomes affect Apple’s demand for microprocessors.
8. Why are automated teller machines (ATMs) relatively more common in countries
with higher labor costs?
9. Explain the meaning of buyer surplus.
10. Buyer surplus applies to consumer demand but not business demand. True or
false?
11. Passengers on a London to Sydney flight compared fares and discovered that
they had paid different prices for the same flight. Explain how this illustrates
that the market demand curve is downward-sloping.
12. The price of mobile telephone calls is 10 cents a minute. Antonella buys 200
minutes a month. Illustrate her demand curve and identify her buyer surplus.
13. “Summer sale: the more you buy, the more you save.” Comment.
14. What is a package deal? How can a broadband service provider use package
deals to increase profit?
15. What is two-part pricing? How can a broadband service provider use two-part
pricing to increase profit?
1. An important issue in economic development is the relationship between fertility

and female literacy. With data from 110 countries, the following linear relationship
between the female literacy rate and the number of births per woman (fertility
rate) was estimated: when the literacy rate is 0%, the fertility rate is 7.63 per
woman; and when the literacy rate is 100%, the fertility rate is 2.42 per woman.
(a) On a figure with female literacy on the vertical axis and fertility on the
horizontal axis, draw a graph of the linear relationship. Referring to the
linear relationship, if the literacy rate is 60%, what would be the fertility
rate (approximately)?
(b) A large cost of having a baby is the time that the mother must invest
in bearing and rearing the child. For a more educated woman, is the
value of this time higher or lower?
(c) In the chart, mark “cost of child” on the vertical axis. Does the trend
line have any relation to a demand curve? Explain.
(d) Give an alternative explanation of the figure: use the fertility rate to
explain female literacy.
2. Sprint is one of the largest mobile service providers in the USA. In March
2011, Sprint had 13.1 million prepaid customers, and 33 million post-paid
Demand 39
customers. Sprint’s prepaid brands include Virgin Mobile USA, Boost Mobile,
and Assurance Wireless. Sprint’s ARPU from prepaid customers was $28 per
month, as compared with $56 per month from post-paid customers.
(a) Explain the meaning of normal and inferior products.
(b) Apply the concepts of normal and inferior products to prepaid and
post-paid mobile telephone service.
(c) Refer to Sprint’s ARPU from prepaid and post-paid customers. Are the
data consistent with your answer to (b)?
3. In fall 2005, an ABC News poll reported that 46% of lower-income Americans
(earning less than $100,000 a year) were angry about gasoline prices, while just
32% of higher-income people were angry. Despite higher gasoline bills, just 22%
of survey respondents said that they had reduced driving. However, 63% said
that they would cut back on driving if gasoline prices rose above $3 per gallon.
(a) Use a demand curve to explain why the proportion of people who
would reduce driving is higher when the price of gasoline is higher.
(Hint: You are free to assume any data necessary to draw the demand.)
(b) Explain how people can adjust to higher gasoline prices by replacing
their cars rather than driving less.
(c) Why were relatively fewer higher-income people “angry” about high
gasoline prices?
4. The Coca-Cola Company markets the Coke brand and manufactures
concentrate for sale to regional bottlers. Coke bottlers mix concentrate with
sweetener and water to produce the soft drink for supermarkets, restaurants,
and other retail outlets. Possible sweeteners include corn syrup and sugar.
Owing to federal restrictions against imports, sugar is relatively more expensive
in the United States than in the rest of the world.
(a) Why do US soft drink bottlers use relatively more corn syrup than
bottlers elsewhere in the world?
(b) Draw a US Coke bottler’s demand for corn syrup. (Hint: You are free to
assume any data necessary to draw the demand.)
(c) Use your figure to explain how the following changes would affect
the Coke bottler’s demand for corn syrup: (i) removal of the federal
restrictions against sugar imports; (ii) a fall in the price of corn syrup;
and (iii) an increase in the sales of Pepsi.
(d) Who benefits and who loses from the federal restrictions against sugar
imports?
5. Historically, Bombardier specialized in producing regional jets, which are
smaller short-range jets with up to 100 seats. In 2008, it secured a launch
customer for the new CSeries, a family of 100- to 149-seat mid-range aircraft,
which is scheduled to enter service in late 2015. The CSeries will reduce fuel
consumption by 20% through use of advanced materials and a more fuel-
efficient engine, the Pratt & Whitney PW1000G.
(a) Draw the market demand for jet aircraft. (Hint: You are free to assume
any data necessary to draw the demand.)
(b) Use your figure to explain how changes in the consumer demand for
air travel affect the airline demand for jet aircraft.
40 2 Demand
(c) If travelers are less sensitive to fare increases, how would that affect
the demand for fuel-efficient aircraft?
(d) How does the price of oil affect the demand for advanced materials?
6. More than half of all residential water connections in Britain are not metered.
Residential customers pay a flat fee regardless of usage. Scottish Water, which
supplies water in Scotland, says that it has no evidence that “installation of
meters encourages lower than normal usage of water.” (Source: “Will switching
to a water meter save money?” Guardian, July 8, 2014.)
(a) Suppose that the Salmond home water supply is not metered, and
the family consumes 10,000 gallons a month. Illustrate the family’s
monthly demand curve for water, assuming that the demand curve is
a straight line, and, if the price is £ 50 per 1,000 gallons, the Salmond
family would consume nothing.
(b) Calculate the total benefit and marginal benefit from water when the
Salmond family consumes 10,000 gallons a month. What is the family’s
buyer surplus?
(c) Suppose that Scottish Water installs a water meter at the Salmond
home and charges a price of £ 5 per 1,000 gallons. How much water
would the family buy and how much would it spend each month?
(d) What is the maximum that Scottish Water could charge the Salmond
family for the consumption in (c)?
(e) Suppose that, with metering, the Salmonds’ neighbors spend more
than the Salmond family on water each month. Does this imply that the
neighbors get more benefit from water?
7. Students at the University of California, Los Angeles, enjoy several privileges.
One is a good education at a low price. Another is California’s unbeatable
weather, and a third is access to discounted movie tickets. Suppose that Alan
buys 12 tickets a year at $7 rather than the full price of $10. By how much does
he gain from the discount scheme?
(a) One answer is that Alan “saves” $10 − $7 = $3 on each ticket, which sums
to a total of $3 × 12 = $36. Explain why this overestimates Alan’s gain.
(b) Using a suitable diagram and the concept of buyer surplus, explain
how much Alan gains from the discount scheme.
(c) Taking account of the discount movie scheme, by how much could the
University raise Alan’s tuition fees?
8. Suppose that the typical buyer of a packaging machine has a straight-line
individual demand curve for packaging materials. At a price of $5 per kilogram
of materials, she would buy zero, while at a price of $1 per kilogram, she would
buy 100,000 kg per year. The buyer plans to use the machine for one year.
Production costs $150,000 for the machine and $1 per kilogram of materials.
(a) On a figure with price per kilogram of materials on the vertical axis and
quantity per year on the horizontal axis, draw the demand curve.
(b) Suppose that the manufacturer sells the machine bundled with 100,000 kg
of material. What is the maximum that the manufacturer can charge for
the bundle? What would be the manufacturer’s profit per customer?
Demand 41
(c) Suppose that the manufacturer sets a two-part pricing policy, comprising
a price for the machine and a price of $2 per kilogram of materials.
What is the maximum price, F, that the manufacturer can charge for the
machine? What would be the manufacturer’s profit per customer?
(d) Suppose that the manufacturer sets a two-part pricing policy, with
price, F, for the machine and a price of $2 per kilogram of materials.
After buying the machine, the customer is influenced by sunk costs
and her demand curve shifts up by $1 at all levels of consumption.
On the figure in (a), draw the new demand curve. What would be the
manufacturer’s profit?
9. China Mobile offers nine price plans for its “Worldwide Connect” service in the
city of Nanjing. The cheapest plan provides 350 “free minutes” of calls for 68
yuan per month, while the next cheapest plan provides 450 “free minutes” of
calls for 88 yuan per month. Under both plans, the price of additional calls is
0.29 yuan per minute.
(a) Lin Jun’s demand curve for mobile calling is a straight line. Two points
on her demand curve are: (i) at a price of 1.29 yuan per minute, a quantity
of 0 minutes; and (ii) at a price of 0.29 yuan per minute, a quantity of
400 minutes. Draw her demand curve.
(b) Suppose that Lin Jun subscribes to the 88 yuan per month plan. (i) How
much calling time would she consume? (ii) What would be her total
benefit? (iii) What would be her buyer surplus (benefit less charges)?
(c) Replicate (b) assuming that Lin Jun subscribes to the 68 yuan per
month plan. Note that she might decide to buy additional minutes
beyond the free minutes.
(d) Which plan should she choose to maximize buyer surplus?

Consider a product that your organization sells.
(a) What customer segments buy the product?
(b) Describe the current pricing policy.
(c) Under the current pricing policy, do any of the customer segments enjoy
buyer surplus?
(d) Explain how you could use the techniques of package-deal or two-part
pricing to extract the buyer surplus and raise profit.
Appendix: Constructing Market Demand by

Horizontal Summation
This chapter introduces the concept of market demand, which shows the quantity
that all buyers will buy at every possible price. One way to construct the market
demand is to ask each potential consumer the quantity that they would buy at every
42 2 Demand
Table 2.3 Market demand
Price ($) (per movie) Joy Max Lucas Market
20 0 0 0 0
19 1 0 0 1
18 2 0 0 2
... ... 0 0 ...
10 10 10 0 20
8 12 14 2 28
... ... ... ... ...
0 20 30 10 60
Price ($ per movie)
20
15
10
0 10 20 30 60
FIGURE 2.7 Market demand curve.

Notes: The market demand curve is the horizantal summation of the individual demand curves.
At every price, the market quantity demanded is the sum of the individual quantities demanded.
possible price. Then, at each price, sum the reported individual quantities to get the
quantity that the market as a whole will buy. Table 2.3 reports the market demand.
Another way to construct the market demand is by horizontal summation of the
individual demand curves. In Figure 2.7, we draw the individual demand curves
of the three potential consumers – Joy, Max, and Lucas. Horizontal summation
means adding the curves in the horizontal direction.
Each individual demand curve shows the number of movies that the consumer
would buy at every possible price. So, at every price, add the quantities that the
three consumers would buy to obtain the quantity that the market as a whole will
buy. Figure 2.7 depicts the market demand as a dashed line.
C H A P T E R
3
Elasticity
LEARNING OBJECTIVES
• Understand the concepts of own-price elasticity, income elasticity,
cross-price elasticity, and advertising elasticity of demand.
• Distinguish price elastic and price inelastic demand.
• Appreciate the intuitive factors underlying the elasticity of demand.
• Appreciate that, if demand is inelastic, the seller can increase profit
by raising the price.
• Understand the impact of adjustment time on the elasticity of
demand for non-durables and durables.
• Appreciate how behavioral biases affect the elasticity of demand.
1. Introduction
In 2005, the online retailer Amazon launched its “Prime” service, which provides
free two-day shipping with no requirement of a minimum purchase, unlimited
photo storage, and exclusive access to movies, music and Kindle books for a fixed
annual fee of $79. In January 2014, when reporting its fourth quarter financial
results, Amazon revealed that it was considering raising the price of the Prime
service by between $20 and $40.1
The investment bank UBS surveyed Amazon Prime subscribers: 58% said that
they would renew if Amazon raised the price by $20, while 24% would renew if
the price rose by $40. On February 12, UBS reduced its rating of Amazon shares
from “Buy” to “Neutral,” and the price of Amazon shares fell.
44 3 Elasticity
Separately, stockbroker Piper Jaffray surveyed 500 Amazon Prime subscribers,

asking how likely they would be to renew their subscription if the price were raised
to $109. Although 66% of respondents answered that they would be unlikely or
highly unlikely to renew, analyst Gene Munster argued that, at most, the highly
unlikely group or about a quarter of subscribers would not renew. Projecting from
prior increases in prices by Netflix and Redbox, Mr Munster concluded that “the
two comparable price increases point to a mid single digit percent decrease in
Prime customers if Amazon were to raise prices.”
In March 2014, Amazon announced a $20 increase for the Prime service. On
the day of the announcement, the price of Amazon shares rose by 87 cents or
0.2%. Wedbush Securities analyst Michael Patcher remarked: “A lot of people say
they would quit, but the truth is most people get used to the convenience . . . . It’s
like higher fees on credit cards. People complain about them bitterly when they go
up, but most never get around to actually changing their service.” Wedbush Secu-
rities estimated that Amazon had 25 million Prime subscribers and that the price
increase would add about $500 million in revenue and operating profit per year.
How would a price increase affect the demand for Amazon’s Prime service?
How would the price increase affect Amazon’s revenues? By how much should
Amazon adjust the price of the Prime service?
To address these questions, we apply the concept of elasticity. The elasticity
of demand measures the responsiveness of demand to changes in an underlying
factor, such as the price of the item, the prices of complementary or substitute
products, buyers’ income, and advertising expenditure. There is an elasticity cor-
responding to every factor that affects demand.
The own-price elasticity of demand measures the responsiveness of the quan-
tity demanded to changes in the price of the item. With the own-price elasticity,
a seller can estimate the impact of an increase or reduction in price on quantity
demanded. The seller can then estimate the impact on buyers’ expenditure and its
own revenue.
Using the own-price elasticity of the demand for the Prime service, Amazon
could estimate the impact of the March 2014 price increase on demand and reve-
nues. A price increase will always reduce the quantity demanded. We show that a
price increase will raise consumers’ expenditure and sellers’ revenue if demand is
price inelastic, but reduce consumers’ expenditure and sellers’ revenue if demand
is price elastic.
Next, we introduce the concepts of income, cross-price, and advertising elastic-
ities of demand. Then we discuss how elasticities vary with the time for adjust-
ment, and, finally, how behavioral biases affect elasticities.
The concept of elasticity is essential to gauging the impact of changes in prices,
incomes, and other factors on demand, consumers’ expenditure, and sellers’ rev-
enue. It is fundamental to the management of both profit-oriented businesses
and non-profit organizations. For instance, just like online retailers Amazon and
Alibaba, hospitals and city transport systems need the own-price elasticity to
gauge the impact of price increases on demand, revenue, and profit.
Elasticity 45
2. Own-Price Elasticity
To gauge the impact of changes in price on quantity demanded, we need a measure

of buyers’ sensitivity to price changes – the own-price elasticity of demand. The
concept of own-price elasticity of demand is so basic that it is
often called simply the price elasticity or demand elasticity. Own-price elasticity of
demand: The percentage
The own-price elasticity of demand is the percentage by which quantity
by which the quantity demanded will change if the price of demanded will change
the item rises by 1%. Equivalently, the own-price elasticity is if price of the item rises
the ratio by 1%.
Proportionate change in quantity demanded

(3.1)
Proportionate change in pr
p ice
or
Percentage change in quantity demanded

. (3.2)
Percentage change in price
Estimation
To estimate the own-price elasticity of demand, collect records of price changes
and the corresponding changes in quantity demanded. Then calculate the own-
price elasticity as the ratio of the proportionate (percentage) change in quantity
demanded to the proportionate (percentage) change in price.
To illustrate, Figure 3.1 represents the demand for cigarettes. Presently, the
price of cigarettes is $1 a pack and the quantity demanded is 1.5 billion packs a
month. According to Figure 3.1, if the price rises to $1.10 per pack, the quantity
demanded would drop to 1.44 billion packs.
The proportionate change in quantity demanded is the change in quantity
demanded divided by the initial quantity demanded. The change in quantity
demanded is 1.44 − 1.5 = −0.06 billion packs, and the initial quantity demanded
is 1.5 billion packs. Hence, the proportionate change in quantity demanded is
−0.06/1.5 = −0.04.
Similarly, the proportionate change in price is the change in price divided
by the initial price. The change in price is $1.10 − $1 = $0.10 per pack, while
the initial price is $1 per pack. Hence, the proportionate change in price
is 0.1/1 = 0.1.
Thus, by equation (3.1), the own-price elasticity of the demand for cigarettes
is −0.04/0.1 = −0.4. Equivalently, in this example, the percentage change in quantity
demanded was −4%, while the percentage change in price was 10%. By equation (3.2),
the own-price elasticity is −4/10 = −0.4.
46 3 Elasticity
Price ($ per pack)
1.1
0 1.44 1.5
Quantity (billion packs a month)
FIGURE 3.1 Calculating own-price elasticity.

Note: The own-price elasticity of the demand for cigarettes is the proportionate change in quantity
demanded divided by the proportionate change in price = −0.04 ÷ 0.1 = −0.4.
Properties
The cigarette example illustrates several properties of the own-price elasticity of
demand. First, as discussed in Chapter 2, demand curves generally slope down-
ward: if the price of an item rises, the quantity demanded will fall. Hence, the
own-price elasticity is a negative number. For ease of interpretation, some analysts
report own-price elasticities as an absolute value, that is, without the negative sign.
When applying the concept, note that the own-price elasticity is a negative number.
Second, the own-price elasticity is a pure number, independent of the units of
measurement. The quantity demanded of cigarettes is measured in billion packs per
month. The proportionate change in quantity demanded, however, is the change
in quantity demanded divided by the initial quantity demanded. Hence, it is a pure
number that does not depend on any units of measurement: the proportionate change
would be the same whether we measure quantity demanded in billions, millions, or
thousands of packs. Likewise, the proportionate change in price is a pure number.
Since the own-price elasticity is the proportionate change in quantity demanded
divided by the proportionate change in price, it is also a pure number. Thus, the
own-price elasticity of demand provides a handy way of characterizing price sen-
sitivity that does not depend on units of measurement. Hence, it can be used to
compare the price sensitivity of the demand for different goods and services.
Third, recall from equation (3.1) that the own-price elasticity is the ratio of the
proportionate change in quantity demanded to the proportionate change in price. If
a very large proportionate change in price causes no change in quantity demanded,
then the elasticity will be 0. By contrast, if an infinitesimal percentage change in
price causes a large change in quantity demanded, then the elasticity will be nega-
tive infinity. Accordingly, the own-price elasticity ranges from 0 to negative infinity.
Elasticity 47
Accuracy
The estimate of the own-price elasticity depends on the calculation of the pro-
portionate change, and specifically the denominator of the proportionate change.
Equation (3.1) uses the initial prices and quantities as the denominator, but the
calculation could also use the average or final prices and quantities. As we consider
smaller and smaller price changes, the estimate of the own-price elasticity will con-
verge to a single number, which is called the “point estimate” of the elasticity.
Evidently, from equation (3.1), the formula using the initial price/quantity as
the denominator of the proportionate change in price/quantity is not workable if
the initial price/quantity is zero. In that case, we should use either the average or
final price/quantity as the denominator.
Also, note that the own-price elasticity is a measure that depends on all factors
that affect demand – including price, income, prices of complementary and sub-
stitute products, and sellers’ advertising. So, changes in any of these may lead to
a change in the own-price elasticity.
In particular, the own-price elasticity may vary along the demand curve, and
so vary with changes in the price itself. Hence, strictly, the own-price elasticity is
accurate only for small changes in the price.
PROGRESS CHECK 3A
Referring to Figure 3.1, suppose that, initially, the price of cigarettes is $1.10 a
pack and the quantity demanded is 1.44 billion packs a month. Then the price
falls to $1 per pack and the quantity demanded rises to 1.5 billion packs. Cal-
culate the own-price elasticity.
AMAZON PRIME: EFFECT OF PRICE INCREASE
Wedbush Securities estimated that, at the original price of $79 a year, Amazon
Prime had 25 million subscribers. According to the UBS survey, if Amazon raised
the price by $20, subscriptions would fall by 42% to 14.5 million. The $20 increase
amounts to a change in price of 20/79 = 25.3%. So, the predicted change in
demand implies that the own-price elasticity of demand is −42/25.3 = −1.66.
Also according to UBS survey, if Amazon raised the price by $40, subscrip-
tions would fall by 76% to 6 million. Hence, between the prices, $99 and $119,
an increase of 20/99 = 20.2%, subscriptions would fall from 14.5 to 6 million,
that is, by 58.6%. Thus, over this range, the predicted change in demand
implies that the own-price elasticity of demand is −58.6/20.2 = −2.90.
Sources: “Will Amazon Prime customers accept price hike? Maybe not,” Forbes, February 12,
2004; Wedbush Securities, “Quick note: Amazon.com (AMZN – NEUTRAL),” March 13, 2014.
48 3 Elasticity
3. Elastic/Inelastic Demand
The demand for an item is described as price elastic or elastic with respect to
price if a 1% increase in price leads to more than a 1% drop in quantity demanded.
Equivalently, demand is price elastic if a price increase causes
Price elastic: A price a proportionately larger reduction in quantity demanded.
increase causes a
If the demand is elastic, the elasticity is less than −1. This
proportionately larger
reduction in quantity means that the absolute value (that is, without the negative
demanded. sign) of the elasticity exceeds 1.
By contrast, the demand is described as price inelastic
or inelastic with respect to price if a 1% price increase causes
Price inelastic: A price less than a 1% drop in quantity demanded. Equivalently,
increase causes a demand is price inelastic if a price increase causes a propor-
proportionately smaller
tionately smaller reduction in quantity demanded.
reduction in quantity
demanded. If the demand is inelastic, the elasticity exceeds −1. This
means that the absolute value (that is, without the negative
sign) of the elasticity is less than 1.
Intuitive Factors
To estimate the own-price elasticity requires information on a change in price and
the corresponding change in quantity demanded. However, changing the price
to estimate the elasticity may be too costly or not practical. As an alternative,
managers can consider several intuitive factors to gauge the own-price elasticity
of demand.
• Availability of direct or indirect substitutes. The fewer substitutes that are

available, the less elastic will be the demand. People who are dependent on
alcoholic drinks or cigarettes feel that there is no other way to satisfy their
needs. Hence, the demand for these items is relatively inelastic.
There are fewer substitutes for a product category than for specific products
within a category. Consider, for instance, the demand for beer compared
with the demand for a particular brand. The particular brand has many
more substitutes than the category as a whole. Accordingly, the demand for
the brand will tend to be more elastic than the demand for the category. This
means that, if beer manufacturers can collectively raise prices by 10%, their
sales will fall by a smaller percentage than if a single manufacturer increases
its price by 10%.
• Buyer’s prior commitments. A person who has bought a particular car becomes
a captive customer for spare parts. Automobile manufacturers understand
this very well. They set relatively higher prices on spare parts than on new
cars. The same applies as well in the software business. Once users have
invested time and effort to learn one program, they become “locked in” for
future upgrades. Whenever there is such a lock-in, demand is less elastic.
Elasticity 49
Typically, commitments unwind over time. For instance, subscribers to

24-month mobile service contracts will be free to switch after 24 months.
Accordingly, the effect of buyer commitments on the own-price elasticity of
demand will diminish over time.
• Benefits/costs of economizing. Buyers have limited time to spend on searching for
better prices, so they focus attention on items that account for relatively larger
expenditures. Parents, for instance, spend more time economizing on diapers
than on cotton buds. Similarly, office managers focus attention on copying
paper rather than on paper clips. Marketing practitioners have given the name
“low involvement” to products that get relatively little attention from buyers.
The balance between the benefit and cost of economizing also depends on
a possible split between the person who incurs the cost of economizing and
the person who benefits. If you bring a damaged car for repair, the repair
manager will surely ask: “Are you covered by insurance?” Car owners who
are covered by insurance care less about price. They get the benefit of the
repair work, while the insurer pays most or all the costs. A car owner who
bargains over the repairs must spend his or her own time, while the insurer
will enjoy most of the saving.
PROGRESS CHECK 3B
What are the intuitive factors that influence the own-price elasticity of demand?
SHARED COSTS: FREQUENT FLYER PROGRAMS
Whenever there is a split between the person who pays and the person who
chooses the product, the demand will be less elastic. In 1981, American
Airlines established its AAdvantage program for frequent flyers. This program
gives free flights and other awards according to the member’s travel on
American Airlines.
The AAdvantage program does not give mileage credit for travel on competing
airlines such as United or Delta. So, it provides members with a strong
incentive to concentrate travel on American Airlines.
The AAdvantage program is especially attractive to travelers, such as
business executives, who fly at the expense of others. Such travelers are
relatively less price sensitive than those who pay for their own tickets.
AAdvantage gives them an incentive to choose American Airlines even if
the fare is higher. Among customers who fly at the expense of others, the
program makes demand relatively less elastic.
AAdvantage was a brilliant marketing strategy, and the other major airlines
soon established their own frequent flyer programs.
50 3 Elasticity
4. Forecasting
The own-price elasticity of demand can be applied to forecast the effect of price
changes on quantity demanded and buyers’ expenditure. Expenditure is related
to the quantity demanded, since expenditure equals the quantity demanded mul-
tiplied by the price. (In Chapter 9 on pricing, we consider the possibility of price
discrimination. With price discrimination, different buyers pay different prices, so
expenditure is not simply quantity demanded multiplied by price.)
The own-price elasticity concept can be applied at the level of an entire market
as well as for individual sellers. From the standpoint of an individual seller, the
quantity demanded is sales, while buyers’ expenditure is revenue. Hence, using the
own-price elasticity of demand, the seller can forecast the effect of price changes
on sales and revenue.
Quantity Demanded
Let us first consider how to use the own-price elasticity of demand to forecast
the effect of price changes on the quantity demanded. Refer to the demand for
cigarettes in Figure 3.1. Suppose that the price is $1 a pack and the quantity
demanded is 1.5 billion packs a month. How would a 5% increase in the price
affect the quantity of cigarettes that buyers demand?
Above, we calculated the own-price elasticity of demand at the $1 price to
be −0.4. By definition, the own-price elasticity is the percentage by which the
quantity demanded will change if the price rises by 1%. Hence, if the price of
cigarettes increases by 5%, then the quantity demanded will change by −0.4 × 5 =
−2%, that is, the quantity demanded will fall by 2%.
To forecast the change in quantity demanded in terms of cigarettes, multiply
the percentage change of −2% by the quantity demanded before the price change.
Accordingly, the 5% price increase would change the quantity demanded by −2% ×
1.5 billion = 30 million packs a month.
As the cigarette example illustrates, the rule for estimating the impact of prices
changes on buyers’ quantity demanded is:
Proportionate change in quantity demanded

= Proportionate change in price × Own-price elasticity of demand. (3.3)
We can also use the above rule to estimate the effect of a reduction in the price
on the quantity demanded. Suppose that the price of cigarettes is initially $1 a
pack and then falls by 5%. The quantity demanded will change by −0.4 × (−5%) =
2%, that is, it will increase by 2%. This example shows that it is important to keep
track of the signs of the own-price elasticity and the price change.
Elasticity 51
Expenditure
Next, let us see how to use the own-price elasticity of demand to estimate the effect
of changes in price on buyers’ expenditure. Buyers’ expenditure equals the quantity
demanded multiplied by the price. Hence, a change in price will affect expenditure
through the price itself as well as through the related effect on quantity demanded.
Generally, the rule for estimating the impact of prices changes on buyers’
expenditure is:
Proportionate change in expenditure = Proportionate change in price

+ Proportionate change in quantity demanded. (3.4)
Consider the effect of an increase in price. By itself, the price increase will tend
to raise the expenditure. The price increase, however, will reduce the quantity that
buyers demand, which would tend to reduce the expenditure. Hence, the net effect
on expenditure depends on which effect is relatively larger.
The concept of own-price elasticity helps to determine whether the price or
quantity effect is relatively larger. Recall that demand is elastic with respect
to price if an increase in price causes a proportionately larger fall in quantity
demanded, while demand is inelastic if a price increase causes a proportionately
smaller fall in quantity demanded. The own-price elasticity enables us to compare
the relative magnitude of changes in price and quantity demanded.
If the demand is price elastic, then the drop in the quantity demanded will be
proportionately larger than the increase in price, and hence, the price increase
will reduce expenditure. Generally, if demand is price elastic, a price increase will
reduce expenditure while a price reduction will increase expenditure.
By contrast, if the demand is price inelastic, the drop in quantity demanded will
be proportionately smaller than the increase in price, and then the price increase
will increase expenditure. Generally, if demand is price inelastic, a price increase
will increase expenditure while a price reduction will reduce expenditure.
To clarify further, substitute from equation (3.3) for the proportionate change
in quantity demanded in (3.4). Then the rule for estimating the impact of prices
changes on buyers’ expenditure simplifies to:
Proportionate change in expenditure

= Proportionate change in price × (1 + Own-price elasticity of demand).
(3.5)
Pricing Strategy
Whenever sales managers are asked to raise prices, their most frequent response
is: “But my sales would drop!” Since demand curves slope downward, it certainly
52 3 Elasticity
is true that a higher price will reduce sales. The real issue is the extent to which
the price increase will reduce sales. Sales managers ought to be thinking about the
own-price elasticity of demand.
To explain, suppose that a manufacturer’s demand is price inelastic at the current
price. What if the manufacturer raises the price? Since demand is price inelastic,
the price increase will lead to a proportionately smaller reduction in the quantity
demanded. The buyers’ expenditure will increase, which means that the manufac-
turer’s revenue will increase.
Meanwhile, owing to the reduction in quantity demanded, the manufacturer
can reduce production, cutting its costs. Since revenues will be higher and costs
will be lower, the manufacturer’s profits definitely will be higher. Accordingly, if
demand is price inelastic, a seller can increase profit by raising price.
As this discussion shows, under the right conditions (inelastic demand), a price
increase can raise profits even though it may cause sales to drop. Therefore, when
setting the price for an item, managers should focus on the own-price elasticity
of demand. Generally, the price should be raised until the demand becomes price
elastic. We will develop this idea further in Chapter 9 on pricing.
PROGRESS CHECK 3C
Suppose that the own-price elasticity of the demand for a particular mobile
phone service provider is −2.5. If the service provider raises price by 7%,
what would be the proportionate effect on quantity demanded and buyers’
expenditure?
NEW YORK TIMES: MAY 2009 PRICE INCREASES
In 2008, the New York Times Media Group earned revenues of $668 million
from circulation, $1.077 billion from advertising, and $181 million from other
sources. Facing competition from new media and free newspapers, the Group
management decided on a strategy to raise circulation prices and trim less
profitable readership. In May 2009, the New York Times planned to raise its
weekday cover price from $1.50 to $2, and its Sunday cover price from $5 to $6.
The price increase was estimated to raise revenue by $40 million.
How reasonable is the estimate of the impact on revenue? We can address
this question by using the rule for calculating the impact of a price increase
on buyers’ expenditure, equation (3.5).
At the current circulation of 1.04 million and price of $1.50, and assuming
300 weekdays a year, the current annual revenue from weekday sales of the
Times is 1.04 × $1.50 × 300 million = $468 million. So, the price increase was
estimated to raise revenue by 40/468 = 8.5%.
Elasticity 53
Focusing on the price increase for weekday papers from $1.50 to $2, the
percentage change in price would be (2.00 − 1.50)/1.50 = 0.50/1.50 = 33%.
Using equation (3.5), the proportionate change in revenue is 0.085 = 0.33 × (1 +
Own-price elasticity of demand). Hence, the own-price elasticity of demand
is 0.085/0.33 − 1 = 0.26 − 1 = −0.74.
For the Times price increase to raise revenue by $40 million, the own-price
elasticity of demand must be −0.74. This assumption on the elasticity seems
quite reasonable.
However, commentators on the price increases did not mention the
impact on advertising revenues. The demand for advertising depends on
the circulation of the newspaper. Assuming that the own-price elasticity of
demand is −0.74, the 33% price increase would have changed circulation by
33% × −0.74 = −24%. The 24% reduction in circulation would have substantially
reduced advertising revenue.
Sources: “New York Times set to increase price,” Financial Times, May 2, 2009; New York
Times Company, Annual Reports 2008 and 2009.
5. Other Elasticities
In addition to price, the demand for an item also depends on buyers’ incomes, the
prices of complementary and substitute items, and sellers’ advertising. Changes in
any of these factors shift the demand curve.
There is an elasticity to measure the responsiveness of demand to changes in
each factor. Managers can use these elasticities to forecast the effect of changes
in these factors. In particular, the elasticities can be used to forecast the effect of
changes in multiple factors that occur at the same time.
The analysis of elasticities of demand with respect to income, the prices of
complementary and substitute items, and sellers’ advertising is quite similar to
that for the own-price elasticity. Accordingly, we will focus on the differences for
the other elasticities.
Income Elasticity
The income elasticity of demand measures the sensitivity
of demand to changes in buyers’ incomes. By definition, the Income elasticity: The
percentage by which
income elasticity of demand is the percentage by which
demand will change if
the demand will change if the buyers’ incomes rise by 1%. buyers’ incomes rise
Equivalently, the income elasticity is the ratio by 1%.
Percentage change in demand

. (3.6)
Percentage change in buyers’ income
54 3 Elasticity
For a normal product, if buyers’ incomes rise, the demand will rise, so the
income elasticity will be positive. By contrast, for an inferior product, if incomes
rise, demand will fall, so the income elasticity will be negative. Thus the sign of the
income elasticity will depend on whether the product is normal or inferior. Hence,
it is important to note the sign of the income elasticity. The income elasticity can
range in value from negative infinity to positive infinity.
Demand is described as income elastic or elastic with respect to income if a
1% increase in income causes more than a 1% change in demand. Demand is
described as income inelastic or inelastic with respect to income if a 1% increase
in income causes less than a 1% change in demand.
The demand for necessities tends to be relatively less income elastic than the
demand for discretionary items. Consider, for instance, the demand for raw food
as compared with restaurant meals. Eating in a restaurant is more of a discretion-
ary item as compared to cooking at home. Accordingly, the demand for raw food
is relatively less income elastic than the demand for restaurant meals.
Cross-Price Elasticity
Just as the income elasticity of demand measures the sensitivity of demand to
changes in income, the cross-price elasticity measures the sen-
Cross-price elasticity: sitivity of demand to changes in the prices of related prod-
The percentage by which
demand will change if
ucts. By definition, the cross-price elasticity of demand
the price of a related item with respect to another item is the percentage by which the
rises by 1%. demand will change if the price of the related item rises by
1%. Equivalently, the cross-price elasticity is the ratio

. (3.7)
Percentage change in price of related item
e
If two products are substitutes, an increase in the price of one will increase the
demand for the other, so the cross-price elasticity will be positive. The more sub-
stitutable are two items, the higher their cross-price elasticity will be. By contrast,
if two products are complements, an increase in the price of one will reduce
demand for the other, and so the cross-price elasticity will be negative. The cross-
price elasticity can range from negative infinity to positive infinity.
Advertising elasticity:
The percentage by which Advertising Elasticity
demand will change The advertising elasticity measures the sensitivity of demand
if sellers’ advertising
expenditure rises by 1%.
to changes in the sellers’ advertising expenditure. By defini-
tion, the advertising elasticity of demand is the percentage
Elasticity 55
by which the demand will change if the sellers’ advertising expenditure rises by 1%.
Equivalently, the advertising elasticity is the ratio

. (3.8)
Percentage change in sellers’ advertisin
i g expenditure
Generally, an increase in advertising will raise the demand, and so the elastic-
ity will be positive. Most advertising is undertaken by individual sellers to pro-
mote their own business. By drawing buyers away from competitors, advertising
has a much stronger effect on the sales of an individual seller than on the market
demand. Accordingly, the advertising elasticity of the demand faced by an individ-
ual seller tends to be larger than the advertising elasticity of the market demand.
Forecasting Multiple Factors

The business environment will often change in conflicting ways. For instance,
incomes may rise, while the prices of substitutes and complements rise as well.
For a normal product, the higher income would raise demand, the higher price of
substitutes would also raise demand, but the higher price of complements would
reduce demand.
What is the net effect on demand? This question can be addressed using the
elasticities with respect to each of the factors affecting demand. Generally, the
percentage change in demand due to changes in multiple factors is the sum of
the percentage changes due to each separate factor.
To illustrate, suppose that the price of cigarettes is $1 per pack and sales are
1.5 billion packs a month. Then the price increases by 5%, while buyers’ incomes
rise by 3%. What would be the net impact on demand?
Suppose that the own-price elasticity of the demand for cigarettes is −0.4 and
the income elasticity of demand is 0.1. Then the 5% increase in price would change
the quantity demanded by −0.4 × 5% = −2%. Further, the 3% increase in incomes
would change demand by 0.1 × 3% = 0.3%. Hence, the net effect of the increases
in price and incomes is to change demand by −2% + 0.3% = −1.7%.
Originally, the quantity demanded of cigarette services was 1.5 billion packs
per month. After the increases in price and incomes, the quantity demanded will
be 0.983 × 1.5 billion = 1.475 billion packs per month. A similar approach can
be used to forecast the effects of changes in other factors, including the prices of
related products and sellers’ advertising expenditures.
PROGRESS CHECK 3D
Refer to Section 2 of this chapter for the properties of own-price elasticity.
What are the corresponding properties of the income elasticity?
56 3 Elasticity
GASOLINE PRICES AND CAR CHOICE
With gasoline prices falling below $3 a gallon, Americans are buying more
SUVs (sport-utility vehicles) and pick-up trucks. Mike Jackson, chief executive
of auto retailer AutoNation, remarked: “Americans just love big.”
Falling gasoline prices have reinforced an existing trend of buying larger
vehicles. Between October 2014 and 2015, the percentage of SUVs and
trucks rose from 68.5% to 72% of Ford’s sales.
Brian Johnson, auto analyst at investment bank Barclays, emphasized that
buying a new SUV could save gasoline: “A new Ford Escape [crossover SUV]
is more fuel efficient than a 10-year-old Camry”.
Indeed, in the United States between 1997 and 2005, consumers adjusted
to higher gasoline prices by scrapping less fuel-efficient cars and buying more
fuel-efficient ones. The cross-price elasticity for fuel economy with respect to
gasoline prices was 0.02 in the short run and 0.20 in the long run.
Source: Shanjun Li, Christopher Timmins, and Roger H. von Haefen, “How do gasoline prices
affect fleet fuel economy?” American Economic Journal: Economic Policy, Vol. 1, No. 2,
August 2009, pp. 113–137.
6. Adjustment Time
We have analyzed the elasticities of demand with respect to changes in price,

income, the prices of related products, and advertising expenditures. In addition,
another factor affects all elasticities: the time available for buyers to adjust.
With regard to adjustment time, it is important to distin-
Short run: Time horizon guish the short run from the long run. The short run is a time
within which buyers horizon within which a buyer cannot adjust at least one item
cannot adjust at least one of consumption. By contrast, the long run is a time horizon
item of consumption. long enough for buyers to adjust all items of consumption.
To illustrate the distinction, consider how Fred commutes
into Chicago. He does not have a car, so he takes the train. To
Long run: Time horizon switch from the train to a car, he needs time to buy or lease a
long enough for buyers car. Accordingly, with regard to Fred’s choice of transport,
to adjust all items of
consumption.
the short run is any period of time shorter than that which he
needs to get a car. The long run is any period of time longer
than that which he needs to get a car.
Let us now discuss the effect of adjustment time on the elasticities of demand,
and how the effect depends on whether the item is durable or non-durable.
Non-durables
Consider an everyday item such as commuter train services. Suppose that one Monday
morning, the train operator announces a permanent 10% increase in fares. Many
Elasticity 57
commuters may have already made plans for that day, so the response to the higher
fare may be quite weak on that day. Over time, however, the response will be stron-
ger: as more commuters acquire cars, the demand for train services will drop.
Generally, for a non-durable good, the longer the time
Non-durable good:
that buyers have to adjust, the bigger will be the response to a
Demand is more elastic in
price change. Accordingly, the demand for such items will be long run than short run.
more elastic in the long run than in the short run. This applies
to all non-durable items, including both goods and services.
Figure 3.2 illustrates the short- and long-run demand for a non-durable item.
Suppose that the current price is $5 and quantity demanded is 1.5 million units.
If the price drops to $4.50, the quantity demanded will rise to 1.6 million units in
the short run and 1.75 million units in the long run.
At the price of $5 and quantity demanded of 1.5 million units, the proportion-
ate change in the price is ($4.50 − $5)/$5 = −$0.50/$5 = −0.1. In the short run, the
proportionate change in the quantity demanded is (1.6 − 1.5)/1.5 = 0.1/1.5 = 0.067.
Accordingly, the short-run own-price elasticity is 0.067/(0.1) = −0.67.
In the long run, the proportionate change in the quantity demanded is (1.75 − 1.5)/
1.5 = 0.25/1.5 = 0.167. Thus, the long-run own-price elasticity is 0.167/(−0.1) = −1.67.
This confirms that the demand is more elastic in the long run than in the short run.
Durables
The effect of adjustment time on the demand for durable goods such as cars is
somewhat different. For both durables and non-durables, buyers need time to
adjust, which causes demand to be relatively more elastic in the long run.
However, for durables only, a countervailing effect causes demand to be rela-
tively more elastic in the short run. This countervailing effect is especially strong
with respect to changes in income.
Price ($ per unit)
4.5
long-run demand
short-run demand
0 1.5 1.6 1.75

Quantity (million units a month)
FIGURE 3.2 Short- and long-run demand for a non-durable item.

Note: If the price drops from $5 to $4.50, the quantity demanded will rise to 1.6 million units in the
short run and 1.75 million units in the long run.
58 3 Elasticity
Consider, for instance, the demand for cars. Most drivers buy cars at intervals
of several years. Suppose that there is a drop in incomes. Then drivers will plan
to keep their cars longer. Some drivers, who were just about to replace their cars,
will put off the decision to do so. So the drop in incomes will cause purchases to
dry up until sufficient time passes that these drivers want to replace their cars at
the new lower income.
However, in the long run, the effect on sales will be more muted: eventually, all
drivers will replace their cars, but less frequently. Thus, the drop in income will
cause demand to fall more sharply in the short run than in the long run.
Similarly, if income rises, drivers will replace their cars more frequently. Some
drivers will find that they want to replace their cars immediately, causing a boom
in purchases. This boom, however, will last only as long as it
Durable good: Demand takes all such drivers to adjust to their new replacement fre-
may be more or less
quency. Thus, the increase in income will tend to cause demand
elastic in long run than
short run, depending on to increase more sharply in the short run than in the long run.
the balance between Accordingly, for a durable good, the difference between short-
time for adjustment and and long-run elasticities of demand depends on a balance between
replacement frequency. the need for time to adjust and the replacement frequency effect.
PROGRESS CHECK 3E
For a non-durable good, explain why the long-run demand is more elastic than
the short-run demand.
DEMAND FOR GASOLINE AND CARS: EFFECT OF TIME
People buy gasoline to fuel their cars. So the demand for gasoline depends
on the ownership of cars and their fuel efficiency. As households change their
cars over time, their demand for gasoline will also change.
A Canadian study estimated the demand for gasoline as measured by the
fuel consumption per unit of distance. The own-price elasticity was −0.029 in
the short run and −0.089 in the long run. Apparently, taking into consideration
the long-run impact of gasoline on car ownership, the demand was three
times more price elastic in the long run than in the short run.
The same study also estimated the demand for car ownership. The income
elasticity of the demand for cars was 0.285 in the short run and 0.391 in the
long run. Among Canadians, the long-run demand for cars was 37% more
income elastic than in the short run. The estimates suggest that the effect of
adjustment time outweighed the replacement frequency.
Source: Philippe Barla, Bernard Lamonde, Luis F. Miranda-Moreno, and Nathalie Boucher,
“Traveled distance, stock and fuel efficiency of private vehicles in Canada: price elasticities
and rebound effect,” Transportation, Vol. 36, No. 4, 2009, pp. 389–402.
Elasticity 59
7. Bounded Rationality
Owing to cognitive limitations and difficulties in self-control, individuals behave

with bounded rationality. Bounded rationality leads to systematic biases in decision-
making that affect the elasticity of demand.
• Sunk-cost fallacy. Once a person has incurred a sunk cost, she may feel
mentally obliged to justify the sunk cost and that mental obligation will affect
her subsequent choices. For instance, having paid the annual subscription
fee for unlimited shipping, a consumer may feel that she must make more
purchases to justify the annual fee. If the subscription fee is higher, the
consumer will feel obliged to spend even more and her demand for shopping
will be less price elastic.
• Anchoring. Facing uncertainty, individuals need information and may
use cues, even irrelevant cues, to guide their choices. Consumers may be
imprecise about their benefit from consumer products. Appreciating this
uncertainty, retailers set list prices to anchor consumer perceptions of the
benefit. To the extent of such anchoring, consumers perceive that prices
above the list price exceed their benefit, while prices below the list price yield
buyer surplus. Accordingly, the consumer demand is price elastic at prices
above the list price and price inelastic at prices below the list price.
KEY TAKEAWAYS
• The own-price elasticity of demand is the percentage by which the quantity

demanded will change if the price of the item rises by 1%.
• The demand for an item is price elastic if a 1% increase in price leads to more
than a 1% drop in quantity demanded, and price inelastic if a 1% increase in
price leads to less than a 1% drop in quantity demanded.
• The demand for an item will be more elastic to the extent that: (i) it has more
direct or indirect substitutes; (ii) the buyer has fewer prior commitments to the
item; and (iii) the benefits of economizing are larger than the costs.
• If demand is inelastic, the seller can increase profit by raising the price.
• The income elasticity of demand is the percentage by which the demand will
change if buyers’ incomes rise by 1%.
• The cross-price elasticity of demand is the percentage by which the demand
will change if the price of a related item rises by 1%.
• The advertising elasticity of demand is the percentage by which the demand
will change if sellers increase advertising expenditure by 1%.
• For non-durables, the demand is more elastic in the long run than in the short
run. For durables, the demand may be more or less elastic in the long run than
in the short run, depending on the balance between time for adjustment and
replacement frequency.
60 3 Elasticity
• Incurring a sunk cost causes demand to be less price elastic.

• Demand is price elastic above the anchor price, and price inelastic below the
anchor price.
REVIEW QUESTIONS
1. Consider a service that you buy frequently. (a) Suppose that the price is 5%
lower. How much more would you buy each year? (b) Calculate the own-price
elasticity of your demand.
2. Explain why the own-price elasticity is a pure number with no units and is
negative.
3. Under what conditions is demand price elastic or price inelastic?
4. Consider the intuitive factors that influence the own-price elasticity of demand.
Apply the factors to gauge the own-price elasticity of demand for air travel
among executives traveling at the expense of their employers.
5. Suppose that the own-price elasticity of the market demand for food is −0.7
and that, as a result of a severe drought, the price of food rises by 10%. Will
expenditure on food rise or fall?
6. The own-price elasticity of the demand for one brand of frozen vegetables
is −1.5. Suppose that the manufacturer reduces price by 5%. What would be
the percentage effect on the volume of sales?
7. Consider a good that you buy frequently. (a) Suppose that your income is 10%
higher. How much more would you buy each year? (b) Calculate the income
elasticity of your demand.
8. Changes in the price of an item may affect the income elasticity of demand.
True or false?
9. Tire manufacturers use both natural and synthetic rubber to produce tires.
Suppose that the cross-price elasticity of demand for natural rubber with
respect to changes in the price of synthetic rubber is negative. Are the two
types of rubber substitutes or complements?
10. Explain why the advertising elasticity of the market demand for beer may be
less than the advertising elasticity of the demand for one particular brand.
11. Suppose that the advertising elasticity of the demand for one brand of cigarettes
is 1.3. If the manufacturer raises advertising expenditure by 5%, by how much
will the demand change?
12. Consider the effect of changes in fares on the quantity demanded of taxi
services. Do you expect demand to be more elastic with respect to fare
changes in the short run or in the long run?
13. Suppose that the income elasticity of the demand for cars is 0.285 in the short
run and 0.391 in the long run. Compare the effect of a 10% rise in incomes on
the demand for cars in the short and long run.
14. How does the sunk-cost fallacy affect the elasticity of demand?
15. Compare the own-price elasticity of the demand for a particular brand of inkjet
cartridges: (a) before the user has bought an inkjet printer; and (b) after the user
has bought the printer.
Elasticity 61
1. At a major French food retailer, the own-price elasticities of the demand for
various brands of pasta were: −1.36 for national brands, −2.16 for private labels,
and −1.85 for low-price brands. At the same retailer, the own-price elasticities of
the demand for various brands of biscuits were: −1.00 for national brands, −1.14
for private labels, and −0.50 for low-price brands. (Source: Fabian Berges,
Daniel Hassan, and Sylvette Monier-Dilhan, “Are consumers more loyal to
national brands than to private labels?” Working Paper, Toulouse School of
Economics, 2009.)
(a) Compare the elasticities of the demand for national brands and private
labels of pasta. Does the difference make sense?
(b) Do national brands or private labels or low-price brands command
more brand loyalty? (Hint: Interpret brand loyalty by the own-price
elasticity.)
(c) Which is more elastic? The demand for pasta or biscuits?
(d) Based on the own-price elasticities, can you make any recommen-
dations on pricing?
2. Between 1995 and 2005, the own-price elasticity of the demand for water
among urban residential users ranged between a high of −0.20 in October and
a low of −0.06 in December. Among urban commercial users, the own-price
elasticity of demand ranged from a high of −0.17 in December to a low of −0.08
in January. (Source: David R. Bell and Ronald C. Griffin, “Urban water demand
with periodic error correction,” Land Economics, Vol. 87, No. 3, August 2011,
pp. 528–544.)
(a) Are the residential and commercial demands for water price elastic or
inelastic?
(b) Do you expect the elasticities of the residential and commercial
demands to be similar or different?
(c) Explain why the own-price elasticity of the demand varies by season.
(d) If urban suppliers of water were to raise their prices by 10%, what
would be the effect on residential and commercial expenditures?
3. In 2008, the New York Times Media Group earned revenues of $668 million
from circulation, $1.077 billion from advertising, and $181 million from other
sources. The Group decided to raise circulation prices and trim less profitable
readership. In May 2009, the New York Times planned to raise its weekday
cover price from $1.50 to $2. The previous year, the Times had raised the price
from $1.25 to $1.50, and circulation fell 3.6% to 1.04 million.
(a) Using the 2008 price and circulation information, calculate the own-
price elasticity of demand for the New York Times weekday edition.
(b) At the current price of $1.50, and assuming 300 weekdays a year, what
is the annual revenue from weekday sales?
(c) Consider the expected 2009 price increase from $1.50 to $2. What is
the percentage change in price?
62 3 Elasticity
(d) Suppose that the expected 2009 price increase from $1.50 to $2 does
indeed yield $40 million in incremental revenue. What is the percentage
change in revenue?
(e) Calculate the price elasticity of demand which would imply the
$40 million increase in revenue. (Hint: Use equation (3.5).)
(f) Compare the elasticities in (a) and (e). Does the difference make intuitive
sense?
4. In the US market for four prescription medicines (analgesic/musculoskeletal,
antilipidemics, gastrointestinal acid reducers, and insomnia remedies), the
elasticity of demand with respect to consumer advertising ranged between 0.13
and 0.19, while the elasticity of demand with respect to physician advertising
was 0.51. The own-price elasticity was −0.67 for drugs advertised to consumers
and −0.73 for drugs not advertised to consumers. (Source: Dhaval Dave and
Henry Saffer, “Impact of direct-to-consumer advertising on pharmaceutical
prices and demand,” Southern Economic Journal, Vol. 79, No. 1, 2012,
pp. 97–126.)
(a) How would a 5% increase in expenditure on advertising to consumers
affect the demand for the four prescription medicines?
(b) What about a 5% increase in expenditure on advertising to
physicians?
(c) Do you expect the same difference between advertising to consumers
vis-à-vis physicians in the demand for over-the-counter (non-prescription)
medicines?
(d) Suppose that a drug manufacturer were to increase advertising.
Explain why it should also raise the price of its drugs.
5. At an Asian mobile service provider, the demand for voice calls had an own-
price elasticity of −0.085 and cross-price elasticity with respect to the price
of short message service (SMS) of −0.078. The demand for SMS had an own-
price elasticity of −0.03 and cross-price elasticity with respect to the price of
voice calls of −0.03. (Source: Youngsoo Kim, Rahul Telang, William B. Vogt,
and Ramayya Krishnan, “An empirical analysis of mobile voice service and
SMS: A structural model,” Management Science, Vol. 56, No. 2, February 2010,
pp. 234–252.)
(a) For which service was the demand more price inelastic?
(b) How would you describe the relation between the demand for voice
calls and for SMS? Are they (i) complements, or (ii) substitutes?
(c) Which is the relatively stronger complement/substitute? (i) SMS for
voice calls, or (ii) voice calls for SMS?
(d) Describe the impact on revenues from (i) voice and (ii) SMS if the
provider were to raise the price of voice calls by 5%.
6. Electric power producers have a choice of various fuels, including oil, natural
gas, coal, and uranium, as well as solar and wind energy. Once an electric
power plant has been built, however, the scope to switch fuels may be very
limited. Since power plants last for 30 years or more, producers must consider
Elasticity 63
the relative prices of the alternative fuels well into the future when choosing a
generating plant.
(a) Do you expect the cross-price elasticity between the demand for wind
power plants and the price of coal to be positive or negative?
(b) Will the cross-price elasticity between the demand for oil-fired power
plants and the price of coal be positive or negative?
(c) Would the demand for fuel be more or less elastic in the long run as
compared to the short run?
7. The demand for automobile travel (measured in total number of miles driven)
depends on the price of gasoline and travel time. The cross-price elasticity
of the demand for automobile travel with respect to the price of gasoline is
estimated to be −0.10 in the short run and −0.29 in the long run. The elasticity
of the demand for automobile travel with respect to travel time is −0.27 in the
short run and −0.57 in the long run. (Source: Victoria Transport Policy Institute,
www.vtpi.org/elasticities.pdf.)
(a) Explain why the demand for automobile travel is more elastic in the
long run than in the short run.
(b) Suppose that the price of gasoline rises by 20% and construction
of new roads reduces travel time by 10%. Calculate the percentage
change in the total number of miles driven in the (i) short run, and
(ii) long run.
8. According to a study of US cigarette sales, when the price of cigarettes was
1% higher, consumption would be 0.4% lower in the short run and 0.75% lower
in the long run. (Source: Gary Becker, Michael Grossman, and Kevin Murphy,
“An empirical analysis of cigarette addiction,” American Economic Review,
Vol. 84, No. 3, June 1994, pp. 396–418.)
(a) Calculate the short- and long-run own-price elasticities of the demand
for cigarettes.
(b) Explain why the demand for cigarettes is more elastic in the long run
than in the short run.
(c) If the government were to impose a tax that raised the price of
cigarettes by 5%, what would be the effect on consumer expenditure
on cigarettes in the (i) short run, and (ii) long run?
9. In the following scenarios, it may (not necessarily) help to consider biases in
individual decision-making.
(a) Your mobile service provider requires a cash deposit of $100 from
every customer. A competing provider does not require any deposit.
Both providers raise the prices of their calling plans by 10%. Which
would experience a larger drop in sales?
(b) The local fitness club prices in two ways. It charges $5 per two-hour
visit. It also offers a membership for $240 a year which includes 80 free
visits. (i) What is the break-even between the annual membership and
per-visit price? (ii) Why do many people buy the membership yet use
the club fewer than 48 times a year?
64 3 Elasticity
(c) One supermarket sets a regular price of €5 for a six-pack of Coca-Cola

and discounts to €3. Another supermarket sets a regular price of €4
for a six-pack of Coca-Cola and discounts to €3. Which supermarket
would experience a larger increase in sales?

Identify a product that your organization sells for which the demand is price
inelastic. Write a memorandum to the chief financial officer of your organization
to recommend an increase in price.
Note
1 The following discussion is based in part on “Will Amazon Prime customers accept price hike?
Maybe not,” Forbes, February 12, 2004; “Piper Jaffray reduces churn estimates for potential
Amazon Prime price hike,” Tech Trader Daily, March 7, 2014; Wedbush Securities, “Quick note:
Amazon.com (AMZN – NEUTRAL),” March 13, 2014.
C H A P T E R
4
Supply
LEARNING OBJECTIVES
• Appreciate why producers supply more at higher prices.
• Appreciate how to decide, in the short run, whether to continue in
business, and if so, the scale of production.
• Distinguish fixed and variable costs in the short run.
• Understand the concepts of marginal cost and marginal revenue.
• Appreciate how to decide, in the long run, whether to continue in
business, and if so, the scale of production.
• Appreciate that, in the long run, businesses can adjust by entering
or exiting the industry.
• Appreciate the concept of seller surplus and apply it in purchasing.
• Apply the concept of price elasticity of supply.
1. Introduction
Founded in Durham, Ontario, in 1899, Durham Furniture produces bedroom

furniture. In 2003, the company completed a new plant at Chesley, Ontario, to
manufacture dining room furniture. The total cost was C$38 million, comprising
C$8 million for the 147,500 square foot factory and C$30 million on equipment.
The new plant included eight modern kilns capable of drying 70,000 board feet of
hardwood every 11 days.1
However, the opening of the Chesley plant coincided with increased competition
from Asia and the appreciation of the Canadian dollar against the US dollar. Then,
66 4 Supply
the Great Recession struck, crimping US demand. In 2008, Durham decided to

exit the dining room category and mothball the plant.
Later in the year, Durham Furniture filed for protection under the Companies’
Creditors Arrangement Act. The company had a net worth of between C$3.5
and C$6.3 million, excluding the Chesley plant and equipment, while owing C$37
million in secured borrowings to the Royal Bank of Canada. The Royal Bank of
Canada agreed to the restructuring of the loans provided that Durham found new
investors.
In January 2012, GRS Wood Products, a Chinese-owned company with two
manufacturing plants in China, purchased the Chesley plant. Mayor Paul Eagleson
of the municipality understood that GRS would use the plant to produce hard-
wood flooring and employ 50 workers.
Ready access to lumber products gives an advantage to Canadian manufac-
turers of furniture. However, the manufacturing of furniture is also fairly labor-
intensive. Furniture manufacturers in Asia benefit from lower labor costs. With
falling trade barriers and costs of transportation, Asian furniture manufacturers
can increase their exports to the United States.
How do changes in the prices of inputs such as lumber affect the furniture
industry? How would the increase in exports of furniture from Asian competitors
affect Canadian manufacturers like Durham? Should new investors put money
into Durham?
To address these questions, we need to understand two key decisions of a busi-
ness: first, whether to continue in operation; and, second, the scale at which to
operate. The first is a decision on participation and depends on whether the busi-
ness would break even, and, in turn, on the total revenue and total (relevant)
cost. The second is a decision on extent and depends on the marginal revenue and
marginal cost.
In this chapter, we study the two key decisions in the short run, when businesses
are restricted in the extent to which they can adjust inputs, and in the long run,
when businesses can freely adjust all inputs and possibly enter or exit. We can
analyze whether new investors should invest in Durham.
The analyses of whether to remain in business and, if so, the scale of business
provide the foundation for the concepts of the supply curve. The supply curve
is the seller-side counterpart to the demand curve. Applying the market sup-
ply curve, we can explain the impact of Asian manufacturers on the US market
for furniture and the prospects for Durham and other Canadian manufacturers.
Further, we can explain the impact of lumber and steel prices on the furniture
industry.
This chapter also presents the concepts of seller surplus and elasticity of sup-
ply. Seller surplus is the seller-side counterpart to buyer surplus. By extracting
seller surplus, managers can reduce the cost of purchasing and raise profit. The
own-price elasticity of supply measures the responsiveness of quantity supplied
to changes in the price of the item. This will tell managers how much of a price
increase will be necessary to meet any desired increase in purchases.
Supply 67
The decision-making techniques and concepts presented here apply to any

industry which is characterized by competition among many small suppliers, each
of which can sell as much as it would like at the market price. They obviously
apply to mining, farming, and fisheries, and also to small-scale manufacturing
and service industries.
2. Short-Run Costs
The two key decisions of a business – whether to continue in operation and the
scale at which to operate – both depend on the time horizon. In Chapter 3, we
introduced the concepts of short run and long run in relation to buyers. The same
concepts apply to sellers as well.
The short run is a time horizon in which a seller cannot Short run: A time horizon
adjust at least one input. In the short run, the business must in which a seller cannot
work within the constraints of past commitments such as adjust at least one input.
employment contracts and investment in facilities and equip-
ment. Over time, however, these commitments expire. The
long run is a time horizon long enough for the seller to adjust Long run: A time horizon
all inputs, including possibly entering or exiting the industry. long enough for the
seller to adjust all inputs,
The difference between the short run and long run depends
including possibly entering
on the circumstances. Consider a factory that has just engaged or exiting the industry.
50 workers on 12-month contracts. The employment of these
workers cannot be adjusted until the expiration of the contracts. Hence, the factory’s
short run is at least 12 months long. By contrast, a factory that hires all workers on
a daily basis can adjust its workforce every day. Similarly, a factory that has pur-
chased manufacturing equipment has committed to a relatively longer horizon
than one that has leased the equipment on a yearly contract.
Fixed and Variable Costs

To determine its scale or rate of production, a business needs to know the cost
of delivering an additional unit of product (for convenience, we use “scale” and
“rate” of production interchangeably). To decide whether to continue in operation,
a business needs to know how shutting down will affect its total costs.
An important factor in both decisions is the distinction between fixed and variable
costs. The fixed cost is the cost of inputs that do not change
Fixed cost: The cost of
with the production rate. By contrast, the variable cost is
inputs that do not change
the cost of inputs that do change with the production rate. with the production rate.
Let us consider the distinction between fixed and vari-
able costs in the context of Luna Plywood, which produces
plywood. Like those of most businesses, Luna’s account- Variable cost: The cost
of inputs that do change
ing records do not classify expenses into fixed and variable. with the production rate.
Rather, the records organize expenses according to the type
68 4 Supply
Table 4.1 Short-run weekly expenses
Weekly Rent ($) Equipment Salaries ($) Wages ($) Cost of Total ($)
production rate leasing ($) supplies ($)
0 2,000 10,000 8,000 2,000 0 22,000
1,000 2,000 10,000 8,000 5,290 1,000 26,290
2,000 2,000 10,000 8,000 8,360 2,000 30,360
3,000 2,000 10,000 8,000 12,160 3,000 35,160
4,000 2,000 10,000 8,000 16,970 4,000 40,970
5,000 2,000 10,000 8,000 22,930 5,000 47,930
6,000 2,000 10,000 8,000 30,150 6,000 56,150
7,000 2,000 10,000 8,000 38,700 7,000 65,700
8,000 2,000 10,000 8,000 48,620 8,000 76,620
9,000 2,000 10,000 8,000 59,960 9,000 88,960
Table 4.2 Analysis of short-run costs
Weekly Fixed Variable Total Marginal Average Average Average

production cost ($) cost ($) cost ($) cost ($) fixed variable cost ($)
rate cost ($) cost ($)
0 22,000 0 22,000
1,000 22,000 4,290 26,290 4.29 22.00 4.29 26.29
2,000 22,000 8,360 30,360 4.07 11.00 4.18 15.18
3,000 22,000 13,160 35,160 4.80 7.33 4.39 11.72
4,000 22,000 18,970 40,970 5.81 5.50 4.74 10.24
5,000 22,000 25,930 47,930 6.96 4.40 5.19 9.59
6,000 22,000 34,150 56,150 8.22 3.67 5.69 9.36
7,000 22,000 43,700 65,700 9.55 3.14 6.24 9.39
8,000 22,000 54,620 76,620 10.92 2.75 6.83 9.58
9,000 22,000 66,960 88,960 12.34 2.44 7.44 9.88
of input: rent, equipment lease, salaries, wages, and payments for supplies. By
interviewing Luna’s management, we can learn the costs required for alternative
short-run production rates. Table 4.1 presents this information.
To distinguish between fixed and variable costs, a business must analyze how
each category of expense varies with changes in the scale of operations. Referring
to Table 4.1, we can perform this analysis for Luna. In the short run, Luna can-
not adjust the size of its facility, equipment or managers’ salaries, so the rent and
equipment leasing do not vary with the production rate. The rent is $2,000, equip-
ment leasing is $10,000 and managers’ salaries are $8,000 whether Luna produces
nothing or 9,000 sheets of plywood a week; hence, they are a fixed cost. Workers’
wages vary with the production rate, but even when Luna produces nothing, it
incurs wages of $2,000. Hence, the wages include a fixed component of $2,000,
while the remainder is variable. Finally, the cost of supplies is completely variable.
In Table 4.2, we assign Luna’s expenses – rent, managers’ salaries, wages, and
cost of supplies – into the two categories of fixed costs and variable costs. As the
production rate increases from nothing to 9,000 sheets of plywood a week, the
Supply 69
fixed cost is always $22,000. By contrast, the variable cost increases from nothing
for no production to $66,960 for 9,000 sheets a week.
Total cost is the sum of fixed cost and variable cost. If we Total cost: The sum of
represent total cost by C, fixed cost by F, and variable cost fixed cost and variable
by V, then cost.
C = F + V. (4.1)
Provided that there are some variable costs, the total cost will increase with oper-
ations. In Luna’s case, referring to Table 4.2, the total cost is $22,000 for no pro-
duction, and rises to $88,960 for production of 9,000 sheets a week.
It is helpful to illustrate the concepts of total, fixed, and variable costs graphi-
cally. In Figure 4.1, the vertical axis represents cost, while the horizontal axis rep-
resents the scale of production. We draw a curve representing variable cost. The
total cost curve is the variable cost curve, shifted up everywhere by the amount of
the fixed cost. In particular, the fixed cost is represented by the height of the total
cost curve at the production rate of zero.
By analyzing its costs as fixed and variable, the management of a business can
understand which cost elements will be affected by changes in the scale of oper-
ations. The distinction between fixed and variable costs is important whether the
business is growing or shrinking. For instance, suppose that management is plan-
ning to reduce costs by downsizing the scale of operations. Downsizing will have
no effect on fixed costs and will only reduce the variable costs. Hence, in a business
whose costs are mostly fixed, downsizing may have relatively little effect on costs.
PROGRESS CHECK 4A
In Figure 4.1, if the fixed cost were higher, how would that affect the total and
variable cost curves?
total cost
80
Cost (thousand $)
variable cost
60
40
20
fixed cost
0 2 4 6 8
Production rate (thousand sheets a week)
FIGURE 4.1 Short-run total cost.

Note: The total cost is the variable cost curve shifted up by the amount of the fixed cost.
70 4 Supply
Marginal Cost
To determine the scale at which it should operate, a business needs to know the cost
of making an additional unit of product. Then the business can see whether selling
the additional unit will add to or subtract from its total profit.
The change in total cost due to the production of an addi-
Marginal cost: The
tional unit is the marginal cost. The marginal cost can be
change in total cost due
to the production of an derived from the analysis of fixed and variable costs.
additional unit. Let us derive the marginal cost in the case of Luna Ply-
wood. Referring to Table 4.2, as the production rate increases
from zero to 1,000 sheets of plywood a week, the total cost
increases from $22,000 to $26,290. The increment of $26,290 − $22,000 = $4,290
is the additional cost of producing the 1,000 sheets. Hence, the marginal cost is
$4,290/1,000 = $4.29 per sheet.
Notice that, as the production rate increases from zero to 1,000 sheets a week,
the fixed cost remains unchanged; only the variable cost increases. Hence, we
can also calculate the marginal cost from the increase in variable cost. Using
this approach, as the production rate increases from zero to 1,000 sheets a week,
the variable cost increases from $0 to $4,290. Therefore, the marginal cost is
$4,290/1,000 = $4.29 per sheet.
Similarly, as the production rate increases from 1,000 to 2,000 sheets a week,
the variable cost increases from $4,290 to $8,360. The marginal cost is now $4,070
for 1,000 sheets, or $4.07 per sheet. With each increase in the production rate, the
marginal cost increases, reaching $12.34 at the rate of 9,000 sheets a week.
In Luna’s case, each additional sheet requires more variable cost than the one
before. We display this information in Table 4.2.
Average Cost
The marginal cost is the cost of producing an additional unit. A related concept
is the average cost, which is the total cost divided by the
Average cost (unit production rate. The average cost is also called the unit cost.
cost): The total cost
Given the scale of operations, the average cost reflects the
divided by the production
rate. cost of producing a typical unit.
Let us derive the average cost in the case of Luna Ply-
wood. Referring to Table 4.2, we can obtain the average cost
as the total cost divided by the production rate. At 1,000 sheets a week, the aver-
age cost is $26,290/1,000 = $26.29 per sheet, while at 2,000 sheets a week, the
average cost is $30,360/2,000 = $15.18 per sheet. The average cost continues to fall
with increases in the production rate until it reaches a minimum of $9.36 at 6,000
sheets a week. Thereafter, it increases with the production rate. At 9,000 sheets a
week, the average cost is $9.88 per sheet.
To understand why the average cost first drops with increases in the production
rate and then rises, recall that the total cost is the sum of the fixed cost and the
Supply 71
variable cost. Let q represent the production rate. Then dividing equation (4.1)
throughout by q, we have
C F V
= + . (4.2)
q q q
In words, the average cost is the average fixed cost plus the average variable cost.
The average fixed cost is fixed cost divided by the production rate. So, if the
production rate is higher, the fixed cost will be spread over more units; hence, the
average fixed cost will be lower. This factor causes the average cost to fall with
increases in the production rate.
The other element in average cost is the average variable cost, which is the variable
cost divided by the production rate. In the short run, at least one input is fixed;
hence, to raise the production rate, the business must combine increasing quanti-
ties of the variable inputs with an unchanged quantity of the fixed input.
The increase in output arising from an additional unit of
an input is called the marginal product from that input. At Marginal product: The
low production rates, there is a mismatch between the vari- increase in output arising
able inputs and the fixed input. Owing to the mismatch, the from an additional unit of
marginal product is low and the average variable cost is high. an input.
With a higher production rate, the variable inputs match the
fixed input relatively better, and the average variable cost is lower.
As more of the variable inputs are added in combination with the fixed input,
there will be a mismatch again. Eventually, there will be a diminishing marginal
product from the variable inputs. This means that the marginal product becomes
smaller with each increase in the quantity of the variable inputs. With a dimin-
ishing marginal product from the variable inputs, the average variable cost will
increase with the production rate.
In Luna’s case, Table 4.2 shows that the average variable cost first drops from
$4.29 to $4.18 per sheet as the production increases from 1,000 to 2,000 sheets
a week. Then the average variable cost rises from $4.18 to $7.44 per sheet as the
production increases from 2,000 to 9,000 sheets a week.
Recall that the average cost is the average fixed cost plus the average variable
cost. While the average fixed cost falls with the production rate, the average vari-
able cost falls and then increases. Accordingly, where the average variable cost
is increasing, the relationship between the average cost and the production rate
depends on the balance between the declining average fixed cost and the increas-
ing average variable cost.
If the fixed cost is not too large and the average variable cost increases suf-
ficiently, the average cost will first decline with the production rate and then
increase. As Table 4.2 shows, this is the case for Luna.
Figure 4.2 graphs the marginal, average, and average variable costs against the
production rate. The vertical axis represents cost per unit of production, while the
72 4 Supply
30
Cost ($ per sheet) 25
20
15
marginal cost
10 average cost
average variable cost
5
0 2 4 6 8
FIGURE 4.2 Short-run marginal, average variable, and average costs.

Notes: The marginal, average variable, and average cost curves are U-shaped. The curves decrease
at low production rates, reach a minimum, and then increase for high production rates.
horizontal axis represents the production rate. The marginal cost curve falls from
$4.29 per sheet at 1,000 sheets a week, to reach a minimum of $4.07 at 2,000 sheets a
week, and rises thereafter. The average variable cost curve falls from $4.29 per sheet
at 1,000 sheets a week, to a minimum of $4.18 per sheet at 2,000 sheets a week, and
then rises. Similarly, the average cost curve falls from $26.29 per sheet at 1,000 sheets
a week, to a minimum of $9.36 at 6,000 sheets a week, and then rises. The graphs of
the marginal, average variable, and average cost curves are each U-shaped.
PROGRESS CHECK 4B
In Figure 4.2, if the fixed cost were higher, how would that affect the average
cost, average variable cost, and marginal cost curves?
Production Technology
In the preceding analysis, we derived the information about costs by asking the seller
for the cost of producing at various rates. Accordingly, at every production level, the
total, average, and marginal costs depend on the seller’s production technology.
This approach has two implications. First, the curves will change with adjust-
ments in the seller’s technology. A manufacturer that discovers a technology
involving a lower fixed cost will lower its average cost curve. A manufacturer that
uses a technology with a lower variable cost will lower its average, average vari-
able, and marginal cost curves.
Second, sellers may use different technologies, and hence have different cost
curves. They may differ in the structure of fixed and variable costs. Some may
have better technologies and hence lower overall costs than others.
Supply 73
FAME: FROM WASTE TO ENERGY
With stocks of fossil fuel depleting and emissions exacerbating global warming,
the worldwide demand for renewable, low-emissions sources of energy is
growing. Fatty acid methyl ester (FAME) is a biodiesel fuel that is aromatic-
free, highly biodegradable, and generates low emissions.
FAME is an attractive source of energy as it can be manufactured from
waste cooking oil. The challenge in producing FAME is that the manufacturing
process requires large quantities of water to remove impurities.
A recent study compared four batch processing methods, using different
catalysts and processes, to produce FAME from waste cooking oil. For a
production scale of 7,269 tons a year, the method with the lowest variable
cost, $391 per ton, required a fixed investment in plant of $8.3 million. The
method with the highest variable cost, $416 per ton, required a relatively lower
fixed investment of $7.99 million.
Source: Tsutomu Sakai, Ayato Kawashima, and Tetsuya Koshikawa, “Economic assessment
of batch biodiesel production processes using homogeneous and heterogeneous alkali
catalysts,” Bioresource Technology, Vol. 100, No. 13, July 2009, pp. 3268–3276.
3. Short-Run Individual Supply
Costs are one aspect of the short-run decisions whether to continue in operation
and how much to produce. The other side to these decisions is revenue. We now
consider the revenues of a business.
In analyzing revenues, we shall assume that the business aims to maximize
profit and that the business is so small relative to the market that it can sell as
much as it would like at the market price. We need the assumption of smallness to
construct individual and market supply, which are the counterparts to individual
and market demand.
Production Rate
Supposing that the price of plywood is $7 per sheet, how much should Luna
produce? Generally, the profit of a business is its total revenue less its total cost,
and in turn, total revenue is the price multiplied by sales.2
In Table 4.3, we show Luna’s cost and revenue at various production rates, with
the assumption that the price is $7. For instance, if sales are 1,000 sheets a week,
then Luna’s total revenue will be $7 × 1,000 = $7,000. If sales are 2,000 sheets a
week, then Luna’s total revenue will be $7 × 2,000 = $14,000. Similarly, we can
calculate the total revenue at other production rates.
74 4 Supply
Table 4.3 Short-run profit
Weekly Variable Total Total Accounting Economic Marginal Marginal

production cost ($) cost ($) revenue profit ($) profit ($) cost ($) revenue
rate ($) ($)
0 0 22,000 0 (22,000) 0
1,000 4,290 26,290 7,000 (19,290) 2,710 4.29 7.00
2,000 8,360 30,360 14,000 (16,360) 5,640 4.07 7.00
3,000 13,160 35,160 21,000 (14,160) 7,840 4.80 7.00
4,000 18,970 40,970 28,000 (12,970) 9,030 5.81 7.00
5,000 25,930 47,930 35,000 (12,930) 9,070 6.96 7.00
6,000 34,150 56,150 42,000 (14,150) 7,850 8.22 7.00
7,000 43,700 65,700 49,000 (16,700) 5,300 9.55 7.00
8,000 54,620 76,620 56,000 (20,620) 1,380 10.92 7.00
9,000 66,960 88,960 63,000 (25,960) (3,960) 12.34 7.00
total cost variable cost

Cost/revenue (thousand $)
total revenue
40.97
loss = $12,970
28
0 1 4 9
FIGURE 4.3 Short-run profit.

Notes: At a production rate of 4,000 sheets a week, the total revenue is $28,000 and the total cost
is $40,970; hence, the vertical difference between total revenue and total cost is a loss of $12,970.
Marginal revenue is represented by the slope of the total revenue line, while marginal cost is
represented by the slope of the total cost curve.
From Table 4.3, the highest accounting profit is a loss of $12,930, which comes
from producing at a rate of 5,000 sheets a week. (Below, we discuss why it makes
sense for Luna to produce “at a loss.”)
We can derive a general rule for the profit-maximizing production rate by illus-
trating cost and revenue with a diagram. Figure 4.3 shows the cost curves from
Figure 4.1, and includes a line to represent Luna’s total revenue at a price of $7.
The line rises at a rate of $7,000 for every increase of 1,000 sheets in the produc-
tion rate. Equivalently, the slope of the line is $7. For instance, one point on the
Supply 75
total revenue line is a production rate of 4,000 sheets a week and revenue of $7 ×
4,000 = $28,000.
Using Figure 4.3, we can measure the difference between the total revenue and
the total cost at any production rate. In the figure, the vertical difference between
the total revenue line and the total cost curve represents the accounting profit.
For instance, at a production rate of 4,000 sheets a week, the height of the total
revenue line is $28,000, while the height of the total cost curve is $40,970. Hence,
the vertical difference is a loss of $12,970.
Generally, to maximize profit, a business should produce at that rate where its
marginal revenue equals its marginal cost. The marginal
revenue is the change in total revenue arising from selling Marginal revenue: The
change in total revenue
an additional unit.
arising from selling an
To explain the rule for maximizing profit, consider Figure 4.3. additional unit.
Graphically, the marginal revenue is represented by the slope
of the total revenue line. Similarly, since marginal cost is the
change in total cost due to the production of an additional unit, the marginal cost
is represented by the slope of the total cost curve.
At a production rate of 1,000 sheets a week, the total revenue line climbs faster
than the total cost curve, or equivalently, the marginal revenue exceeds the mar-
ginal cost. Then an increase in production will raise the profit. Wherever the
marginal revenue exceeds the marginal cost, Luna can raise profit by increasing
production.
By contrast, at a production rate of 9,000 sheets a week, the total revenue climbs
more slowly than the total cost curve, or equivalently, the marginal revenue is less
than the marginal cost. Then a reduction in production will increase profit. Wher-
ever the marginal revenue is less than the marginal cost, Luna can raise profit by
reducing production.
Thus, Luna will maximize profit at the production rate where its marginal reve-
nue equals marginal cost. At that point, the total revenue line and the total cost
curve climb at exactly the same rate. Hence, a small change in production (either
increase or reduction) will affect both total revenue and total cost to the same
extent. Accordingly, it is not possible to increase profit any further.
For a small seller, which can sell as much as it would like at the market price, the
rule for profit-maximizing production can be expressed in
another way. By definition, the marginal revenue is the Profit-maximizing
change in total revenue arising from selling an additional scale of production:
The scale where marginal
unit. For a business that can sell as much as it would like at revenue equals marginal
the market price, the change in total revenue arising from cost.
selling an additional unit is exactly equal to the price. Hence,
the marginal revenue equals the price of the output.
In Figure 4.4, we draw the marginal and average cost curves from Figure 4.2
and also include a line representing the marginal revenue and the price. Where the
price exceeds the marginal cost, Luna can increase profit by raising production.
By contrast, where the price is less than the marginal cost, Luna can increase
76 4 Supply
Cost/revenue ($ per sheet)
marginal cost
average cost
7
marginal revenue = price
5
break-even
price
FIGURE 4.4 Short-run production rate.

Note: Given the price of $7, the seller maximizes profit by producing at the rate of 5,000 sheets a
week, where marginal cost equals the price.
profit by cutting production. Therefore, Luna will maximize profit by producing

at 5,000 sheets a week, a rate at which its marginal cost just balances the price.
Break-Even Analysis
Should the business continue in operation? To make this decision, the business
needs to compare the profit from continuing in production with the profit from
shutting down. An essential consideration in this comparison is the structure of
fixed and variable costs.
Suppose that a business continues production. Let the total revenue from the
profit-maximizing production rate be R, while the fixed cost is F and the variable
cost is V. Then the maximum profit is R − F − V.
Now suppose that the business shuts down. Clearly, this will reduce its total
revenue to 0. How will the shutdown affect its costs? We assume that the entire
fixed cost, F, of the business is also sunk in the short run.
A sunk cost is one that has been committed and cannot be
Sunk cost: A cost that
avoided. Costs which are not sunk are avoidable.
has been committed and
cannot be avoided. By assumption, the entire fixed cost, F, is also sunk. This
means that, even if the business shuts down, it must still pay
the fixed cost, F. In contrast to the fixed cost, the variable
cost is avoidable. Hence, if the business shuts down, it need not incur any variable
cost. Thus, if it shuts down, the profit will be the zero revenue minus the fixed cost,
that is, −F.
The business should continue in production if the maximum profit from continu-
ing in production is at least as large as the profit from shutting down. Algebraically,
this break-even condition is
R − V − F ≥ −F. (4.3)
Supply 77
which simplifies to
R ≥ V. (4.4)
The business should continue in production so long as its total revenue covers the
variable cost.
Given that the fixed cost is sunk, it is not relevant to the decision whether to con-
tinue in production. Hence, as generally outlined in Chapter 1, the short-run deci-
sion whether to continue in production resolves into comparing the total revenue
with the total (relevant) cost. Since sunk costs are not relevant, the total relevant
cost is just the variable cost.
Recall that total revenue is the price multiplied by sales, R = p × q. Divide the
break-even condition (4.4) throughout by sales (which is also the production rate)
to obtain
V
p≥ . (4.5)
q
Hence an equivalent way of stating the short-run break- Short-run break-even:

even condition is that the price must cover the average vari- Total revenue covers
able cost. variable cost, or price
To summarize, in the short run, a business (which can covers average variable
sell as much as it would like at the market price) maximizes cost.
profit by:
• if the total revenue covers the variable cost, producing at the rate where the
marginal cost equals the price;
• if the total revenue does not cover the variable cost, shutting down.
How does the short-run break-even analysis apply to Luna Plywood? Suppose
that Luna continues in production. Then, from Table 4.3, by producing at a rate
of 5,000 sheets a week, Luna will operate at a loss of $12,930. By assumption,
Luna’s entire $22,000 fixed cost is also sunk. Thus, if Luna shuts down, its profit
will be the zero revenue minus the $22,000 fixed cost, that is, a loss of $22,000.
Clearly, Luna is better off continuing in production.
Another way to make this decision simply ignores the fixed cost – because it is
sunk and thus not relevant. Table 4.3 shows the variable cost and total revenue. If
Luna produces 5,000 sheets a week, it will earn a total revenue of $35,000, while
its variable cost would be $25,930. Since the total revenue exceeds the variable
cost, Luna should continue in production. Thus, it makes sense for Luna to pro-
duce “at a loss” – the reason is that the “loss” includes a sunk cost that should not
be considered in the decision-making.
Using the concept of economic profit, the correct break-even analysis is obvious.
In calculating economic profit, sunk costs should be ignored. So, the economic
78 4 Supply
profit from producing 5,000 sheets a week is the total revenue of $35,000 less the
variable cost of $25,930. Hence, the economic profit is $9,070, which is positive,
hence Luna should continue in operation.
ALASKA MARINE HIGHWAY SYSTEM: $22 MILLION BAILOUT
In Alaska, the state ferry service, officially called the Alaska Marine Highway
System, provides ferry transport for passengers and vehicles within the state
as well as to Canada and the lower United States. In 2005, owing to financial
losses, the state ferry service required a government bailout of $22 million.
Governor Frank Murkowski blamed the ferry’s losses in part on increases
in scheduled sailings, particularly in winter. The director of the Alaska Marine
Highway System, Robin Taylor, disagreed: “Even when a vessel is tied up,
expenses are still incurred as a captain and crew must be on board and line
handlers must be available. When those expenses are coupled with the loss
of revenue, the more economical choice would be to operate the vessel rather
than to tie it up.”
Sources: Alaska State Senate Finance Committee hearing, March 2, 2006; Alaska Public
Radio, October 17, 2005.
Individual Supply Curve

Using the rule for profit-maximizing production, we can determine how much a
business (which can sell as much as it would like at the price) should produce at
various prices for its output. The rule is that it should produce at the rate that bal-
ances its marginal cost with the price, provided that the price covers the average
variable cost.
Referring to Figure 4.4, if the price of the plywood is $8 rather than $7 a sheet,
Luna should expand production to the rate where the new price equals the
marginal cost. Indeed, by varying the price, we can trace out the quantity that
Luna should supply at every possible price. This is the information needed to
construct Luna’s individual supply curve. The individual
Individual supply curve: supply curve is a graph showing the quantity that a seller
A graph showing the
quantity that a seller will
will supply at every possible price.
supply at every possible The individual supply curve is identical with the portion of
price. the seller’s marginal cost curve that lies above the average vari-
able cost curve. To expand production, the seller must incur
a higher marginal cost. So the seller will expand production
only if it receives a higher price. Accordingly, the individual supply curve slopes
upward.
The individual supply curve shows how a seller should adjust its produc-
tion in response to changes in the price of its output. Hence, the effect of any
Supply 79
original marginal cost

marginal cost with

lower input price
7
marginal revenue = price
0 5 5.6
FIGURE 4.5 Lower input price.

Note: With a lower input price, the marginal cost curve will shift downward; hence the seller will
increase the production rate from 5,000 to 5,600 sheets.
change in the output price will be represented by a movement along the supply
curve.
PROGRESS CHECK 4C
Using Figure 4.4, show the quantity that Luna should produce at a price of
$7.50 per sheet.
Input Demand
We can now explain the individual seller’s demand for inputs. We derived the
seller’s marginal cost from its total cost, which in turn was derived from the esti-
mates of the expenses on rent, salaries, wages, and other supplies needed at vari-
ous production rates. These estimates depend on the prices of the various inputs.
Suppose, for instance, that Table 4.1 assumes a wage of $10 per hour. Then the
calculations in Tables 4.2 and 4.3 and Figures 4.3 and 4.4 are based on a wage of
$10 per hour. What if the wage is $9 per hour? Then we must go back to adjust
Tables 4.1–4.3 and Figures 4.3 and 4.4 using the new wage rate.
Intuitively, as we show in Figure 4.5, the marginal cost curve will shift down-
ward. The profit-maximizing production rate increases from 5,000 to 5,600 sheets
a week. From the new production rate of 5,600 sheets, we can go back to deter-
mine the corresponding quantity of the labor input. With a higher rate of produc-
tion, the quantity of labor demanded will also be higher.
By varying the wage rate, we can determine the quantity of labor demanded at
every possible wage rate. This provides the individual seller’s demand for labor.
As just shown, the quantity demanded will be higher at a lower wage; hence, the
demand curve for labor will slope downward. The same method can be used to
derive the individual seller’s demand for every other input.
80 4 Supply
OIL: TO PRODUCE OR NOT TO PRODUCE?
In the oil industry, production is the activity of extracting crude oil from the
ground. As the price of oil fluctuates on world markets, oil producers must
consider whether to enter into production, continue in production, suspend
production, or exit the industry.
Following the Great Recession, the price of West Texas Intermediate crude
oil fell from $100 per barrel in 2008 to $62 per barrel in 2009. With the economic
recovery, the price of oil climbed steady to over $98 per barrel in 2013.
The high price of oil and improvements in production technology have
boosted US production of oil, particularly from shale deposits. Between 2008
and 2013, US production rose by almost half from 302.3 to 446.2 million tonnes.
Owing to this increase in US production, coupled with slowing economic
growth, energy conservation, and greater use of renewable energy, the price
of oil peaked in 2014 and fell to around $50 per barrel.
In parallel, the number of active rotary production rigs in the United States
fell from a peak of over 1,700 in late 2008 to a low of below 1,200 in mid-2009.
The number of active rigs then rose steadily to over 2,000 in 2011 and then
fluctuated between 1,800 and 2,000 thereafter.
Sources: BP Statistical Review of World Energy, 63rd edition, June 2014; Baker Hughes,
North America Rotary Rig Count, November 7, 2014.
4. Long-Run Individual Supply
In the short run, a business must work within the constraints of past commit-
ments such as employment contracts and investment in facilities and equipment.
Over time, however, contracts expire and investments wear out. With sufficient
time, all inputs become avoidable. A long-run planning horizon is a time frame far
enough into the future that all inputs can be freely adjusted. Then the business
will have complete flexibility in deciding on inputs and production.
How should a business make two key decisions – whether to continue in oper-
ation and how much to produce – in the long run? To address these decisions, we
first analyze the long-run costs and then look at the revenue.
Long-Run Costs
Let us analyze long-run costs in the context of Luna Plywood. We ask the man-
agement to estimate the costs of producing at various rates when all inputs are
avoidable. Suppose that Table 4.4 presents the expenses classified into rent, sala-
ries, wages, and cost of supplies.
In the long run, Luna can adjust the managers’ employment according to the pro-
duction rate, and so their salaries vary from $2,500 a week at zero production up to
Supply 81
Table 4.4 Long-run weekly expenses
Weekly Rent Equipment Salaries Wages Cost of Total

production ($) leasing ($) ($) ($) supplies ($)
rate ($)
0 1,000 1,000 2,500 0 0 4,500

1,000 1,000 1,000 5,000 790 1,000 8,790
2,000 1,000 1,000 7,500 2,610 2,000 14,110
3,000 1,000 1,000 10,000 5,570 3,000 20,570
4,000 1,000 1,000 12,500 9,760 4,000 28,260
5,000 1,000 1,000 15,000 15,220 5,000 37,220
6,000 1,000 1,000 17,500 22,030 6,000 47,530
7,000 1,000 1,000 20,000 30,210 7,000 59,210
8,000 1,000 1,000 22,500 39,820 8,000 72,320
9,000 1,000 1,000 25,000 50,890 9,000 86,890
Table 4.5 Analysis of long-run costs
Weekly Total Marginal Average

production rate cost ($) cost ($) cost ($)
0 4,500
1,000 8,790 4.29 8.79
2,000 14,110 5.32 7.06
3,000 20,570 6.46 6.86
4,000 28,260 7.69 7.07
5,000 37,220 8.96 7.44
6,000 47,530 10.31 7.92
7,000 59,210 11.68 8.46
8,000 72,320 13.11 9.04
9,000 86,890 14.57 9.65
$25,000 at a production rate of 9,000 sheets a week. Similarly, the wages are nothing
at zero production and rise to $50,890 at a production rate of 9,000. The cost of sup-
plies rises from nothing at zero production to $9,000 at a production rate of 9,000.
Extracting the relevant information from Table 4.4, we can compile the
long-run marginal and average costs in Table 4.5 and graph them in Figure 4.6.
As this example shows, even in the long run, there may be fixed costs (but these
are not sunk).
Production Rate
How much should the business produce in the long run? The general rule that
we derived for short-run production also applies. To maximize profit, a business
should produce at that rate where its marginal cost equals the marginal revenue.
In the long run, we use the long-run marginal cost. For a business that can sell as
much as it would like at the market price, the marginal revenue equals the price
82 4 Supply
Cost/revenue ($ per sheet) marginal

cost
average
cost
marginal revenue =
7.00
price
6.86
break-even
price
0 3.0 3.4
FIGURE 4.6 Long-run production rate.

Note: Given a price of $7, the seller maximizes profit by producing at the rate of 3,400 sheets a week,
where the long-run marginal cost equals the price.
Table 4.6 Long-run profit
Weekly Total Total Economic Marginal Marginal

production cost ($) revenue ($) profit ($) cost ($) revenue ($)
rate
0 4,500 0 (4,500)
1,000 8,790 7,000 (1,790) 4.29 7.00
2,000 14,110 14,000 (110) 5.32 7.00
3,000 20,570 21,000 430 6.46 7.00
4,000 28,260 28,000 (260) 7.69 7.00
5,000 37,220 35,000 (2,220) 8.96 7.00
6,000 47,530 42,000 (5,530) 10.31 7.00
7,000 59,210 49,000 (10,210) 11.68 7.00
8,000 72,320 56,000 (16,320) 13.11 7.00
9,000 86,890 63,000 (23,890) 14.57 7.00
of its output. Hence, the rule to maximize profit is to produce at that rate where
marginal cost equals the price.
For Luna Plywood, Table 4.6 shows the long-run cost, revenue, and profit. The
column for economic profit shows that profit reaches a maximum at a produc-
tion rate of around 3,000 sheets a week. Referring to Figure 4.6, the precise prof-
it-maximizing production rate is 3,400. The marginal cost at a production rate of
3,400 is $7 per sheet. Since the price is $7 per sheet, this confirms that the produc-
tion rate of 3,400 maximizes profit.
Break-Even Analysis
Should the business continue in operation? In the long run, a business should
continue in production if the maximum profit from continuing in production is
Supply 83
at least as large as the profit from shutting down. In the long run, all costs are
avoidable. Hence, if the business shuts down, it will incur no costs and so its profit
from shutting down would be zero.
Let R − C represent the maximum profit from continuing in production. Then
the business should continue in production if
R − C ≥ 0, (4.6)
which simplifies to
R ≥ C. (4.7)
This break-even condition says that the business should continue in production so
long as total revenue covers total cost.
Since total revenue is price multiplied by sales, R = p × q, we can divide the
break-even condition by sales (equal to the production rate) to obtain
C
p≥ . (4.8)
q
Thus, an equivalent way of stating the long-run break-

even condition is that the price must cover the average Long-run break-even:
Total revenue covers
cost.
total cost, or price covers
To summarize, in the long run, a business (which can average cost.
sell as much as it would like at the market price) maximizes
profit by:
• if the total revenue covers the total cost, producing at the rate where the
marginal cost equals the price;
• if the total revenue does not cover the total cost, shutting down.
Referring to Table 4.5, Luna’s lowest average cost is $6.86. It attains this cost
at a production rate of 3,000 sheets a week. Hence, if the price of plywood falls
below $6.86, then Luna should go out of business.
Individual Supply Curve

A seller maximizes profit by producing at the rate where its long-run marginal
cost equals the price of the output. By varying the price, we can determine the
quantity that the seller will supply at every possible price of the output. Further,
the seller should remain in business only if the price covers the average cost. Thus,
the seller’s long-run individual supply curve is that part of its long-run marginal
cost curve which lies above its long-run average cost curve.
84 4 Supply
Short and Long Run

Referring to Tables 4.2 and 4.5, the average cost of production is higher in the
short run than in the long run. The reason is that, in the long run, the seller has
more flexibility in optimizing inputs to changes in the production rate. Accord-
ingly, it can produce at a relatively lower cost than in the short run, when one or
more inputs cannot be adjusted.
Note, however, that the short-run average cost includes the average fixed cost,
which, being sunk, is not relevant. So, to compare the short- and long-run break-
even conditions, the relevant comparison is between the short-run average vari-
able cost and the long-run average cost.
PROGRESS CHECK 4D
Referring to Table 4.5, if the long-run market price of plywood is $10.31 per
sheet, how much should Luna produce?
DURHAM: CONTINUE IN BUSINESS?
Durham Furniture produces bedroom furniture at Durham, Ontario, Canada.

In 2003, the company invested C$38 million in a new plant at Chesley, Ontario,
to manufacture dining room furniture. However, the initiative coincided with
increased competition from Asia and the appreciation of the Canadian dollar
against the US dollar, followed by the US economic downturn.
In 2008, Durham mothballed the Chesley plant and filed for protection
under the Companies’ Creditors Arrangement Act. The company had a net
worth of between C$3.5 and C$6.3 million, excluding the Chesley plant and
equipment. However, the company owed C$37 million in secured borrowings
to the Royal Bank of Canada. The Bank supported a restructuring provided
that new investors invested in the company.
Why would new investors put money in Durham Furniture? The long-run
break-even condition for a business is that the (long-run) price covers the
average cost. The new investors must expect that, in the long run, demand
would recover and that growth in supply from Asia and elsewhere would not
outstrip the growth in demand. Then the long-run price would appreciate to
cover the average cost.
Source: “Durham Furniture plant may be sold,” Sun Media, January 30, 2009.
5. Seller Surplus
We derived the individual supply curve by asking the seller how much it would
supply at every possible price. Another way to interpret the supply curve is that it
Supply 85
marginal cost
individual seller surplus
c b
7
marginal revenue =
price
d
4.29
a
0 1 5
FIGURE 4.7 Individual seller surplus.

Notes: The marginal cost of producing 1,000 sheets is $4.29. At a price of $7, the seller receives a
surplus of ($7 – $4.29) × 1,000 = $2,710 for that production. At the $7 price, the seller will produce
5,000 sheets a week. The individual seller surplus is the shaded area dbc between the price line and
the marginal cost curve.
shows the minimum price that the seller will accept for each unit of production.
Using this approach, we can explain how sellers benefit or suffer from changes
in the price of output. This also motivates a concept that is essential to effective
management of purchasing.
In the case of Luna, referring to Figure 4.7, the marginal cost of producing 1,000
sheets a week is $4.29. This is the minimum price that Luna will accept for the first
1,000 sheets. At a market price of $7, however, Luna receives $7 for each sheet pro-
duced. The difference of (7 − 4.29) × 1,000 = $2,710 is a surplus for the seller. Indeed,
the seller will get a surplus on every unit up to the marginal unit produced.
Generally, the seller surplus is the difference between a
Seller surplus: The
seller’s total revenue from some quantity of production and the difference between a
seller’s avoidable cost of producing that quantity. To illustrate, seller’s total revenue
consider Luna’s short-run seller surplus at the price of $7. from some quantity
Referring to Figure 4.7, the total revenue is represented of production and the
by the area of the rectangle 0abc under the price line up to seller’s avoidable cost of
producing that quantity.
the production of 5,000 sheets a week. The variable cost is
the area 0abd under Luna’s marginal cost curve up to 5,000
sheets. The shaded area dbc between the price line and the marginal cost curve
represents the seller surplus.
In the short run, the seller surplus equals total revenue less variable cost
(assuming that the fixed cost is sunk). In the long run, the seller surplus equals
total revenue less total cost (assuming that the fixed cost is avoidable).
The concept of seller surplus is central to the management of purchasing. Note
that the minimum that a seller would be willing to accept for some quantity of
86 4 Supply
production is the avoidable cost. Any extra payment would yield seller surplus. So
the ideal for a purchasing manager is to pay each supplier just the avoidable cost,
leaving the supplier with zero surplus. Essentially, the ideal would be to buy up the
seller’s marginal cost curve.
PROGRESS CHECK 4E
In Figure 4.7, show how an increase in the price of plywood to $9 per sheet
would affect Luna’s seller surplus.
6. Elasticity of Supply
Consider a typical issue for an industry analyst. If the price of plywood and wages
increase by 5% and 10% respectively, what would be the effect on the supply of
plywood? The analyst could answer this question with the supply curve. In prac-
tice, however, analysts seldom have sufficient information to construct the entire
supply curve.
Another response to the question applies the elasticities of supply, which mea-
sure the responsiveness of supply to changes in underlying factors such as the
prices of the item and inputs. The elasticities of supply are the supply-side coun-
terparts to the elasticities of demand, introduced in Chapter 3.
The price elasticity is a handy way of comparing the sensitivity of the sellers of
different items to changes in price. For instance, a food manufacturer that processes
fruits and vegetables will want to know how sensitive suppliers of the fruits and
vegetables are to changes in prices. The manufacturer can address this question by
comparing the price elasticities of the supplies.
Price Elasticity
Price elasticity of
The price elasticity of supply measures the responsiveness of
supply: The percentage by the quantity supplied to changes in the price of the item. By
which the quantity supplied definition, the price elasticity of supply is the percentage
will change if the price of by which the quantity supplied will change if the price of the
the item rises by 1%. item rises by 1%. Equivalently, the price elasticity is the ratio
Percentage change in quantity supplied

. (4.9)
Percentage change in price
Let us calculate the price elasticity of Luna’s supply of plywood. Referring

to Figure 4.8, at a price of $7, Luna produces 5,000 sheets a week. If the price
increases to $8, Luna would increase production to 5,800 sheets. The percentage
change in quantity supplied is the change in quantity supplied divided by the
Supply 87
Individual
supply curve
8
Cost/price ($ per sheet)
0 5 5.8
Production rate (thousand sheets per week)
FIGURE 4.8 Price elasticity of supply.

Notes: The percentage change in quantity supplied is 16%, while the percentage change in price is
14.3%. Hence, the price elasticity of supply is 16/14.3 = 1.12.
initial quantity supplied. The change in quantity supplied is 5,800 − 5,000 = 800
sheets, hence the percentage change in quantity supplied is 800/5,000 = 16%.
Similarly, the percentage change in price is the change in price divided by the
initial price. The change in price is $8 − $7 = $1, hence, the percentage change in
price is 1/7 = 14.3%. Accordingly, the price elasticity of Luna’s short-run supply
is 16/14.3 = 1.12.
Intuitive Factors
To estimate the price elasticity requires information on a change in price and the
corresponding change in quantity supplied. However, changing the price to esti-
mate the elasticity may be too costly or not practical. As an alternative, managers
can consider two intuitive factors to gauge the price elasticity of supply.
• Available production capacity. One factor is the available production

capacity. A seller with considerable excess capacity will step up production
in response to even a small increase in price. So supply will be relatively
elastic. On the other hand, if capacity is tight, the seller may not increase
production by much for even a substantial price increase. Then supply will
be relatively inelastic.
• Adjustment time. In the short run, some inputs may be costly or impossible
to change. Consequently, the marginal cost of production will be high. For
instance, a factory wishing to step up production quickly may have to pay
overtime rates to workers. Since overtime rates are higher than regular wage
rates, the marginal cost of expanding production will be relatively high. With
sufficient time, however, the factory could hire more workers at the regular
88 4 Supply
wage. Accordingly, in the long run, the marginal cost will increase more gently.
So, generally, the long-run supply will be more elastic than the short-run supply.
PROGRESS CHECK 4F
Under what intuitive conditions would supply be inelastic?
7. Market Supply
To address questions such as the impact of entry of new manufacturers and

changes in the prices of wood on the furniture market, we need to understand the
market supply. The market supply curve of an item is a graph showing the
quantity that all sellers will supply at every possible price.
Market supply curve:
A graph showing the The market supply is the seller-side counterpart to the mar-
quantity that all sellers will ket demand, introduced in Chapter 2. Together, the supply
supply at every possible and demand constitute a market.
price. The market supply curve is constructed in a similar way to
the individual supply curve. To construct the market supply
for an item, ask each potential seller for the quantity that it would supply at every
possible price. Then, at each price, add the reported individual quantities to get
the quantity that the market as a whole will supply.3
Short and Long Run

The long-run market supply differs from the short-run market supply in one essen-
tial way. In the long run, every business will have complete flexibility in deciding
on inputs and production. This flexibility implies that existing sellers can leave the
industry, and new sellers can enter. The freedom of entry and exit is the essential
difference between the short and long run.
For a seller to break even in the long run, its total revenue must cover its total
cost. If a seller’s total revenue does not cover the total cost, then the seller should
leave the industry. Hence, the seller’s individual supply will fall to zero. This depar-
ture will reduce the market supply, hence raise the market price and the profits of
the other sellers. Sellers that cannot cover their total costs will leave the industry
until all the remaining sellers break even.
By contrast, an industry where businesses can make (economic) profits, in the
sense that total revenue exceeds total cost, will attract new entrants. Each of the new
entrants will contribute its individual supply and so add to the market supply. The
increase in the market supply will push down the market price and hence reduce
the profits of all the existing sellers.
Accordingly, in the long run, when there is a change in the market price, the
quantity supplied will adjust in two ways: first, all existing sellers will adjust their
Supply 89
quantities supplied along their individual supply curves; and second, some sellers
may enter or leave the market. Therefore, for any change in price, the long-run
market supply is more elastic than the short-run market supply.
Properties
The properties of the market supply curve are similar to those of the individual
supply curve. As each seller’s marginal cost increases with production, the market
supply curve slopes upward. Equivalently, at a higher price, the market as a whole
will supply a larger quantity.
The market supply depends on other factors that affect individual demand – in
particular, the prices of inputs. So, for instance, the market supply of furniture
depends on the price of wood. If the price of wood is higher, the marginal cost
of producing furniture will be higher, and so sellers will produce less. This implies
that the market supply curve will shift to the left.
The market supply curve also depends on the sellers in existence. If Asian fur-
niture manufacturers enter the US furniture market, then, at every possible price,
there will be a larger quantity supplied. This means that the market supply curve
will shift to the right.
The market seller surplus is the difference between the sellers’ total revenue and
the sellers’ avoidable cost. Graphically, it is the area between the price line and the
market supply curve.
In general, the effect of a change in the price of the output is represented by a
movement along the market supply curve. By contrast, a change in the price of
any input will cause a shift of the entire market supply curve.
US FURNITURE SUPPLY
The US supply of furniture combines supply by domestic, Canadian, and Asian

manufacturers. Furniture manufacturing is relatively intensive in wood and labor.
Canada’s ample forests give it a natural advantage in the furniture industry.
With the advantage of substantially lower labor costs, Asian manufacturers
pose direct and indirect challenges to Canadian furniture. Asian manufacturers
can directly challenge the Canadian industry in both fully assembled and
ready-to-assemble furniture. In this regard, the Asian manufacturers have
increased US supply at the lower-cost end of the supply curves for fully
assembled and ready-to-assemble furniture.
Moreover, Asian manufacturers pose an indirect challenge to Canadian and
US domestic manufacturers of ready-to-assemble furniture. Fully assembled
furniture from Asia is cheap enough to compete with ready-to-assemble
furniture made in Canada and the United States.
90 4 Supply
KEY TAKEAWAYS
• To the extent that marginal costs increase with production, sellers will only
increase production for higher prices.
• In the short run, a business (which can sell as much as it would like at the
market price) maximizes profit by: if the total revenue covers the variable cost,
producing at the rate where the marginal cost equals the price; if the total
revenue does not cover the variable cost, shutting down.
• Fixed cost is the cost of inputs that do not change with the production rate,
while variable cost is the cost of inputs that do change with the production rate.
• Marginal cost is the change in total cost due to the production of an additional
unit, while marginal revenue is the change in total revenue arising from selling
an additional unit.
• In the long run, a business (which can sell as much as it would like at the market
price) maximizes profit by: if the total revenue covers the total cost, producing
at the rate where the marginal cost equals the price; if the total revenue does
not cover the total cost, shutting down.
• In the long run, businesses can adjust by entering or exiting the industry.
• Seller surplus is the difference between the seller’s total revenue from some
quantity of production and the seller’s avoidable cost of producing that quantity.
• Managers can raise the profit from purchasing by extracting the sellers’ surplus.
• The price elasticity of supply is the percentage by which quantity supplied will
change if price of the item rises by 1%.
REVIEW QUESTIONS
1. Explain the distinction between the short run and long run. How is this related
to the distinction between fixed and variable costs?
2. Comment on the following statement: “It costs our factory an average of $5 to
produce a shirt. I cannot accept any order for less than $5 per shirt.”
3. Farmer Axel’s fixed cost of growing corn is higher than that of Farmer Julia. What, if
anything, does this imply for the difference in their marginal costs of growing corn?
4. Advertising is an important input into the marketing of shampoo. Explain the
meaning of the marginal product of advertising for a manufacturer of shampoos.
5. Under what conditions would a seller’s marginal revenue equal the market
price of its product?
6. Presently, Jupiter Oil is producing 2,000 barrels of crude oil a day. The market
price is $15 per barrel. Its marginal cost is $20 per barrel. Explain how the
company can increase its profit.
7. Does the following analysis under- or overestimate the change in profit? The
price of eggs is 60 cents per dozen and Farmer Luke produces 10,000 dozen
eggs a month. If the price of eggs rises to 70 cents a dozen, Luke’s profit will
increase by $1,000.
8. Explain why the following two conditions for short-run break-even are equivalent:
(i) revenue covers variable cost; and (ii) price covers average variable cost.
9. Some companies continue in business even though they are losing money. Are
they making a mistake?
Supply 91
10. Explain why the following two conditions for long-run break-even are equivalent:
(i) revenue covers total cost; and (ii) price covers average cost.
11. Explain the differences between the short-run and long-run decisions to
continue in business.
12. How would (a) a higher price of lumber, and (b) lower wages affect the supply
of furniture?
13. For a given increase in price of the product, will the increase in seller surplus be
smaller or larger if the supply is more elastic? (Hint: Draw two supply curves,
one more elastic than the other.)
14. If the supply curve slopes upward, the price elasticity of supply will be positive.
True or false?
15. Consider the price elasticity of the market supply of taxi service. Do you expect
supply to be relatively more elastic with respect to the fare in the short run or
long run?
1. A retail bank’s sources of funds include savings, time, and checking (current)
deposits. In June 2011, Hang Seng Bank, a leading Hong Kong bank, quoted
interest rates of 0.01% on savings accounts, 0.15% on a 12-month time deposit,
and nothing on checking accounts.
(a) Suppose that the bank incurs an additional 0.2% cost to administer
checking accounts, and 0.1% cost for savings accounts and time
deposits. List the three sources of funds in ascending order of annual
cost per dollar of funds.
(b) Suppose that the bank has $2 billion of savings deposits, $5 billion of time
deposits, and $3 billion of checking deposits. On a diagram with amount
of funds in billions of dollars (from 0 to 10 billion) on the horizontal axis
and cost in millions of dollars on the vertical axis, illustrate the (i) average
variable cost of funds, and (ii) marginal cost of funds.
2. Table 4.1 shows the weekly expenses of Luna Plywood. Suppose that wages
are 5% higher, increasing the cost of labor by 5%.
(a) By recalculating Table 4.1, explain how the increase in wages will affect
the (i) average variable cost, and (ii) average cost.
(b) Luna’s management says that the wage increase has raised its break-
even price by 5%. Do you agree?
(c) Supposing that the market price remains at $7, how should Luna adjust
production?
3. Ole Kirk Kristiansen, founder of the Lego Group in Billund, Denmark,
invented the Lego brick, made from acrylonitrile butadinese styrene. In 2005,
Lego outsourced the production of Duplo bricks to Flextronics, a contract
manufacturer with a plant in Hungary. Then, in 2006, Lego sold its Lego brick
factory at Kladno, Czech Republic, to Flextronics. Lego retained production of
only the more complicated Technic and Bionicles products in Billund.
(a) Wages are higher in Denmark than in Hungary. Compare the equipment
and labor costs of producing Duplo bricks in Denmark and Hungary.
92 4 Supply
(b) How did the sale of the Kladno factory affect Lego’s fixed relative to
variable costs of producing Lego bricks?
(c) How did the sale of the Kladno factory affect Lego’s short- and long-
run break-even for Lego bricks? (Hint: State any assumption on the
price which Flextronics charges to manufacture Lego bricks.)
4. Barrick Gold’s two most productive gold mines are Cortez and Goldstrike in
Nevada, USA. Table 4.7 reports the selling prices and costs for the two mines.
The “average cash cost” includes operating cost, royalties, and taxes, while
the “average cost” includes the cash cost as well as amortization.
(a) Suppose that Barrick Gold operates each mine at the scale where
the (upward-sloping) marginal cost equals the price of gold. The
marginal cost curve shifts with past production, the prices of inputs,
and production technology. Using the data from Table 4.7, illustrate
the shifts in Goldstrike’s marginal cost curve, selling price, and profit-
maximizing scale of production between 2010 and 2012. (Hint: Barrick
Gold reports only one point on each marginal cost curve. You are free
to assume any other data necessary to draw the figures.)
(b) Use Barrick’s 2012 data to compare the (i) short-run break-even
conditions for Cortez and Goldstrike; and (ii) the long-run break-even
conditions for the two mines.
(c) If the price of gold falls to $600 per ounce, how should Barrick adjust
production at the two mines?
5. In Alaska, the state ferry system transports passengers and vehicles. With the
ferry system incurring a financial deficit, Governor Frank Murkowski blamed
losses in part on increased winter sailings. However, management’s view was
that: “Even when a vessel is tied up, expenses are still incurred as a captain
and crew must be on board and line handlers must be available.” In 2005,
the ferry system raised fares by 17%, and revenue fell. However, the following
year, the system cut fares and registered an 18% increase in winter revenue.
(Sources: Alaska State Senate Finance Committee hearing, March 2, 2006;
Alaska Public Radio, October 17, 2005.)
Table 4.7 Barrick Gold
Cortez Goldstrike
2010 2011 2012 2010 2011 2012
Production 1,141 1,421 1,370 1,239 1,088 1,174

(thousand ounces)
Selling price ($ per 1,228 1,578 1,669 1,228 1,578 1,669
ounce)
Average cash cost 244 245 282 475 511 541
($ per ounce)
Average cost ($ per 452 426 503 569 593 629
ounce)
Source: Barrick Gold Corporation, year-end reports and year-end mine statistics.
Supply 93
(a) Categorize the following as fixed or variable costs in the short run:
(i) depreciation of ships; (ii) salaries and wages; (iii) fuel expenses.
(b) How can you infer that the demand for ferry service is elastic?
(c) Supposing that the ferry system aims to minimize losses, under what
condition should the system increase winter sailings?
6. Japanese researchers compared various methods, using different catalysts
and processes, of producing fatty acid methyl ester (FAME) from waste cooking
oil at a scale of 7,269 tons a year. With the CaO-D method, the average fixed
cost is $232 per ton and average variable cost is $391 per ton (including the
cost of 3,226 kilograms of methanol). With the KOH-D method, the average
fixed cost is $225 per ton and average variable cost of $416 per ton (including
the cost of 3,970 kilograms of methanol).
(a) The price of FAME will fluctuate in the future according to demand and
supply. Which method has the lower break-even price in the: (i) short
run; (ii) long run?
(b) You are setting up a factory to produce FAME. Explain how to use
the technique of net present value to choose between the CaO-D and
KOH-D methods.
(c) The original estimates assumed that the price of methanol is $0.455
per kilogram. How would an increase in the price of methanol affect
the choice between the two methods?
7. The US supply of furniture combines supply by domestic, Canadian, and Asian
manufacturers. Furniture is sold in two formats – fully assembled, and ready-to-
assemble in kits which the consumer must assemble at home. Manufacturing
of furniture is relatively labor-intensive.
(a) How do wages affect the relative cost of manufacturing fully assembled
compared to ready-to-assemble furniture?
(b) Asian manufacturers benefit from lower wages. Using suitable
diagrams, illustrate how the entry of Asian manufacturers would
affect the US supply of: (i) fully assembled; and (ii) ready-to-assemble
furniture.
(c) Using either of your diagrams for (b), show how a reduction in the cost
of shipping would affect the supply of furniture.
(d) Which do lower shipping costs affect relatively more: Asian exports of
fully assembled or ready-to-assemble furniture?
8. In July 2012, Johnson Service Group decided to close 100 of its 460 dry
cleaning shops in Britain. At the outlets targeted for closure, sales had declined
by 2.7% in the first half of the year. However, Johnson did not know whether
the fall in demand for dry cleaning was due to continued weakness in the
macroeconomy or consumers switching to machine-washable suits. (Source:
“Johnson closing 100 shops as dry cleaning feels the pinch,” The Independent,
July 5, 2012.)
(a) What is the long-run break-even condition for a dry cleaner?
(b) Which of the following affects the long-run price of dry cleaning
relatively more: (i) macroeconomic weakness; or (ii) a shift in consumer
preference toward machine-washable suits?
94 4 Supply
(c) Using a suitable figure, illustrate how the Johnson Service Group’s
decision affected the supply of dry cleaning services. (Hint: You are
free to assume any data necessary to draw the figures.)
(d) For a competing dry cleaner, was the Johnson Service Group’s decision
good or bad news?
9. A controversial issue in managing climate change is the effect of taxes on
gasoline. Higher taxes would reduce the after-tax price received by gasoline
producers. The price elasticity of the supply of gasoline has been estimated
to be 2.0. (Source: David Austin and Terry Dinan, “Clearing the air: The
costs and consequences of higher CAFE standards and increased gasoline
taxes,” Journal of Environmental Economics and Management, Vol. 50, No. 3,
November 2005, pp. 562–582.)
(a) Explain why the elasticity of supply with respect to price is positive.
(b) Suppose that a tax on gasoline reduces the after-tax price of gasoline
by 5%. By how much would suppliers reduce gasoline production?
(c) Do you expect the impact of the tax on the production of gasoline to be
smaller or larger in the short run as compared to the long run? Explain
your answer.

Consider the various products that your organization sells. Identify products for
which your organization is planning to reduce production.
(a) Compare the assumptions about the market underlying the original
production plan and the actual outcomes.
(b) Are the key differences in (a) of a short- or long-run nature?
(c) Should your organization immediately cut production or gradually reduce
production over time?
Appendix: Constructing Market Supply
This chapter has introduced the concept of market supply, which shows the quan-
tity that all sellers will supply at every possible price. Here, we discuss in detail the
method by which to construct the market supply curve from the individual supply
curves of the various sellers.
Generally, the market supply curve is the horizontal sum of the individual sup-
ply curves. At any particular price, each seller’s individual supply curve shows the
quantity that the seller will supply. The sum of these quantities is the quantity
supplied by the market as a whole. By varying the price, we can get the informa-
tion needed to construct the market supply curve.
Short Run
The short-run supply curve of an individual seller is the portion of its marginal
cost curve that lies above its average variable cost curve. Hence, the market supply
Supply 95
Luna Venus Market
Price ($ per sheet)

Price ($ per sheet)
Price ($ per sheet)

8 8 8
7 7 7
0 5 5.8 0 8 9.2 0 13 15
Production rate Production rate Production rate
FIGURE 4.9 Market supply.

Notes: The market supply curve is the horizontal summation of the individual supply curves. At a price
of $7 per sheet, the market quantity supplied is 5,000 + 8,000 = 13,000 sheets. At a price of $8 per
sheet, the market quantity supplied is 5,800 + 9,200 = 15,000 sheets.
curve begins with the seller that has the lowest average variable cost. The market
supply curve then gradually blends in sellers with higher average variable cost.
In Figure 4.9, we assume that there are just two producers of plywood: Luna
Plywood and Venus Manufacturing. We draw the individual supply curves of the
two producers and sum these horizontally to obtain the market supply curve. For
instance, at a price of $7, the market as a whole will supply 5,000 + 8,000 = 13,000
sheets a week. At a price of $8, the market will supply 5,800 + 9,200 = 15,000
sheets.
Long Run
The long-run supply curve of an individual seller is the portion of its marginal
cost curve that lies above its average cost curve. Hence, the market supply curve
begins with the seller that has the lowest average cost. The market supply curve
then gradually blends in sellers with higher average cost.
Notes
1 The following analysis is based, in part, on “Durham Furniture plant may be sold,” Sun Media,
January 30, 2009; “Chesley wood plant to reopen,” Homegoodsonline, January 30, 2012.
2 Here, we assume that every unit is sold at the same price – a policy of uniform pricing. In Chap-
ter 9, we consider the possibility of price discrimination, where various units are sold at different
prices. With price discrimination, total revenue is not simply price multiplied by sales.
3 The appendix to this chapter discusses in detail the method by which to construct the market
supply curve from the individual supply curves of the various sellers.
5 C H A P T E R
Market Equilibrium
LEARNING OBJECTIVES
• Appreciate the impact of excess supply on the market price.
• Appreciate the impact of excess demand on the market price.
• Apply the price elasticities of demand and supply to predict the
impact of shifts in supply on market price and production.
• Apply the price elasticities of demand and supply to predict the
impact of shifts in demand on market price and production.
• Appreciate the short- and long-run effects of shifts in demand on
market price.
• Appreciate the short- and long-run effects of shifts in demand on
production.
1. Introduction
As the leading source of energy, oil underpins economic growth throughout the
world. Oil from major exporting countries in the Middle East, West Africa, and
the Gulf of Mexico is shipped by tankers to consumers in the West and Asia.
With the Great Recession, the demand for oil, and consequently the demand
for tanker services, fell precipitously. In just 12 months, between 2008 and 2009,
freight rates for very large crude carriers (VLCCs), with capacities of around
300,000 tons, collapsed from a record high of $88,400 to $28,000 a day. Rates for
Suezmax tankers, with capacities of around 160,000 tons, fell from $67,200 to
$25,900 a day. In tandem, the utilization rate of the overall tanker fleet, includ-
ing VLCCs, Suezmax, and smaller tankers, fell from 91% to 84%. The industry
Market Equilibrium 97
analysts Platou remarked: “This indicates an overcapacity not seen since the
1980s.”1
In 2010, following the economic recovery led by China and other Asian econ-
omies, the demand for tanker services began to recover. VLCC rates increased to
$34,800 a day, while Suezmax rates rose to $28,000 a day. Utilization of the overall
tanker fleet rose to 86%.
The tanker industry is a global industry but quite fragmented. It includes the
fleets of integrated oil producers such as BP, ExxonMobil, and Saudi Aramco,
as well as independent owner-operators. In December 2010, total tonnage of the
industry amounted to 451 million deadweight tons (dwt).
The shipping tycoon John Fredriksen controls Frontline, a large independent
owner-operator, with a fleet of 17.6 million dwt comprising 44 VLCCs and 21
Suezmaxes. Frontline owns 12 VLCCs and 12 Suezmaxes, and operates the
other vessels under charter, spot rentals, and other contractual arrangements. The
average age of its tankers is 11 years, compared to the industry average of 9 years.
Frontline emphasized in its 2010 Annual Report: “The tanker industry is highly
cyclical, experiencing volatility in profitability, vessel values and freight rates. Freight
rates are strongly influenced by the supply of tanker vessels and the demand for
oil transportation.”
Nevertheless, the company must plan for the future. When facing declining
freight rates and falling utilization, it must decide whether to lay up or scrap ves-
sels. When facing rising freight rates and increasing utilization, it must decide
whether to charter new vessels and/or to place orders with shipyards for new ves-
sels. Frontline must also respond to changes in wages, interest rates, and the costs
of other inputs.
Looking into the future, Frontline’s management must make short- and long-run
decisions – whether to remain in business, and the scale of operations (how many
vessels to operate). To address each of these decisions, management must under-
stand demand, supply, and the interaction between the two sides of the market.
At first glance, the need to understand both demand and supply may seem sur-
prising. Worldwide economic recession and recovery affect the demand for oil and
thus the demand for tanker services. They do not affect the supply of tanker services.
Yet, as we shall see, management cannot make correct decisions unless they
consider both sides of the market. The same applies to many other managerial
issues. Although the initial change affects only one side of a market, it is necessary
to consider the interaction with the other side to make the best decisions.
This chapter combines the earlier analyses of demand and supply to under-
stand how they interact in competitive markets. The central concepts are the role
of price in communicating information to buyers and sellers, and market equilib-
rium of demand and supply.
Within the demand–supply framework, we can explain the effects of global
economic recession and recovery on the market for tanker service. When demand
falls, the market will be in excess supply and price will fall. When demand rises, the
market will be in excess demand and price will increase. For Frontline, with global
98 5 Market Equilibrium
economic recovery, the key to investment decisions is the increase in demand and
the short- and long-run elasticities of supply.
The demand–supply framework is the core of managerial economics. It can
be applied to address managerial decision-making in a wide range of markets,
including both goods and services, consumer as well as industrial products, and
domestic and international markets.
2. Perfect Competition
Generally, demand–supply analysis applies to markets in which competition is

very keen in the particular sense of perfect competition. The following conditions
define perfect competition:
• The product is homogeneous.

• There are many buyers, each of whom purchases a quantity that is small
relative to the market supply.
• There are many sellers, each of whom supplies a quantity that is small
relative to the market demand.
• New buyers and sellers can enter freely, and existing buyers and sellers can
exit freely.
• All buyers and all sellers have symmetric information about market conditions.
In practice, very few markets exactly satisfy the conditions for perfect competition.
This means that demand–supply analysis does not precisely apply. Nevertheless,
this method of analysis is still useful: many of the managerial implications are
similar even if a market is not perfectly competitive.
Homogeneous Product
Competition in a market where products are differentiated is not as keen as that in
a market where products are homogeneous. Gold is a homogeneous commodity.
Gold mined in Canada is a perfect substitute for gold from Australia, Brazil, and
South Africa. If a Canadian gold producer tried to sell its gold at even 1% above
the prevailing world market price, absolutely no one would buy. By the same token,
if the Canadian producer offered its gold at 1% less than the market price, it would
be swamped with orders. The price of gold from any source is exactly the same.
In contrast, mineral water is not homogeneous. Water from different sources
has different chemical composition and hence different taste and therapeutic
effect. A mineral water producer can raise its price by 1% without worrying that
all of its consumers would switch to other brands. Likewise, if it reduced its price
by 1%, its sales would increase, but by only a limited degree. Consequently, there
is no uniform price for mineral water: different manufacturers may charge dif-
ferent prices. In general, competition among manufacturers of mineral water is
relatively weaker than competition among gold mines.
Many Small Buyers

The second condition for perfect competition is that there are many buyers, each
of whom purchases a quantity that is small relative to the market supply. In such a
market, no buyer can get a lower price than others; hence, all buyers face the same
price and all buyers compete on the same level playing field.
In a market where some buyers have market power, the buyers pay different
prices, and the same buyer might even pay different prices for different units of
the same product. Thus, when some buyers have market power, it is not possible
to construct a market demand curve. (This is because constructing the demand
curve requires an assumption that each buyer can buy as much as they would like
at the market price.)
In the market for cotton, there are countless buyers of cotton, ranging from
Indian villagers to Paris designers. Each buyer’s purchases are very small relative
to the world supply. Every buyer pays the same world price.
By contrast, the demand for the Chinese star anise herb is dominated by the
pharmaceutical manufacturer, Roche, which uses it to produce the flu vaccine,
Tamiflu. Roche has market power in the demand for the herb. As a result, it can
buy at better prices than smaller buyers.
Many Small Sellers

The third condition for perfect competition is that there are many sellers, each of
whom supplies a quantity that is small relative to the market demand. In such a
market, no seller can get a higher price than others; hence, they all face the same
price and compete on the same level playing field.
By contrast, when some sellers have market power, they receive different prices.
Then it would not be possible to construct a market supply curve. (This is because
constructing the supply curve requires an assumption that each seller can supply
as much as they would like at the market price.)
Compare the markets for dry cleaning and cable television. In any city, there
are many dry cleaners, none of which has market power. By contrast, there would
be one or two providers of cable television, each of which has substantial market
power. Accordingly, the market for dry cleaning is more competitive than that for
cable television.
Free Entry and Exit

The fourth condition for perfect competition is that new buyers and sellers can
enter freely, and existing buyers and sellers can exit freely. In particular, this means
that no technological, legal, or regulatory barriers constrain entry or exit. To
explain this condition, let us focus on free entry by new sellers and free exit by
existing sellers. The logic for free entry and exit among buyers is quite similar.
Consider a market with free entry and exit. As we saw in Chapter 4 on market
supply, if the market price rises above a seller’s average cost, then new sellers will
be attracted to enter. This will add to the market supply and bring down the price.
Hence, with free entry and exit, the market price cannot stay above a seller’s aver-
age cost for very long. The market will be very competitive.
To illustrate, compare the markets for telephone service and telemarketing.
Telephone service providers must make huge investments to build networks. By
contrast, a telemarketer can set up with a few telephone lines and staff. As a result,
there is much more competition among telemarketers.
The degree of competition also depends on barriers to exit. Suppose that a
telephone service provider must pay the government a compensation fee to cease
service. It must consider this exit cost when deciding whether to enter the market.
Therefore, the higher the exit costs, the less likely new providers are to enter the
market; hence, the less competitive the market will be.
Symmetric Information
The fifth condition for perfect competition is that all buyers and all sellers have
symmetric information about market conditions such as prices, available substi-
tutes, and technology. With symmetric information, every seller or buyer will be
subject to intense competition. If, for instance, a new supplier offers a key input at
a lower price, then every producer will immediately get the same lower price. No
seller can enjoy the privilege of secret information.
Dry cleaning service is a mature industry. Information about service quality is
transparent and evenly available among buyers and sellers.
By contrast, information is not symmetric in the market for medical services.
Patients rely on their doctors for advice. In this market, sellers (doctors) have
better information than buyers (patients). There may be differences in informa-
tion on the supply side as well as between the two sides of the market. Doctors
who attend continuing education and follow the latest journals may provide better
advice and treatment than others.
Markets where there are differences in information among buyers, or among
sellers, or between buyers and sellers are not as competitive as those where all
buyers and all sellers have symmetric information.
PROGRESS CHECK 5A
What are the conditions for a market to be in perfect competition?
3. Market Equilibrium
Market equilibrium: The For markets that meet the conditions for perfect com-
price at which the quantity petition, we can apply the demand–supply framework. In
demanded equals the this framework, the concept of market equilibrium unifies
quantity supplied.
demand and supply. Market equilibrium is the price at
which the quantity demanded equals the quantity supplied.
Market equilibrium is the basis for all analyses of how changes in demand or
supply affect market outcomes.
Demand and Supply

To introduce the concept of market equilibrium, consider the market for tanker
services. Shippers generate the demand, and tanker lines provide the supply. On a
graph with the price of tanker service in dollars per ton-mile on the vertical axis
and the quantity of tanker service in billions of ton-miles a year on the horizontal
axis, we can draw the demand for and the supply of tanker service. (One ton-mile
represents the carriage of one ton over a distance of one mile.)
Suppose that the equilibrium in the market for tanker service is at a price of
$200 per ton-mile and quantity of 10 billion ton-miles a year. Referring to Figure 5.1,
the demand curve shows that, at the price of $200, buyers want to purchase a total
of 10 billion ton-miles a year. The supply curve shows that, at the price of $200, sellers
want to supply a total of 10 billion ton-miles a year. The quantity that buyers want
to purchase exactly balances the quantity that sellers want to supply.
At the market equilibrium, there is no tendency for price, purchases, or sales
to change. The price will not tend to change because the quantity demanded of
10 billion ton-miles just balances the quantity supplied of 10 billion ton-miles.
Purchases will not tend to change because, at the price of $200, buyers (shippers)
maximize benefits less expenditure by purchasing 10 billion ton-miles. Further,
sales will not tend to change because, at the price of $200, sellers (tanker lines)
maximize profits by supplying 10 billion ton-miles.
For a market to be in equilibrium, both the quantity demanded and the quan-
tity supplied must be the result of voluntary choices by buyers and sellers, respec-
tively. Neither buyers nor sellers may face rationing or other restrictions.
Excess supply supply

a
Price ($ per ton-mile)
220
b
200 equilibrium
180
Excess demand demand
0 7 8 10 12 13
Quantity (billion ton-miles a year)
FIGURE 5.1 Market equilibrium.

Notes: At the price of $200 per ton-mile, the quantity demanded is 10 billion ton-miles a year and
the quantity supplied is the same. When the price is $220 per ton-mile, the excess supply is 5 billion
ton-miles a year. When the price is $180 per ton-mile, the excess demand is 5 billion ton-miles a year.
What happens when the market is not in equilibrium? Then, generally, the mar-
ket price will tend to change in such a way as to restore equilibrium. The price
signals information and provides an incentive for buyers and sellers to converge
to equilibrium.
Excess Supply
One way in which the market can be out of equilibrium is when the price exceeds
the equilibrium level. Suppose that the market price is $220 per ton-mile. Then,
referring to the demand curve in Figure 5.1, buyers would cut back purchases to
7 billion ton-miles a year. Referring to the supply curve, sellers would increase
quantity supplied to 12 billion ton-miles a year.
Hence, at a price of $220 per ton-mile, the quantity supplied would exceed the
quantity demanded by 12 − 7 = 5 billion ton-miles a year. In more colorful terms,
there would be too many ships chasing too few customers. The amount by which
the quantity supplied exceeds the quantity demanded is the
Excess supply: The excess supply. So, at a price of $220, there would be an
amount by which the excess supply of 5 billion ton-miles a year.
quantity supplied exceeds
In a situation of excess supply, the market price will tend to
the quantity demanded.
fall. Tanker lines would compete to clear their extra capacity
and the market price would drop back toward the equilib-
rium level of $200.
From Figure 5.1 it is clear that, if the price were even higher, at $250 per ton-
mile, the excess supply would be even larger than the excess supply at a price of
$220. Generally, the higher is the price above the equilibrium level, the larger will
be the excess supply.
Excess Demand
Another way in which the market can be out of equilibrium is when the price is
below the equilibrium level. Then the market price will tend to rise. In the market
for tanker service, suppose that the price is $180 per ton-mile, and that, at the
price of $180, buyers would purchase 13 billion ton-miles a year, and sellers would
supply 8 billion ton-miles a year. Then the quantity demanded would exceed the
quantity supplied by 13 − 8 = 5 billion ton-miles a year.
The amount by which the quantity demanded exceeds the quan-
Excess demand: The tity supplied is the excess demand. At a price of $180 per ton-
amount by which the mile, there would be excess demand of 5 billion ton-miles a year.
quantity demanded Faced with this excess demand, buyers would compete for the
exceeds the quantity limited capacity and the market price would rise toward the
supplied.
equilibrium level of $200. Generally, the lower is the price below
the equilibrium level, the greater will be the excess demand.
PROGRESS CHECK 5B
In Figure 5.1, illustrate the excess demand if the price is $160 per ton-mile.
Mark the quantities demanded and supplied.
4. Supply Shift
In general, changes in wages, interest rates, and the cost of other inputs, or gov-
ernment policies will shift the demand, supply, or both. Even if the change super-
ficially appears to affect only one side of the market, it is essential to analyze the
effects on the other side as well. Any analysis that ignores the other side of the
market could be seriously incomplete.
In this section, we consider the effect of a change that shifts the supply curve.
Specifically, how would a reduction in wages affect the price of tanker service?
Let us apply the demand–supply framework to address this question. We start
from the equilibrium before the change in wages. Suppose that, before the change
in wages, the price of tanker service was $200 per ton-mile and the quantity pur-
chased was 10 billion ton-miles a year.
Equilibrium Change
The supply curve of tanker service does not explicitly show wages. Accordingly,
any change in wages will shift the entire supply curve. Suppose that the reduction
in wages causes the entire supply curve of tanker service to shift downward by
$6 per ton-mile. We represent this shift in Figure 5.2.
The entire supply curve shifts down because the reduction in wages affects
sellers’ marginal costs whatever the quantity that they supply. Another way of
looking at the impact of the change in wages is that it shifts the supply curve to
the right: at every possible price, sellers want to supply more.
The change in wages, however, does not affect the demand for tanker service.
Referring to Figure 5.2, the new supply curve crosses the unchanged demand curve
at a new equilibrium point d, where the price is $196 per ton-mile. The fall in price
from $200 to $196 increases the quantity demanded from 10 to 10.4 billion ton-
miles a year. On the new supply curve, at the price of $196, the quantity supplied
is 10.4 billion ton-miles a year. The price of $196 is the new market equilibrium.
a
original supply
b
200 $6
196 new supply
d
demand
0 10 10.4
FIGURE 5.2 Supply shift.

Notes: When wages fall, the entire supply curve shifts downward by $6. At the new equilibrium, the
price is $196 per ton-mile and the quantity is 10.4 billion ton-miles per year.
When the supply curve shifts down by $6, the equilibrium price falls by only $4.
Generally, a shift in the supply curve will change the equilibrium price by no more
than the amount of the supply shift. What determines the change in price? It
depends on the price elasticities of demand and supply.
Price Elasticity of Demand

Intuitively, on the demand side, if buyers are extremely insensitive to price, then
they will not buy more; hence, the price will fall by the entire $6. If, however, they
are very sensitive to price, then the shift in supply would result in no change to the
equilibrium price.
To illustrate, consider Figure 5.3(a), which depicts an extremely inelastic demand.
This means that buyers are completely insensitive to the price: they purchase the
(a) Extremely inelastic demand (b) Extremely elastic demand

demand
original supply
original supply
200 b 200 demand b $6
$6
194 new supply
new supply
0 10 0 10 10.6
Quantity (billion ton-miles a year) Quantity (billion ton-miles a year)
(c) Extremely inelastic supply (d) Extremely elastic supply

original and new
a supply
b original supply
200 b 200
$6 new supply
194
demand
demand
0 10 0 10 11
Quantity (billion ton-miles a year) Quantity (billion ton-miles a year)
FIGURE 5.3 Price elasticities of demand and supply.

Notes:
(a) When demand is extremely inelastic, the supply curve shift reduces the price by exactly $6 but
does not affect quantity.
(b) When demand is extremely elastic, the supply curve shift does not affect the price but raises the
quantity to 10.6 billion.
(c) When supply is extremely inelastic, the supply curve does not shift.
(d) When supply is extremely elastic, the supply curve shift reduces the price by exactly $6 and raises
the quantity to 11 billion.
same quantity regardless of the price. Accordingly, when the supply curve shifts,
the buyers do not change their behavior – they continue to purchase exactly the
same quantity. In Figure 5.3(a), when the supply curve shifts down by $6, the
equilibrium price drops by exactly $6 to $194 per ton-mile.
Figure 5.3(b) depicts the other extreme, which is an extremely elastic demand.
This means that buyers are extremely sensitive to price. When the supply curve
shifts, the buyers soak up all the additional quantity supplied. Consequently, the
equilibrium price does not change at all. In Figure 5.3(b), when the supply shifts
down by $6, the equilibrium price remains unchanged at $200 per ton-mile. The
new equilibrium quantity is 10.6 billion.
Comparing Figures 5.3(a) and (b), we see the relationship between the price
elasticity of demand and the outcome of a shift in supply. Generally, if the demand
is more elastic, then the change in the equilibrium price resulting from a shift in
supply will be smaller.
Price Elasticity of Supply

Intuitively, on the supply side, if sellers are very insensitive to price, then the
reduction in cost will not induce them to sell more; hence, the price of tanker ser-
vice will not change at all. If, however, sellers are very sensitive to price, then the
shift in supply would cause the price to fall by the entire $6.
To illustrate, consider Figure 5.3(c), which depicts an extremely inelastic supply.
This means that sellers are completely insensitive to the price: they provide the
same quantity regardless of the price. In particular, if their costs change, they will
not change the quantity supplied. In Figure 5.3(c), tanker lines supply 10 billion
ton-miles a year whatever the market price. Consequently, the change in fuel cost
does not change the equilibrium price.
Figure 5.3(d) depicts the other extreme, which is an extremely elastic supply.
This means that, essentially, the marginal cost of production is constant. Accord-
ingly, if the cost of an input changes, the marginal cost changes by the same
amount at all production levels. Then the equilibrium price changes by exactly
the same amount. In Figure 5.3(d), when the supply curve shifts down by $6, the
equilibrium price drops by exactly $6 per ton-mile. The quantity supplied rises to
11 billion ton-miles a year.
Comparing Figures 5.3(c) and (d), we see the relationship between the price
elasticity of supply and the outcome of a shift in supply. Generally, if the supply
is more elastic, then the change in the equilibrium price resulting from a shift in
supply will be larger.
PROGRESS CHECK 5C
Which of (a) and (b) is true? For a given downward shift of the market supply
curve, the drop in the equilibrium price will be larger if: (a) the demand is more
price elastic; or (b) the supply is more price elastic.
THE DEMAND FOR FRENCH PRODUCTS: FOIE GRAS

AND BUTTER
France exports foie gras and butter. The supplies of French products to world
markets depend on the exchange rate between the euro and other world
currencies. If the euro becomes more expensive, the supply curves of French
products to world markets will be higher. If, however, the euro becomes
cheaper, then the supply curves will be lower.
Suppose that the euro becomes 10% more expensive, causing the supply
curves of French foie gras and butter to shift up by 10%. How will these shifts
affect the prices of foie gras and butter on world markets?
There are few substitutes for French foie gras; hence, the demand is relatively
inelastic. The upward shift of the supply curve will result in a relatively large
increase in the world price. By contrast, there are many close substitutes for
French butter, so that the demand is relatively elastic. Accordingly, the upward
supply shift curve will result in a relatively small increase in the world price.
TANKERS AND INTEREST RATES
As tankers cost up to $100 million and operate for 20 years or more, the
demand for tankers depends on financing. So the demand for tankers and, in
turn, the supply of tanker services are sensitive to interest rates.
In the wake of the Great Recession, national central banks boosted the money
supply, leading to lower interest rates. To the extent that they also reduced long-
run interest rates, the supply of tanker services would have shifted downward.
This would have further depressed the rates for tanker services.
5. Demand Shift
We have just shown that, to understand the impact of a shift in supply, it is crucial
to consider the interaction between supply and demand. Our next application
begins with a change that shifts demand. For a complete understanding of the
outcome, however, it is necessary to consider the supply side as well.
How would an increase in oil shipments affect the price and quantity of tanker
service? Figure 5.4 shows the original equilibrium at point b, with a price of $200
and quantity of 10 billion ton-miles a year.
Suppose that the demand rises by 1 billion ton-miles at all price levels. Accord-
ingly, in Figure 5.4, the entire demand curve shifts to the right by 1 billion ton-miles.
The increase in oil shipments, however, does not directly affect the supply of
tanker service. So the supply curve does not change.
supply
a 1 billion
f
b
200
1 billion new demand
original demand
c
0 10 10.8
FIGURE 5.4 Demand shift.

Note: A 1 billion ton-mile increase in demand shifts the demand curve to the right and results in a
higher price and larger quantity.
Referring to Figure 5.4, the new demand curve crosses the unchanged supply
curve at a new market equilibrium (point f ). The new equilibrium has a higher
price and a larger quantity of tanker service. By how much will the price rise and
by how much will the quantity of tanker service increase? They depend on the
price elasticities of both demand and supply.
PROGRESS CHECK 5D
In Figure 5.4, show how a 2 billion ton-mile reduction in demand would affect
the supply curve and the equilibrium price.
DEMAND AND SUPPLY ON VALENTINE’S DAY
People buy greeting cards and roses throughout the year. As Valentine’s
Day approaches, the demand for both cards and roses increases. Applying
demand–supply analysis, we expect the prices of both products to rise. The
price of roses, however, always increases much more sharply than the price of
greeting cards. Why?
Consider the price elasticities of supply in the two markets. The supply of
greeting cards on Valentine’s Day is much more elastic than the supply of roses.
Greeting cards can be stored, so manufacturers can easily step up production
and prepare larger stocks ahead of Valentine’s Day. This means that the supply of
cards is relatively elastic; hence, an increase in demand has little effect on price.
By contrast, roses are perishable. Only roses maturing around Valentine’s
Day will be suitable for that day. It is relatively costly to increase the quantity
supplied just for Valentine’s Day. This means that the supply is relatively inelastic,
and consequently, the increase in demand causes the price to increase sharply.
Source: B. Peter Pashigian, “Demand and supply on Valentine’s Day,” in Price Theory and
Applications, New York: McGraw-Hill, 1995, p. 19.
6. Adjustment Time
The impact of shifts in demand and supply on the market equilibrium depends
on the price elasticities of demand and supply. The elasticities vary with the time
horizon. Accordingly, shifts in demand and supply may have different short-run
and long-run effects.
To illustrate these differences, suppose that, originally, the market for tanker
service was in short- and long-run equilibrium, with a price of $200 per ton-mile
and quantity of 10 billion ton-miles. What are the short-run and long-run effects
of an increase in demand by 1 billion ton-miles?
Short-Run Equilibrium
Figure 5.5 depicts the short-run market equilibrium at a price of $200 per ton-mile.
Figure 5.5(a) shows the cost and demand curves of an individual seller. The short-
run supply curve of any individual seller is that portion of its short-run marginal
cost curve that lies above its short-run average variable cost curve.
By the assumption of perfect competition, each seller supplies a quantity that is
small relative to the market. Equivalently, it has a small market share. Hence, the
demand facing the seller is extremely elastic at the $200 market price. In the short
run, the seller maximizes profit by operating at the point where its short-run mar-
ginal cost equals the market price. In Figure 5.5(a), the profit-maximizing scale is
100 million ton-miles a year.
(a) Individual seller (b) Market
short-run
marginal short-run
cost short-run supply

1 billion
average
variable cost c
220 220
200 200 a
price
short-run
demand
0 100 105 0 10 12
Quantity (million ton-miles a year) Quantity (billion ton-miles a year)
FIGURE 5.5 Short-run market equilibrium.

Notes:
(a) An individual seller operates at 100 million ton-miles a year, where the short-run marginal cost
equals the market price.
(b) The market is in equilibrium at a price of $200, where the short-run demand crosses the short-run
supply.
(a) Individual seller (b) Market
new long-run

average cost 1 billion long-run
supply
d
210 210
200 200 a
original long-
run average
cost long-run
demand
0 100 0 10 13
Quantity (million ton-miles a year) Quantity (billion ton-miles a year)
FIGURE 5.6 Long-run market equilibrium.

Notes:
(a) An individual seller operates at 100 million ton-miles a year, where the long-run marginal cost
equals the market price.
(b) The market is in equilibrium at a price of $200, where the long-run demand crosses the long-run
supply.
Figure 5.5(b) shows the short-run market equilibrium at point a. At the

$200 equilibrium price, the short-run market demand curve crosses the short-run
Long-Run Equilibrium
Figure 5.6 depicts the long-run market equilibrium at a price of $200 per ton-mile.
Figure 5.6(a) shows the cost and demand curves of an individual seller. The long-
run supply curve of any individual seller is that portion of its long-run marginal
cost curve that lies above its long-run average cost curve.
As each seller has a small market share, it faces a demand that is extremely elas-
tic at the $200 market price. In the long run, it maximizes profit by operating at the
point where its long-run marginal cost equals the market price. By Figure 5.6(a),
the profit-maximizing scale is 100 million ton-miles a year.
Figure 5.6(b) shows the long-run market equilibrium at point a. At the
$200 equilibrium price, the long-run market demand curve crosses the long-run
Demand Increase
Starting from the short- and long-run equilibria, we suppose that the demand
curve shifts to the right by 1 billion ton-miles. For simplicity, we assume that the
short- and long-run demand curves are the same.
First consider the new short-run equilibrium. Referring to Figure 5.5(b), the
shift in demand will move the short-run market equilibrium to point c, with a
higher price of $220. At the same time, referring to Figure 5.5(a), every seller
expands its operations to the scale of 105 million ton-miles, where its short-run
marginal cost equals the new market price of $220. This means operating service
capacity more intensively – for instance, by delaying routine maintenance on ships
and postponing the crews’ annual vacations.
Generally, the extent to which a seller expands its operations depends on the
slope of its short-run marginal cost curve. If the short-run marginal cost curve is
steep, then the price increase will not lead the seller to expand operations by very
much. By contrast, if the short-run marginal cost curve is gentle, then the price
increase will induce a large expansion in operations. The slope of the short-run
marginal cost curve depends on such factors as the availability of excess produc-
tion capacity and the cost of overtime relative to standard wage rates.
Next consider the new long-run equilibrium. In a long-run horizon, there is
enough time for all costs to become avoidable, for new sellers to enter the market,
and for existing sellers to leave. Accordingly, as shown in Chapter 4, the market
supply tends to be more elastic in the long run than in the short run.
In the tanker service market, the increase in demand raises the market price and
hence each seller’s profits. Over the long run, this will induce existing sellers to
expand – acquire new ships and hire more crew – and also will attract new sellers
to enter the market. The industry will expand along the long-run market supply
curve. Referring to Figure 5.6(b), the shift in demand will move the long-run market
equilibrium to point d with a price of $210.
Figure 5.6(a) shows the new long-run equilibrium for an individual seller.
Although the price is higher than in the original equilibrium, higher input prices
result in higher marginal and average cost curves. Accordingly, in the new long-run
equilibrium, each individual seller just breaks even. No other sellers will wish to
enter the industry, and no seller will wish to leave.
Figure 5.7 depicts both the short- and long-run market equilibria. The original
equilibrium is point a, the new short-run equilibrium is point c, and the new long-
run equilibrium is point d. The price in the new long-run equilibrium is lower than
in the new short-run equilibrium but higher than in the original equilibrium. The
quantity in the new long-run equilibrium is higher than in the new short-run equilib-
rium, which in turn is higher than in the original equilibrium. The basic reason for
these differences is that the supply is more elastic in the long run than the short run.
In the new long-run equilibrium, there will be more sellers than in the new short-
run equilibrium or the original equilibrium. The higher price attracts new sellers to
enter and supports a larger number of sellers; hence, the industry becomes larger.
Demand Reduction
We have considered the short- and long-run impacts of an increase in demand.
We can apply the same approach to study the effects of a reduction in demand.
Figure 5.8 illustrates a 1 billion ton-mile reduction in the demand for tanker
service. The original equilibrium is at point a.
1 billion short-run supply
c long-run supply
220
d
210
200
a
1 billion
new demand
original demand
0 10 12 13
FIGURE 5.7 Demand increase: short and long run.

Notes: Following an increase in demand, the new short-run equilibrium is at point c and the long-
run equilibrium is at point d. The price rises more in the short run than in the long run. The quantity
increases less in the short run than in the long run.
1 billion short-run supply

long-run supply
a
200
f
190
170
e
1 billion
new demand original demand

0 10
FIGURE 5.8 Demand reduction: short and long run.

Notes: Following a reduction in demand, the new short-run equilibrium is at point e and the long-run
equilibrium is at point f. The price falls more in the short run than in the long run. The quantity drops
less in the short run than in the long run.
The reduction in demand will move the short-run market equilibrium to point e,
with a lower price of $170. Those sellers whose average variable cost exceeds the
price will shut down. In the tanker market, this means laying up their ships. Those
sellers for whom the price covers their average variable cost will continue in busi-
ness. Each will cut back operations to the scale where its short-run marginal cost
equals the new market price of $170.
The extent of the cutback depends on two factors. One factor is the extent of
sunk costs. If a seller has many prior commitments, most costs are sunk. It will con-
tinue to produce so long as the price covers its average variable cost. In this case, the
price reduction will lead to a relatively minor cutback in operations. Generally, in an
industry where production involves substantial sunk costs, the reduction in demand
will translate into a relatively large drop in price and a small reduction in quantity.
The second factor is the slope of the seller’s short-run marginal cost curve. If
the short-run marginal cost curve is steep, then the price reduction will not induce
the seller to cut back operations by very much. By contrast, if the short-run mar-
ginal cost curve is gentle, then the price reduction will have a relatively larger
impact on quantity.
In the long run, there is enough time for all costs to become avoidable, for new
sellers to enter the market, and for existing sellers to leave. Referring to Figure 5.8,
the shift in demand will move the long-run market equilibrium to point f, with a
price of $190. For some sellers, the long-run price is below their average total cost.
These will exit the industry, which means scrapping their ships and dismissing all
workers. The entire industry will shrink along the long-run market supply curve.
In the new long-run equilibrium, there will be a smaller number of sellers, and
each will exactly break even with average total cost equal to the market price.
Referring to Figure 5.8, the price in the new long-run equilibrium is higher than
in the new short-run equilibrium but lower than in the original equilibrium. Fur-
ther, the quantity in the new long-run equilibrium is less than in the new short-
run equilibrium, which in turn is less than in the original equilibrium. The basic
reason for these differences is that the market supply is more elastic in the long
run than in the short run.
Short and Long Run

In the short run, some costs are sunk, and, with less adjustment time, supply is
less price elastic. By contrast, in the long run, all costs are avoidable and, with
more adjustment time, supply is more price elastic.
Generally, in response to shifts in demand:
• The market price will change more in the short run than the long run.
Specifically, if there is an increase in demand, the market price will increase
more in the short run than in the long run. By contrast, if there is a reduction
in demand, then the market price will fall more in the short run than in the
long run. Generally, the market price tends to overshoot – adjusting relatively
more in the short run and then reversing toward the initial price in the long run.
• Production will adjust more over the long run than in the short run. If there is
an increase in demand, production will increase somewhat in the short run and
more in the long run. Likewise, if there is a reduction in demand, production
will fall somewhat in the short run and more in the long run. Generally, the
long-run adjustment in production amplifies the short-run adjustment.
We can apply the same approach to consider the short- and long-run effects of
shifts in supply. Note that demand is less price elastic in the short run than in the
long run (except possibly for durables).
PROGRESS CHECK 5E
In Figure 5.8, suppose that the short-run supply is less elastic and the long-run
supply is more elastic. How would that affect the difference between short- and
long-run prices?
FRONTLINE
Frontline is a large independent owner-operator of tankers, with a fleet of

17.6 million dwt comprising 44 VLCCs and 21 Suezmaxes. Frontline owns
12 VLCCs and 12 Suezmaxes, and operates the other vessels under charter,
spot rentals, and other contractual arrangements.
With the Great Recession, the demand for tanker services fell sharply.
Between 2008 and 2009, freight rates for VLCCs collapsed from a record
high of $88,400 to $28,000 a day. Rates for Suezmax tankers fell from $67,200
to $25,900 a day. The utilization rate of the overall tanker fleet fell from 91%
to 84%.
How should Frontline adjust its fleet to the fall in freight rates? Suppose that
the fall in demand was permanent (which it was not).
Over time, fleet operators would lay up vessels and others would even scrap
their ships. So, at every price, there would be less capacity. Referring to Figure
5.8, the long-run supply would be more elastic than the short-run supply.
Hence, in the long run, freight rates would recover. Indeed, by 2010, VLCC
rates increased to $34,800 a day, while Suezmax rates rose to $28,000 a day.
For long-run decisions whether to scrap vessels or acquire new ships,
management should distinguish short-run from long-run freight rates. It is the
long-run rates that matter.
KEY TAKEAWAYS
• If the market price exceeds equilibrium, there will be excess supply and the
price will tend to fall.
• If the market price falls below equilibrium, there will be excess demand and
the price will tend to rise.
• A shift in supply will affect the market price and quantity to an extent that
depends on the elasticities of both demand and supply.
• A shift in demand will affect the market price and quantity to an extent that
depends on the elasticities of both demand and supply.
• A shift in demand will lead to a larger change in price in the short run than in
the long run.
• A shift in demand will lead to a smaller change in production in the short run
than in the long run.
REVIEW QUESTIONS
1. To what extent does the market for dry cleaning meet the conditions for perfect
competition?
2. To what extent does the market for plywood meet the conditions for perfect
competition?
3. How would the requirement for a government license affect the competitiveness
of an industry?
4. If the market is in excess supply, what will happen to the price?
5. If the market is in excess demand, what will happen to the price?
6. Explain why the effect of an increase in consumer incomes on the market price
of clothing depends on the price elasticity of supply.
7. Under what conditions of the price elasticities of demand and supply would
an upward shift in demand (due to buyers getting more benefit) have a larger
effect on the price?
8. Under what conditions of the price elasticities of demand and supply would
an upward shift in demand (due to buyers getting more benefit) have a larger
effect on the quantity consumed?
9. Consider an increase in consumer incomes that shifts the demand to the right
(at every price, consumers want to buy more). Explain why the effect on the
production of cars depends on the price elasticity of demand.
10. Under what conditions on the price elasticities of demand and supply would a
reduction of supply (shift in supply curve to the left) have a larger effect on the
price?
11. Under what conditions on the price elasticities of demand and supply would a
reduction of supply (shift in supply curve to the left) have a larger effect on the
quantity produced?
12. Explain why the effect of an increase in wages of waiters and kitchen staff on
the market price of restaurant meals depends on the price elasticity of demand.
13. Suppose that demand increases. Why does the price increase more in the
short run than in the long run?
14. Suppose that demand increases. Why does production increase more in the
long run than in the short run?
15. How does the difference in short- and long-run effects of an increase in demand
depend on the difference between short- and long-run price elasticities?
1. Between 2008 and 2009, freight rates for VLCCs collapsed from $88,400 to
$28,000 a day. The utilization rate of the overall tanker fleet fell from 91% to
84%. Between 2009 and 2010, as the world economy recovered, VLCC rates
rose to $34,800 a day, and utilization of the overall tanker fleet rose to 86%.
(a) Using relevant demand and supply curves, illustrate the shift in short-
run equilibrium in tanker services between 2008 and 2009.
(b) Using relevant demand and supply curves, illustrate the shift in short-
run equilibrium between 2009 and 2010.
(c) Comment on the price elasticity of the short-run supply of tanker services.
2. Producers of television sets outsource production to contract manufacturers in
China and elsewhere. As of November 2014, the exchange rate of the Chinese
yuan to the US dollar was 16 US cents to 1 yuan.
(a) Explain how the appreciation of the Chinese yuan from 16 to 18 US
cents would affect the cost of supplying TVs to the United States.
(b) Suppose that the Chinese yuan rises by 10% against the US dollar.
Which of the following are plausible explanations of why the US retail
price of Chinese-made TVs will rise by less than 10%: (i) the wholesale
cost accounts for only part of retailers’ costs; (ii) American retail
demand for Chinese-made TVs is inelastic; (iii) American retail supply
of Chinese-made TVs is inelastic?
3. Seasonal changes can affect demand and supply. In the market for fresh fruit
and vegetables, the supply varies with the season. By contrast, in the market
for heating oil, the demand varies with the season.
(a) Using suitable demand and supply curves, explain how prices of fruit
and vegetables would vary over the four seasons of the year.
(b) Using suitable demand and supply curves, explain how the price of
heating oil would vary through the seasons.
4. Using Figure 5.3 as a basis, construct a set of four diagrams to show the effect
of an increase in the demand for tanker service on the market price when:
(a) demand is extremely inelastic; (b) demand is extremely elastic; (c) supply is
extremely inelastic; and (d) supply is extremely elastic.
5. Manufacturers of packaging use inputs of wood pulp and wastepaper.
In Sweden, the estimated price elasticity of the demand for wastepaper is −0.8 in
the short run and −1.7 in the long run. The estimated price elasticity of the supply
is 0.6 in the short run and 0.4 in the long run. (Source: Anna Mansikkasalo,
Robert Lundmark, and Patrik Söderholm, “Market behavior and policy in the
recycled paper industry: A critical survey of price elasticity research,” Forest
Policy and Economics, Vol. 38, January 2014, pp. 17–29.)
(a) Consider a government policy that reduces the price of wastepaper to
manufacturers by 5%. How will this affect the quantity demanded in
the short and long run?
(b) Consider a government policy that increases the price of wastepaper
to sellers by 5%. How will this affect the quantity supplied in the short
and long run?
(c) Are price incentives an effective way of increasing the recycling of
wastepaper?
6. The Japanese consume relatively more fish and less meat than people in other
developed countries. In 1995, Worldwatch Institute President Lester R. Brown
pronounced: “[I]f the Chinese were to consume seafood at the same rate as the
Japanese do, China would need the annual world fish catch.” (Source: Lester
R. Brown, Who Will Feed China?, New York: Norton, 1995, p. 30.)
(a) On a single figure, draw the world demand and supply of fish. (Hint:
You are free to assume any data necessary to draw the curves.)
(b) Using your figure, explain how increases in the Chinese demand for
fish would affect the world market.
(c) How would increases in the Chinese demand for fish affect Japanese
consumption of fish?
(d) Is it likely that China would consume the entire world fish catch?
7. The Kurdistan region of Iraq, estimated to have 45 billion barrels of oil reserves,
has been exporting crude oil by truck through Turkey. In late 2013, the regional
government completed a 20-inch diameter connection to the main Iraq–Turkey
pipeline. The connection can convey 125,000 barrels of oil per day to the port
of Ceyhan in Turkey. (Source: “Iraqi Kurds, Turkey to double oil export pipeline
capacity,” Bloomberg, August 20, 2014.)
(a) Compare the fixed and marginal costs of transporting oil by (i) truck
and (ii) pipeline.
(b) Which involves more sunk cost: supply by truck or pipeline?
(c) Draw the short-run supply curve of crude oil from Kurdistan through Turkey.
On the supply curve, mark the supplies by truck and pipeline. (Hint:
Consider whether the pipeline or trucks supply oil more inelastically.)
(d) Suppose that the Kurdistan government levies a tax on oil transported by
the pipeline to pay for security. How would the tax affect the supply curve?
(e) The demand for Kurdistan oil fluctuates with global market conditions.
Would the price of Kurdistan oil be more volatile if supplied by truck or
pipeline?
8. With January 1, 2000 heralding a new millennium, many predicted shortages of
lobsters. New England and Canadian wholesalers amassed stockpiles. Boston
dealer James Hook ordered 675,000 kilograms, 50% more than the previous year.
The anticipated shortages, however, did not materialize. In early December, with
just weeks to the New Year, the wholesale price of lobster sank 12% to $11.70 per
kilogram (Source: “Lobster dealers net meager sales on New Year celebration
stockpile,” Asian Wall Street Journal, December 29, 1999, pp. 1 and 7.)
(a) On a suitable diagram, draw the long-run supply of lobster for New
Year’s Eve. In gauging the price elasticity of supply, note that lobster
can be stockpiled for over 6 months.
(b) Illustrate the effect of an increase in demand from 1998 to 1999. How
would the increase in demand affect the price? How would the price
effect depend on the price elasticity of supply?
(c) Processors have developed a method to freeze whole lobsters in a plastic
sleeve of brine that provides a quality almost equal to the fresh animal.
How would this new technology affect the elasticity of long-run supply?
9. In January 2009, Vaughan-Bassett, a manufacturer of bedroom furniture,
suspended work at its Elkin, North Carolina, plant, laying off 400 workers. Eight
months later, with sales up by 10%, the company reopened part of the Elkin plant
to finish furniture produced at its main plant at Galax, Virginia. Vaughan-Bassett

easily found workers to staff the reopened plant. (Source: “Vaughan-Bassett
reopens part of Elkin factory,” Time Warner Cable News, November 16, 2014.)
(a) How do sales of new and existing residential housing affect the demand
for furniture?
(b) How would the appreciation of the Chinese yuan against the US dollar
affect the North Carolina furniture industry?
(c) In the face of falling demand, how should Vaughan-Bassett decide
between suspending (“mothballing”) a plant and closing it permanently?
(d) Using suitable short-run demand and supply curves, illustrate the shifts
in the market equilibrium between January and September 2009.

Consider some good or service that your organization supplies in a competitive
market. You have experienced a surge in demand, which is expected to be
permanent. Prices have risen sharply and senior management is proposing new
investment to capitalize on the higher prices. Write a memorandum analyzing
the business case for the new investment, focusing on the short- and long-run
increase in prices.
Note
1 The following discussion is based in part on data from Frontline Ltd, Annual Report 2010; and
Platou Report, 2011.
6 C H A P T E R
Economic Efficiency
LEARNING OBJECTIVES
• Appreciate economic efficiency as a benchmark for maximizing
value.
• Apply the conditions for economic efficiency.
• Appreciate the invisible-hand role of price in competitive markets.
• Apply transfer pricing and outsourcing in decentralization of an
organization.
• Appreciate the incidence of changes in demand or supply on buy-
ers and sellers.
• Appreciate the incidence of intermediation on buyers and sellers.
1. Introduction
The Port Authority of New York and New Jersey manages John F. Kennedy
International, Newark Liberty International, and LaGuardia airports. In 2008,
Newark served 35.4 million passengers, as compared with 47.8 million at Ken-
nedy and 23.1 million at LaGuardia.
Continental Airlines, with 72% of takeoff and landing slots, dominates flights
at Newark. The remaining 28% of slots are spread among other US and interna-
tional carriers. By contrast, LaGuardia (where the largest carrier is US Airways
with 32%) and Kennedy (where the largest carrier is Delta Airlines with 31%) are
less concentrated.
The Port Authority charges airlines landing fees based on aircraft weight.
Larger aircraft pay higher fees than smaller planes for the same slot. The fees are
an average of $6 per passenger and do not vary with the time of day.
Economic Efficiency 119
During peak hours, the demand for takeoffs and landings at Newark exceeds
capacity.1 With two closely-spaced parallel runways, Newark is particularly sus-
ceptible to increased delays when weather conditions reduce visibility.
Between 2000 and 2007, average daily operations (landings and takeoffs) at
Newark decreased by 3% from 1,253 to 1,219. However, on-time arrivals fell from
71% to 62%, and delays in arrival of one hour or more rose from 54 to 93 per
day. In 2007, the proportion of flights arriving on time at Newark was the second
worst among the top 35 US airports.
In October 2008, the Federal Aviation Administration (FAA) presented a
10-year plan limiting scheduled takeoffs and landings to 81 per hour and estab-
lishing an auction for landing and takeoff slots. Carriers with 20 or fewer daily
slots would be assigned the same number of slots for 10 years. Carriers with more
than 20 daily slots would be assigned 20 slots plus 90% of slots in excess of 20 for
10 years, but had to return the remaining 10% to the FAA. The FAA expected
that it would take back 96 slots out of a total of 1,219. It planned to auction them
over 5 years.
The FAA estimated that the plan would reduce the average delay by 23%
and increase buyer and seller surplus by $839 million over the 10-year period
2009–2019. However, the Port Authority, airline associations, Continental Air-
lines, and US Airways resisted the FAA plan. Finally, in May 2009, by order of
the US Court of Appeals, the FAA abandoned the plan and sought other ways to
relieve congestion at Newark.
This chapter introduces the concept of economic efficiency and explains how
to identify inefficiency. Understanding economic inefficiency is fundamental to
every manager because it points to opportunities to increase value added and
profit. Wherever the allocation of resources is not economically efficient, there is
a way to add value and make money by resolving the inefficiency. This is a simple
yet very powerful rule.
Next, we show that competitive markets allocate scarce resources in an eco-
nomically efficient way. Market prices communicate information and provide
incentives for users and suppliers to maximize the value added from scarce
resources.
Takeoff and landing slots at an airport with limited runway capacity are a
scarce resource. However, if the slots are allocated by administrative rule, the
allocation of resources might not be economically efficient. Indeed, the FAA esti-
mated that a reallocation could increase buyer and seller surplus by $839 million
over 10 years.
This chapter presents two important applications of economic efficiency. One
is to decentralization within an organization and the use of transfer pricing. The
other is intermediation in the context of competitive markets. The analysis of
intermediation explains the impact of transportation costs, brokerage fees, and
taxes on buyers and sellers.
Economic efficiency is an important benchmark in settings of market power and
imperfect markets as well as competitive markets. Chapters 9, 12, 13, and 14 apply
120 6 Economic Efficiency
economic efficiency to pricing, externalities, asymmetric information, and inter-

nal organization, respectively. In all of these contexts, the concept of economic
efficiency helps managers to identify and exploit opportunities to increase value
added and profit.
2. Benchmark
An allocation of resources is economically efficient if no reallocation of resources

can make one person better off without making another person worse off. To
appreciate economic efficiency as a benchmark, consider a situation that is not
economically efficient. Then, by some reallocation of resources, it is possible to
add value – specifically, to make one person better off without making another
person worse off. Clearly, the original situation should be avoided.
From the viewpoint of a manager, if a situation is not economically efficient,
then there will be some way to add value, that is, increase benefit by more than
cost. There will also be some way to take some of the value added in profit. So,
economic efficiency is a very useful benchmark for management in profit-oriented
and non-profit organizations.
Conditions for Economic Efficiency

The concept of economic efficiency is a useful benchmark.
Economic efficiency: The definition, however, is difficult to apply in practice. In
practice, it is much easier to use a set of three conditions that
• Same marginal benefit
are based on users’ benefits and suppliers’ costs. An alloca-
for all users.
• Same marginal cost for tion of resources is economically efficient if: all users achieve
all suppliers. the same marginal benefit; all suppliers operate at the same
• Marginal benefit equals marginal cost; and marginal benefit equals marginal cost.
marginal cost. Let us review these three conditions in the context of pas-
senger airlines providing a service between Hong Kong and
New York. Suppose that the route from Hong Kong is served by several airlines.
• Equal marginal benefit. The first condition for economic efficiency is that
all users receive the same marginal benefit. Compare two passengers on a
Hong Kong–New York flight. Max, whose father is a pilot, gets unlimited
free travel, while Maria must pay for her flight. Max flies until his marginal
benefit is zero, while Maria’s marginal benefit equals the fare that she pays.
If society reallocated a seat from Max to Maria, Max’s loss would be less
than Maria’s gain, so society as a whole would be better off. This shows that
an allocation of resources is economically efficient only if all users achieve
the same marginal benefit.
• Equal marginal cost. The second condition for efficiency is that all suppliers
must operate at the same marginal cost. Suppose that, owing to better
management, lower labor costs, and higher fuel efficiency, one airline provides
the service at a marginal cost which is 10% lower than the other airlines. Society
as a whole could reduce the overall cost of airline travel while maintaining
the number of flights by expanding the services of the lower-cost airline and
shrinking the services of the other airlines. This shows that an allocation of
resources is economically efficient only if all suppliers operate at the same
marginal cost.
• Marginal benefit equals marginal cost. The fi nal condition for efficiency
links users with suppliers: for a resource allocation to be economically
efficient, the users’ marginal benefit must equal the suppliers’ marginal
cost. If the marginal benefit is less than the marginal cost, society
overall could add value by reducing the number of fl ights. Likewise, if
the marginal benefit exceeds the marginal cost, then society could add
value by increasing the number of fl ights. Thus, an allocation of resources
is economically efficient only if the users’ marginal benefit equals the
suppliers’ marginal cost.
Consumer Sovereignty
The concept of economic efficiency assesses resource allocations in terms of each
individual user’s evaluation of their benefit. The principle of taking individual
users’ evaluations as given is that of consumer sovereignty. For instance, if some
people like heavy metal, while others like opera, the concept of economic effi-
ciency takes these preferences as a given in assessing the efficiency of resource
allocation. So, an economy in which some consumers like heavy metal but which
produces only opera is not efficient, while an economy in which some consumers
like opera but which produces only heavy metal is also inefficient.
Technical Efficiency
Technical efficiency means providing an item at the min-
imum possible cost. Technical efficiency alone, however, Technical efficiency:
Provision of an item at
does not imply that scarce resources are being well used. For
minimum cost.
instance, an airline may be providing service at the minimum
possible cost. But these may be flights that no one wants.
The concept of economic efficiency extends beyond technical efficiency. For
economic efficiency, the production of the item must be such that the marginal
benefit equals the marginal cost.
PROGRESS CHECK 6A
Explain the difference between economic efficiency and technical efficiency.
FREE STORAGE, UNLIMITED SERVICE – FOR HOW LONG?
Many Web portals offer a limited quantity of free service to entice consumers.
Initially, Google’s Gmail offered 1 gigabyte of free storage. In competition
with Microsoft’s Hotmail and Yahoo! mail, Gmail increased its free storage
to over 7.5 gigabytes. Not to be outdone, in April 2012, Microsoft’s OneDrive
expanded its free offering to 25 gigabytes.
However, these free services are costly to provide. In 2013, the data centers
that support computing services and storage consumed an estimated 910
million megawatt-hours of electricity.
So how do service providers control the cost of free services? One way is
to close accounts that have been dormant for specified periods of time. In
its terms of service, Gmail clearly states that: “Google may terminate your
account ... if you fail to login to your account for a period of nine months.”
Source: “America’s data centers consuming and wasting growing amounts of energy,”
National Resources Defense Council, December 4, 2014.
3. Adam Smith’s Invisible Hand
[H]e intends only his own gain, and he is ... led by an invisible hand to
promote an end which was no part of his intention.2
Although published over 200 years ago, Adam Smith’s

Invisible hand: Market insight is no less valid today. In a competitive market, buy-
price guides buyers ers and sellers, each acting independently and selfishly, will
and sellers, acting
independently and
channel scarce resources into economically efficient uses.
selfishly, to channel The invisible hand that guides the multitude of buyers and
scarce resources into sellers is the market price. This invisible hand is a simple and
economically efficient uses. practical way of achieving economic efficiency.
Competitive Market
Consider how the invisible hand works in the market for airline services between
Hong Kong and New York. Demand comes from travelers and supply from air-
lines. Figure 6.1 shows the market equilibrium with a price of $2,000 and quantity
of 1.1 million seats a year.
On the demand side, as explained in Chapter 2, each person will buy up to the
quantity such that their marginal benefit equals the price of travel, $2,000, and
this is true for every buyer. In a perfectly competitive market, all buyers face the
same $2,000 price; hence, their respective marginal benefits will be equal. This is
the first condition for economic efficiency.
supply
Price ($ per seat)
a
2,000
demand
Quantity (million seats a year)
FIGURE 6.1 Air travel market.

Notes: In equilibrium, the demand crosses the supply at a price of $2,000. Each consumer purchases
up to the quantity where the marginal benefit is $2,000. Each provider supplies the quantity where
marginal cost is $2,000.
What about the airlines? On the supply side, as shown in Chapter 4, each airline
will expand up to the scale where the marginal cost equals the price, $2,000. This
scale of operations maximizes profit. Again, in a perfectly competitive market, all
airlines face the same $2,000 price. Thus, with each airline selfishly maximizing
profits, every airline will operate at the same marginal cost. This is the second
condition for economic efficiency.
So all buyers balance marginal benefit with price and all airlines balance mar-
ginal cost with price. But, in market equilibrium, all buyers and all airlines face the
same $2,000 price. Therefore, marginal benefit and marginal cost must be equal.
This is the third condition for economic efficiency. Thus, a perfectly competitive
market satisfies all three requirements for economic efficiency.
This example illustrates the power of Adam Smith’s invisible hand. The market
price guides multiple buyers and sellers, each acting independently and selfishly,
to achieve economic efficiency.
Market System
The market price performs two roles to achieve economic efficiency:
• The price communicates the necessary information. It tells buyers how

much to purchase and tells sellers how much to supply.
• The price provides a concrete incentive for each buyer to purchase the quantity
that balances marginal benefit with the market price: by purchasing this
quantity, the buyer achieves the maximum net benefit. Similarly, the price
provides a concrete incentive for every seller to supply the quantity that
balances marginal cost with the market price: by supplying this quantity,
the seller maximizes its profit.
The term market system or price system describes an economic system in

which freely moving prices guide the allocation of resources.
Market (price) system: The term recognizes the key role of prices in a market
An economic system in
which freely moving prices
system. The role of the invisible hand in achieving economic
guide the allocation of effi ciency is the intellectual foundation of the market
resources. system.
PROGRESS CHECK 6B
Explain the two functions of price in a market system.
NEWARK LIBERTY: INVISIBLE HAND DISABLED
The Federal Aviation Administration regulates air services in the United

States. A major issue for the FAA is delays in airline operations. In 2007, the
proportion of flights arriving on time at Newark Liberty International Airport
was the second worst among the top 35 US airports.
In October 2008, the FAA presented a 10-year plan to limit scheduled
operations to 81 per hour and establish an auction for landing and takeoff
slots. Given limited takeoff and landing slots, it is economically efficient to
allocate the slots to the users with the highest marginal benefit. Indeed, the FAA
estimated that the plan would reduce the average delay by 23% and increase
buyer and seller surplus by $839 million over the 10-year period 2009–2019.
By contrast with administrative fiat (or historical legacy), an auction is more
likely to assign slots to the airlines serving travelers with the highest marginal
benefit. The airlines whose passengers were willing to pay the most for the flights
would bid the highest. These airlines would win the auction and get the slots.
However, the Port Authority of New York and New Jersey, which manages
the airport, Continental, the dominant airline at Newark, and others opposed
the FAA. By court order, the FAA was forced to abandon its plan. The invisible
hand was disabled even before it could get to work.
Source: “Congestion management rule for John F. Kennedy International Airport and Newark
Liberty International Airport: Final regulatory evaluation,” US Federal Aviation Administration,
Office of Aviation Policy and Plans, Aircraft Regulatory Analysis Branch, October 10, 2008.
4. Decentralized Management
The concept of economic efficiency applies not only across the economy but also
within an organization. Consider a bank with two divisions – commercial banking
and personal banking. Each division takes deposits and makes loans. The bank
must decide how to allocate funds across the two divisions.
The concept of economic efficiency guides the bank in the use of its limited
funds. Applying the concept, the allocation of funds will be efficient if all users
achieve the same marginal benefit, all suppliers operate at the same marginal cost,
and every user’s marginal benefit balances every supplier’s marginal cost.
In the context of the bank and its limited funds, the users are the lending units
and the suppliers are the deposit-taking units of the two divisions. The first con-
dition for efficiency is that all users receive the same marginal benefit. This means
that each of the bank’s lending units must get the same profit from an additional
dollar of funds. If one lending unit could get more profit than another, the bank
should switch some funds to the lending unit that gets the higher profit. Then the
bank’s overall profit will be higher.
The second condition for efficiency is that all suppliers must operate at the same
marginal cost. If one deposit-taking unit can produce funds at a lower marginal cost
than another, then the bank should direct the lower-cost unit to produce more and
the higher-cost unit to produce less. This would increase the bank’s overall profit.
The final condition for efficiency is that the marginal benefit must balance the
marginal cost. If the marginal benefit of funds to the lending unit is less than
the marginal cost of producing the funds, then the bank should cut back deposit-
taking. The reduction in cost would be greater than the reduction in benefit, so
overall profit would rise. By contrast, if the marginal benefit of funds is greater
than the marginal cost, then the bank should increase deposit-taking. The bank
will maximize profit when the marginal benefit equals the marginal cost.
Internal Market
Just as the invisible hand works to achieve economic efficiency across the entire
economy, it can also work within an individual organization. How should the
bank organize the production and use of funds?
One approach is central planning: the bank headquarters can collect informa-
tion about deposit-taking and lending unit costs and revenues from all lending
units, and then decide the level of funds each unit should accept and how much
each lending unit should lend.
Suppose that there is a competitive market for funds. Then an alternative is for
the bank to decentralize management of deposit-taking in the following way. The
bank can direct the managers of every deposit-taking unit to maximize profit
and sell funds at the market price, whether to the company’s own lending units or
outside buyers. Similarly, the bank can direct every lending unit to maximize
profit and allow them the freedom to procure funds from
any source, whether it be one of the bank’s own deposit-tak-
ing units or external to the bank. Transfer price: The sale
price of an item within an
As these sales are a transfer within the same organization, organization.
the corresponding price is called a transfer price. The bank
should set the transfer price for funds equal to the market price. With the decen-
tralized policy, each lending unit will buy funds up to the point that its marginal
benefit balances the market price.
Since all lending units face the same market price, their marginal benefits will
be equal. Similarly, each deposit-taking unit will produce up to the point that its
marginal cost balances the market price. As all deposit-taking units face the same
market price, their marginal costs will be equal.
Since the deposit-taking and lending units face the same market price, marginal
benefit will be equal to marginal cost. Thus, the decentralized policy achieves the
three conditions for economic efficiency within the same organization. Essentially,
by decentralizing the management of funds, the bank is establishing an internal
market that is integrated with the external market.
Outsourcing
An organization should follow two general rules when decentralizing control over
an internal resource. First, if there is a competitive market for the item, the trans-
fer price should be set equal to the market price. If there is no competitive market
for the item, then the appropriate transfer price is more complicated. We discuss
transfer pricing more generally in Chapter 7.
The second general rule is that producing units should be allowed to sell the
product to outside buyers and consuming units should be allowed to buy the prod-
uct from external sources. Recall from Chapter 1 that outsourcing is the purchase
of services or supplies from external sources.
To explain why the right to outsource is crucial, suppose that the bank requires
all lending units to source funds internally. Then the bank’s deposit-taking units
would have market power, and as Chapter 8 explains, they would charge a price
above the competitive market level. As a result, the lending units would no longer
secure the economically efficient quantity of funds. A similar argument shows
why it is necessary to allow producing units to sell the product to outside buyers.
Any organization that uses resources or products for which there are compet-
itive markets can apply decentralization to achieve internal economic efficiency.
Energy producers can apply the technique to manage the production and use of
crude oil and natural gas, and automobile manufacturers can apply it to manage
production and use of components.
PROGRESS CHECK 6C
Jupiter Electronics manufactures semiconductors and consumer electronics.
Its semiconductors are inputs into the production of electronics. Explain how
the company can use decentralization to ensure that its consumer electronics
division will make efficient use of semiconductors produced by the semicon-
ductor division.
SINOPEC: TRANSFER PRICING
China Petroleum & Chemical Corporation (Sinopec) is a vertically integrated

producer of oil and chemicals. It operates in four lines of business: exploration
and production, refining, marketing and distribution, and chemicals. The
company’s policy is to set inter-segment transfer prices according to “the
market price or cost plus an appropriate margin.”
With a network of over 30,000 stations, Sinopec’s marketing and distribution
division is China’s largest and the world’s second largest. In 2013, the division
earned operating income of 35 billion yuan on revenues of 1486 billion yuan
and assets of 274 billion yuan.
In early 2014, Sinopec announced that it would restructure the marketing
and distribution division and open it to private investment. A key concern for
investors is the transfer prices for supplies of refined petroleum. If Sinopec
refineries set high transfer prices, the marketing and distribution division may
be uncompetitive in the retail market. However, if the investors bargain hard to
reduce transfer prices, then Sinopec refineries may incur losses.
Source: Sinopec, Annual Report, 2014.
5. Intermediation
An important application of the demand–supply framework is to understand the

impact of the costs of retailing, distribution, transportation, brokerage, and other
forms of intermediation on the market for the final good or service.
For instance, manufacturers of industrial products must decide whether to include
the cost of shipment in the price to the customer. One approach is to set an
ex-works or free on board (FOB) price and leave the customer to pay the freight,
while the alternative is to charge a price including freight. Using the demand–
supply framework, we will show that both pricing methods have exactly the same
impact on the manufacturer and customer.
Freight-Inclusive Pricing
We begin by considering the cement market with freight-
CF price: Includes the
inclusive pricing. A price that includes freight is called cost
cost of delivery to the
including freight (CF). buyer.
In the market for cement, the buyers are building contractors
and the sellers are cement manufacturers. Suppose that all
manufacturers set CF prices that include a freight cost of 25 cents per bag. The
market price is $4.50 per bag and buyers purchase 1 billion bags a year. Figure 6.2
illustrates the market equilibrium at point a. In this equilibrium, the marginal
Price ($ per bag)
25 cents CF supply
a 25 cents ex-works supply

4.50
b
CF demand
ex-works demand
0 1
Quantity (billion bags a year)
FIGURE 6.2 Pricing and freight cost.

Notes: If all manufacturers switch to ex-works pricing, the supply curve shifts down by 25 cents and
the demand also shifts down by 25 cents. The equilibrium quantity remains unchanged at 1 billion
bags.
benefit of cement is $4.50 per bag and the marginal cost (inclusive of freight) is
also $4.50 per bag.
Free on Board Pricing

Next, suppose that all cement manufacturers switch to FOB
FOB price: Does not
pricing. An FOB price does not include the freight cost: it
include the cost of
delivery to the buyer. literally means “the price at the gate to the works.”
Since manufacturers no longer incur the freight cost of
25 cents, the switch will shift the entire supply curve of cement
down by 25 cents. In Figure 6.2, the entire supply curve shifts down because each
manufacturer’s marginal cost of supplying cement is reduced whatever the quan-
tity that it actually supplies.
The switch to FOB pricing also affects the market demand. With FOB pricing,
each buyer must pay 25 cents a bag to obtain the cement. In Figure 6.2, this can
be represented by shifting the entire retail demand curve down by 25 cents. Since
each buyer must pay 25 cents in freight for every bag, the buyer’s willingness to
pay will be 25 cents lower at all quantities. This means that the entire demand
curve will shift down by 25 cents.
There is another way to confirm that the switch to FOB pricing will shift down
the demand curve. Consider the buyer of the 1 billionth bag. By the original
demand curve, that buyer would be willing to pay exactly $4.50 for that bag. If,
however, the buyer must incur a 25-cent freight cost, it will now be willing to
pay $4.50 − $0.25 = $4.25 for the 1 billionth bag. The same 25 cent reduction in
willingness to pay applies to all the inframarginal units as well. Hence, the entire
demand curve shifts down by 25 cents.
In Figure 6.2, the new demand and supply curves cross at point b. Relative to
the original equilibrium at point a, the price is lower. The new demand curve is the
original demand curve shifted down by 25 cents. Likewise, the new supply curve
is the original supply curve shifted down by 25 cents. Hence, the new equilibrium
point b must be vertically below the original equilibrium point a, and the vertical
distance a must be 25 cents.
Thus, in the new equilibrium, each buyer pays $4.25 to the seller and $0.25 in
freight, making a “total price” of $4.25 + $0.25 = $4.50, which is exactly the price
under freight-inclusive pricing. Further, the quantity of sales is exactly the same
in the old and new equilibria. Generally, the price and sales are the same whether
the sellers do or do not include the freight cost in their prices.
Further, in the new equilibrium, the marginal benefit of cement is $4.50 per bag,
while the marginal cost (inclusive of freight) is also $4.50 per bag. Accordingly,
from the viewpoint of economic efficiency, the new and old equilibria are identical.
ONLINE RETAIL: FREE(?) SHIPPING
Some online retailers include free shipping while others charge for shipping.
In November 2014, R.T. Edwards offered the Samsung UA65HU9000 65-inch
high definition LED TV for A$4,238 with free shipping, while Exeltek offered
the same TV for $4,095 with a shipping charge of $145 to Perth, Western
Australia.
Demand–supply analysis predicts that, whether online retailers charge for
shipping or provide free shipping, the net price to consumers would be the
same. Indeed, Exeltek’s price plus shipping was A$4,240, which is almost
identical to R.T. Edwards’s price.
Source: getprice.com.au (accessed November 17, 2014).
6. Incidence
When demand or supply shifts, the consequent change in the price for a buyer or
seller is called the incidence on that party. In the freight
example, when manufacturers switch from freight-inclusive Incidence: The change in
to FOB pricing, the market price drops by 25 cents to $4.25; price for a buyer or seller
resulting from a shift in
hence, the net effect on buyers is zero.
demand or supply.
This shows that, although the switch in pricing method
requires buyers to “pay” the freight cost, there is no net effect
after we consider adjustments in both demand and supply. Equivalently, whether
manufacturers set prices that do or do not include the freight cost, the incidence is
the same. The incidence does not depend on which side – buyer or seller – initially
pays the freight cost.
In fact, the incidence of the freight cost depends only on the price elasticities of
demand and supply. This analysis reflects common sense. If sellers pay the freight
cost, the buyers will be willing to pay a higher price. By contrast, buyers that must
pay the freight cost will insist on paying less to sellers.
The distinction between receiving or paying an amount of money and the inci-
dence of the receipt or payment is a fundamental economic concept. We have
applied this distinction in the decision whether to include freight in pricing. The
distinction is also important for understanding the effect of retailing and distri-
bution costs, brokerage fees, and government taxes.
Consider, for instance, the impact of free shipping in e-commerce. Does it mat-
ter whether the vendor offers free shipping or the consumer pays for shipping?
Using similar demand–supply analysis to that applied to freight costs, we can
show that, regardless of whether vendors or consumers pay the shipping costs, the
market price and quantity will be the same.
The analysis of incidence must be qualified if buyers or sellers are subject to
biases in decision-making. The behavioral biases due to sunk costs, status quo,
and anchoring may affect the incidence of receipts or payments. For instance,
with regard to free shipping, consumers might anchor on the free shipping, and so
demand for vendors that offer free shipping would be higher. On the other hand,
consumers might anchor on the price of the product (and give insufficient atten-
tion to shipping charges), and then demand for vendors that charge separately for
shipping would be higher.
PROGRESS CHECK 6D
In Figure 6.2, draw a more elastic supply curve passing through equilibrium a.
How does this affect the difference between freight-inclusive and FOB pricing?
7. Taxes
Governments depend on tax revenues to support public services such as national

defense, administration of justice, and public health. Some taxes are levied on con-
sumers, others on businesses, and some are levied on both. For instance, the US
government imposes an air travel tax on airlines. By contrast, some Asian govern-
ments impose the air travel tax on the passenger. What difference does this make?
From the viewpoint of demand and supply, taxes are like a cost of intermedi-
ation. So, we can apply the analysis above to understand the effect of taxes on
market price and quantity. Specifically, suppose that the US government levies a
$10 tax on air travel between Chicago and Paris.
Buyer’s and Seller’s Price

We assume that the market is perfectly competitive. The demand comes from busi-
ness and leisure travelers, while American and foreign airlines provide the supply.
Since this market is subject to a tax, it is necessary to make one change to the
usual demand–supply analysis. We must distinguish the price that the buyer pays
(buyer’s price) from the price that the seller receives (seller’s price). The seller’s
price is the buyer’s price minus the amount of the tax.
We can draw the demand and supply curves as in Figure 6.3. Suppose that,
initially, there is no tax on airline tickets and the equilibrium is at point b, with a
price of $800. Since there is no tax, $800 is the buyer’s price as well as the seller’s
price. At the price of $800, airlines sell 920,000 tickets a year.
Now the federal government requires airlines to pay a tax of $10 on each ticket.
Referring to Figure 6.3, we can represent this graphically in one of three equiva-
lent ways: (a) shift the supply curve vertically up by $10, showing the market from
buyers’ point of view; (b) shift the demand curve vertically down by $10, showing
the market from airlines’ point of view; or (c) shift neither demand nor supply, but
show the $10 tax as a wedge between the demand and supply curves.
As a result of the tax, there will be a new equilibrium at point e in each diagram,
with a price of $804 and quantity of 900,000 tickets a year. The price is higher and
the quantity of travel is smaller.
(a) Buyer’s point of view (b) Seller’s point of view
supply
Price ($ per ticket)
e $10 supply
804 804
b e
800 b
800
794 h 794
$10 h
demand
demand
0 900 920 900 920

Quantity (thousand tickets a year) Quantity (thousand tickets a year)
(c) Wedge between buyer’s price and seller’s price

supply
804 e
b
800
794
h
demand
900 920
Quantity (thousand tickets a year)
FIGURE 6.3 Air travel tax.

Notes: The tax of $10 per ticket (a) raises the marginal cost of supplying air travel, (b) reduces the
willingness to pay by $10, or (c) drives a $10 wedge between buyers’ willingness to pay and the
marginal cost of air travel. In all these cases the buyers end up paying $804, the sellers keep $794,
and the quantity is reduced to 900,000 tickets a year.
Now that there is a tax, the buyer’s price differs from the seller’s price. The
buyer’s price is $804, while the seller’s price is the buyer’s price less the $10 tax, or
$794. In the new equilibrium, the seller’s price of $794 is lower than the original
seller’s price with no tax, which was $800.
Tax Incidence
So what is the difference between levying a tax on airlines and a tax on travelers?
Obviously, there may be administrative and perhaps psychological differences.
Since there are relatively fewer airlines than passengers, it may be less costly to
collect the tax from airlines. A tax on passengers may be more salient to travelers,
and may have a larger effect when they vote in elections.
Aside from administrative and psychological differences, however, we claim
that the effect of a tax will be the same whether it is collected from the buyers or
the sellers. Incidence depends only on the price elasticities of demand and supply.
The side of the market that is relatively less sensitive to price changes will bear the
relatively larger portion of the tax. Incidence is the same whether it is collected
from the buyers or the sellers.
PROGRESS CHECK 6E
In Figure 6.3, draw in a more inelastic demand curve. How does this affect the
incidence of tax on travelers relative to airlines?
MIGRANT WORKERS – EARN MORE, PAY MORE
The government of Singapore levies a tax on the employment of foreign workers.

Some have called for the government to convert the tax into a contribution to a
savings fund. The savings would benefit the foreign workers at the end of their
employment in Singapore when they return to their home country.
Let us use demand and supply to analyze the likely effects of the proposed
savings fund. Wages in Singapore are substantially higher than in the foreign
workers’ home countries, creating a huge excess supply. In market equilibrium,
foreign workers pay intermediaries to secure Singapore jobs. Among Bangladeshi
construction workers, the average fee was S$7,256.
With millions of people keen to work in Singapore, it is reasonable to
stipulate that the supply of foreign labor is fairly elastic. Hence, the conversion
of the tax into a savings fund would just add to the excess supply of foreign
workers. How would the market equilibrate? Through workers paying more in
placement fees to intermediaries. The workers themselves might not benefit
at all.
Source: Transient Workers Count Too, “Worse off for working?” Singapore, August 12, 2012.
KEY TAKEAWAYS
• Resource allocation is economically efficient if: (a) all users achieve the same
marginal benefit; (b) all suppliers operate at the same marginal cost; and (c)
marginal benefit equals marginal cost.
• In competitive markets, the invisible hand (price) leads to economic efficiency
by communicating information and providing incentives.
• For economic efficiency within an organization, set the transfer price equal to
the market price and allow outsourcing.
• The incidence of changes in demand or supply on buyers and sellers depends
only on the price elasticities of demand and supply.
• Whether the buyer or seller pays for the cost of intermediation does not affect
the incidence of the cost.
REVIEW QUESTIONS
1. The government of the Soviet Union subsidized bread. In 1987, the Soviet leader
Mikhail Gorbachev complained that children were kicking bread in football games.
Comment on the economic efficiency of the production and consumption of bread.
2. The external auditor of the local school system has found that some schools
pay 20% more for cleaning services than other schools. Which condition of
economic efficiency is being violated?
3. In a World War II camp, every prisoner of war received an identical parcel
containing cigarettes, chocolate, pen, and writing paper. Comment in terms of
economic efficiency.
4. In some countries, self-employed people can evade income tax while people
employed by the government and large organizations must pay income tax in
full. Comment on the economic efficiency of the labor market.
5. Consider a competitive finance market. Explain how the invisible hand ensures
that the allocation of investment funds is economically efficient.
6. To limit inflation, some governments impose controls on increases in the prices
of rice and other essential foods. Explain the impact on economic efficiency in
the market for rice.
7. Explain the meaning of outsourcing. What is the opposite of outsourcing?
8. An integrated oil producer both produces crude oil and refines the oil for sale
as gasoline, diesel, and other products in the retail market. In this context,
explain the concept of a transfer price.
9. A real estate group operates a chain of department stores using its own
buildings. Why should the group charge rent to the department store?
10. Explain the difference between CF pricing and FOB pricing.
11. Buyers of residential property must pay a percentage of the house price as a
commission to the agent. In this context, explain the difference between the
payment and incidence of the brokerage fee.
12. Consider two otherwise identical e-commerce retail markets. In one, vendors
offer free shipping, while in the other, vendors charge for shipping. What is the
difference for the price (including shipping charges) that buyers pay?
13. Referring to Figure 6.3(a), under what condition of the price elasticity of demand
would the incidence of the tax on consumers be the lowest?
14. Travelers can reduce costs for airlines by booking tickets through online
channels rather than travel agents. How would you assess whether airlines or
consumers benefit from the lower cost?
15. Free Duty offers tax-free sales of liquor and tobacco products at Hong Kong
International Airport. Do you expect any difference in the pre-tax prices of
liquor and cigarettes between the airport and the city?
1. Tickets to popular sporting events, such as the FIFA World Cup and the Super
Bowl, often sell out. Devoted fans must either spend long hours waiting in
line for a limited supply of tickets or pay a premium price to scalpers or touts.
Scalpers buy tickets to resell.
(a) When tickets sell out, which condition(s) for economic efficiency might
not be satisfied?
(b) Do scalpers improve economic efficiency?
(c) If the ticket agencies auctioned the tickets rather than selling them for
fixed prices, what would be the impact on scalpers?
2. Household incomes and the cost of living are higher in urban than rural areas.
A non-profit group of hospitals operates across urban and rural areas. The
group sets the same prices and pays the same salaries for doctors and other
professional staff at all hospitals.
(a) Management has noticed that there are long waiting lists for treatment
at its urban hospitals. Can you explain this problem?
(b) The company has had great difficulty in recruiting professional staff for
its urban hospitals. Can you explain this problem?
(c) What advice would you give to management?
3. LaGuardia Airport is one of three major airports that serve New York City. The
heavy demand for flights to and from the airport has caused systematic delays.
The entry of new carriers and expansion of service by existing carriers would
increase congestion. In 2000, the airport management decided to allocate a
limited number of new takeoff and landing slots among the various airlines by
lottery. Incumbent carriers were allowed to retain existing slots.
(a) With a limited number of takeoff and landing slots allocated by lottery,
which condition for economic efficiency might be violated?
(b) Comment on the following measures in terms of economic efficiency:
(i) giving the incumbent carriers ownership over their slots and allowing
them to lease or sell their slots; (ii) auctioning the right to new slots.
4. Jupiter Mines produces silver from several mines. Until recently, corporate
headquarters set production targets for each mine based on average production
costs. Then management consultants recommended a new policy: each mine
must aim to maximize profits given the prevailing price of silver.
(a) Under Jupiter’s old production policy, which condition(s) for economic
efficiency might not be satisfied?
(b) Does the new production policy improve economic efficiency?

(c) Explain the role of price under the new policy.
5. China Petroleum & Chemical Corporation (Sinopec) is an integrated producer
of oil and chemicals. Sinopec’s refining division sells most of its production
to the marketing and distribution division. The company’s policy is to set
inter-segment transfer prices according to the market price or cost plus an
appropriate margin. In early 2014, Sinopec announced that it would partially
privatize the marketing and distribution division.
(a) How does the transfer price of refined products affect the profits of the
marketing and distribution division?
(b) China is a large country with refineries sourcing crude oil from diverse
sources, including foreign countries. Would the market prices of
refined products be the same in all places?
(c) Suppose that Sinopec allows the marketing and distribution division
to outsource its supply of refined products. How would that affect the
refining division’s sales to the marketing and distribution division?
6. Some marketing consultants argue that cents-off coupons are more effective in
reducing retail prices than cuts in wholesale prices. They believe that retailers would
absorb cuts in wholesale prices instead of passing the price cuts to consumers.
By contrast, they believe that consumers would benefit fully from coupons.
(a) Consider the retail market for a brand of shampoo. Using relevant
demand and supply curves, illustrate the market equilibrium at a price
of $4 per bottle and a quantity of 500 million bottles a year. (Hint: You
are free to assume any data necessary to draw the curves.)
(b) Suppose that the manufacturer cuts the wholesale price by 25 cents.
Draw a new retail supply curve, shifted down by 25 cents. Identify the
new equilibrium quantity and price.
(c) Suppose, instead, that the manufacturer distributes 25-cent coupons
and that all consumers use these coupons. Draw a new retail demand
curve, shifted down by 25 cents. Identify the new equilibrium quantity
and price.
(d) Compare the equilibria under (b) and (c). Do you agree with the marketing
consultants?
7. Online retailers differ in charges for shipping. In November 2014, R.T. Edwards
offered a Samsung 65-inch high definition LED TV for A$4,238 with free
shipping, while Exeltek offered the same TV for A$4,095 with shipping charge
of A$145 to Perth, Western Australia.
(a) Using relevant demand and supply curves, explain whether it matters
for consumers if the retailer offers free shipping or charges for
shipping. (Hint: You are free to assume any data necessary to draw
the curves.)
(b) If consumers view R.T. Edwards and Exeltek as equivalent (in terms of
quality of service), how should their prices for the same TV compare?
Are the prices consistent with your answer in (a)?
(c) If consumers are biased in decision-making by anchoring, how would
that affect your answer in (a)?
8. The Internet has drastically reduced the cost of intermediary services of travel
agencies, real estate brokers, and investment advisors. Consider the market
for air travel. Suppose that, with booking through travel agencies, the market
equilibrium price is $500 per ticket, including a $25 cost of intermediation. The
quantity bought is 2 million tickets a year. With Internet bookings, however, the
intermediation cost falls to $2 per ticket.
(a) Using suitable demand and supply curves, illustrate the original equilibrium
with booking through travel agencies. Represent the intermediation cost
by shifting the supply curve.
(b) Illustrate the new equilibrium with online booking.
(c) Under what conditions of demand and supply would consumers benefit
the most from online bookings?
9. Mainland Chinese visitors are among the most free-spending shoppers at
Hermes, Louis Vuitton, Prada, and other luxury stores in Hong Kong. The
government of China levies value-added tax (VAT) at 17% on the retail sales
of imported goods within China. Shenzhen, on the border with Hong Kong, is
China’s highest-income city.
(a) Suppose that Chinese tourists can avoid VAT when bringing luxury
items back to China. How would this affect the demand among Chinese
for luxury goods: (i) in Hong Kong, and (ii) within China?
(b) If the demand among Chinese for luxury goods within China is very
elastic (owing to ready access to shops in Hong Kong), what would be
the relation of prices in Shenzhen to those in Hong Kong?
(c) To what extent would your analysis in (a) and (b) also apply to consumer
goods such as shampoo and baby diapers?

Consider some input that your organization uses to produce the goods and
services that it sells.
(a) Are all users of the input getting the same marginal benefit from the input?
(b) Compare the marginal benefit from use of the input with the marginal cost
of producing the input.
(c) How could your organization increase the efficiency of the use of the
input?
Notes
1 This discussion is based in part on “Congestion management rule for John F. Kennedy Inter-
national Airport and Newark Liberty International Airport: Final regulatory evaluation,” US
Federal Aviation Administration, Office of Aviation Policy and Plans, Aircraft Regulatory
Analysis Branch, October 10, 2008.
2 Adam Smith, The Wealth of Nations, 2nd edition, London: W. Strahan and T. Cadell, 1778, Vol. II,
Book IV, p. 35 (first published in 1776).
PA R T
II
Market Power
C H A P T E R
7
Costs
LEARNING OBJECTIVES
• Understand opportunity costs and apply in business decisions.
• Apply the concept of opportunity cost to the cost of capital and
transfer pricing.
• Appreciate that sunk costs should be ignored in business decisions.
• Understand economies of scale and apply in business decisions.
• Understand economies of scope and apply in business decisions.
• Understand the experience curve and apply in business decisions.
• Recognize and avoid behavioral biases in cost decisions.
1. Introduction
The Canadian aerospace manufacturer, Bombardier, is well-established in the pro-

duction of regional jets. It long aspired to expand into large jets and announced
the development of the CSeries in 2004. However, it did not actually commence
development until 2008, upon securing a letter of interest for 60 planes from
Deutsche Lufthansa. The new CSeries is scheduled to enter service in 2015. It will
reduce fuel consumption by 20% through the use of advanced materials and a
more fuel-efficient engine, the Pratt & Whitney PW1000G.1
COMAC of China and Irkut of Russia have begun development of large jets.
Acutely conscious of the competitive threat, Airbus and Boeing are under pres-
sure to respond. In December 2010, Airbus launched a new version of one of its
large jets, the A320neo, with a “new engine option” offering better fuel economy.
140 7 Costs
In March 2011, Airbus announced that it would raise production of the A320
family from 34 to 36 units per month.
In June 2011, at the Paris Air Show, Boeing had still not decided its strategy –
whether to replace the Boeing 737 with a completely new model, or, like Airbus,
offer a new version. Earlier, Boeing announced an increase in 737 production from
31.5 units to 42 units per month.
Why was Bombardier so cautious about commencing development of the
CSeries? For Airbus, what are the advantages of meeting the competition with
the A320neo rather than a completely new plane? How would Airbus’s increase in
A320 production affect its unit costs?
To address these questions, we introduce a framework for understanding costs.
Within a single period of production, costs are either sunk (have been committed
and cannot be avoided) or avoidable (see Figure 7.1). Practically, the most readily
available information on profit is the accounting measure. The concept of economic
profit is related to but differs from accounting profit. Specifically, economic profit dif-
fers from accounting profit by excluding opportunity cost and including sunk cost:
Economic profit = Accounting profit − Opportunity cost + Sunk cost. (7.1)
Strategically, looking forward, businesses must take care

Economic profit:
before incurring sunk costs. Bombardier estimates the cost
Accounting profit less
opportunity cost plus of developing the CSeries to be $2.5 billion, or a quarter of
sunk cost. the company’s market capitalization. The company is pru-
dent to be cautious about sinking so large an amount.
Among the avoidable costs, some are not reported in
accounting statements but are relevant to management decisions. These are oppor-
tunity costs, which must be forgone to continue with the current course of action.
The avoidable costs can be analyzed in two ways. One way is fixed and variable
costs. Fixed costs do not vary with the scale of production, while variable costs
do. Which costs are fixed and which are variable depends on the technology of
Costs
Sunk Avoidable
(including opportunity costs)
Fixed Joint
Variable Not joint
FIGURE 7.1 Costs in a single period of production.

Notes: The division of costs into sunk and avoidable depends on past commitments and the current
planning horizon. The division of avoidable costs into fixed and variable and into joint and not joint
depends on the production technology.
Costs 141
production. Fixed costs are an essential reason for economies of scale. If fixed
costs are substantial in the manufacturing of Airbus jets, then Airbus can reduce
its unit costs by increasing the rate of production.
Joint costs do not vary with the number of products, and are an essential reason
for economies of scope. They explain, in part, why Airbus chose to launch a new
version of an existing plane, the Airbus 320neo, instead of a completely new plane.
By doing so, it could exploit economies of scope across the existing and new planes.
Over time, the cost of production may fall with cumulative production according
to an experience curve. Indeed, the cost of airplane manufacturing falls substan-
tially with cumulative production. Given the cost-reducing effect of experience, it
would not be economic to produce a small number of planes. This also explains
why Bombardier waited until securing substantial interest from Lufthansa before
developing the CSeries.
For decisions in any context – for profit, non-profit, and government – managers
need to understand costs from more than a pure accounting perspective. They can
apply the framework and analyses of this chapter to make more effective deci-
sions in investment, performance evaluation, outsourcing, and pricing.
2. Opportunity Cost
Analyses of costs usually begin with accounting statements. These, however, do

not always provide the appropriate information for effective business decisions. It
is often necessary to look beyond conventional accounting statements.
To illustrate, suppose that Luna Biotech has accumulated cash of $10 million.
It is considering a $10 million research and development project to develop a new
drug to treat a rare disease. The new drug is expected to generate a profit contri-
bution of $20 million. (Profit contribution is revenues less variable cost, excluding
R&D expenditures.) Luna Biotech has not yet commenced R&D.
Meanwhile, an independent scientist has already developed a similar drug to
treat the same disease. The scientist has offered to sell her invention to Luna Bio-
tech for $2 million. The scientist’s drug would be just as good and also yield a
profit contribution of $20 million. How should Luna choose between commenc-
ing R&D on its own drug and buying the scientist’s invention?
Alternative Courses of Action

Table 7.1 presents a projected income statement for the R&D project. The profit
contribution is $20 million for an R&D expense of $10 million. So, the projected
Table 7.1 Conventional income statement ($ million)
Profit contribution 20
R&D expense 10
Profit 10
142 7 Costs
Table 7.2 Income statement showing alternatives ($ million)
Commence Cancel R&D

R&D and buy drug
Profit contribution 20 20
R&D expense 10 0
External purchases 0 2
Profit 10 18
Table 7.3 Income statement showing opportunity cost ($ million)
Commence R&D
R&D expense 10
Opportunity cost 18
Profit −8
profit is $10 million for an investment of $10 million. The return on investment,
10/10 = 100%, looks very good.
However, the income statement overlooks a significant cost of continuing the
R&D project. A proper evaluation of performance should consider the alterna-
tive uses of Luna’s investment funds. Specifically, the scientist is willing to sell her
drug, which would be just as profitable, to Luna Biotech for $2 million.
The income statement, as conventionally presented, does not present the
revenues and costs from alternative courses of action. Table 7.2 presents
an expanded income statement that explicitly shows the alternatives. This
makes it very clear that Luna should cancel its own R&D and buy the scien-
tist’s drug.
Identifying Opportunity Cost

By commencing its own R&D, Luna forgoes the opportunity to buy the scientist’s
drug, which would yield a profit of $18 million. The opportunity
Opportunity cost: What cost of the current course of action is what must be forgone
must be forgone from the from the best alternative course of action. In Luna’s case, there
best alternative course of
is only one alternative to its own R&D and the opportunity
action.
cost is $18 million.
We can apply the concept of opportunity cost to present
the revenues and costs of continuing the R&D project in another way. Following
equation (7.1), this method includes opportunity costs among the costs of the
business. Table 7.3 presents a single income statement, showing both the R&D expense
and the opportunity cost. The (economic) profit from commencing R&D is −$8
million, which is a loss. This approach leads to the same decision as in Table 7.2,
which explicitly shows the two alternative courses of action.
Costs 143
So, there are two ways to uncover relevant costs: explicitly consider the alter-
native courses of action or use the concept of opportunity cost. When applied
correctly, both approaches lead to the correct decision.
In Luna’s case, there is one alternative to the existing course of action. Where
there is more than one alternative, the explicit approach still works well. The
opportunity cost approach, however, becomes more complicated: first identify the
best of the alternatives, and then charge the profit contribution from that alterna-
tive as the opportunity cost of the current course of action.
Conventional methods of cost accounting focus on the costs of the course of
action that management has adopted. They do not consider the revenues and costs
of alternative courses of action; hence, they ignore costs that are relevant but do
not involve cash outlays. One reason for these omissions is that alternative courses
of action and opportunity costs change with the circumstances and hence are
more difficult to measure and verify. Conventional methods of cost accounting
focus on easily verifiable costs, and so overlook opportunity costs.
PROGRESS CHECK 7A
If the scientist demands $12 million for her drug, what is Luna’s opportunity
cost of commencing R&D and what is the right decision?
FREE LUNCH: DAVID HOCKNEY VERSUS WARREN BUFFET
The following story is apocryphal. The billionaire investor, Warren Buffet, and the
famous artist, David Hockney, had lunch at the Smith & Wollensky restaurant.
At the end of a fine meal, David Hockney reached for the bill: “Let me pay. I’ll
write a personal check, draw a few squiggles, and sign it. The manager won’t
ever cash it. She will display it as a work of art. And we’ll have a free lunch.”
Warren Buffet would not agree. “No, allow me to write a personal check.
Remember, the name Warren Buffet is good as gold. The manager won’t cash
my check. She can use it just like money and our lunch will be free.”
Who was right? The correct answer is that it was neither. By drawing a few
squiggles, David Hockney was adding to the world’s stock of his works. The
increase in the supply would reduce the price that he could get for future
works. David Hockney would not get a free lunch – he was bartering a picture
for a lunch.
Warren Buffet was also wrong. Each check that he wrote would add to his
stock of debts and, ultimately, reduced his creditworthiness. He could not
issue an unlimited number of checks. So, by creating a check, Warren Buffet
was exchanging a lunch for a reduction in his creditworthiness. In believing
that they could get a free lunch, both David Hockney and Warren Buffet
overlooked opportunity cost.
144 7 Costs
Opportunity Cost of Capital

Conventional accounting methods require the expensing of interest payments but
do not require the expensing of expected dividends. Consequently, a business that
is partly financed by debt will appear to be “less profitable” than an otherwise
identical business that is completely financed by equity. Indeed, many loss-making
businesses have returned to profit by persuading their creditors to convert their
loans to shares.
However, from a managerial perspective, equity capital is not costless; it has
an opportunity cost. The shareholders of a business would like management to
earn a rate of return on equity that at least matches the return from other invest-
ments with the same risk profile. Businesses that evaluate performance in terms
of accounting profit will tend to be biased in favor of capital-intensive activities.
The appropriate mix between debt and equity as sources of investment funds
is a deep and complicated issue of corporate finance. It is beyond the scope of a
managerial economics book. Our point here is that a complete measure of business
performance should take account of the opportunity cost of equity capital.
One way to evaluate business performance without distortion by debt–equity
capital structure is EBITDA, which is earnings before interest, taxes, depreciation,
and amortization. As EBITDA does not deduct interest, it is not affected by the
debt–equity capital structure. Hence, evaluation will be neutral with respect to
capital intensity.
3. Transfer Pricing
As introduced in Chapter 6, a transfer price is the price for the sale of an item
within an organization. Suppose that Luna Biotech com-
Transfer price: The price prises two divisions – manufacturing and marketing. The
for the sale of an item
manufacturing division produces drugs by cell and tissue
within an organization.
culture for the marketing division which sells the drugs to
hospitals and clinics. How should it price the cultures?
Figure 7.2 shows the manufacturing division’s marginal cost of culturing a
batch of drugs. It also shows the marketing division’s marginal benefit from the
culture, which reflects the additional profit generated by one more batch of drugs.
Luna’s overall profit would be maximized at a quantity of 16,000 batches per
month, where the marketing division’s marginal benefit equals the manufacturing
division’s marginal cost.
Suppose that Luna sets the transfer price at $2,000 per batch, and allows the
marketing division to buy as many batches as it wants at that price. Comparing
its marginal benefit with the transfer price, the marketing division will buy 16,000
batches per month, which is exactly the quantity that maximizes Luna’s overall
profit.
Generally, the organization will maximize its entire profit by setting the transfer
price of an internally produced input equal to its marginal cost. Equivalently, the
Costs 145
Cost/benefit ($ per thousand batches)

marginal cost to
manufacturing division
marginal benefit to
marketing division
0 16 20
Production rate (thousand batches a month)
FIGURE 7.2 Transfer price.

Notes: Luna’s manufacturing division has a maximum capacity of 20,000 batches a month. Luna
maximizes its overall profit at a production rate of 16,000 batches a month, where the manufacturing
division’s marginal cost equals the marketing division’s marginal benefit.
profit-maximizing transfer price is the marginal cost of the input. Recall that
the marginal cost is the change in total cost due to the pro-
duction of an additional unit. There are two important spe- Profit-maximizing
transfer price: The
cial cases, which we discuss below.
marginal cost of the input.
Perfectly Competitive Market

One special case is where there is a perfectly competitive market for the input. We
discussed this case in Chapter 6 when considering decentralized management. Rather
than measure the marginal cost, it is simpler to set the transfer price equal to the
market price. In a perfectly competitive market, a profit-maximizing business would
produce the input at a rate where its marginal cost equals the market price. Hence,
the transfer price (set at the market price) will also be the marginal cost.
Full Capacity
The other special case is where the (upstream) division that supplies the input is
operating at full capacity. Then the marginal cost of the input is not well defined.
For instance, referring to Figure 7.2, if Luna’s manufacturing division is produc-
ing at a rate of 20,000 batches a month, the marginal cost curve is vertical and
hence the marginal cost of a chip is not defined.
In this case, the transfer price should be set equal to the opportunity cost of the
input, which is the marginal benefit that the input provides to the current user. To
understand why, suppose that Luna reallocates one batch away from the market-
ing division to an external customer. Since the manufacturing division’s total pro-
duction remains the same, there will be no effect on its production costs. However,
146 7 Costs
the reallocation will result in an opportunity cost – the reduction in profit for the
marketing division.
To ensure that the alternative use of the batch raises overall profit, the transfer
price should be set equal to the marketing division’s marginal benefit from that
batch. The alternative user will only buy the batch if the benefit exceeds the trans-
fer price, and hence only if its benefit exceeds the marketing division’s benefit.
Accordingly, this rule will maximize the profit of the entire organization.
PROGRESS CHECK 7B
How should an organization set the transfer price for an internally produced
input when: (i) the market for the input is competitive; and (ii) production of
the input is at full capacity?
4. Sunk Costs
Since conventional accounting statements do not present alternative courses of

action, they may fail to reveal costs which are important to effective business deci-
sions. Another shortcoming of conventional accounting statements is that they
present some costs which are not relevant to effective business decisions and
should be ignored.
As introduced in Chapter 4, a sunk cost is a cost that
Sunk cost: Cost that has been committed and so cannot be avoided. Since sunk
has been committed and
costs cannot be avoided, they are not relevant to business
cannot be avoided.
decisions and managers should ignore them.
To appreciate the concept of sunk costs and why they
should be ignored, let us slightly modify the example of Luna Biotech. Suppose
that Luna had agreed to pay $10 million to an external contractor for the R&D to
develop the new drug. The new drug is expected to generate a profit contribution
of $20 million.
Meanwhile, the independent scientist has sold her drug to a competing man-
ufacturer, which has filed a patent, and will soon commence sales. Faced with
this competition, Luna has reduced the expected profit contribution from its own
drug from $20 million to $8 million. Should Luna cancel the R&D project?
Alternative Courses of Action

Table 7.4 presents a projected income statement for the R&D project. The profit
contribution is $8 million for an R&D expense of $10 million. So the projected
profit is −$2 million, that is, a $2 million loss for an investment of $10 million.
Apparently, Luna should cancel the R&D.
However, the income statement overlooks the fact that Luna would incur a
substantial part of the R&D expense even if it canceled the project. A proper
Costs 147
Table 7.4 Conventional income statement ($ million)
R&D expense 10
Profit −2
Table 7.5 Income statement showing alternatives ($ million)
Continue R&D Cancel R&D
Profit contribution 8 0
R&D expense 10 6
Profit −2 −6
Table 7.6 Income statement omitting sunk costs ($ million)
Continue R&D
R&D expense 4
Profit 4
evaluation of the decision to continue the R&D project should consider the profit
from canceling the project.
Table 7.5 lays out the revenues and costs associated with the alternative courses
of action. If Luna continued with the R&D, its profit contribution would be
$8 million. Subtracting the R&D expense of $10 million, Luna’s profit would
be −$2 million, that is, a loss.
By contrast, if Luna canceled the R&D, its profit contribution would be zero.
Suppose that, if Luna cancels the R&D, it must still pay the R&D contractor
$6 million. Hence, if Luna decided to cancel the R&D, its profit would be −$6 million.
Evidently, Luna should continue with the R&D. Continuation yields a loss which
is smaller than the loss from cancelation.
Identifying Sunk Costs

By canceling the R&D, Luna does not save the entire R&D budget of $10 million.
Part has already been spent or committed, and is now sunk, and so should be
ignored in any business decision. Another way in which Luna can correctly decide
whether or not to continue the launch applies equation (7.1) – this uses a single
income statement that omits sunk costs and includes only avoidable costs.
Table 7.6 presents this information, showing only avoidable costs rather than
cash outlays. The profit contribution is $8 million. The avoidable cost of the
R&D project is just $4 million. If Luna continues with the R&D, it will earn an
(economic) profit of $4 million. Accordingly, the correct decision is to continue
with the R&D.
148 7 Costs
We have shown two ways of dealing with sunk costs: explicitly consider the
alternative courses of action or remove all sunk costs from the income statement.
When applied correctly, both approaches lead to the same business decision.
In Luna’s case, there is one alternative to the existing course of action. Where
there is more than one alternative, the explicit approach still works well. The
sunk cost approach, however, becomes more complicated. Which costs are sunk
depends on the alternative at hand. Accordingly, it is easier to consider the alter-
native courses of action explicitly.
Conventional methods of cost accounting focus on the cash outlays associated
with the course of action that management has adopted. These methods report
all costs that involve cash outlays, even sunk costs. To make effective business
decisions, managers must look beyond conventional accounting statements and
ignore sunk costs.
Commitments and Planning Horizon

The extent of sunk costs depends on two factors: past commitments and the
planning horizon. Suppose that Luna Biotech’s contract with the external R&D
provider specifies six months’ notice of termination. Then, from the current
standpoint, the R&D expenditure is sunk for a six-month planning horizon but
not beyond.
If Luna’s contract were different, the sunk cost would also be different. Sup-
pose that the contract specifies only three months’ notice of termination. Then,
from the current standpoint, the R&D expenditure is sunk for only the next three
months. For planning beyond the third month, the expense is avoidable.
This example also illustrates how the extent to which a cost is sunk depends on
the planning horizon. Generally, the longer the planning horizon, the more time
there will be for past commitments to unwind and hence the greater will be man-
agement’s freedom of action.
Chapter 4 distinguishes between short-run and long-run planning horizons.
The short run is a time horizon in which at least one input cannot be adjusted. By
contrast, the long run is a time horizon long enough that all inputs can be freely
adjusted. Consequently, in a short-run planning horizon there will be some sunk
costs, while in a long-run horizon there will be no sunk costs.
Strategic Implications
Generally, managers should ignore sunk costs and consider only avoidable costs.
Sunk costs, once incurred, are not relevant for investment, pricing, or any other busi-
ness decision. Managers who consider sunk costs may stumble into serious mistakes.
If a substantial portion of costs are sunk, then the avoidable costs are relatively
low. For the participation decision, this implies that the break-even revenue will be
relatively low. The business should continue in operation even with relatively low
revenues – provided that the revenues cover the avoidable costs.
Costs 149
For the extent decision (how much to produce, at what scale to operate), sub-
stantial sunk costs may imply that marginal costs will be relatively low. To that
extent, the business should price relatively low and aim to serve larger demand.
From a prospective viewpoint, managers should be very careful before commit-
ting to costs that will become sunk, since such commitments cannot be reversed.
In Chapter 10, we discuss how businesses can exploit investments in sunk costs as
a way to strategically influence the behavior of competitors.
PROGRESS CHECK 7C
Suppose that, if Luna canceled the R&D contract, the R&D expenditure
would be $1 million. What is Luna’s cost of continuing R&D and what is the
right decision?
DAMPIER–BUNBURY PIPELINE: YOUR LOSS, MY PROFIT
The 1,530-kilometer Dampier–Bunbury pipeline transports natural gas from

the vast North Western Shelf gas fields into the state of Western Australia.
The pipeline is the state’s largest. In 1998, the state sold the pipeline to Epic
Energy, a joint venture of US companies El Paso and Dominion Resources,
for A$2.4 billion. Epic financed the purchase with A$1.85 billion of loans from
a consortium of 28 banks.
The investment, however, did not perform to expectation, and, in October
2003, Epic decided to sell the pipeline. Epic blamed the gas regulator for
imposing unacceptably low tariffs. However, Australian Competition and
Consumer Commissioner Edward Willett remarked that Epic had overpaid for
the pipeline by A$1 billion.
In April 2004, with Epic’s pipeline company failing to meet its debt obligations,
the consortium of banks forced the company into receivership. In August, the
receivers named a consortium led by Diversified Utility and Energy Trusts
(DUET) as their preferred buyer at a price of A$1.86 billion. The consortium
also included Western Australia’s largest gas retailer, Alinta, and the pipeline’s
largest customer, aluminum manufacturer, Alcoa, each with a 20% share. DUET
Chief Executive Peter Barry remarked that “with the appropriate acquisition
structure” the pipeline would yield very attractive returns.
How could the pipeline be unprofitable for Epic Energy while being “attractive”
to DUET? One possible reason is that Epic (and its bankers) would not ignore the
sunk cost of Epic’s investment. If, as Willett said, Epic had overpaid by A$1 billion,
then the true market value of the pipeline was A$1.86 billion with A$1.85 billion of
debt outstanding, and the net equity of the pipeline was just A$10 million.
Sources: “Receivers looming for Epic Energy,” Sydney Morning Herald, April 12, 2004; “Joint
venture to buy Australian pipeline,” International Herald Tribune, September 1, 2004.
150 7 Costs
5. Economies of Scale
For effective business decisions, managers must identify all relevant costs and
appreciate how those costs vary with business decisions – specifically, scale, scope,
and experience. A fundamental issue for any organization is whether to operate
on a small scale or large scale.
Large-scale production means mass marketing and relatively low pricing; by
contrast, small-scale production is associated with niche marketing and relatively
high pricing. So, how do costs depend on the scale or rate of production? (We
shall treat the scale and rate of production as synonymous.)
To address this question, recall the distinction between fixed and variable costs
introduced in Chapter 4. The fixed cost is the cost of inputs that do not change with
the production rate. So the fixed cost supports the production of multiple units of
output. The variable cost is the cost of inputs that change with the production rate.
The distinction between fixed and variable costs applies in the short and long run.
Fixed and Variable Costs in the Long Run

To illustrate the distinction between long-run fixed and variable costs, consider
the production of a newspaper, the Daily Globe. The production process begins
when the printing department receives a photographic negative containing the
text of the forthcoming edition. The negative is “burned” on to an aluminum
plate, which is then mounted on electric-powered printing presses. The presses can
be fed a continuous flow of newsprint and ink to produce the newspaper.
Table 7.7 reports the daily expenses for production rates up to 90,000 copies
a day, in the four categories of labor, printing press, ink and paper, and electric
power. Table 7.8 assigns each of the costs of newspaper production into the two
categories of fixed and variable costs.
Table 7.7 Daily expenses for newspaper production
Daily production Labor Printing Ink and Electric Total

(thousands) ($) press ($) paper ($) power ($) ($)
0 5,000 1,000 0 200 6,200

10 5,000 1,500 1,200 300 8,000
20 5,000 2,000 2,400 400 9,800
30 5,000 2,500 3,600 500 11,600
40 5,000 3,000 4,800 600 13,400
50 5,000 3,500 6,000 700 15,200
60 5,000 4,000 7,200 800 17,000
70 5,000 4,500 8,400 900 18,800
80 5,000 5,000 9,600 1,000 20,600
90 5,000 5,500 10,800 1,100 22,400
Costs 151
Table 7.8 Analysis of fixed and variable costs
Daily Fixed Variable Total Marginal Average Average Average

production cost ($) cost ($) cost ($) cost ($) fixed variable cost ($)
(thousands) cost ($) cost ($)
0 6,200 0 6,200
10 6,200 1,800 8,000 0.18 0.62 0.18 0.80
20 6,200 3,600 9,800 0.18 0.31 0.18 0.49
30 6,200 5,400 11,600 0.18 0.21 0.18 0.39
40 6,200 7,200 13,400 0.18 0.16 0.18 0.34
50 6,200 9,000 15,200 0.18 0.12 0.18 0.30
60 6,200 10,800 17,000 0.18 0.10 0.18 0.28
70 6,200 12,600 18,800 0.18 0.09 0.18 0.27
80 6,200 14,400 20,600 0.18 0.08 0.18 0.26
90 6,200 16,200 22,400 0.18 0.07 0.18 0.25
Production involves a fixed cost of $6,200. Indeed, a substantial fixed cost is

a distinctive feature of the newspaper industry. The industry has given the name
first copy cost to the fixed cost. It is the cost of producing just one copy a day.
By contrast, as the print run increases from 0 to 90,000 copies a day, the variable
cost rises from nothing to $16,200. By distinguishing between fixed and variable
costs, the management of a business can understand which cost elements will be
affected by changes in the scale of production.
Marginal and Average Costs

Applying the analysis of fixed and variable costs, we can see how costs depend
on the scale of production. Chapter 4 introduced the concepts of marginal and
average costs. The marginal cost is the change in total cost due to the production
of an additional unit. The average (or unit) cost is the total cost divided by the
production rate or scale.
Let us study the marginal and average costs of production of the Daily Globe.
Table 7.8 shows that, as the print run increases from 0 to 90,000 copies a day, the
total cost of production increases from $6,200 to $22,400.
The marginal cost of the first 10,000 copies is $8,000 − $6,200 = $1,800, or
$1,800/10,000 = 18 cents per copy. The marginal cost is constant at 18 cents at
all scales of production. Recall from Chapter 4 that the marginal cost equals the
rate of change of the variable cost. For the Daily Globe, the average variable cost
remains constant at 18 cents per copy. Hence, the marginal cost is also constant
at 18 cents per copy.
Dividing total cost by the scale of production, we can obtain the average cost.
The average cost drops from 80 cents at a scale of 10,000 copies a day to 25 cents
at 90,000 copies a day. To understand why the average cost decreases with the
scale of production, recall that the average cost is the average fixed cost plus the
152 7 Costs
Marginal/average cost ($ per unit) 0.9
0.8
0.7
0.6
0.5
0.4
0.3 average cost
0.2 marginal cost

0.1
0 10 20 30 40 50 60 70 80 90
Production rate (thousands a day)
FIGURE 7.3 Economies of scale.

Notes: The marginal and average variable costs are identical and do not change with the scale of
production. The average cost decreases with the scale of production.
average variable cost. The average fixed cost is the fixed cost divided by the pro-
duction scale. With a larger scale of production, the fixed cost will be spread over
more units of production and the average fixed cost will be lower.
The average variable cost is constant at 18 cents per copy. Therefore, the average
cost declines as the scale of production increases. Figure 7.3 graphs the marginal,
average variable, and average costs against the scale of production. The marginal
and average variable cost curves are identical and flat. The average cost curve
slopes downward.
A business for which the average cost decreases with the scale of production is
said to exhibit economies of scale or increasing returns
Economies of scale to scale. With economies of scale, the marginal cost will be
(increasing returns to
scale): Average cost
lower than the average cost. Since the marginal unit of pro-
decreases with the scale duction costs less than the average, any increase in produc-
of production. tion will reduce the average. Therefore, the average cost curve
slopes downward.
Intuitive Factors
The basic reason for economies of scale is either fixed inputs, that is, inputs that can
support any scale of production, or inputs that increase less than proportionately
with the scale of production. Consider fixed inputs. At a larger scale, the cost of the
fixed inputs will be spread over more units of production, so that the average fixed
cost will be lower. If the average variable cost is constant or does not increase very
much with the scale of production, then the average cost will fall with the scale.
Costs 153
Any business with a strong element of composition, design, or invention has

substantial fixed inputs. For instance, the cost of developing a new pharmaceuti-
cal is fixed (with respect to the quantity produced). Regardless of the production
rate, the development cost will remain the same. Similarly, the cost of preparing
the computer code for a software package is fixed. It is the same whether the
publishers produces 1 million copies or only one. Accordingly, there are strong
economies of scale in these industries. Indeed, for pharmaceuticals, software, and
other knowledge-intensive industries, the marginal production cost is tiny com-
pared with the average cost.
The other basic reason for economies of scale is inputs that increase less than
proportionately with the scale of production. This explains scale economies in
transportation. To double the capacity of a tanker or pipeline requires less than
double the material (to be precise, the proportion is 2 ).
PROGRESS CHECK 7D
Using the data in Table 7.8, draw the average fixed cost in Figure 7.3.
Diseconomies of Scale
A business where the average cost increases with the scale of production is said to
exhibit diseconomies of scale or decreasing returns to
scale. A business will have diseconomies of scale if the fixed Diseconomies of scale
cost is not substantial and the variable cost rises more than (decreasing returns to
scale): Average cost
proportionately with the scale of production.
increases with scale of
To illustrate diseconomies of scale, consider a hairdress- production.
ing salon. The salon does not involve a significant fixed cost.
The main variable cost is labor. To the extent that additional
workers are less productive, the cost of labor rises more than proportionately with
the scale of production.
The average cost is the average fixed cost plus the average variable cost. For the
salon, the average cost initially decreases with the scale because of the decreasing
average fixed cost. Since the variable cost rises more than proportionately with the
scale of production, the average variable cost is increasing.
Hence, there is a scale at which the decreasing average fixed cost is outweighed
by the increasing average variable cost. Then the average cost reaches a min-
imum and rises with further increases in the scale. The average cost curve is
U-shaped.
The relation between average cost and the scale of production influences the
structure of the industry. If there are economies of scale, a business operating
154 7 Costs
on a relatively large scale will achieve a lower average cost than smaller-scale
competitors. Large-scale production means mass marketing and relatively low
pricing.
An industry where individual producers have economies of scale tends to be
concentrated, with a few producers serving the entire market. There are strong
economies of scale in providing broadband service. The essential input is the net-
work. With the network in place, the provider can serve additional customers at
relatively low marginal cost. Owing to the strong economies of scale, most com-
munities are served by a small number of broadband provid-
ers. In Chapter 8, we analyze the extreme case of a monopoly,
In an industry with where there is just one producer.
economies of scale, large By contrast, in an industry with diseconomies of scale, the
producers dominate
and the industry will be
management should aim at a relatively small scale. Small-
concentrated. scale production is associated with niche marketing and
relatively high pricing. Industries where individual produc-
ers have diseconomies of scale tend to be fragmented. The
extreme case is the model of perfect competition in Chapter 5, where there are
many producers, none of whom can influence the market demand.
Sunk and Fixed Costs

In popular as well as professional parlance, the term “fixed cost” is often used in
two different senses: a cost that cannot be avoided once incurred (properly called
sunk cost); and the cost of inputs that do not change with the production rate
(properly called fixed cost).
Referring to Figure 7.1, it is important to distinguish sunk from fixed costs
because the two concepts have very different implications for business decisions.
Managers should ignore sunk costs that have been incurred, as these cannot be
avoided. By contrast, the presence of fixed costs tends to give rise to economies of
scale, and so management should aim to operate on a large scale. (The confusion
between the two concepts arises because economics tends to assume that, in the
short run, all fixed costs are sunk.)
Some fixed costs become sunk once incurred. Consider, for instance, a manu-
facturer of sports shoes. The manufacturer must pay for the design of the shoes.
The design can support the production of any number of shoes. So, the cost of
design is a fixed cost. Moreover, once the manufacturer has committed to the
design, the cost cannot be avoided. Hence, once incurred, the design cost is a
sunk cost.
Sunk costs, however, are not fixed in the sense of supporting any scale of oper-
ations. Having designed the shoes, the manufacturer needs one set of molds (left
and right) to begin production of shoes. The cost of making the molds is sunk,
once incurred. If the demand for the shoes is sufficiently high, the manufacturer
might have to invest in a second production line. Then it will need a second set of
Costs 155
molds, which requires an additional investment. So, the cost of making molds is
not a fixed cost. Rather, it depends on the scale of production.
Not all fixed costs become sunk. Wireless telecommunications providers need
a government license to operate on the electromagnetic spectrum. Supposing that
the license fee is a lump sum, the fee is a fixed cost of providing service. However,
if the license is transferable, the fee need not be sunk. Only the part of the fee that
cannot be recovered from resale is sunk.
ECONOMIES OF SCALE IN TANKER CONSTRUCTION
In 1833, Marcus Samuel opened a shop in London to sell seashells. While

procuring shells in the Caspian Sea, Marcus Samuel’s son spotted a new
business opportunity – the export of kerosene from Russia to the Far East.
This business grew into Shell Transport and Trading.
Historically, oil had been transported in wooden barrels on cargo ships.
In 1892, Marcus Samuel Junior conceived the idea of building a ship in the
shape of a tank. This became the world’s first oil tanker.
An oil tanker is like a pipeline in the sense that its capacity increases with the
cross-sectional area, and hence the square of the radius of the cross section,
but the material and construction costs increase with the circumference, and
hence the radius of the cross section.
Figure 7.4 shows the price of new tankers per deadweight ton for four
standard tanker sizes. Clearly, larger tankers are cheaper, presumably because
they are less costly to build.
1000
900
Price per deadweight ton ($)
800
700
600
500
400
300
200
100
0
0 50 100 150 200 250 300 350
Size (thousand deadweight tons)
FIGURE 7.4 New tanker prices: January 2006.

Sources: Shell Oil Company, “Shell heritage,” www.shellenergy.com, September 7, 2004; Lloyd’s
Shipping Economist, February 2004; Platou Report 2005, www.platou.com.
156 7 Costs
AIRBUS: INCREASING PRODUCTION
In March 2011, Airbus announced that it would raise production of the A320
family from 34 to 36 units per month. Subsequently, Boeing announced a
progressive increase in production of the 737 from 31.5 units to 42 units per
month.
How would these changes affect the costs of production? The impact of
higher rates of production on average costs depends on the proportion of
fixed and variable costs and the impact on average variable costs. Both Airbus
and Boeing depend on external contractors for many essential components,
including avionics, composites, and titanium parts.
Opinions differ on the ability of suppliers to increase the rate of production,
and so the impact on the average variable costs of Airbus and Boeing. Robert
P. Barker, President of Parker Aerospace, stressed: “It’s very manageable for
us to take on these added quantities in our factories . . . . We can go from 35
actuators a month to 42 pretty easily over a period of time.”
By contrast, Donald Majcher, Vice President at the Ohio Aerospace Institute in
Cleveland, Ohio questioned: “In recent years you’ve seen a consolidation in the
supply base . . . . The question is, how much of an increase can it now support?”
Source: “Can suppliers keep up with aircraft orders?” Aviation Week, July 7, 2011.
6. Economies of Scope
Besides scale, another fundamental strategic issue for any

Economies of scope:
business is whether to offer many different products or focus
The total cost of production
is lower with joint than it is on a single item. The answer to this question depends in part
with separate production. on the relation between cost and the scope of production.
There are economies of scope across two products if the
total cost of production is lower when two products are pro-
Diseconomies of duced together than when they are produced separately.
scope: The total cost Conversely, there are diseconomies of scope across two
of production is higher
products if the total cost of production is higher when two
with joint than it is with
separate production. products are produced together.
Joint Cost
Consider how costs depend on the scope of production through the following
example. Suppose that the management of the Daily Globe is considering whether
to launch an afternoon paper, the Afternoon Globe. Table 7.9 shows three cate-
gories of expenses required to produce the Daily Globe and the Afternoon Globe,
assuming a print run of 50,000 copies a day for each paper.
Costs 157
Table 7.9 Expenses for two products
Organization Daily Labor Printing Ink, paper, Total

production ($) press ($) electric cost ($)
(’000) power ($)
Separate production
Daily Globe 50 5,000 3,500 6,700 15,200
Afternoon Globe 50 5,000 3,500 6,700 15,200
Combined production 100 10,000 3,500 13,400 26,900
If the two newspapers are printed in separate facilities, then the total production
cost is $15,200 a day for each paper, or $30,400 for the two papers. If, however,
both newspapers are printed in the same facility, then the total cost of producing
the two newspapers is $26,900. The cost of producing both newspapers in the
same facility is 11.5% lower than if they were produced separately.
What explains the difference in cost? The key is that the same printing press can
be used in the night to print the morning paper, and in the late morning to print
the afternoon paper.
To produce the Daily Globe by itself, the publisher must spend $3,500 a day on
the printing press. Likewise, to produce the Afternoon Globe by itself, the pub-
lisher must spend $3,500 a day on the printing press. To produce both newspapers
from the same facility, however, the publisher spends $3,500 only once.
The expense of the printing press is a joint cost of the morn-
ing and afternoon newspapers. The joint cost is the cost of Joint cost: The cost
inputs that does not change with the scope of production. The of inputs that does not
change with the scope of
joint cost supports the production of multiple products. Econ-
production.
omies of scope arise wherever there are significant joint costs.
Where two products are linked by economies of scope, it will be relatively cheaper to
produce the products together. Then a supplier of both items can achieve a relatively
lower cost than competitors that specialize in one or the other product. Subject to
conditions of market demand and competition, the management should offer both
products. Multi-product suppliers dominate industries with economies of scope.
Broadband service and cable TV provide an important example of scope econ-
omies. Broadband service can be provided through a wire network to potential
subscribers. Similarly, cable TV requires a wire network connecting potential sub-
scribers. In this case, the cost of building and maintaining the network is a significant
joint cost. Consequently, there are very substantial economies of scope across the
broadband and cable TV businesses. Combined providers of broadband and cable
TV service can deliver the services at relatively lower cost than specialized services.
Economies of scope in advertising and promotion are essential for the strategy
of brand extension in marketing. When Sony spends $1 million to advertise Sony, it
158 7 Costs
promotes the sales of every Sony-branded product, including computers, TV sets,

and game consoles. Accordingly, the expenditure on advertising a brand is a joint
cost of marketing all the products marked with the brand.
In an industry with This joint cost gives rise to economies of scope in advertising
economies of scope,
businesses supply
and promotion. Through a brand extension, the owner of an
multiple products. established brand can introduce new products at relatively
lower cost than a competitor with no established brand.
Diseconomies of Scope
There are diseconomies of scope across two products if the total cost of pro-
duction is higher when the two items are produced together than when they are
produced separately. Diseconomies of scope arise where joint costs are not signif-
icant and making one product increases the cost of making the other in the same
facility.
Where diseconomies of scope prevail, it will be relatively cheaper to produce
the various items separately. Hence, specialized producers can achieve relatively
lower costs than competitors that combine production. In such circumstances, the
management should aim for a narrow scope and focus on one product.
PROGRESS CHECK 7E
Referring to Table 7.9, suppose that, with combined production, the expenses
on the printing press were $7,000 a day. Are there economies or diseconomies
of scope?
BANKING AND INSURANCE: DISECONOMIES OF SCOPE?
During the 1990s and into the early 2000s, banks and insurers worldwide rushed
to merge. In 1998, Citibank combined with the insurer, Travelers Group, in one
of the world’s largest mergers to form Citigroup. In 2001, the German insurer,
Allianz, acquired Dresdner Bank.
The French bancassurance model promised higher profits through boosting
demand and economies of scope in service provision. The underlying theory
was that the same institution could provide clients with a wide range of
services and products, including loans, deposits, investments, and insurance.
However, in reality, the economies of scope were difficult to realize.
Consumers make deposits and use many other banking services on a routine
basis. By contrast, investments and insurance are relatively big-ticket items
which consumers buy infrequently. So marketing and supporting investments
and insurance requires different skills than those needed for banking services.
In 2002, just four years after the merger, Citigroup spun off Travelers
Property and Casualty insurance to its shareholders. In 2008, Allianz sold
Dresdner to Commerzbank for €5.5 billion.
Costs 159
7. Experience Curve
We have discussed how the cost of production varies with the scale and scope of
production within any time period. The cost of production may also vary with
experience – as measured by accumulated production – over time.
Experience matters especially in industries characterized by relatively short pro-
duction runs. In aerospace manufacturing, as engineers and workers gain experience
in production, they devise new ways to reduce cost, including developing better
tools and faster processes. Even in semiconductor manufacturing, a large-scale
process, experience matters as the rate of defects falls with accumulated produc-
tion. In surgery, with accumulated experience, the medical professionals become
more proficient individually and as teams.
Accordingly, the unit cost of production falls with accu-
mulated experience, which motivates the concept of the expe- Experience curve:
rience curve, which is also called the learning curve. Typically, The unit (average) cost
experience is measured by cumulative production, hence the of production falls with
cumulative production
experience curve shows how the unit (average) cost of over time.
production falls with cumulative production over time.
Figure 7.5 illustrates an experience curve, where the unit
(average) cost of production is indexed to 100 for the first unit produced. The unit
cost falls most sharply with the initial units of cumulative production. Assuming
a 20% cost reduction for every doubling of cumulative production (that is, a learn-
ing percentage of 80%), by the fourth unit of cumulative production the unit cost
drops to 64, by the 32nd unit to 32.77, etc.
The learning percentage determines the rate at which the unit cost falls with
cumulative production. The lower the learning percentage, the higher the rate
120.00
100.00
80.00
Unit cost
60.00
40.00
20.00
0.00
0 200 400 600 800 1000 1200
Cumulative production
FIGURE 7.5 Experience curve.

Notes: The experience curve shows how the unit cost of production falls with cumulative production
over time. An 80% learning percentage is shown, and the unit cost of production is indexed to 100
for the first unit produced.
160 7 Costs
of cost reduction from doubling cumulative output. The learning percentage

depends on the particular technology and process of manufacturing, and so var-
ies with product and industry.
The experience curve is distinct from economies of scale. The experience curve
relates cumulative production over preceding periods to production costs in one
period. By contrast, economies of scale relate the scale of production within one
period to production costs in the same period.
In any industry where production costs are subject to a substantial experience curve
relative to cumulative sales, it is crucial to forecast cumulative production accurately.
Accurate forecasting of cumulative production is crucial for both planning invest-
ments and setting prices. The challenge is especially great to the extent that sales,
and hence cumulative production, depend on pricing and competitors’ strategies.
For instance, referring to Figure 7.5, if the manufacturer plans for cumulative
production of 500 units, its unit (average) cost would be about 17. Hence, it can
break even with a price of 17 or more. By contrast, if the manufacturer plans
for cumulative production of 100, the unit cost would be about 29, which is 70%
higher than with cumulative production of 500 units.
The experience curve also motivates a strategy of cutting price to increase sales
and so gain more cumulative production. The business that gets ahead in cumu-
lative production can then outcompete competitors on costs. So, the experience
curve can justify a strategy of competing for market share.
PROGRESS CHECK 7F
On Figure 7.5, draw a new experience curve with a smaller reduction (lower
than 20%) in unit cost for every doubling of cumulative production.
BOMBARDIER: WAITING FOR LUFTHANSA
The cost of aircraft manufacturing is strongly affected by accumulated

experience. Airbus, Boeing, and other airplane manufacturers must set prices
for aircraft based on projections of forecast cumulative sales. If the sales fall
short of target, unit costs will be higher than planned, and the manufacturer
may incur a substantial loss on the plane.
Bombardier is well established in the production of regional jets. Long
aspiring to expand into large jets, it announced the CSeries in 2004. Bombardier
estimated the cost of development to be $2.5 billion, with part of the cost being
born by the governments of Canada and the United Kingdom, and suppliers.
Much of the development cost would be sunk, once incurred. In January
2006, with no significant orders, Bombardier scaled back plans and reallocated
resources to produce the CRJ1000 regional jet.
Costs 161
Then, in July 2008, Bombardier secured a letter of interest for 60 CSeries planes,
including 30 options, from Deutsche Lufthansa. Bombardier then committed to
development, with commercial service to begin in 2013. Given the experience
curve and that the costs of development would be substantially sunk once incurred,
Bombardier was prudent in delaying commitment until securing a big order.
Source: Richard Tortoriello, “Aerospace & defense,” Standard & Poor’s Industry Surveys,
February 10, 2011; “Bombardier jet strains to take off,” Wall Street Journal Online, June 16,
2011; “Bombardier CSeries,” Wikipedia (accessed July 25, 2011).
8. Bounded Rationality
Being human, managers, like consumers, are subject to bounded rationality in

decision-making. In decisions with respect to costs, managers are particularly
prone to status quo bias, the sunk-cost fallacy, and what I call the “fixed-cost
fallacy.” The techniques of this chapter are even more pertinent to the extent that
managers fall prey to these biases in their costing decisions.
• Status quo bias. Human beings tend to be biased toward the status quo,
perhaps because of inertia. To the extent that the status quo involves
opportunity costs, the status quo bias reinforces any systematic failure to
account for opportunity costs. The result is an even stronger bias toward
continuing with the status quo, rather than taking an alternative course of
action, which might actually be more profitable.
• Sunk-cost fallacy. One human response to sunk costs is to rationalize the
sunk costs by increasing usage or consumption. An (in)famous example is
the Anglo-French supersonic airliner, Concorde. Having invested heavily,
Britain and France continued to pour additional resources into the
project long after it was clear that Concorde was not commercially viable.
Their actions exemplified the sunk-cost fallacy. This human tendency to
rationalize costs that are already sunk by additional expenditures results in
overinvestment.
• Fixed-cost fallacy. Related to the sunk-cost fallacy, but resulting in the opposite
outcome, is what might be called the “fixed-cost fallacy.” This is the tendency,
which is common, for managers to treat fixed costs as variable. The mistake
is to set a target production rate and allocate the fixed cost to each unit of
production. Essentially, this allocation increases the perceived variable cost
and so results in underproduction relative to the profit-maximizing level.
KEY TAKEAWAYS
• For effective decision-making, consider alternative courses of action, take account

of opportunity costs and ignore sunk costs.
162 7 Costs
• The opportunity cost is what must be forgone from the best alternative course
of action.
• Businesses financed by equity should take account of the opportunity cost of
equity capital.
• To maximize profit, set the transfer price equal to the marginal cost of the input.
• Commit to sunk costs with caution as they cannot be reversed.
• Economies of scale arise from fixed costs, which support the production of
multiple units of output.
• With economies of scale, businesses should produce on a large scale and the
industry will tend to be concentrated.
• Economies of scope arise from joint costs, which support production of multiple
products.
• With economies of scope, businesses should produce multiple products.
• The experience curve shows that unit (average) cost of production falls with
cumulative production over time at a rate according to the learning percentage.
• With the experience curve, it is important to forecast cumulative production
and the business can gain from increasing market share.
• Managers should take care to avoid behavioral biases in decisions with respect
to costs.
REVIEW QUESTIONS
1. A salesman buys lunch for a potential client. Why is this lunch not free for the client?
2. A social enterprise provides free primary school education in West Africa. It
charges fees to just cover costs. The enterprise is completely financed by
European charities. Compare the school’s value added (according to the
definition in Chapter 1) and EBITDA.
3. There is a perfectly competitive market for lumber. How much should Saturn’s
residential development group pay the building materials division for lumber?
4. Luna Biotech’s manufacturing division uses a unique patent-protected process
to culture drugs. Presently, the division is operating at full capacity. How should
it price its services to Luna’s marketing division?
5. Mercury Transport is wholly equity-financed, while Jupiter Trucking has
borrowed from banks to finance its business. The two businesses are otherwise
identical. Jupiter is losing money, while Mercury is profitable. Compare the two
businesses in terms of EBITDA.
6. “Our costs are very high because of the huge pensions of our retirees.” Are the
pensions of retired employees relevant for forward-looking business decisions?
7. In which situation are sunk costs more significant: (a) Tre Stagioni, which has a
permanent staff of two chefs and five waiters; or (b) Campus Deli, which relies
mainly on part-time workers, hired on a monthly basis?
8. The most substantial cost in family medicine practice is human resources. To treat
twice as many patients, a clinic will probably need twice as many doctors, nurses,
and other professional staff. Does this business have economies of scale?
9. Explain the difference between fixed cost and sunk cost.
10. What are the strategic implications of economies of scale?
11. Explain the difference between economies of scale and economies of scope.
Costs 163
12. What are the strategic implications of economies of scope?

13. Draw an experience curve with a learning percentage of 100%.
14. What are the strategic implications of the experience curve?
15. Explain the fixed-cost fallacy in the context of producing a newspaper.
1. The Great Recession of 2007–2009 severely depressed world trade, the demand
for tanker services, and the daily rental rate for tankers. Herbjorn Hansson, CEO
of Nordic American Tanker Shipping, remarked: “Those who have a lot of debt
are suffering . . . . If you’re collecting $10,000 a day and you have a cash break-
even of $25,000 a day, that’s a $15,000 a day loss.” By contrast, Hansson felt
that his company was well placed to survive the downturn as it carried no debt.
Its ships could make a profit with rates of $10,000 a day. (Source: “Oil tanker
owners see rates fall,” Financial Times, May 5, 2009.)
(a) Suppose that the operating costs of a tanker are $8,000 a day, and,
for a debt-financed tanker, the interest cost is $17,000 a day. Given
revenues of $10,000 a day, construct a conventional income statement.
(b) Suppose that the operating cost of a laid-up tanker is $1,500 a day.
Construct an income statement showing two alternative courses of
action – continuing operations and laying up the vessel. Should the
tanker owner lay up the ship?
(c) In discussing the situation of tanker owners financed by debt, what
mistake did Hansson make?
(d) By 2010, the world demand for oil and tanker rates had recovered. From
a long-run perspective, tankers that are equity-financed yield higher
profit than debt-financed tankers. (i) Why do equity-financed tankers
yield higher profit? (ii) Explain a better way to compare the performance
of equity- and debt-financed tankers.
2. Consider a distributor of luxury American and European cars in China. In
January 2011, it ordered shipments of high-end BMWs from Germany at
€50,000 each and Cadillac SUVs from the United States at $40,000 each. By
July 2011, the Chinese yuan had depreciated by 3% from 8.93 yuan against
the euro and appreciated by 2% from 6.58 yuan against the US dollar.
(a) For purposes of pricing, what is the relevant cost of the BMWs?
(b) For purposes of pricing, what is the relevant cost of the Cadillacs?
(c) Use the examples in (a) and (b) to explain the concepts of opportunity
cost and sunk cost.
3. Following a collapse of sales in the wake of the subprime financial crisis, the US
automobile manufacturer Chrysler LLC filed for bankruptcy in April 2009. It then
sold its brands and various other assets to Fiat SpA. Table 7.10 lists the estimated
recovery value of Chrysler’s property, plant, and equipment in the event of liquidation.
(a) Which of the following best describes the difference between book
value and estimated recovery value: (i) opportunity cost; (ii) sunk cost;
or (iii) joint cost? Explain your answer.
164 7 Costs
Table 7.10 Chrysler LLC liquidation ($ million)
Assets Book value Recovery value
Low estimate High estimate
Assembly plants 2,205 110 220

Stamping plants 1,129 113 226
Powertrain plants 3,513 352 702
Tooling 1,337 − 67
Furniture, fixtures, etc. 487 5 24
Source: Chrysler LLC, Preliminary Hypothetical Liquidation Analysis – Orderly Liquidation,
January 31, 2009.
(b) Use the low and high estimates of recovery value to calculate the
average recovery value.
(c) Define specificity as the ratio of book value less average estimated
recovery value to book value, in percentage terms. Calculate the
specificity of each category of assets.
(d) Explain the relation between sunk costs and specificity.
(e) Stamping equipment is heavy machinery used to produce metal
parts such as car bodies. Tooling is equipment designed to produce
particular models of cars. Explain why Chrysler’s tooling has a higher
specificity than the stamping plants.
4. Elevators generally break down at random times and for different reasons.
Elevator maintenance contractors must have trained service personnel to
provide routine and emergency service. Shan On Elevator, Hong Kong, has
200 service personnel to maintain 1,000 elevators.
(a) Suppose that Shan On has received a contract to maintain an
additional 1,000 elevators. Do you expect that Shan On will need to
double service personnel? Why or why not?
(b) Does the example in (a) illustrate economies of scale or scope?
(c) Escalators and elevators use quite different technology and parts. But many
clients operate both escalators and elevators. Are there any economies of
scope for a contractor to maintain both escalators and elevators?
5. Generators of electric power produce electricity in large-scale plants with
capacities of hundreds of megawatts. However, when electricity is transmitted
over cables, resistance causes loss of energy. The loss of energy increases
with the distance of transmission.
(a) Draw a graph to illustrate the average cost of generating electricity at
different scales of production (megawatt hours per month).
(b) Draw a graph to illustrate the average cost of transmitting electricity
for different distances.
(c) If a producer of electricity increases the scale of generation to reduce
average cost of generation, what will be the effect on the average cost
of transmission?
(Hint: You are free to assume any data necessary to draw the curves.)
Costs 165
6. Airbus and Boeing dominate the large commercial jet aircraft industry. Each
company manufactures jet aircraft for commercial use by passenger and cargo
airlines as well as for military use by national air forces. In February 2011, Boeing
won a US Air Force tender for 179 aerial refueling tankers at an estimated price of
$35 billion. The Boeing tanker is based on Boeing’s 767 wide-body commercial
aircraft.
(a) Which one of the following concepts best explains Boeing’s adaptation
of the Boeing 767 to supply an aerial tanker: (i) economies of scale;
(ii) economies of scope; (iii) experience curve? Explain your answer.
(b) Following the initial tender for 179 tankers, the US Air Force is expected
to buy additional tankers. Apply relevant cost concepts to explain
Boeing’s advantage in competing for follow-on orders.
(c) “The more an aircraft manufacturer spends on R&D, the higher the
price that it must charge to recover its investment.” Comment.
7. Punch Taverns owns over 3,500 pubs (bars) in Britain. The group borrowed
heavily to grow, but was hit by a ban on smoking in bars, higher taxes on beer,
and the Great Recession. In October 2014, the group negotiated with creditors
to exchange £0.6 billion of debt for equity. The restructuring reduced the group’s
debt by a quarter to £1.8 billion and left the original shareholders with just 15% of
the shares. Table 7.11 reports the group’s income statement for 2014. (Sources:
“Punch Taverns wins final restructuring approval,” Financial Times, October 7,
2014; Punch Taverns PLC, Preliminary Results, August 23, 2014.)
(a) Explain which concept of cost can justify Punch Tavern’s expansion strategy.
(b) Which items in the income statement would the restructuring affect?
(c) How would the restructuring affect Punch Tavern’s EBITDA?
(d) If the income statement took account of the opportunity cost of equity
capital, would the restructuring improve the group’s financial performance?
8. In August 2004, Infineon announced a deal with Taiwan’s Winbond Electronics
to build a new dynamic random-access memory (DRAM) factory. Infineon
executive Ralph Heinrich hoped to secure a quarter of the global DRAM market:
“the fight to survive in the DRAM industry depends largely on size – since the
more chips a company churns out, the lower the cost per chip.” DRAMs are
cut from circular wafers of semiconductors. Wafers of 300 millimeter (mm)
diameter potentially yield more than twice as many DRAMs as 200 mm wafers.
A 300 mm wafer fabrication facility costs more to build and set up than a
Table 7.11 Punch Taverns (£ million)
2014
Revenue 448
Operating costs and joint venture profit –270
EBITDA 178
Depreciation, amortization, impairment etc. –55
Operating profit 123
Finance income and costs –363
Profit before tax –240
166 7 Costs
200 mm facility. (Source: “Infineon’s deal with Winbond reaffirms outsourcing

strategy,” Asian Wall Street Journal, August 10, 2004.)
(a) Assume that the variable costs of manufacturing 200 mm and 300 mm
wafers are the same. Explain why the economies of scale in manufacturing
DRAMs from 300 mm wafers are larger than with 200 mm wafers.
(b) Infineon produces DRAMs from multiple factories, some of which are
joint ventures with Taiwanese manufacturers. Does the cost per DRAM
depend on the total quantity produced by the entire company or each
individual factory?
(c) When a wafer fabrication facility is first commissioned, the percentage
of output that meets product standards (the “yield”) tends to be low.
Engineers then fine-tune the manufacturing process to increase the
yield. What principle of cost does this illustrate?
9. In April 2004, Boeing launched the new 787 Dreamliner with 50 firm orders
from All Nippon Airways of Japan. Boeing aimed to secure 200 firm orders
by December. However, by December 2004, Boeing had only 52 orders. Then
Airbus introduced the A350, a derivative of the existing A330, enhanced with
a new wing, more fuel-efficient engines, and other new technologies. Airbus’s
Chief Commercial Officer, John Leahy, predicted that the A350 would draw
Boeing customers and so “put a hole in Boeing’s Christmas stocking.” (Source:
“350: Airbus’s counter-attack,” Flight International, January 25, 2005.)
(a) Draw a timeline to mark when a manufacturer incurs the costs of
development and production.
(b) How would the costs of developing the 787 Dreamliner vary with the
total quantity manufactured?
(c) Referring to Figure 7.5, compare Boeing’s average cost with cumulative
production of 50 and of 200 units. (Note that, in Figure 7.5, the average cost
is not absolute but rather indexed to 100 with production of the first unit.)
(d) Suppose that the price of a Boeing 787 is $120 million and that Boeing
would just break even on the costs of development ($10 billion) and
manufacturing with cumulative production of 200 units. How much
would Boeing lose with cumulative production of 50 units?

Consider your organization’s various lines of business. For each line of business,
consider the revenues and costs from alternative uses of the resources –
people, property, and funds. Is every line of business maximizing profit?
Note
1 This discussion is based, in part, on Richard Tortoriello, “Aerospace & defense,” Standard & Poor’s
Industry Surveys, February 10, 2011; “Boeing likely to boost 737, 777 production rates,” ATWOn-
line, March 18, 2010; “Bombardier jet strains to take off,” Wall Street Journal Online, June 16, 2011;
“Airbus and Boeing call end to ‘duopoly,’” Financial Times, June 21, 2011; “Airbus-Boeing duopoly
holds narrow-body startups at bay at Paris Air Show,” Bloomberg, June 23, 2011.
C H A P T E R
8
Monopoly
LEARNING OBJECTIVES
• Appreciate how to gain market power.
• For a seller with market power, identify the scale of production/sales
that maximizes profit.
• Appreciate how to adjust sales to changes in demand and costs.
• For a seller with market power, identify the levels of advertising and
R&D expenditure that maximize profit.
• Appreciate that sellers with market power restrict sales to raise
margins and profit.
• Apply the incremental margin percentage to measure market
power.
• For a buyer with market power, identify the scale of purchases that
maximizes profit.
1. Introduction
Atorvastatin, marketed by Pfizer under the brand name Lipitor, inhibits the liver
enzyme, HMG-CoA reductase, and so reduces the level of low-density lipoprotein
cholesterol in the human body. In 2010, Lipitor was Pfizer’s best-selling drug, with
sales revenue of $10.7 billion, or 15.7% of the company’s total revenue.1
Lipitor faced competition from other statins – particularly simvastatin. In June
2006, Merck’s US patent on simvastatin expired and Merck cut the price of its
branded simvastatin, Zocor. Moreover, Pfizer’s US patent on atorvastatin was
initially due to expire in June 2011.
168 8 Monopoly
In 2003, the Indian generic drug manufacturer, Ranbaxy Laboratories, filed

an Abbreviated New Drug Application (ANDA) with the US Food and Drug
Administration (FDA) for a generic version of atorvastatin. The Hatch-Waxman
Act provides six months of exclusivity to the first generic manufacturer approved by
the FDA. The six-month period of generic exclusivity begins immediately after the
expiry of the patent of the original drug. Typically, the exclusive generic manufac-
turer will price its drug at 70–80% of the price of the patented drug. Once the generic
exclusivity expires and open competition ensues, the price will fall even further.
Following its ANDA, Ranbaxy sued Pfizer to establish the expiry of Pfizer’s US
patents. In June 2008, the two companies settled: Pfizer dropped its opposition to
Ranbaxy’s generic atorvastatin and Ranbaxy agreed to delay selling the generic
in the United States until November 2011, giving Pfizer an additional five months
of exclusivity. Pfizer also agreed to Ranbaxy selling the generic atorvastatin in
Australia, Canada, and specific European countries.
Ranbaxy expected to earn $600 million in profit from atorvastatin during the six
months of generic exclusivity. This profit was significant for a company with $1.9
billion in revenue and $459 million in earnings. Pfizer itself prepared to manufac-
ture a generic version of atorvastatin for distribution by Watson Pharmaceuticals.
Blockbuster drugs like atorvastatin require millions of dollars of investment in
research and development. For every blockbuster drug, many molecules fail in the
market and many more do not make it beyond the laboratory. Yet R&D is key for
branded pharmaceutical manufacturers to differentiate themselves from generic
manufacturers. How much should Pfizer spend on R&D?
Lipitor already competed with simvastatin and other statins. Generic competi-
tion was imminent in the market for atorvastatin in the United States and Europe.
Pfizer had to decide how to manage the competition. How much should it spend
on advertising? At what scale should Pfizer produce the branded drug, Lipitor,
and the generic version? How would generic production of atorvastatin affect the
market for the ingredients in the production of the drug?
To address these questions, we must understand the behavior of buyers or sell-
ers that have the power to influence market conditions. A buyer or seller that can
influence market conditions is said to have market power.
Market power: The A buyer with market power can influence market supply – in
ability of a buyer or seller particular, the price and quantity supplied. A seller with mar-
to influence market ket power can influence market demand – in particular, the
conditions.
price and quantity demanded.
Economic profit is the difference between revenue and cost.
Monopoly: One seller in Businesses can use their market power to increase revenue and
a market. reduce cost, and so increase profit. Here, we ask how to build
market power and how to use it. For simplicity, we focus on
markets in which there is either just one seller or one buyer. If
Monopsony: One buyer there is only one seller in a market, that seller is called a
in a market.
monopoly. If there is only one buyer in a market, that buyer
is called a monopsony.
Monopoly 169
This chapter begins with the sources of market power. Then we analyze how a
profit-maximizing monopoly determines its scale of production and price. This
then shows how the producer should adjust production and price in response to
changes in demand and costs. Applying this analysis, we can address the issues of
how Pfizer should adjust production of Lipitor and at what scale to produce the
generic atorvastatin.
Next, we consider how much a monopoly should spend on advertising and R&D,
and explain how competing sellers can benefit by restricting competition among
themselves. These analyses explain how much branded pharmaceutical manufac-
turers like Pfizer should spend on advertising and R&D.
Finally, we consider monopsony and analyze how a monopsony that maximizes
net benefit would set the scale and price of purchases. This explains how the expiry
of the patents on atorvastatin would affect the market for the ingredients in the
production of the drug.
2. Sources of Market Power
Market power has two ingredients – one is barriers to com-

petition, and the other is the elasticity of demand or supply. Sources of market
For a monopoly or monopsony to exist, competitors must be power:
deterred or prevented from entering the market to compete • product differentiation;
for the business. So, one source of market power is the barri- • intellectual property;
ers that keep competitors out. • economies of scale,
scope, and experience;
However, even a monopoly cannot increase profit unless it
• regulation.
can influence demand conditions – in particular, by raising
price. Likewise, a monopsony cannot increase profit unless it
can influence supply conditions. So, the other ingredient to market power is the
elasticity of demand or supply.
Here, we focus on the seller’s market power, while noting that the ingredients for
the buyer’s market power are quite symmetric. Sellers can reduce the price elastic-
ity of demand and create barriers to competition in four ways.
Product Differentiation
Chapter 3 presented the intuitive conditions for demand to be inelastic with respect
to price. Here, we emphasize one factor – product differentiation. To the extent that
the differentiation appeals to buyers, it would increase the demand and reduce the
price elasticity of demand, and so contribute to market power.
Broadly, sellers can differentiate their products in four ways:
• Design provides the potential buyer’s first impression of a product. Hence, an

obvious way to differentiate products is by distinctive design – the appearance,
form, and feel. Appealing design can transform utilitarian products into
distinctive offerings that buyers value and prefer.
170 8 Monopoly
• Function is, besides design, the other aspect of the day-to-day benefit that the
product provides to the user.
• Distribution channels make products available to buyers at a time and place of
their convenience. So another way to differentiate products is through distribution
channels. Manufacturers of luxury products use exclusive distribution to build
brand image and demand. By contrast, producers of mass consumer goods use
intensive distribution to provide wide and timely availability.
• Advertising and promotion introduce buyers to products and communicate
the brand image. Hence, yet another way to differentiate products is through
advertising and promotion that influence and sustain buyers’ preferences.
Intellectual Property
Product differentiation builds, in part, on innovation. To encourage particular forms
of innovation, society may award the innovator a period of exclusive use through
intellectual property. Innovators may be able to exclude competitors and so con-
tribute to market power by establishing intellectual property over their innovations.
• A patent gives the owner an exclusive right to an invention for a specified

period of time. Pfizer’s patent on atorvastatin provided the manufacturer with
a monopoly until 2011.
• Copyright provides exclusivity over published expressions for a specified
period of time. Microsoft’s copyright over the Windows operating system
and Office application suite provides the software publisher with a monopoly
over the software.
• A trademark provides exclusivity over words or symbols associated with a
good or service. Trademarks are the basis for branding and advertising and
promotion. Pfizer’s Lipitor trademark complements its patent, and Microsoft’s
Windows trademark complements its copyright.
• Trade secrecy provides exclusivity over information that is not generally
known and that provides commercial advantage. The scope of secrecy extends
to business information such as customer lists and technical information that
possibly cannot be patented, such as algorithms. Google famously protects
its technologies through secrecy.
Economies of Scale, Scope, and Experience

A third source of market power is economies of scale, scope, and experience. In an
industry characterized by economies of scale, scope, or experience, an incumbent
producer will have a cost advantage over potential competitors. By establishing a
sufficient cost advantage, the incumbent producer may be able to deter entry by
competitors and so gain market power.
The combined provision of broadband and cable television illustrates econ-
omies of scope. Both services depend on a network of cables from the service
Monopoly 171
provider to subscribers. Owing to economies of scope, a combined provider of

broadband and cable television can achieve lower costs than specialized providers.
Further, the cable network also exhibits economies of scale – it involves large
fixed costs and relatively low variable costs. Consequently, the broadband and
cable television industries tend to be dominated by a few providers, each of which
provides both services.
Regulation
Finally, for various reasons, including economic or social policy, the government
may decide to limit competition, and in the extreme, allow only a single producer.
A producer with the government license is shielded from competition by law, and
so gains market power.
An important economic reason for such regulation is the presence of large fixed
costs in production. So, for example, most governments limit competition in the
distribution of electricity, natural gas, and water. The policy helps to avoid dupli-
cation of the fixed costs of the distribution network.
Governments may limit competition in particular markets for social reasons.
Examples include sports and mass media, retailing of alcohol and tobacco, and
gambling.
PROGRESS CHECK 8A
What are the two ingredients of market power?
APPLE: PRODUCT DIFFERENTIATION AND INNOVATION
Most manufacturers of personal computers produce utilitarian machines and

market their products on function and price. Apple famously took a different
approach. From the first Apple computer to the latest Mac, Apple founder Steve
Jobs emphasized design as much as function. Apple has consistently priced
the Mac above other brands of personal computers and earned higher margins.
Apple has also pioneered differentiation through exclusive distribution to
connect directly with consumers. In 2001, Apple opened its first dedicated
retail stores in Tyson’s Corner, Virginia, and Glendale, California. To staff the
stores, Apple carefully selects passionate people who are keen to collaborate
and ready to learn.
And, of course, Apple reinforces its brand through advertising and creates
new products through R&D. In 2013, Apple spent $1.1 billion on advertising
and $4.5 billion on R&D – amounting to 0.6% and 2.6% respectively of its
$170.9 billion sales revenue.
Source: Apple Inc., Annual Report 2013.
172 8 Monopoly
COKE: 17 YEARS WAS TOO SHORT
In 1886, pharmacist John Pemberton sold the first glass of Coca-Cola for
5 cents at Jacobs’ Pharmacy in downtown Atlanta. The Coca-Cola Company
famously did not patent its invention. Applying for a patent would require
disclosure of the formula for the beverage. And, upon the expiry of the patent
(after 17 years), competitors would be free to produce the “real thing.” So,
Coca-Cola has relied on secrecy to protect the formula.
There are many competing colas in the retail market. Besides its secret
formula, Coca-Cola advertises heavily to reinforce its brand and differentiate
its products. In 2010, Cola-Cola spent $2.9 billion or 8.3% of its $35.1 billion
sales revenue on advertising. The advertising supports a price premium over
competing beverages.
Source: Coca-Cola Company, Annual Report 2010.
3. Profit Maximum
Having gained market power, how should a seller use it? Here, we consider the
scale of production and price. Suppose that Venus Pharmaceutical has a monop-
oly over a drug, Gamma-1, that cures bone-marrow cancer. Venus faces the two
basic business decisions. One is participation – should it produce the drug? The
other is extent – how much should it produce and sell (and how should it price
the drug)?
Supposing that Venus decides to produce the drug, we first analyze the prof-
it-maximizing sales and price. Then we consider whether Venus should produce
the drug at all.
The essence of market power is that the seller faces a demand curve that slopes
downward. Venus Pharmaceutical, being a monopoly, faces the market demand
curve. Unlike a perfectly competitive seller, a monopoly has to consider how its
sales will affect the market price.
Given the market demand curve, a monopoly can either decide how much to
sell and let the market determine the price at which it is willing to buy that quan-
tity, or set the price and let the market determine how much it will buy. If the
monopoly tries to set both sales and price, the sales and price may be inconsistent
in the sense that, at that price, the market wants to buy more or less than the
quantity the monopoly is selling. In terms of a graph, inconsistency means that
the monopoly is choosing a combination of sales and price off the demand curve.
Accordingly, a monopoly can set either sales or price, but not both.
Let us focus on the decision on sales. For simplicity, we ignore inventories and
hence production equals sales. So, scale of production and sales are equivalent.
To analyze the profit-maximizing sales, we need to know how Venus’s sales affect
its revenues and costs.
Monopoly 173
Revenue
First, consider the relationship among price, sales, and revenue. Table 8.1 shows
the demand for Gamma-1. Specifically, the second column shows, for every price,
the quantity that Venus expects to sell. The quantity demanded increases by
200,000 units for every $10 reduction in price. Using this information, we can then
calculate Venus’s total revenue for every price, which is price multiplied by sales.2
From the total revenue, we can then calculate marginal revenue, which is the
change in total revenue arising from selling an additional unit.
To sell additional units, Venus Pharmaceutical must reduce its price. So, when
increasing sales by one unit, Venus will gain revenue from
selling the additional (or marginal) unit, but it will lose reve- Inframarginal units:
Units sold other than the
nue on the inframarginal units. The inframarginal units are
marginal unit.
those units sold other than the marginal unit. Venus would
have sold the inframarginal units without reducing the price.
For example, referring to Table 8.1, to increase sales from 200,000 to 400,000 units,
Venus must reduce the price from $190 to $180. Hence, Venus will gain revenue
of $180 × 200,000 = $36 million on the additional units, but lose $(190 − 180) ×
200,000 = $2 million on the inframarginal 200,000 units that it could have sold
at $190. Thus, Venus’s revenue for the additional 200,000 units is $36 million −
$2 million = $34 million, which means that marginal revenue is $170 per unit.
In general, the marginal revenue from selling an additional unit will be less than
the price of that unit. The reason for this is that the marginal revenue is the price
of the marginal unit minus the loss of revenue on the inframarginal units.
The difference between the price and the marginal revenue depends on the price
elasticity of demand. If demand is very elastic, then the seller need not reduce
the price very much to increase sales; hence, the marginal revenue will be close to the
Table 8.1 Monopoly revenue, cost, and profit
Price Sales Total revenue Marginal Total cost Marginal Profit

($) (million) ($ million) revenue ($) ($ million) cost ($) ($ million)
200 0.0 0.00 50.00 −50.00

190 0.2 38.00 190 64.20 71 −26.20
180 0.4 72.00 170 78.40 71 −6.40
170 0.6 102.00 150 92.60 71 9.40
160 0.8 128.00 130 106.80 71 21.20
150 1.0 150.00 110 121.00 71 29.00
140 1.2 168.00 90 135.20 71 32.80
136 1.28 174.08 76 140.88 71 33.20
135 1.30 175.50 71 142.30 71 33.20
134 1.32 176.88 69 143.72 71 33.16
130 1.4 182.00 64 149.40 71 32.60
120 1.6 192.00 50 163.60 71 28.40
110 1.8 198.00 30 177.80 71 20.20
100 2.0 200.00 10 192.00 71 8.00
90 2.2 198.00 −10 206.20 71 −8.20
174 8 Monopoly
price. If, however, demand is very inelastic, then the seller must reduce the price sub-
stantially to increase sales; so the marginal revenue will be much lower than the price.
The marginal revenue can be negative, if the loss of revenue on the inframar-
ginal units exceeds the gain on the marginal unit. Table 8.1 shows that, if Venus
cut the price from $100 to $90, sales would increase from 2.0 to 2.2 million units.
The change in revenue, however, is −$2 million for the additional 200,000 units,
which means that marginal revenue is −$10 per unit.
PROGRESS CHECK 8B
If demand is extremely elastic, what will be the difference between the price
and the marginal revenue?
Costs
We have considered the relation among price, sales, and revenue. The other side
to profit is cost. Table 8.1 also shows data for Venus Pharmaceutical’s production
costs. It reports only avoidable costs. From the total cost at a zero production
scale, we can infer that production requires a fixed cost of $50 million.
Total cost increases with the scale of production. Table 8.1 shows Venus’s mar-
ginal cost, which is the change in total cost due to the production of an additional
unit. The change in total cost arises from change in the variable cost. For simplic-
ity, the marginal cost is $71 per unit at all scales of production.
Profit-Maximizing Scale
With information on both revenue and cost, we can calculate Venus’s profit at
every possible price and quantity. Profit is total revenue less total (fixed and vari-
able) cost. The last column of Table 8.1 reports profit at each quantity of sales.
Looking down the column, we see that Venus’s maximum profit is $33.2 million.
It achieves this profit with a price of $135 and sales of 1.3 million units.3
Another way to identify the profit-maximizing scale
Profit-maximizing scale of production is the scale at which the marginal revenue
of production: The scale
at which the marginal
balances the marginal cost. At the price of $135, Venus’s
revenue balances the sales are 1.3 million units. The marginal revenue is $71 per
marginal cost. unit (note that, to calculate the marginal revenue more pre-
cisely, we drilled down to a smaller increment between prices
around the profit-maximizing level). The marginal cost is also $71 per unit. So, at
the profit-maximizing scale, the marginal revenue equals the marginal cost.
This suggests a general rule: to maximize profit, a monopoly should produce at
a scale where its marginal revenue balances its marginal cost. This rule applies to
any seller and not only a monopoly.
Let us illustrate the profit-maximizing price and operation scale with a dia-
gram. Figure 8.1 shows Venus Pharmaceutical’s demand, marginal revenue, and
marginal cost curves. The demand curve shows, for every price, the quantity that
Monopoly 175
demand (marginal benefit)

Price ($ per unit)
135
b
a e marginal cost
71 c
d marginal revenue
0 1.2 1.3 1.4

Quantity (million units a year)
FIGURE 8.1 Monopoly production scale.

Notes: The marginal revenue and marginal cost curves cross at the quantity of 1.3 million units. From
the demand (marginal benefit) curve, the profit-maximizing price is $135. At 1.2 million units, the
marginal revenue exceeds the marginal cost.
the market will buy. Equivalently, it shows, for every quantity of purchases (on
the horizontal axis), the maximum price (on the vertical axis) that the market will
pay for that quantity.
The marginal revenue curve shows, for every quantity (on the horizontal axis),
the marginal revenue (on the vertical axis). As we have explained, for every quan-
tity, the marginal revenue is less than the price. Accordingly, at all quantities, the
marginal revenue curve lies below the demand curve.
The marginal revenue and marginal cost curves cross at the quantity of 1.3 million
units. From the demand curve, we see that the price at that quantity is $135. This
is the profit-maximizing price.
Let us understand why a seller maximizes profit at a production scale where
marginal revenue balances marginal cost. Suppose that Venus produces at a scale,
such as 1.2 million units, where the marginal revenue exceeds the marginal cost.
Then, if Venus increases production by 100,000 units, its revenue will increase by
more than its cost; indeed, it will increase its profit by the area shaded bca.
By contrast, suppose that Venus produces at a scale such as 1.4 million units,
where the marginal revenue is less than the marginal cost. Then, if Venus cuts
production by 100,000 units, its revenue will fall by less than its cost; hence it will
increase its profit by the area ade. Generally, a seller will maximize profit by pro-
ducing at a scale where its marginal revenue balances its marginal cost.
Profit maximum:
Break-Even Analysis
• Produce if total revenue
What about the participation decision: should Venus pro- covers total cost.
duce Gamma-1 at all? We have already analyzed the profit- • Produce at scale where
marginal revenue
maximizing scale of production or sales, assuming that equals marginal cost.
Venus does produce the drug. So Venus should produce the
176 8 Monopoly
drug if the total revenue covers the total cost, with both revenue and cost calcu-
lated at the profit-maximizing scale.
Profit contribution: Total

revenue less variable cost. Profit Measures
Later in this chapter, we discuss how much a seller with mar-
Incremental margin: ket power should spend on advertising and R&D. That dis-
Price less marginal cost. cussion applies two measures of profit. One is the profit
contribution, which is total revenue less variable cost. The other
Incremental marginal is the incremental margin, which is the price less the mar-
percentage: The ratio ginal cost. Further, the incremental marginal percentage
of incremental margin to is the ratio of the incremental margin (price less marginal
price. cost) to the price.
PROGRESS CHECK 8C
Suppose that, at the current scale of production, Venus’s marginal revenue is
less than its marginal cost. How should management adjust its production?
BIG PHARMA VERSUS GENERICS
Branded pharmaceutical manufacturers such as Pfizer invest billions of dollars

in R&D to discover and develop new drugs. Of every 10,000 compounds
investigated, only five are tested in clinical trials, and, of those, only one
receives approval for use among patients.
R&D expenditures are fixed costs in the sense that they do not vary with
scale of production. To cover the R&D expenditures, Pfizer needs a large
profit contribution. This comes from either a high incremental margin on each
unit of sales or large scale of production or both.
Generic manufacturers such as Ranbaxy follow a different strategy. They
jump in after the patents of branded manufacturers expire and focus R&D
on developing generic equivalents. With lower R&D expenditures, they incur
relatively lower fixed costs, and so can be profitable with smaller incremental
margin and production.
In 2010, Pfizer earned revenues of $67.8 billion, with a gross margin of 76%,
and spent $9.4 billion or 13.9% of revenues on R&D. By contrast, Ranbaxy
earned revenues of 66.7 billion Indian rupees ($1.5 billion), with a gross margin
of less than 67%, and spent 5.0 billion rupees or 9.5% of revenues on R&D.
Sources: Pfizer, Inc, Annual Report 2010; Ranbaxy Laboratories Ltd, Annual Report 2010;
Pharmaceutical Research and Manufacturers of America, “Drug discovery and development,”
www.phrma.org.
Monopoly 177
4. Demand and Cost Changes
How should a seller with market power respond to changes in demand and cost?
Generally, when there is a change in either demand or cost, the profit-maximizing
adjustment to sales depends on both the marginal revenue and marginal cost curves.
The seller should adjust the sales until its marginal revenue equals its marginal cost.
Demand Change
Suppose, for instance, that Venus Pharmaceutical increases advertising, and so
boosts demand for Gamma-1. How should Venus adjust its sales? To address this
question, Figure 8.2 shows the new demand curve. From the new demand curve,
we can calculate the new marginal revenue curve.
The new marginal revenue curve lies further to the right. Since the upward-sloping
marginal cost curve does not change, the new marginal revenue curve crosses the
marginal cost curve at a larger scale. Specifically, the two curves cross at a scale of
1.4 million units, and the new profit-maximizing price is higher, at $140.
Only the demand has changed. However, to identify the new profit-maximizing
scale and price, we need both the new marginal revenue and the original marginal
cost.
Marginal Cost Change

We can use a similar approach to understand how a seller with market power
should respond to a change in the marginal cost. Suppose, for instance, that,
new demand (marginal benefit)

Price ($ per unit)
140
f
marginal cost
71
new marginal
revenue
0 1.4
FIGURE 8.2 Demand increase.

Notes: The new marginal revenue curve and original marginal cost curve cross at the quantity of 1.4
million units a year. The new profit-maximizing price is $140.
178 8 Monopoly
demand (marginal benefit)

Price ($ per unit)
marginal
revenue
130
f marginal cost
71
51 new marginal
cost
0 1.4
FIGURE 8.3 Reduction in marginal cost.

Notes: The marginal revenue curve and new marginal cost curve cross at the quantity of 1.4 million
units a year. The new profit-maximizing price is $130.
relative to the data reported in Table 8.1, the marginal cost is $20 lower at all
scales of production. Should Venus reduce the price of the drug by $20 as well?
To address this question, consider Figure 8.3, which shows Venus’s marginal
revenue and the new marginal cost. The marginal revenue and the new marginal
cost cross at a larger scale of 1.4 million units. The new profit-maximizing price
is lower, at $130.
Relative to the original maximum (Figure 8.1), Venus now maximizes profits by
cutting its price by $10, which is less than the fall in the marginal cost. Further,
although the change was only in the marginal cost, Venus must consider the mar-
ginal revenue as well as the new marginal cost to obtain the new profit-maximizing
sales and price.
Fixed-Cost Change
We should stress that the seller’s profit-maximizing sales and price do not depend
in any way on the fixed cost (so long as it is not too large). Recall that a seller with
market power maximizes profit by producing at the scale where its marginal reve-
nue equals its marginal cost. Changes in the fixed cost will not affect the marginal
cost curve; hence, they will not affect the profit-maximizing sales level.
If, however, the fixed cost is so large that the total cost exceeds total revenue,
then the business should shut down. In Venus’s case, referring to Table 8.1, if the
fixed cost exceeds $84 million, the company should close.
The principle that the profit-maximizing sales and price do not depend on the
level of fixed costs is crucial in knowledge-intensive industries such as media and
publishing, pharmaceuticals, and software. In these industries, production is char-
acterized by relatively high fixed costs and low variable costs.
Monopoly 179
PROGRESS CHECK 8D
In Figure 8.3, show how Venus should adjust sales if marginal cost increases by
$20 at all scales of production.
5. Advertising
Any seller with market power can influence the demand for its products through
advertising. Advertising can shift out the demand curve as well as cause it to be
less price elastic. In the analysis of advertising, for simplicity, we take price as
given. However, in practice, it is important to bear in mind that the seller should
simultaneously maximize on price, advertising, and other influences on demand.
Profit-Maximizing Advertising
By shifting out the demand curve and causing it to be less price elastic, advertising
can raise sales. The increase in sales will affect total revenue and variable cost.
Accordingly, the benefit of advertising is to increase the profit contribution. To
maximize profit, Venus should advertise up to the level that the marginal benefit
of advertising equals the marginal cost.
Using the concepts of the incremental margin percentage and the advertising
elasticity of demand, we can derive a simple rule for the profit-maximizing level of
advertising. Recall from Chapter 3 that the advertising elasticity of demand is the
percentage by which the demand will change if the seller’s advertising expenditure
rises by 1%, other things equal.
Advertising–Sales Ratio
When the marginal benefit equals the marginal cost of adver-
Profit-maximizing
tising, the advertising–sales ratio (the ratio of the adver- advertising–sales
tising expenditure to sales revenue) equals the incremental ratio: Incremental margin
margin percentage multiplied by the advertising elasticity of percentage multiplied by
demand. This provides a simple rule for the profit-maximizing advertising elasticity of
level of advertising expenditure. demand.
Strictly, the rule stipulates the ratio of advertising expen-
diture to sales revenue. So the ratio should be called the advertising–revenue ratio.
However, in practice, it is usually called the advertising–sales ratio.
We can apply the advertising–sales ratio rule to determine the profit-maximizing
level of advertising expenditure for Venus’s new drug Gamma-1. Recall that, with
the demand and costs in Table 8.1, the profit-maximizing price is $135 per unit
and the marginal cost is $71. This means that the incremental margin percentage
is $(135 − 71)/$135 = 0.474. Suppose that, at the price of $135, the advertising
180 8 Monopoly
elasticity of the demand is 0.26. Then the profit-maximizing advertising–sales

ratio is 0.474 × 0.26 = 0.123, or 12.3%.
At the $135 price, Gamma-1 revenue is $135 × 1.3 million = $175.5 million.
Hence, the profit-maximizing advertising expenditure is 0.123 × $175.5 million =
$21.6 million.
The rule for advertising expenditures implies that, if the incremental margin
percentage is higher, then the seller should spend relatively more on advertising.
The reason is that each dollar of advertising produces relatively more benefit.
Accordingly, when the incremental margin percentage is higher, the seller should
increase advertising. This means that, whenever a seller raises its price or reduces
its marginal cost, it should also increase advertising expenditure. By contrast, if
a seller reduces its price or raises its marginal cost, it should reduce advertising
expenditure.
Further, the rule for advertising expenditures implies that, if either the adver-
tising elasticity of demand or the sales revenue is higher, then the seller should
spend relatively more on advertising. Essentially, a higher advertising elasticity
of demand or sales revenue means that the influence of advertising on buyer
demand is relatively greater. In these circumstances, it makes sense to advertise
more.
PROGRESS CHECK 8E
Suppose that the profit-maximizing scale of production for Gamma-1 is
1.3 million units. At that scale, the price is $135 per unit, the marginal cost
is $71, and the advertising elasticity of demand is 0.14. How much should
Venus spend on advertising?
PFIZER: MANAGING COMPETITION
Atorvastatin, marketed by Pfizer under the brand name Lipitor, reduces the
level of low-density lipoprotein cholesterol in the human body. In 2010, Lipitor
was Pfizer’s best-selling drug, with sales revenue of $10.7 billion, or 15.7% of
the company’s total revenue.
However, in November 2011, Ranbaxy Laboratories introduced a generic
version of atorvastatin. During a six-month period of generic exclusivity in the
United States, Ranbaxy expected to set price at between 70% and 80% of
the price of the patented Lipitor.
With Ranbaxy’s generic taking part of the demand for Lipitor, it would have
made sense for Pfizer to reduce its own production. Ranbaxy would draw
price-sensitive patients who had a relatively weak preference for Lipitor. High
sales of Lipitor could only be maintained if the remaining patients, with a
strong preference for Lipitor, formed a relatively large group.
Monopoly 181
In response to the entry of generic simvastatin, Pfizer mounted an

aggressive marketing campaign. Pfizer prepared to introduce its own generic
to compete with other generics. The cost of maintaining the production line is
fixed, so any additional profit contribution, whether from branded Lipitor or a
generic, adds to profit.
Sources: “Pfizer deal with Ranbaxy means a delay for generic form of Lipitor,” New York Times,
June 19, 2008; “The War over Lipitor,” CNN Money, May 6, 2011; “Atorvastatin,” Wikipedia
(accessed July 12, 2011).
6. Research and Development
A seller with market power can also influence demand through research and devel-
opment. Especially in knowledge-intensive industries, R&D drives the pipeline of
new products and refreshes existing products.
How much should a business invest in R&D? The principles are the same as for
advertising. R&D shifts out the demand curve and causes it to be less price elastic.
The benefit of R&D is to increase the profit contribution.
A simple rule for the profit-maximizing level of R&D expen-
R&D elasticity: The
diture applies the concept of the R&D elasticity. The R&D
percentage by which
elasticity of demand is the percentage by which demand will demand will change if the
change if the seller’s R&D increases by 1%. The R&D elastic- seller’s R&D increases
ity of demand depends on two factors: one is the effectiveness by 1%.
of R&D in generating new products and enhancing existing
products, and the other is the effect of new and enhanced products on demand.
The rule for profit maximization is to spend on R&D up to the level where the
R&D–sales ratio (the ratio of the R&D expenditure to sales
revenue) equals the incremental margin percentage multi- Profit-maximizing
R&D–sales ratio:
plied by the R&D elasticity of demand. Incremental margin
By this rule, when the incremental margin percentage is percentage multiplied by
higher (higher price or lower marginal cost), the seller should R&D elasticity of demand.
increase R&D expenditure relative to sales revenue. Con-
versely, when the incremental margin percentage is lower (lower price or higher
marginal cost), the seller should reduce R&D expenditure relative to sales revenue.
Further, if either the R&D elasticity of demand or the sales revenue is higher,
then the seller should increase R&D, and if either the R&D elasticity of demand
or the sales revenue is lower, then the seller should reduce R&D.
PROGRESS CHECK 8F
If price is higher while marginal cost is lower, how should R&D expenditure
be adjusted?
182 8 Monopoly
SPENDING ON ADVERTISING AND R&D
Table 8.2 presents the amounts that various manufacturers of foods, personal
consumer products, information technology (IT), and telecom equipment
spend on advertising and R&D. Manufacturers of foods and personal
consumer products spend relatively more on advertising, while producers of
IT and telecom equipment spend relatively more on R&D.
Table 8.2 Advertising and R&D, 2013
Industry/ Currency Sales Advertising Advertising– R&D R&D–

company revenue sales ratio sales ratio
Foods and personal consumer products

China RMB million 43,357 2,710 6.3% 57 0.1%
Mengniu
Proctor and US$ million 83,062 9,236 11.1% 2,023 2.4%
Gamble
Unilever € million 49,797 6,832 13.7% 1,040 2.1%
IT and telecom equipment
Apple US$ million 170,910 1,100 0.6% 4,475 2.6%
Microsoft US$ million 77,849 2,600 3.3% 10,411 13.4%
Nokia € million 12,709 n.a. 2,619 20.6%
ZTE RMB million 75,234 568 0.8% 7,384 9.8%
7. Market Structure
Monopoly, the case of a single seller, is one extreme of a range of market structures.
At the other extreme lies perfect competition, where there are numerous sellers,
each of whom is too small to affect market conditions. By comparing monopoly
with perfect competition, we can understand how production and price depend
on the competitive structure of the market.
Effects of Competition
Consider the market for trucking service between two cities. Assume that, in the
long run, the provision of the service involves no fixed cost and the marginal cost
is constant at 30 cents per pound of freight. Let us compare production and price
when the trucking industry is perfectly competitive and when there is a monopoly.
First, suppose that the industry is perfectly competitive. Since provision requires
only a constant marginal cost of 30 cents per pound, all truckers will be willing to
supply unlimited service at 30 cents per pound. Hence, the market supply will be
perfectly elastic at 30 cents per pound. Given the market demand, the supply will
Monopoly 183
(a) Perfect competition (b) Monopoly
Price (cents per pound)

Price (cents per pound)
60
marginal
cost
30 30
supply
demand demand
marginal revenue
0 10 0 5
Quantity (million pounds a year) Quantity (million pounds a year)
FIGURE 8.4 Market structure.

Notes:
(a) Under perfect competition, competition drives the market price down toward the long-run average
cost of 30 cents. Production is 10 million pounds a year.
(b) Under a monopoly, the monopoly restricts production below the competitive level and sets a
higher price of 60 cents to obtain larger profit.
balance demand at a price of 30 cents. Figure 8.4(a) illustrates the market equilib-
rium. The sales and production will be the quantity demanded at a price of 30 cents,
say, 10 million pounds a year. In equilibrium, each trucker earns zero profit.
Next, suppose that the trucking industry is a monopoly. The monopoly will pro-
duce at a scale that balances marginal revenue and the marginal cost of 30 cents.
Since the marginal revenue curve lies below the demand curve, marginal revenue
equals marginal cost at a quantity of less than 10 million pounds a year. Accord-
ingly, the monopoly will set the price above 30 cents. Suppose that the monopoly
price is 60 cents and sales are 5 million pounds a year. Figure 8.4(b) depicts the
monopoly price and sales. The monopoly will enjoy profits of $(0.60 − 0.30) × 5 =
$1.5 million a year.
The trucking example illustrates several general principles. First, a monopoly
restricts production below the competitive level and, in so doing, can extract a
relatively higher margin and thus larger profit. By contrast, competition drives
the market price down toward the long-run average cost and results in more pro-
duction. Further, the profit of a monopoly exceeds what would be the combined
profit of all the sellers if the same market were perfectly competitive.
Potential Competition
We have just seen that competition will push down the market price toward the
long-run average cost. It is worth emphasizing that, under specific conditions,
184 8 Monopoly
even potential competition suffices to keep the market price close to the long-run
average cost.
Consider a market in which sellers can enter and exit at no cost. Such a market
is called perfectly contestable. A monopoly in a perfectly
Perfectly contestable
contestable market cannot raise its price substantially above
market: Sellers can enter its long-run average cost.
and exit at no cost. To understand why, suppose that Jupiter Trucking has a
monopoly on freight transport between the two cities. How-
ever, other truckers can easily switch their trucks among routes. So, if Jupiter
raises its price above the long-run average cost, other truckers can profit by enter-
ing the route between the two cities. They would quickly enter to compete for a
share of the market. The resulting increase in supply would drive the market price
down toward the long-run average cost.
So, in a perfectly contestable market, even a monopoly cannot raise its price
substantially above the long-run average cost. The degree to which a market is
contestable depends on the extent of barriers to entry and exit. Earlier in this
chapter, we reviewed barriers to entry. To the extent that there are barriers to
entry, it will be more difficult for competing sellers to enter and, hence, it will be
easier for a monopoly to raise its price above the long-run average cost.
The degree to which a market is contestable also depends on the extent of
barriers to exit. Recall that, if Jupiter Trucking raises its price above long-run
average cost, it might attract other truckers to enter. These other truckers are
lured by the attraction of temporary profits, made possible by Jupiter’s rela-
tively high price. Once Jupiter lowers its price back toward long-run average
cost, these other truckers will leave. But their brief presence in the market will
have been profitable.
Now suppose that these other truckers must incur liquidation costs to exit the
market. When deciding whether to enter the market, these other truckers must
consider these exit costs. The higher are such exit costs, the less likely these other
truckers are to enter when Jupiter raises its price. This illustrates how barriers to
exit affect the degree to which a market is contestable.
Measuring Market Power

Having understood the potential effect of market power on production and price,
it is worth considering how to measure market power. In a perfectly competitive
market, every seller produces at a scale where its marginal cost equals the market
price; hence, its incremental margin percentage is zero. By contrast, a seller with
market power restricts sales to raise its price above its marginal cost. The more
inelastic is the market demand, the more the seller can raise its price above its
marginal cost.
We measure market power by the incremental margin percentage. This measure
can be used to compare market power across different markets. Some drugs cost
hundreds of dollars per dose, while others cost less than a dollar. It would not
Monopoly 185
make sense to directly compare their prices or incremental margins. Even if the
market for the expensive drugs were almost perfectly competitive, the price would
be a few dollars above the marginal cost. This difference would exceed the incre-
mental margin for the cheap drugs. Hence, to compare market power, it makes
more sense to use the incremental margin percentage.
The incremental margin percentage also captures the impact of potential com-
petition. If, owing to the presence of potential competitors, a monopoly sets a
price close to its marginal cost, then the incremental margin percentage will also
be relatively low.
VALUE OF A GAMBLING DUOPOLY
The state of Victoria, Australia, legalized slot machines in 1992. The state
government issued master licenses to Tattersall’s and Tabcorp for each to
operate 13,750 slot machines at clubs and hotels across the state. The state
also licensed Crown Casino for 2,500 machines in its casino. The master
licenses expired in 2012.
By 2010, Tattersall’s and Tabcorp operated 26,682 slot machines at
514 clubs and hotels. Revenues from slot machines, more properly called
“player loss,” were $2.6 billion, or more than half of all gambling revenues
in Victoria.
In April 2008, the government announced that it would not renew the master
licenses. Instead, it would open applications to individual clubs and hotels for
licenses valid for 10 years from 2012. The market value of Tattersall’s and
Tabcorp fell by A$2.8 billion in one day.
The government issued the new licenses in two rounds. In the first round,
236 clubs with existing slot machines bought licenses for 8,712 machines. The
license fee was a percentage of the historical revenue and averaged A$42,014.
In the second round, licenses for 4,838 machines in clubs and 13,750 in hotels
were auctioned for an average of A$14,134 and A$39,686, respectively. The
government received a total of A$980 million for the new licenses.
Source: Victorian Auditor-General, Allocation of Electronic Gaming Machine Entitlements,
Melbourne: Victorian Government Printer, June 2011.
8. Monopsony
A seller with market power will restrain sales to raise its price and so increase
profit. What about a buyer with market power? How would its business deci-
sions differ from those of a perfectly competitive buyer? For simplicity, we dis-
cuss a situation with a single buyer, that is, a monopsony. Since there are close
parallels between monopoly and monopsony, we will focus on the important
differences.
186 8 Monopoly
Benefit and Expenditure

Suppose that a key input into the production of Gamma-1 is an Indonesian herb.
Venus is the only buyer of this herb; hence, it is a monopsony. By contrast, many
growers produce the herb. Each grower is too small to affect market conditions,
so the supply of the herb is perfectly competitive.
Since the herb is a key input into Venus’s manufacturing process, it provides
a benefit that can be measured as the revenue generated less the costs of other
associated inputs. The herb, however, must be bought from Indonesian growers.
Venus’s expenditure is the market price of the herb multiplied by the quantity
purchased. Accordingly, Venus’s net benefit from the herb is its benefit less expen-
diture. We suppose that Venus’s objective is to maximize its net benefit.
At what quantity of purchases will Venus maximize its net benefit? Referring to
Figure 8.5, Venus’s benefit depends on the quantity of its purchases: we suppose
that the marginal benefit of a small quantity is very high and that the marginal
benefit falls with the scale of purchases.
Also referring to Figure 8.5, the supply curve shows, for every quantity, the
price at which competitive sellers will provide the herb. Equivalently, the supply
curve represents the monopsony’s average expenditure for every possible quantity
of purchases. Since the price must be higher to induce a greater quantity of sup-
ply, the average expenditure curve slopes upward.
The marginal expenditure is the change in expenditure
Marginal expenditure: resulting from an increase in purchases by one unit. For the
The change in expenditure average expenditure curve to slope upward, the marginal
resulting from an increase
expenditure curve must lie above the average expenditure
in purchases by one unit.
curve and slope upward more steeply.
marginal expenditure
x
400
Price ($ per ton)
supply (average
y expenditure =
350
marginal cost)
273
t z
marginal benefit
u
v
0 6,000 8,000
Quantity (tons a year)
FIGURE 8.5 Monopsony purchasing.

Notes: The marginal benefit and marginal expenditure curves cross at a quantity of 6,000 tons.
Reading from the supply curve, the price at that quantity is $273 per ton. The supply curve also shows
the marginal cost. At 6,000 tons, the marginal benefit exceeds the marginal cost.
Monopoly 187
Maximizing Net Benefit

We can now state the following rule: a buyer with market power will maximize its
net benefit by purchasing the quantity at which its marginal benefit equals its
marginal expenditure.
To explain this rule, consider a scale of purchases Net benefit maximum:
Purchase at a scale
where Venus’s marginal benefit exceeds its marginal
such that marginal
expenditure. Then, if Venus steps up purchases, its benefit benefit equals marginal
will increase by more than its expenditure; hence it will expenditure.
obtain a larger net benefit. By contrast, if marginal bene-
fit is less than marginal expenditure, Venus should reduce purchases: its benefit
will drop by less than its expenditure. Venus will exactly maximize net benefit
when it purchases the quantity where its marginal benefit balances its marginal
expenditure.
Referring to Figure 8.5, the quantity that maximizes net benefit is 6,000
tons. At that quantity, the price is $273 per ton. Notice that the price of $273
per ton is less than the buyer’s marginal benefit. By contrast, if the demand
side were competitive and the marginal benefit curve represented the mar-
ket demand, the equilibrium price would be $350 per ton and the quantity
would be 8,000 tons. This illustrates a general point: a monopsony restricts
purchases to get a lower price and increase its net benefit above the competi-
tive level.
PROGRESS CHECK 8G
In Figure 8.5, shade the area that represents Venus’s total expenditure on the
herb.
GAVI: CHARITABLE MONOPSONY
Established with a US$750 million pledge from the Bill & Melinda Gates
Foundation, the Global Alliance for Vaccines and Immunizations (Gavi) aims to
increase vaccination of children in the world’s poorest countries. By 2014,
Gavi had extended vaccination to 440 million children, saving 6 million
lives.
Gavi cleverly applies economic principles to carry out its mission. It consolidates
purchases of vaccines and uses its buying power to get lower prices. In addition,
it arranges large advance purchases to encourage manufacturers to keep open
production lines for proven vaccines and to stimulate R&D to develop new
vaccines.
Source: www.gavi.org (accessed December 1, 2014).
188 8 Monopoly
KEY TAKEAWAYS
• Gain market power by limiting competition and making demand less price
elastic.
• To maximize profit, produce if total revenue covers total cost, and produce at
the scale where marginal revenue equals marginal cost.
• When demand or costs change, adjust production to the scale where marginal
revenue equals marginal cost.
• To maximize profit, spend on advertising to the level where the advertising–sales
ratio equals the incremental margin percentage multiplied by the advertising
elasticity of demand.
• To maximize profit, spend on R&D to the level where the R&D–sales ratio equals
the incremental margin percentage multiplied by the R&D elasticity of demand.
• Sellers with market power restrict sales to raise margins and profit.
• Measure market power by the incremental margin percentage.
• To maximize profit, purchase at the scale where marginal benefit equals marginal
expenditure.
REVIEW QUESTIONS
1. By way of an example, explain how product differentiation contributes to market

power.
2. What are the major forms of intellectual property?
3. Explain how economies of scale can contribute to market power.
4. For a seller with market power, why is marginal revenue less than or equal to
price?
5. True or false? A seller with market power can either set the price and let the
market decide how much to buy, or set the quantity to sell and let the market
decide the price, but not set both price and quantity.
6. A software publisher has priced a new database program such that its marginal
revenue is more than its marginal cost. Advise the company how to raise its
profit.
7. Why should a seller take account of both the marginal revenue and the marginal
cost when considering how to adjust price following a change in costs?
8. Why should a seller take account of both the marginal revenue and the marginal
cost when considering how to adjust price following a change in demand?
9. The profit-maximizing price for a new electronic device is $100. At that price,
the advertising elasticity of demand is 0.1 and sales are 500,000 units a year.
The marginal cost of production is $40 per unit. How much should the publisher
spend on advertising?
10. For a medical device, the advertising to sales ratio exceeds the incremental
margin percentage multiplied by the advertising elasticity of demand. How can
the manufacturer increase profit?
11. What factors affect the R&D elasticity of demand?
12. Explain the profit-maximizing rule for R&D expenditure relative to sales
revenue.
Monopoly 189
13. For a monopoly, the incremental margin percentage would be infinite. True or
false?
14. A buyer with market power restrains its purchases to reduce the market price.
True or false?
15. Compare purchases and price with a monopsony and perfectly competitive
buyers.
1. Eli Lilly owns the patent to Xigris, which is the only approved drug for the treatment
of sepsis. Sepsis is a severe illness caused by a bacterial infection which may
lead to the failure of multiple organs. Bayer manufactures aspirin, which is not
covered by patent, and is one of several drugs that relieve the symptoms of the
common cold.
(a) Which company has relatively more market power: Eli Lilly over treatments
for sepsis, or Bayer over drugs for relieving the common cold? Explain
your answer.
(b) How is the difference between price and marginal revenue related to
the price elasticity of demand?
(c) Compare the incremental margin percentage at profit maximum for
Lilly’s Xigris and Bayer’s aspirin.
2. Table 8.1 describes the demand and costs for Venus Pharmaceutical’s Gamma-1
drug. Suppose that the costs are a fixed cost of $60 million and a constant
marginal cost of $50 per unit. The demand remains the same.
(a) Prepare a new table of revenues and costs according to the new data.
(b) What are the profit-maximizing scale of production and price?
(c) At that production scale, what are the marginal revenue, marginal cost,
and incremental margin percentage?
3. Apple outsources manufacturing of iPhones and iPads to Foxconn, with 800,000
workers at factories in Shenzhen, Chengdu, and elsewhere in China. Faced with
increased competition for labor, Foxconn has raised wages and increased benefits
for its workers. In April 2012, Samsung sued Apple for violating various Samsung
patents to produce mobile phones. (Sources: “Foxconn to raise salaries 20%
after suicides,” Financial Times, May 28, 2010; “Samsung sues Apple on patent-
infringement claims as legal dispute deepens,” Bloomberg, April 22, 2011.)
(a) Using a suitable diagram, explain how Apple should set the production
and price of the iPhone to maximize profit. (Hint: You are free to assume
any data necessary to draw the diagram.)
(b) Using your diagram in (a), explain how Apple should adjust its produc-
tion and price if Foxconn raises its prices for contract manufacturing.
(c) Suppose that Apple must pay Samsung a royalty on each mobile
device that it produces. How should Apple adjust its production and
price in response to the royalty?
(d) How would you change your answer to (b) if Apple must pay Samsung
a lump sum in damages rather than a royalty per unit produced?
190 8 Monopoly
Table 8.3 Google profit and loss, 2010–2013
$ millions 2013 2012 2011 2010
Revenue 59,825 50,175 37,905 29,321

Cost of goods sold 25,858 20,634 13,188 10,417
Gross profit 33,967 29,541 24,717 18,904
R&D 7,952 6,793 5,162 3,762
Selling, general, and 12,049 9,988 7,813 4,761
administrative expenses
Operating income 13,966 12,760 11,742 10,381
4. Atos Origin, Coca-Cola, Eastman Kodak, General Electric, John Hancock

Financial Services, Lenovo Group, McDonald’s, Panasonic, Samsung, and
Visa paid a total of $866 million to the International Olympic Committee to be
global sponsors for the years 2004–2008. The sponsorship period included
the 2006 Winter Olympics in Turin and the 2008 Summer Olympics in Beijing.
By contrast, total sponsorship for the years 2000–2004, including the Salt
Lake City winter games and the Athens summer games, amounted to $666
million. (Source: “For Olympic sponsors, it’s on to Beijing,” International Herald
Tribune, August 31, 2004.)
(a) Compare the benefit from Olympics sponsorship for global brands
such as Kodak and Samsung relative to regional and local brands.
(b) Considering the relative sizes of the Greek and Chinese consumer
markets, explain why sponsors paid more for the 2004–2008 Olympics
than the 2000–2004 Olympics.
(c) Atos Origin’s customers are primarily other businesses, while Lenovo’s
market is mainly within China. Compare the benefit from Olympic
sponsorship for these two companies with the benefit for other
sponsors.
5. At Google’s 2013 third quarter earnings call, analyst Carlos Kirjner remarked,
“the perception by people outside the company is that Google spends a material
amount in long-term R&D that will not generate revenue in the next two years or
so.” CEO Larry Page countered that, with respect to large innovations, “I think
you overestimate short-term and underestimate long-term.” He then pointed
to how Google had transformed the concept of self-driving cars from being
far-fetched to inevitable.
(a) Explain the formula for the profit-maximizing level of R&D relative to
sales revenue in terms of the R&D elasticity of demand and incremental
margin percentage.
(b) Interpret the discussion between analyst Carlos Kirjner and CEO Larry
Page in terms of the R&D elasticity of demand.
(c) Referring to Table 8.3, calculate the ratio of R&D to sales in the years
2010–2013.
(d) Approximate the incremental margin percentage by the ratio of gross
profit to revenue. Suppose that the R&D elasticity of demand in the
Monopoly 191
years 2010–2013 is 0.2. Calculate the R&D expenditure that would

have maximized Google’s profit in each of the years 2010–2013.
(e) Compare your calculations in (d) with Google’s actual R&D expenditure.
6. In 1992, the state of Victoria, Australia, issued 10-year master licenses to
Tattersall’s and Tabcorp for each to operate 13,750 slot machines at clubs and
hotels. Eventually, they set up 26,682 machines at 514 locations. In 2008, the
government decided to replace the master licenses with individual transferable
10-year licenses. The market value of Tattersall’s and Tabcorp fell by A$2.8
billion in one day. Subsequently, the government issued 27,290 new licenses
for fees totaling A$980 million. (Source: Victorian Auditor-General, Allocation
of Electronic Gaming Machine Entitlements, Melbourne: Victorian Government
Printer, June 2011.)
(a) Consider a club that has acquired one of the new 10-year licenses.
Using a suitable diagram, explain how the club should set the price of
gambling to maximize profit. (Hint: You are free to assume any data
necessary to draw the diagram.)
(b) How should the club take account of the once-only license fee in
its decisions: (i) whether to continue in business; and (ii) its scale of
operations? How does it matter that the license is transferable?
(c) Comparing the master and individual licensing systems, what effect
do you expect on the quantity and price of gambling?
7. The National Collegiate Athletic Association (NCAA) regulates competitive sports
at member colleges and universities. The NCAA restricts the pay of student
athletes (generally limited to the full cost of their education) and requires
student athletes to attend full-time programs of study.
(a) What market power does the NCAA have, and what are its source(s)?
(b) In April 2014, football players at Northwestern University voted in a secret
ballot on whether to form a union. Explain why the NCAA vigorously
opposed the vote.
(c) If the US government were to forbid the NCAA from restricting pay
to student athletes, how would that affect: (i) the earnings of student
athletes; and (ii) professional sports leagues?
8. Cricket is India’s top spectator sport. Under Indian law, private broadcasters
must share any coverage of Indian cricket matches with the national television
and radio broadcasters, Doordarshan and All India Radio. However, the law
does not require national broadcasters to share their cricket telecasts with
private channels. (Source: “DD may get a blank cheque,” Times of India,
August 13, 2004.)
(a) How would the Indian law affect the ability of a private television
channel to differentiate itself from Doordarshan?
(b) How would the law affect the amount that a private television channel
would bid for rights to broadcast Indian cricket matches?
(c) How would the law affect Doordarshan’s degree of market power
relative to: (i) television viewers; and (ii) the organizers of Indian cricket
matches?
192 8 Monopoly
(d) Some predicted that, owing to the law, only national broadcasters
would bid for rights to broadcast Indian cricket matches, and private
broadcasters would not bid. Do you agree?
9. Some automobile parts, such as batteries and tires, wear out with use and
must be replaced frequently. Suppliers of these parts sell their products both
as original equipment to auto manufacturers and as replacement parts to car
owners. By contrast, supplies of air bags and ignition systems sell mainly to
auto manufacturers.
(a) Assess the power of automobile manufacturers over suppliers of
(i) batteries and tires as compared to (ii) air bags and ignition systems.
(b) For products like batteries and tires, do you expect prices to be higher
in the original equipment market or the replacement market?
(c) Suppose that the supply of batteries is perfectly competitive. Using an
appropriate diagram, explain how an automobile manufacturer would
determine the quantity of batteries to buy.

Consider some good or service that your organization produces. If your
organization produced and sold 5% more, what would be the effect on revenue
and cost? If your organization produced and sold 5% less, what would be the
effect on revenue and cost? Is your organization maximizing profit?
Notes
1 The following discussion is based, in part, on: “Pfizer deal with Ranbaxy means a delay for
generic form of Lipitor,” New York Times, June 19, 2008; “The War over Lipitor,” CNN Money,
May 6, 2011; “Atorvastatin,” Wikipedia (accessed July 12, 2011).
2 Chapter 9 introduces the concept of price discrimination. With price discrimination, different
units are sold at different prices, so total revenue is not simply price multiplied by sales.
3 According to Table 8.1, Venus can earn the same profit of $33.2 million with a price of $136 and
sales of 1.28 million. This ambiguity is due to our considering increments of sales of 200,000
units. The rule presented in this section – choosing the scale of production where marginal
revenue equals marginal cost – is more precise, and essentially considers sales in infinitesimal
increments.
C H A P T E R
9
Pricing
LEARNING OBJECTIVES
• Apply uniform pricing.
• Appreciate how price discrimination can increase profit beyond uni-
form pricing.
• Understand complete price discrimination.
• Apply direct segment discrimination.
• Apply indirect segment discrimination.
• Appreciate the choice between alternative pricing policies.
1. Introduction
Founded in Australia in 1920 as Queensland and Northern Territory Air Services,

the Qantas Group is a leading international carrier. The Qantas Group operates
the Qantas and Jetstar airlines. In the year 2012–2013, the group earned A$13.7
billion in revenue and its seat factor (capacity utilization) was 79.3%. Its fleet com-
prised 312 aircraft, with an average age of 7.9 years.1
On January 6, 2015, passengers traveling in economy class on Qantas flight
QF401 from departing Sydney at 6 am to Melbourne who compared tickets might
have been surprised at the differences among their fares. The fully flexible fare
was A$600, the Flexi Saver fare was A$365, while the Red e-Deal fare was A$245.
If they had traveled half an hour later, at 6:30 am, they could have got the Red
e-Deal for just A$155.
Each fare is subject to different conditions. The fully flexible fare is fully refund-
able. The Flexi Saver is not refundable but allows changes up to the time of the
194 9 Pricing
flight, subject to fees. The Red e-Deal is not refundable but allows changes up to
the day before the flight, subject to fees.
Until April 2014, Qantas offered a 20% discount on flexible fares to seniors
aged 60 and over, and to children traveling with adults. However, the airline did
not allow any discount on Red e-Deal fares.
Pricing is crucial for Qantas, as for any airline. Travelers vary widely in the
amount that they are willing to pay for the same flight. Moreover, airline opera-
tions involve substantial fixed costs and relatively low variable costs. Consequently,
airlines have invested substantial amounts in reservations and yield management
systems to adjust fares and allocate seats.
Since variable costs are low, why doesn’t Qantas fill all its seats and earn more
profit? Why did it offer senior discounts? Why is the Red e-Deal cheaper for the
6:30 am flight? Why does Qantas set different conditions for changes in the Flexi
Saver and Red e-Deal fares?
This chapter ties threads from previous chapters on demand, elasticity, costs,
and monopoly to analyze how a seller with market power should set prices to
maximize profit. We first apply the price elasticity of demand and marginal cost.
Our rule will explain why Qantas doesn’t fill all its seats. Essentially, the reason
is that, to fill all the empty seats, it must cut fares to a point that would reduce
overall profit.
Next, we show how to increase profit by setting different prices that realize dif-
ferent margins from various market segments. This depends on the extent of the
seller’s information about the individual buyer’s demand. We consider situations
where the seller has complete information as well as those where the seller has
limited information.
Where the seller has sufficient information to directly identify consumer seg-
ments, it can apply direct segment discrimination. This explains why Qantas
offered discounts for seniors. By doing so, it might raise profit.
However, even if the seller lacks sufficient information or the ability to dis-
criminate directly, it might be able to apply indirect segment discrimination. This
explains why Qantas sets different prices for different flights, and imposes differ-
ent conditions for change in the Flexi Saver and Red e-Deal fares. By doing so, it
can earn more profit from travelers who prefer particular times, and from travelers
who might change their plans.
Any seller with market power can use the techniques of pricing presented in this
chapter to better achieve its objectives and, in particular, increase profit. Non-profits
can use pricing techniques to increase production and serve more customers.
2. Uniform Pricing
Whenever managers are asked why they do not set a higher price, the most fre-
quent response is: “Because I would lose sales.” This does not, however, answer
the question. Unless the demand is completely inelastic, a higher price will always
result in lower sales. The real issue is how the increase in price will affect the profit
Pricing 195
of the business. As we will show, the answer depends on the price elasticity of
demand and the marginal cost.
Here, we begin with the simplest pricing policy – uniform Uniform pricing: The
same price for every unit
pricing, where the seller charges the same price for every of the product.
unit of the product.
Price Elasticity
Suppose that Mercury Airlines offers air service and the cost of service comprises
a constant marginal cost of $80 per passenger. How should Mercury price the
service? For simplicity, we suppose that the profit contribution is large enough to
exceed the fixed cost, so Mercury should continue in operation.
Recall from Chapter 3 that demand is elastic if a 1% price increase causes the
quantity demanded to drop by more than 1%, and inelastic if a 1% price increase
causes the quantity demanded to fall by less than 1%. Generally, if the demand is
inelastic, an increase in price will lead to a higher profit. Accordingly, a seller that
faces an inelastic demand should raise the price.
Profit-Maximizing Price
Profit-maximizing
Indeed, the seller should raise price until the demand is elas- price: The incremental
tic. In the price elastic range, what price maximizes the sell- margin percentage equals
er’s profit? Chapter 8 identified the profit-maximizing sales the reciprocal of the
and price by the rule that marginal revenue equals marginal absolute value of the price
cost. Figure 9.1 shows the profit-maximizing sales and price elasticity of demand.
for Mercury Airlines.
400
Price ($ per seat)
240
80 marginal cost
marginal revenue demand

0 2,500 5,000
Quantity (seats a week)
FIGURE 9.1 Uniform pricing.

Note: At the profit maximizing quantity of sales, the marginal revenue equals the marginal cost.
Equivalently, the incremental marginal percentage equals the reciprocal of the absolute value of the
price elasticity of demand.
196 9 Pricing
Managers usually do not readily have information about marginal revenue. Typ-
ically, however, they have better information about the price elasticity of demand.
So, it is more convenient to have a pricing rule based on elasticity.
A rule equivalent to marginal revenue being equal to marginal cost is that the
incremental margin percentage equals the reciprocal of the absolute value of the
price elasticity of demand. So a seller maximizes profit by setting a price where
1
incremental margin percentage = − . (9.1)
price elasticity of demand
The price elasticity of demand is negative; hence, the minus sign on the right-hand
side of the pricing rule ensures that the entire right-hand side is positive.
Let us apply the rule to Mercury’s pricing. Suppose that the price elasticity of
demand is −1.5. Then, for Mercury to maximize its profit, the incremental margin
percentage must be 1/1.5 = 2/3. Representing the price by p, and recalling that the
marginal cost is $80, the rule implies that
p − 80 2
= . (9.2)
p 3
By solving this equation, we find that p = 240. Hence, the price that maximizes
Mercury’s profit is $240 (as Figure 9.1 shows, we get the same price if we look
for the sales where the marginal revenue equals the marginal cost). At the price
of $240, the quantity demanded is 2,500 seats per week. Hence, Mercury’s total
revenue is 240 × 2,500 = $600,000 a week. Mercury’s total cost is 80 × 2,500 =
$200,000. Thus, its profit contribution is $400,000 a week.
The price elasticity may vary along a demand curve. Further, the marginal cost
may change with the scale of production. Accordingly, determining the profit-
maximizing price involves a series of trials with different prices until a price is
found such that the incremental margin percentage equals the reciprocal of the
absolute value of the price elasticity.
Demand and Cost Changes

The pricing rule shows how a seller should adjust its price when there are changes
in the price elasticity of demand or marginal cost. Consider changes in the price
elasticity. If the demand is more elastic, then the price elasticity will be a larger
negative number. So, by the rule, the seller should aim for a lower incremental
margin percentage.
For instance, suppose that, in Mercury’s case, the price elasticity is −2 rather
than −1.5. Then the profit-maximizing incremental margin percentage will be
1/2 = 50%. Letting the price be p, we have (p − 80)/p = 0.50, which implies that the
profit-maximizing price is $160.
Pricing 197
By contrast, if the demand were less elastic, say, with an elasticity of −1.33, then
the profit-maximizing incremental margin percentage would be 1/1.33 = 75%. Again,
representing the price by p, we would then have (p − 80)/p = 0.75, which implies that
the profit-maximizing price would be $320.
Next, let us consider changes in the seller’s marginal cost. In our original exam-
ple, the price elasticity was −1.5, while the marginal cost was $80. Suppose that
the marginal cost is lower at $60. How should Mercury adjust its price? Using
the pricing rule, the profit-maximizing price must satisfy (p − 60)/p = 1/1.5, which
implies that p = 180. Notice that, although the marginal cost is $20 lower, the
profit-maximizing price is $60 lower.
Similarly, we can show that, if the marginal cost is higher, Mercury should not
raise its price by the same amount. The reason is that Mercury must consider the
effect of the price change on the quantity demanded.
These examples demonstrate that the way a seller should adjust its price to
changes in either the price elasticity or the marginal cost depends on both the
price elasticity and the marginal cost. In particular, this means that a seller
should not necessarily adjust the price by the same amount as a change in mar-
ginal cost.
PROGRESS CHECK 9A
In the case of Mercury Airlines, suppose that the price elasticity of demand
is −2, while the marginal cost is $70 per seat. Calculate the price that maxi-
mizes profit.
Common Misconceptions
A common mistake in pricing is to set the price by marking up average cost.
Cost-plus pricing is problematic. In businesses with economies of scale, the
average cost depends on the scale of production. So, to apply cost-plus pric-
ing, the seller must make an assumption about the scale. But the sales and
production scale depend on the price, hence cost-plus pricing leads to circular
reasoning.
Moreover, cost-plus pricing gives no guidance as to the appropriate mark-up
on average cost. Should a seller apply the same or different mark-ups to different
products? Suppose that a seller wants to set the mark-up to maximize profits.
Then the seller must go back to considering the price elasticity of demand and the
marginal cost.
Another common mistake is to believe that the profit-maximizing price depends
only on the price elasticity. To illustrate the correct approach, consider the mini-
bars provided by many hotels. The mini-bar has considerable market power, espe-
cially in the early hours of the morning, when it would be inconvenient if not
hazardous to venture out of the hotel for a beverage.
198 9 Pricing
Suppose that the price elasticities of demand for Heineken beer and Coca-Cola
in the mini-bar are the same. Should the hotel set the same price for both items?
Absolutely not. The hotel should set the same incremental margin percentage on
the two items. Since the marginal cost of Heineken beer is higher than that of
Coca-Cola, the hotel should set a higher price for the Heineken.
A common mistake in the pricing of services is to aim to fully utilize the avail-
able capacity. Capacity utilization is a useful metric of cost efficiency. However,
to achieve 100% utilization means increasing sales. In turn, with uniform pric-
ing, that means cutting the price and losing revenue on inframarginal buyers. So,
increasing capacity utilization may lead to lower profit!
PRICE ELASTICITY: WHO IS THE CUSTOMER?
Whenever a damaged car arrives at a repair shop, one of the first questions
that the car owner must answer is: “Do you have insurance for the damage?”
Why does the repair shop care whether the owner has insurance coverage?
Automobile repair is a case where there are two persons on the demand
side: the car owner who makes the buying decision and the insurer who pays
the bill. The owner of a damaged car will be relatively less sensitive to the
price of repairs. Indeed, the owner may ask the repair shop to fix some other
outstanding damage at the insurer’s expense.
Generally, demand is less sensitive to price whenever there is a split
between the party that makes the buying decision and the party that pays
the bill. Auto repair shops understand and exploit this split. However, the car
owner will be concerned about the price of repairs to the extent that his or her
future insurance premium or policy renewal depends on past claims.
3. Complete Price Discrimination
The previous section introduced a rule for uniform pricing: set the price so that
the incremental margin percentage equals the reciprocal of the absolute value of the
price elasticity of demand. In Figure 9.2, we illustrate the profit-maximizing uni-
form price for Mercury Airlines. On closer examination, however, we can show
that uniform pricing does not yield the maximum possible profit. This suggests
that we should look for better pricing policies.
Shortcomings of Uniform Pricing

Recall that the demand curve for a product also reflects the marginal benefits of the
various buyers. At the price of $240 per seat, the benefit of a flight for the marginal
buyer is just equal to the price. For all the other (inframarginal) buyers, who account
Pricing 199
400 a
gain in profit from complete

price discrimination
Price ($ per seat)
240 b
d
c marginal cost
80 g
e
marginal revenue
f demand
0 2,500 5,000
Quantity (seats a week)
FIGURE 9.2 Complete price discrimination.

Notes: With complete price discrimination, the seller prices each unit at the buyer’s benefit and sells
a quantity such that the marginal benefit equals the marginal cost. The increase in profit over uniform
pricing is the shaded area abd plus the shaded area bec.
for 2,499 seats, the benefit exceeds the price. Each of these inframarginal buyers
enjoys some buyer surplus. The market buyer surplus is the area abc in Figure 9.2.
With uniform pricing, the inframarginal buyers do not
pay as much as they would be willing to pay. This suggests Shortcomings of
that, by taking some of the buyer surplus, Mercury could uniform pricing:
increase its profit. • It does not extract the
Another shortcoming of uniform pricing is that it sells an entire buyer surplus.
economically inefficient quantity. (Recall from Chapter 6 on • It does not provide the
economic efficiency that the allocation of an item is economi- economically efficient
cally efficient if the marginal benefit equals the marginal cost.) quantity.
The marginal buyer derives a benefit of $240, while the
marginal cost is only $80. This economic inefficiency identifies an opportunity for
profit. By providing the service to everyone whose marginal benefit exceeds the
marginal cost, Mercury can increase profit.
PROGRESS CHECK 9B
What are the two shortcomings of uniform pricing?
Price Discrimination
Ideally, Mercury should sell each seat at the respective buyer’s benefit. Referring
to Figure 9.2, this would be like selling down the market demand curve. Then
Mercury would earn a higher incremental margin from buyers with higher benefit
and a smaller margin from buyers with lower benefit.
200 9 Pricing
Any pricing policy under which a seller sets prices to earn different incremental
margins on various units of the same or a similar product is called price discrimination.
(This term simply means setting different prices, and is not
Price discrimination: intended to have any negative connotation. An equivalent term
Pricing so as to earn
would be “price differentiation.”)
different incremental
margins on various units Complete price discrimination is the pricing policy
of the same or a similar which prices each unit at the buyer’s benefit and sells a quan-
product. tity such that the marginal benefit equals the marginal cost.
This policy is called “complete” because it charges every
Complete price buyer the maximum that they are willing to pay for each
discrimination: Pricing unit. Hence, the policy leaves each buyer with no surplus.
each unit at the buyer’s To illustrate complete price discrimination, consider Mer-
benefit and selling cury Airlines’ pricing of air service. Referring to Figure 9.2,
a quantity such that
marginal benefit equals
the demand curve is a straight line with a slope of −320/5,000 =
marginal cost. −0.064. This means that the first traveler is willing to pay 400
− 0.064 = $399.936 for a seat, the second traveler 400 − 2 ×
0.64 = $399.872, and so on. Hence, under complete price discrimination, Mercury
should charge the first traveler $399.936, the second traveler $399.872, and so on.
Mercury should not stop selling at the 2,500th seat. The reason is that the 2,501st
traveler derives a benefit of 240 − 0.064 = $239.936, which exceeds Mercury’s marginal
cost of $80. This means that Mercury can raise its profit by selling a seat to that traveler.
Indeed, Mercury should sell up to the quantity where the marginal benefit just
equals the marginal cost. Referring to Figure 9.2, this balance occurs at a quantity
of 5,000 seats a week. The 5,000th traveler is willing to pay exactly $80 for a seat,
which is Mercury’s marginal cost. If Mercury tried to sell beyond 5,000 seats, it
would make a loss on additional units. Under complete price discrimination, the
buyer of the 5,000th seat is the marginal buyer.
With complete price discrimination, Mercury’s total revenue is the area 0fca
under the demand curve from the quantity of 0 up to 5,000 seats a week. This area
is (400 + 80)/2 × 5,000 = $1.2 million a week. As for costs, Mercury’s total cost is
area 0fcg, which is 80 × 5,000 = $0.4 million a week. Hence, with complete price
discrimination, Mercury’s profit contribution is $800,000 a week.
By contrast, in the preceding section, we showed that Mercury’s maximum profit
contribution with uniform pricing would be $400,000 a week. So, Mercury earns a
higher profit with complete price discrimination than with uniform pricing.
Under a policy of complete price discrimination, the seller should sell each
unit for the benefit that it provides its buyer and sell the quantity where the buy-
er’s marginal benefit just equals the marginal cost. Complete price discrimination
resolves the two shortcomings of uniform pricing:
• By pricing each unit at the buyer’s benefit, the policy extracts all the buyer
surplus.
• It provides the economically efficient quantity; hence, it exploits all opportunity
for additional profit through increasing sales.
Pricing 201
In the example of Mercury Airlines, the policy of complete price discrimination

enables Mercury to extract higher prices for the 2,499 seats that would be inframar-
ginal under uniform pricing. This increase in profit contribution is represented by the
area adb in Figure 9.2. Second, with complete price discrimination, Mercury would sell
2,500 more seats. These additional seats raise the profit by the area bec in Figure 9.2.
The total increase in profit contribution is the sum of the areas adb and bec.
Economic Efficiency
We have given a motivation for complete price discrimination as a way to increase
profit. It is interesting to note that it also achieves economic efficiency. In this
sense, maximizing profit is aligned with the social goal of economic efficiency,
that is, allocating resources so that no person can be better off without making
another person worse off.
To the extent that price discrimination achieves economic efficiency, it is useful
for non-profit and government organizations such as hospitals, museums, and
universities. By applying price discrimination, they may be able to expand their
service to more people. For instance, by charging more to customers who are will-
ing to pay more, a non-profit can use the additional revenue to provide service to
poorer customers.
Information and Resale

Under complete price discrimination, the seller charges each buyer a different
price for each unit of the product. To implement complete price discrimination:
• The seller must know each potential buyer’s individual demand curve. It is not
enough to know the price elasticities of the individual demand curves. Rather,
the seller must know the entire individual demand curve of each potential
buyer.
• The seller must be able to prevent customers from buying at a low price and
reselling to others at a higher price.
Typically, it is more difficult to resell services, especially personal services, than

goods. For instance, it is more difficult to resell medical treatment than pharma-
ceuticals, and it is more difficult to resell tax planning advice than tax preparation
software. Accordingly, price discrimination is relatively more widespread in ser-
vices than goods and is especially common in personal services.
PROGRESS CHECK 9C
In Figure 9.2, the profit contribution from complete price discrimination
exceeds the profit contribution from uniform pricing by areas adb and bec.
Calculate the dollar values of these areas.
202 9 Pricing
DOES THE DOCTOR REALLY NEED TO KNOW

YOUR OCCUPATION?
Price discrimination is common in medical services. Doctors treat patients

on an individual basis. A doctor’s first step in treatment is always to record
the patient’s history. This routinely includes questions about the patient’s
occupation, employer, home address, and insurance coverage. This information
is very useful in gauging a patient’s ability and willingness to pay as well as the
patient’s health.
To the extent that a patient is paying her or his own bill, the doctor can use
this information to charge different prices to various patients for the same
treatment. The result is close to complete price discrimination. Indeed, the
healthcare scholars Victor Fuchs and Alan M. Garber remarked approvingly:
“Medicine has a long and generally honorable history of price discrimination.
Doctors have provided free or heavily discounted care to the needy, and
drug companies have charged lower prices to those less able to pay full
price.”
Source: Victor Fuchs and Alan M. Garber, “Medical innovation: promises & pitfalls,” Brookings
Review, Vol. 21, No. 1 (Winter 2003), pp. 44–48.
SELLING DOWN THE DEMAND CURVE:

SALESFORCE.COM
With broadband Internet service pervasive and reliable, software publishers are
mimicking automobile manufacturers and movie studios. Auto manufacturers
offer leases to drivers who do not wish to buy. Similarly, movie studios sell
videos to rental stores that cater to low-benefit viewers. Not every potential
customer derives enough benefit to justify outright purchase.
The same principle applies to software. Systems for customer relationship
management (CRM) can be very expensive. Prices are out of reach for many
businesses. Yet the marginal cost of software is low. Enter the application
service provider (ASP) to offer software “on tap” through the Internet. The
ASP charges by usage and so reaches down the demand curve to low-benefit
users.
Marc Benioff founded Salesforce.com in 1999 to provide CRM through
the Internet to clients of all sizes. By 2011, Salesforce.com served 97,700
customers worldwide, earning revenues of $1.78 billion and gross profit of
$1.33 billion. The company’s market value was $21.3 billion.
Sources: Salesforce.com; Yahoo! Finance
Pricing 203
4. Direct Segment Discrimination
To implement complete price discrimination, a seller must know the entire indi-
vidual demand curve of each potential buyer. What if the seller does not have so
much information? A seller without sufficient information to
price on an individual basis may still be able to discriminate Segment: A significant
cohesive group of buyers
among segments of buyers. A segment is a significant cohe- within a larger market.
sive group of buyers within a larger market.
Homogeneous Segments
Suppose that Mercury Airlines transports adults and seniors at the same marginal
cost of $80. Then Mercury can divide the market into two
segments according to age. We give the name direct segment Direct segment
discrimination:
discrimination to the policy of setting different incremental Pricing to earn different
margins for each identifiable segment. incremental margins from
In Mercury’s case, there are two identifiable segments: each identifiable segment.
adults and seniors. Suppose that adults are willing to pay
exactly $360 for travel, while seniors are willing to pay just
$90. The willingness to pay of both segments exceeds Mercury’s marginal cost,
which is $80. Accordingly, Mercury Airlines should set the regular adult fare
at $360 and the senior fare at $90. Mercury would earn incremental margins of
360 − 80 = $280 from each adult passenger, and 90 − 80 = $10 from each senior.
In this simple scenario, Mercury is able to achieve complete price discrimination
through direct segment discrimination. However, as discussed next, if the buyers
within each segment are heterogeneous and Mercury lacks sufficient informa-
tion to identify sub-segments, then direct segment discrimination will not achieve
complete price discrimination.
Heterogeneous Segments
What if adults differ in their willingness to pay, or seniors differ in their willing-
ness to pay? Then the profit-maximizing pricing policy depends on whether Mer-
cury can identify sub-segments within the broader segments of adults and seniors
and prevent resale within the sub-segments.
If Mercury does not have sufficient information to identify such sub-segments or
cannot prevent resale within segments, it has two choices for pricing within the adult
and senior segments. One is to apply uniform pricing within each segment. The other
is to apply indirect segment discrimination within each segment. Here, we focus on
within-segment uniform pricing, leaving indirect segment discrimination to Section 6.
Within each segment, applying the rule for uniform pricing, the price should be
such that the incremental margin percentage equals the reciprocal of the absolute
value of the price elasticity of demand.
204 9 Pricing
(a) Adults’ demand (b) Seniors’ demand
400
demand
Price ($ per seat)
Price ($ per seat)

200
240
marginal
revenue 140
demand
80 80
marginal cost marginal
cost
0 2,500 0 937 marginal revenue
Quantity (seats a week) Quantity (seats a week)
FIGURE 9.3 Direct segment discrimination.

Note: The demand for adult seats is relatively less elastic, so the seller should set a relatively higher
incremental margin percentage on adult seats.
Let the adults’ and seniors’ demands be as shown in Figure 9.3. Consider first
the demand from adults. Suppose that the profit-maximizing price is a. Through
a process of trial and error, we find that, at the price a, the price elasticity of
demand is −1.5. Accordingly, Mercury should set the price a so that the incre-
mental margin percentage is 1/1.5 = 67%. This means (a − 80)/a = 0.67; hence,
the price of an adult fare is a = $240. Suppose that, at this price, the quantity
demanded is 2,500 seats.
Next, consider the demand from seniors. Seniors derive lower marginal benefits;
hence, their demand curve is lower than that of adults. Suppose that the profit-
maximizing price is s. Through trial and error, we find that, at the price s, the price
elasticity of demand is −7/3. Hence, Mercury should set the senior fare so that the
incremental margin percentage is 3/7. Then (s − 80)/s = 3/7, which implies that the
senior fare is s = $140. Suppose that, at this price, the quantity demanded is 937 seats.
In this example, the adult demand is less elastic. Therefore, Mercury should set
a relatively higher incremental margin percentage on regular adult fares. Mercu-
ry’s profit contribution from the adult segment is (240 − 80) × 2,500 = $400,000.
Further, its profit contribution from the senior segment is (140 − 80) × 937 =
$56,220. So, its total profit contribution with direct segment discrimination is
$456,220 a week. By contrast, its profit contribution with complete price discrim-
ination is $800,000 a week, while its profit contribution with uniform pricing is
$400,000 a week.
Generally, in a policy of direct segment discrimination coupled with uniform
pricing within segments, the seller should set prices in the following way. Get a
relatively lower incremental margin percentage from the segment with the more
elastic demand and a relatively higher incremental margin percentage from the
segment with the less elastic demand.
Pricing 205
Implementation
To implement direct segment discrimination, the seller must identify and be able
to use some identifiable and fixed buyer characteristic that segments the market.
The characteristic must be fixed; otherwise, a buyer might switch segments to take
advantage of a lower price.
The demand for movies and theme parks varies with such buyer characteristics
as income, occupational status, and age. Movie theaters and theme parks cannot
observe a customer’s income, but they can check a customer’s age and whether a
customer is a student.
Age is a characteristic that fits the conditions for direct segment discrimination.
It is easy to identify and impossible to change. A middle-aged adult cannot buy a
senior citizen’s ticket. Accordingly, movie theaters and theme parks set prices so
as to extract lower margins from senior citizens and children. Assuming that the
marginal cost of serving all patrons is the same, the result is lower prices for senior
citizens and children and higher prices for middle-aged adults.
The other condition for direct segment discrimination is that the seller must be
able to prevent consumers getting lower prices from reselling to those targeted
with higher prices. Movie theaters mark their tickets as “not transferable.” To the
extent that theaters can prevent buyers from reselling tickets, they can discrimi-
nate effectively.
PROGRESS CHECK 9D
Referring to Figure 9.3, suppose that the marginal cost is $100. Use the graph
to illustrate the new prices for adult and senior fares.
HEINZ KETCHUP: “NOT FOR RETAIL SALE”
The market for ketchup comprises a retail consumer segment and an

institutional segment. Institutional customers include restaurants, catering
services, airlines, schools, and even prisons. Institutions order larger quantities
and often employ professional purchasing staff who aim to secure better deals.
Typically, the institutional demand is more price elastic than the retail demand.
Ketchup manufacturers supply retail consumers through supermarkets
and grocery stores. To the extent that manufacturers can prevent institutional
customers from reselling ketchup to retail stores, they can implement direct
segment discrimination. This means setting prices for a lower incremental
margin from institutional customers. It is no coincidence that bottles of Heinz
ketchup served in restaurants are marked “not for retail sale.”
206 9 Pricing
NATIONAL UNIVERSITY HOSPITAL: YOUR IDENTITY

CARD, PLEASE
The National University Hospital is a tertiary referral healthcare and teaching

facility that is affiliated with the National University of Singapore. Like other
public-sector hospitals, the National University Hospital charges different
prices to Singaporeans and foreigners.
Subsidized rates, which range from S$27 to S$31 for the first consultation
and from S$25 to S$28 for follow-up consultations, are available only to
Singapore citizens and permanent residents. The rates vary by the seniority
of the attending doctor. Foreigners must pay the unsubsidized, private rates,
which range from S$66.90 to S$85.95 for the first consultation and from
S$43.10 to S$62.14 for follow-up consultations.
5. Location
To the extent that a product is costly to transport and the seller can identify a
buyer’s location, the seller can discriminate on the basis of the buyer’s location.
Generally, there are two ways of pricing to buyers in different locations.
One way is to set a common price to all buyers that does not
FOB price: Does not include delivery. Such a price is called ex-works or free on
include the cost of
delivery to the buyer.
board (FOB). Each buyer pays the FOB price plus the cost of
delivery to its respective location. With FOB pricing, the dif-
ferences among the prices at various locations are exactly the
CF price: Includes the cost
differences in the costs of delivery to those locations.
of delivery to the buyer.
The alternative way of pricing to different locations is
to set prices that include delivery. The price is called cost
including freight (CF). With CF pricing, the differences among the prices at
various locations need not correspond to the differences in the cost of delivery to
the respective locations.
FOB and CF Pricing

To explain FOB and CF pricing (with uniform pricing in each market), suppose that
Jupiter Bikes sets a price of $350 for a racing bike in its domestic market and the
cost of freight to Japan is $30. Then the FOB price would be $350 + $30 = $380.
With an exchange rate of 100 to the dollar, the FOB price in Japanese currency
would be 38,000.
FOB pricing, however, ignores the differences between the price elasticities of
demand in the various markets. Suppose that Jupiter can implement direct seg-
ment discrimination across the two markets. Then Jupiter should apply the anal-
ysis presented in the previous section to set a CF price. The incremental margin
Pricing 207
percentage in Japan should equal the reciprocal of the absolute value of the price
elasticity of demand.
If Japanese demand is more elastic than domestic demand, then Jupiter should
set prices so that its incremental margin percentage is lower in Japan than in the
home market. A lower margin, however, does not necessarily mean a lower price,
because Jupiter’s marginal cost of supplying the Japanese market is higher owing
to the cost of freight.
By contrast, if the Japanese demand is less elastic than domestic demand, then
Jupiter should set prices so that its incremental margin percentage is higher in
Japan than in the domestic market. Taking into account the higher marginal cost
of supplying the Japanese market, this definitely means that Jupiter’s price in
Japan should exceed that in the domestic market.
With CF pricing, the difference in the prices between the two markets will sim-
ply be the result of the different incremental margin percentages and the different
marginal costs of supplying the two markets. In particular, the price difference
need not necessarily be the cost of freight from the domestic to the foreign market.
Earlier we showed that, generally, direct segment discrimination provides more
profit than uniform pricing. Similarly, CF pricing yields more profit than FOB
pricing – because it takes account of differences in the price elasticity of demand.
Parallel Imports
For discrimination by buyer’s location, the various buyers must not be able to
adjust location to take advantage of price differences. For most goods, the seller
can control only the location at which it sells the product and cannot directly
monitor the buyer’s location.
If the difference between the prices of a product in two markets exceeds the
transportation cost, retailers, and even consumers, might buy the item in one mar-
ket and ship it to another. Such parallel imports are a particular challenge for
lightweight high-price items such as pharmaceuticals and cosmetics.
Manufacturers deal with parallel imports in several ways. One is to customize
the product to fit the market. English consumers are less likely to buy medicines
labeled in Swahili. Another way is to limit sales to the sources of parallel imports.
Manufacturers estimate the potential demand in low-price markets and limit sales
to such markets accordingly. With durable goods, manufacturers may restrict war-
ranty service to the country of purchase. This would discourage consumers from
buying the good away from home.
PROGRESS CHECK 9E
Suppose that the price elasticity of demand for Jupiter bikes in Japan is −2.5
and the marginal cost including freight is ¥30,000. Calculate Jupiter’s CF price
in Japan. If Jupiter’s domestic price is $350, and the exchange rate is ¥100 to $1,
what is the difference between the Japanese and domestic prices?
208 9 Pricing
WALL STREET JOURNAL ASIA: PRINT AND

INTERNET EDITIONS
The Wall Street Journal Asia is printed daily in Hong Kong, Singapore, and
Tokyo for same-morning delivery. The newspaper charges widely varying
prices in the three cities. In December 2014, the prices for an 8-week
subscription were HK$310 ($40) in Hong Kong, ¥17,820 ($143) in Japan, and
S$72 ($56) in Singapore.
A business newspaper is particularly suited to discrimination by location.
Few executives would trade off a lower price for old news. Hence, the Wall
Street Journal Asia can maintain a price in Tokyo that is more than triple that
in Hong Kong.
By contrast, it is relatively difficult to pinpoint the location of an Internet
subscriber. Accordingly, the price of a digital subscription is the same $8 for
eight weeks in all three cities. Generally, retailers of items that are both sold
and delivered over the Internet have difficulty discriminating by location.
Source: Wall Street Journal Asia, www.wsj.com (accessed December 2, 2014).
6. Indirect Segment Discrimination
For a policy of direct segment discrimination, the seller must be able to identify
some fixed buyer characteristic that divides the market into segments with different
demand curves. A seller may know that specific segments have different demand
curves but cannot find a fixed characteristic with which to discriminate directly.
Under these circumstances, the seller may still be able to discriminate on price,
but indirectly.
Suppose that the demand for Mercury Airlines’ service comprises two segments –
business and leisure travelers. Each segment is homogeneous. As Table 9.1 shows,
the benefit of travel to business travelers is $501, and the benefit to leisure travelers
is $201.
Ideally, Mercury would apply direct segment discrimination – charging the
business travelers $500 and the leisure travelers $200. But how can it distinguish
business from leisure travelers? The airline’s check-in staff might ask every traveler
to declare their purpose of travel. Then every passenger would claim to be going on
vacation. Alternatively, Mercury’s staff might check whether passengers are wear-
ing business attire or casual clothes. But what if business travelers dress casually?
Structured Choice
Let us now consider another method of price discrimination, which uses an indirect
means to discriminate. Consider the travelers’ plans in more detail. Vacationers can
fix their plans well in advance. By contrast, business travelers may need to change
Pricing 209
Table 9.1 Indirect segment discrimination in air services
Benefit ($) Price ($) Marginal Profit contribution

cost ($) per traveler ($)
Business Leisure
travelers travelers
Unrestricted fare 501 201 500 80 420

Restricted fare 101 181 180 80 100
their plans to fit work requirements. So, business and leisure travelers are differen-
tially sensitive to fees for changes in flight bookings.
Mercury could offer two fares: a more expensive, unrestricted fare with no fees
for changes; and a cheaper, restricted fare with a fee and limitations on changes.
As Table 9.1 shows, the benefit of the restricted fare to business travelers is $101,
and the benefit to leisure travelers is $181. Both business and leisure travelers get
more benefit from unrestricted than restricted fares. However, business travelers
get relatively more benefit from unrestricted fares and relatively less benefit from
restricted fares.
Suppose, for instance, that Mercury prices the unrestricted fare at $500, and
the restricted fare at $180. How would the business travelers choose? They would
not buy the restricted fare, as the price exceeds their benefit. They would buy the
unrestricted fare, and get a buyer surplus of $501 − $500 = $1. As for the leisure
travelers, they would not buy the unrestricted fare, as the price exceeds their bene-
fit. They would buy the restricted fare, and get a buyer surplus of $181 − $180 = $1.
Table 9.1 shows the airline’s profit contribution from each traveler.
The two segments – business and leisure travelers – have different demand curves.
The airline, however, has no way to directly identify the segments. So instead it
structures a choice between unrestricted and restricted fares to exploit the differen-
tial sensitivity of business and leisure travelers to fees for
changes. We give the name indirect segment discrimination Indirect segment
discrimination:
to the policy of structuring a choice for buyers so as to earn Structuring choice for
different incremental margins from each segment. buyers to earn different
Indirect segment discrimination uses product attributes to incremental margins from
discriminate indirectly among the various buyer segments. each segment.
Essentially, the product attributes are a proxy for the buyer
attributes. To determine the profit-maximizing prices, the seller
must consider how buyers with different attributes substitute among the various
choices. Accordingly, the seller must not price any product in isolation. Rather, it
must set the prices of all products together.
Implementation
To implement indirect segment discrimination, the seller must control some vari-
able to which buyers in the various segments are differentially sensitive. The seller
210 9 Pricing
can then use this variable to structure a set of choices that will discriminate among
the segments.
Since indirect segment discrimination allows each buyer a choice of products,
the seller obviously cannot direct buyers to particular products. The other condi-
tion for indirect segment discrimination is that buyers must not be able to circum-
vent the discriminating variable.
Suppose, for instance, that Mercury were to allow travelers holding restricted
fares to change their flights without any fee. Then business travelers would switch
from unrestricted to restricted fares. Such switching would undermine the segment
discrimination. In practice, airlines strictly enforce the conditions of restricted
fares.
PROGRESS CHECK 9F
Referring to Table 9.1, suppose that the business traveler’s benefit from the
restricted fare is $401, and the leisure traveler’s benefit from the restricted fare
is $101. Calculate the business traveler’s buyer surplus from the unrestricted
and restricted fares, and the leisure traveler’s buyer surplus from the unre-
stricted and restricted fares. Can Mercury Airlines implement indirect segment
discrimination?
QANTAS: RED E-DEAL, FLEXI SAVER, OR FULLY FLEXIBLE?
Price discrimination is widespread in airline travel. Airlines are careful to

match passengers to tickets (indeed, with increases in airport security, the
government lends a helping hand). So, passengers cannot resell tickets to one
another.
Fares vary by day of the week and time of day, and by the changes allowed and
the fees for changes. On flight QF401 from Sydney to Melbourne, scheduled
for January 6, 2015, Qantas offered three fares – a fully flexible fare of A$600,
a Flexi Saver fare of A$365, and a Red e-Deal fare of A$245.
The more expensive the fare, the more flexibility it allows. The fully flexible
fare is fully refundable. The Flexi Saver is not refundable but allows changes
up to the time of the flight, subject to fees. The Red e-Deal is not refundable
but allows changes up to the day before the flight, subject to fees. Obviously,
the more flexible fares target business travelers, and the more restricted fares
are aimed at leisure travelers.
Source: Qantas Airlines.
Pricing 211
7. Selecting the Pricing Policy
Generally, the pricing policy that yields the most profit is complete price discrimi-
nation. This, however, also requires the most information. The next most profitable
pricing policy is direct market segmentation. With direct segment discrimination,
the seller discriminates directly on some fixed attributes of the buyer. The seller
must be able to identify each buyer segment and prevent one segment from buying
the product targeted at another segment.
The next most profitable pricing policy is indirect segment discrimination,
which works indirectly through product attributes rather than directly through
buyer attributes. Indirect segment discrimination is less profitable than direct seg-
ment discrimination for two reasons.
• Buyer benefit. To induce buyers with different attributes to choose different

products, the seller may resort to designing low-end products in a less appealing
way. So, the products provide less benefit to buyers. For instance, with indirect
discrimination, airlines deliberately impose unattractive conditions on restricted
fares. The main purpose is to make them unappealing to business travelers.
• Cost. Indirect discrimination may involve relatively higher costs. For instance,
consumer products manufacturers use cents-off coupons to indirectly
discriminate among consumers with different price elasticity. Coupons impose
costs on manufacturers, retailers, and consumers.
The least profitable pricing policy is uniform pricing. This involves no discrim-
ination at all. However, it is also the simplest, requiring the least information.
Table 9.2 ranks the various pricing policies in order of profitability and informa-
tion requirement.
Technology
The rapid development of information technology, especially the Internet, has had
profound but conflicting effects on price discrimination. Information technology
both facilitates and impedes price discrimination.
Table 9.2 Pricing policies: profitability and information
Profitability Policy Information and administration
lowest Uniform pricing lowest

Indirect segment discrimination
Direct segment discrimination
highest Complete price discrimination highest
212 9 Pricing
The explosion in consumer usage of the Internet has enabled marketers to collect
large volumes of detailed information about consumer preferences. In addition,
falling costs of computing power and storage have lowered the cost of storing,
analyzing, and applying the consumer information. So sellers can better design
and target offers to particular segments.
For instance, airlines and supermarkets apply information technology to man-
age loyalty programs, tracking consumer purchases and targeting members with
special offers. Online auctions enable sellers to fine-tune prices according to each
individual buyer’s willingness to pay. With computerized production and online
delivery, producers can create multiple versions of a product and target different
customer segments.
On the other hand, the falling costs of computing power and storage have also
promoted the growth of consumer-oriented search services. Continuously prowl-
ing the Internet, these services help consumers to compare products and prices,
thus identifying the best offer and circumventing price discrimination.
Cannibalization
We mentioned earlier that sellers may deliberately design low-end products in an
unappealing way. They fear cannibalization, that is, high-benefit segments buying
the item aimed at low-benefit segments. Cannibalization reduces the profitability
of indirect segment discrimination.
Cannibalization occurs when buyers switch from high
incremental margin products to low incremental margin prod-
Cannibalization:
Buyers switch from ucts. Some examples of cannibalization are business travelers
high incremental flying on restricted fares, high-income consumers redeeming
margin products to coupons, and wealthy families buying basic sedans rather
low incremental margin than high-end luxury cars.
products. The fundamental reason for cannibalization is that the seller
cannot discriminate directly, and hence must rely on a struc-
tured choice of products to discriminate indirectly. To the extent that the discrim-
inating variable does not perfectly separate the buyer segments, cannibalization
will occur.
There are several ways to mitigate cannibalization. One is to use product design.
Upgrading the high-margin item would make it relatively more attractive, and
hence, less likely to be cannibalized. Degrading the low-margin item would make
it less attractive, and thus less likely to cannibalize the demand for the high-mar-
gin item.
Further, the products can be designed with multiple discriminating variables.
For instance, airlines specify multiple conditions for restricted fares, including
minimum and maximum stay, limits on stopovers at intermediate destinations,
and penalties for cancelation or changes in itinerary. Each of these conditions
helps to reduce the degree to which the restricted fare would cannibalize the
demand for the unrestricted fare.
Pricing 213
Finally, cannibalization can be mitigated by controlling availability. Limiting

the availability of the low-margin item would make it less attractive. Airlines, for
instance, limit the number of seats allocated to lower fares.
PROGRESS CHECK 9G
Explain why indirect segment discrimination yields less profit than direct seg-
ment discrimination.
THE HERMITAGE: MUSEUM PRICING
Founded by Empress Catherine II (the Great), the Hermitage in St Petersburg,

Russia, is one of the world’s great art museums. Following the October
Revolution, the Soviet government opened the Hermitage to the public.
Subsequently, the Soviet government substantially expanded the museum
by incorporating the Winter Palace of the former imperial family and adding
works seized from private art collectors.
The Hermitage’s pricing policy combines direct and indirect segment
discrimination. General admission is free for all students and children and
Russian pensioners. The price is 350 roubles for Russian adults, and 400
roubles (equivalent to $6) for foreign adults.
Through its website, the Hermitage encourages visitors to buy tickets
through the Internet at a price of $17.95 and so “avoid a long line at the
ticketing office.” The Hermitage also offers free admission to all visitors on
the first Thursday of each month.
Source: Hermitage Museum, www.hermitagemuseum.org (accessed July 20, 2011).
KEY TAKEAWAYS
• To maximize profit with uniform pricing, set the price so that the incremental
margin percentage equals the reciprocal of the absolute value of price elasticity
of demand.
• Price discrimination can increase profit by taking buyer surplus and providing
a quantity closer to economically efficient.
• Complete price discrimination charges a different price for each unit of the
product.
• Direct segment discrimination sets prices to earn different incremental margins
from each segment.
• Indirect segment discrimination structures a choice for buyers to earn different
incremental margins from each segment.
• Location is one profitable basis for segment discrimination.
214 9 Pricing
• The ranking of pricing policies from most to least profitable is complete price
discrimination, direct segment discrimination, indirect segment discrimination,
and uniform pricing.
REVIEW QUESTIONS
1. Many supermarkets sell both branded and own-label merchandise. A

supermarket estimates that the demand for its own-label cola is less elastic
than the demand for Coca-Cola. Should it set a higher price for own-label cola?
2. Sol Electric manufactures low-energy light bulbs. The marginal production
cost is $2 per unit, while the price elasticity of demand is −1.25. With uniform
pricing, what is the profit-maximizing price?
3. Using a new manufacturing process, Saturn Tire has reduced the marginal
cost of a tire from $50 to $40. Should it reduce the selling price of a tire by $10?
4. Book publishers typically set prices by the number of pages multiplied by a
standard price per page. Comment on this pricing policy.
5. How does complete price discrimination increase profit as compared with
uniform pricing?
6. What conditions are necessary to implement complete price discrimination?
7. Give an example of direct segment discrimination. Discuss whether the
example meets the conditions for such discrimination.
8. Explain the difference between FOB and CF prices.
9. On which of the following products would it be easier to discriminate by the
buyer’s location: newspapers or scientific journals? Explain your answer.
10. Give an example of indirect segment discrimination. Discuss whether the
example meets the conditions for such discrimination.
11. Typically, car rental agencies charge much higher prices for gasoline than
nearby gas stations. Explain how this indirectly segments between drivers who
are paying for the rental themselves and those who are renting at the expense
of others.
12. How can consumer goods manufacturers use coupons to discriminate on price?
13. Rank the various pricing policies in terms of profit.
14. How does information technology affect a seller’s ability to discriminate on
price?
15. What is cannibalization and how can it be managed?
1. In November 2014, Morgan Stanley, Citigroup, Deutsche Bank, and JP Morgan

led the syndication of a three-year $300 million loan to Alibaba. The interest
rate on the loan was set at the London Interbank Offer Rate (Libor) plus a
spread of 52 basis points (0.52%). Banks source funds from demand, savings,
and time deposits, as well as the interbank market. However, Libor is relatively
higher than interest rates on deposits. (Source: “Alibaba smashes Asian bond
records,” IFR Asia, November 22, 2014.)
Pricing 215
Table 9.3 DEWA: electricity rates
Residential/commercial customers Industrial customers
Monthly usage (kWh) Fils per kWh Monthly usage (kWh) Fils per kWh
0−2,000 23 0−10,000 23
2,001−4,000 28 10,001 and above 38
4,001−6,000 32
6,001 and above 38
Source: Dubai Electricity and Water Authority.
(a) Does Libor reflect a typical bank’s average or marginal cost of

funds?
(b) For purposes of pricing, which is relevant – average or marginal cost?
(c) Explain the banks’ pricing policy in terms of the incremental margin
percentage and the price elasticity of demand.
2. Doctors routinely ask patients for personal information such as occupation,
employer, home address, and insurance coverage.
(a) How do the following factors affect the scope for price discrimination
in medical services? (i) Doctors treat patients on an individual basis
and it is physically impossible to transfer medical treatment from one
person to another. (ii) Characteristics such as occupation and home
address are quite fixed.
(b) Explain how, if doctors use price discrimination, they can treat more
patients than if they use uniform pricing.
3. Dubai Electricity and Water Authority (DEWA) supplies electricity and water.
In 2009, DEWA produced 30,056 GWh of electricity, of which 29.25% was
used by residential customers, 44.53% by commercial customers, and 8.45%
by industrial customers. Table 9.3 presents rates for electricity as of January
2011.
(a) When applying direct segment discrimination coupled with uniform
pricing within segments, how should a seller set prices?
(b) Assuming that DEWA sets prices to maximize profits, do the rates
suggest that industrial demand is more or less elastic than residential
and commercial demand?
(c) A necessary condition for price discrimination is no resale of the good.
In this regard, what is the challenge to DEWA?
(d) DEWA can inspect customer premises for technical reasons. Does this
help to address the challenge in (c)? Explain your answer.
4. Up to 2 million US consumers buy pharmaceuticals from online Canadian
pharmacies, where prices are substantially lower than in the United States. Monthly
sales reached a peak of $43.5 million in early 2004. Then US pharmaceutical
manufacturers restricted supplies to Canadian wholesalers that sold to online
pharmacies. In response, Universal Drugstore and other Canadian pharmacies
sourced drugs from wholesalers in Australia, New Zealand and Britain. (Source:
“Kinks in Canada drug pipeline,” New York Times, April 7, 2006.)
216 9 Pricing
(a) By considering price elasticities of demand, and production and shipping

costs, explain why US pharmaceutical manufacturers set higher prices
in the United States than in Australia and New Zealand.
(b) How do parallel imports affect US pharmaceutical manufacturers?
(c) Compare the problem of parallel imports for drugs administered by
medical professionals relative to other drugs.
5. Every year, Heinz sells 650 million bottles and over 13 billion packets of ketchup
worldwide (Source: H.J. Heinz Company). The demand side of the ketchup
market comprises a retail segment and an institutional segment. Institutions
order larger quantities and may employ professional purchasing staff. Retail
consumers are supplied through supermarkets and grocery stores.
(a) Explain why institutional demand for ketchup is likely to be more
price elastic than retail demand. How would Heinz like to apply direct
segment discrimination?
(b) If Heinz supplies both institutional customers and retail distribution
channels through wholesalers, explain how the wholesalers might
undermine direct segment discrimination.
(c) Compare the problem in (b) for ketchup in bottles as compared with
packets.
(d) Why does Heinz mark ketchup sold to restaurants “not for retail sale”?
6. In 2010, Google earned revenues of $29.3 billion, of which 96% derived from
advertising. Google’s AdWords is a service that places advertisements in
Google search pages and websites. In a continuous auction, Google scores
each advertisement by the advertiser’s maximum bid per click multiplied by the
click-through rate. Google displays the advertisements in decreasing order of
score. Advertisers pay for each click according to the score of the next highest
advertisement divided by their own click-through rate. (Source: Austin Rachlin,
“Introduction to the ad auction,” adwords.blogspot.com.)
(a) Explain Google’s AdWords auction in terms of complete price discrimination.
Discuss whether Google meets the conditions for such discrimination.
(b) Explain Google’s AdWords auction in terms of indirect segment
discrimination. What induces advertisers who value the advertising
space more to bid higher?
(c) Considering the consumer demand for Google search and Google
websites, explain why Google takes account of both the advertiser’s
bid and click-through rate in allocating advertising space.
7. The Chinese Visa Application Center in London charges applicants for visas
an application service fee and a visa fee. The application service fee is £35.25
for processing within four working days and £47 for processing within three
working days. The single-entry visa fee is £30, £65, and £20 for citizens of
the United Kingdom, United States, and other countries respectively. The
corresponding double-entry visa fees are £45, £65, and £30. (Source: Chinese
Visa Application Center, London, January 5, 2010.)
(a) Explain the Visa Center’s pricing policies in terms of direct segment
discrimination. Discuss whether issuance of visas meets the conditions
for such discrimination.
Pricing 217
(b) Explain the Visa Center’s pricing policies in terms of indirect segment
discrimination between tourists and business travelers. Discuss whether
issuance of visas meets the conditions for such discrimination.
(c) Should the Chinese government set the same visa fees in all foreign
countries?
8. Qantas sells tickets through conventional travel agents, online intermediaries,
its own telephone call center, and its own website. Typically, an airline’s cost
per transaction is lowest for bookings through its own website and highest for
bookings through conventional travel agents. Qantas’s cheapest fare, the Red
e-Deal, is only available for online booking.
(a) How does technology affect an airline’s ability to discriminate on price?
(b) Considering Qantas’s cost of bookings and travelers’ elasticity of demand,
explain why the airline offers the Red e-Deal only through online booking.
(c) For flight QF401 departing Sydney on January 6, 2015 at 6 am for
Melbourne, the Red e-Deal fare was A$245. For the next flight, departing at
6:30 am, the Red e-Deal fare was A$155. Explain the difference in pricing.
(d) Explain why Qantas limits the number of seats sold by the Red e-Deal.
9. Founded by Empress Catherine II (the Great), the Hermitage in St Petersburg,
Russia, is one of the world’s great art museums. General admission is free for
all students and children, and Russian pensioners. The price is 350 roubles for
Russian adults, and 400 roubles (equivalent to $6) for foreign adults. Through its
website, the Hermitage encourages visitors to buy tickets through the Internet at
a price of $17.95 and so “avoid a long line at the ticketing office.” The Hermitage
also offers free admission to all visitors on the first Thursday of each month.
(a) Explain how the Hermitage uses direct segment discrimination. Discuss
whether Hermitage pricing meets the conditions for such discrimination.
(b) Explain how the Hermitage uses indirect segment discrimination. Discuss
whether Hermitage pricing meets the conditions for such discrimination.
(c) Comment on the free admission on the first Thursday of each month
with regard to: (i) maximizing the number of visitors served for a given
budget; and (ii) cannibalization.

Consider your organization’s various lines of business. For each line of business,
consider the revenues and costs from alternative uses of the resources –
people, property, and funds. Is every line of business maximizing profit?
Note
1 The following discussion is based, in part, on Qantas Annual Report 2013 and “Qantas slugs
families, elderly on fares by cancelling discounts,” The Australian, April 30, 2014.
10 C H A P T E R
Strategic Thinking
LEARNING OBJECTIVES
• Appreciate strategic situations.
• Apply games in strategic form to situations with simultaneous
moves.
• Appreciate the use of randomization in competitive situations.
• Distinguish zero-sum and non-zero-sum games.
• Apply games in extensive form to situations with sequential moves.
• Plan strategic moves and conditional strategic moves, both threats
and promises.
• Appreciate strategy in repeated situations.
1. Introduction
The market for narrow-body medium-range commercial aircraft has been very
profitable for Airbus and Boeing. However, manufacturers from Brazil, Canada,
China, and Russia are poised to challenge Airbus’s A320 family and Boeing’s 737.
Indeed, at the Paris Air Show in June 2011, Jim Albaugh, CEO of Boeing Com-
mercial Airplanes, conceded: “The days of the duopoly with Airbus are over.”1
In 1969, Empresa Brasileira de Aeronáutica (Embraer) was founded to man-
ufacture military and agricultural aircraft. In 2000, it was privatized and listed
on the São Paulo and New York stock exchanges. Embraer is now a well-reputed
manufacturer of regional jets (aircraft with up to 120 seats). While Bombardier,
COMAC, and Irkut have committed to build large planes, Embraer hesitated. At
the Paris Air Show in June 2011, Frederico Curado, CEO, remarked: “Going up
against Boeing and Airbus in head-to-head competition is really tough, not only
Strategic Thinking 219
because of their size, but because of their existing product line and industrial
capacity .... They can have a very quick response and literally flood the market.”
The government of China established the Commercial Aircraft Corporation
of China (COMAC) to spearhead a national plan to produce commercial jets.
COMAC is owned by the state-owned Assets Supervision and Administration
Commission and various state-owned enterprises. In 2006, COMAC launched the
C919. In November 2010, COMAC announced 100 orders, mainly from Chinese
airlines. The C919 is scheduled to fly in 2015 and begin commercial deliveries
in 2018.
As national champion in aviation manufacturing, COMAC is the successor to
the Aviation Industry Corporation of China. COMAC is indirectly owned and
controlled by the government of China. By contrast, Embraer is a publicly listed
company.
How does the difference in ownership between COMAC and Embraer affect
the decisions of the two manufacturers? Why do airplane manufacturers heavily
publicize new orders, especially of new models under development? Given the
established positions of Airbus and Boeing, and the commitment to enter by
Bombardier, COMAC, and Irkut, should Embraer have proceeded?
The situation among Airbus, Boeing, Bombardier, and COMAC is strategic.
A strategic situation is one where the parties consider
interactions with one another in making decisions. Airbus Strategic situation:
Where the parties
and Boeing watch each other closely, and both pay great consider interactions with
attention to the plans of COMAC and Embraer, and vice one another in making
versa. A strategy is a plan for action in a strategic situation. decisions.
This chapter considers how to organize thinking about stra-
tegic decisions, choose among alternative strategies, and make Strategy: A plan for
more effective strategic decisions. The chapter is based on a action in a strategic
set of principles to guide strategic thinking called game theory. situation.
The first set of principles is the model of games in strategic
form, which applies to situations where parties choose strat-
egies at the same time. COMAC and Embraer can each apply a game in strategic
form in its decision whether to produce a narrow-body jet. The same model
applies to competition in markets with few sellers. It explains why competing sell-
ers tend to cut price, although collectively they could raise profit by avoiding price
competition.
The second set of principles is the model of games in extensive form, which
applies to situations where parties act in sequence. Applying a game in strategic
form, Embraer can appreciate that, through government support and the commit-
ment of orders for 100 planes, COMAC has established a first-mover advantage.
Managers can apply the same model to plan strategic moves and conditional stra-
tegic moves, both threats and promises.
The ideas and principles of game theory provide an effective guide to strategic
decision-making in many situations. Corporate financiers apply game theory in
leveraged buyouts and takeovers. Unions apply game theory in bargaining with
220 10 Strategic Thinking
employers. And, of course, game theory is useful to any business with market
power in deciding competitive strategy.
PROGRESS CHECK 10A

The nail manufacturing industry is almost perfectly competitive. Your com-
pany is deciding whether to manufacture nails. Explain why this decision is not
strategic.
2. Nash Equilibrium
To introduce a framework for situations where parties choose strategies at the

same time, consider the following example. Jupiter Gasoline and Saturn Fuel
operate adjacent gasoline stations. Each station independently decides its price.
Should Jupiter maintain or cut price? What about Saturn?
The situation between Jupiter and Saturn is clearly strategic. Consumers are
price sensitive and would switch to the station offering a lower price. Jupiter’s
sales and profit depend on Saturn’s price, and likewise, Saturn’s sales and profit
depend on Jupiter’s price. How should Jupiter act?
Let us try to clarify Jupiter’s position in the following way. Jupiter has a choice
of two strategies – maintain price or cut price. Likewise, Saturn has a choice of
the same two strategies. Accordingly, there are four possible outcomes: Jupiter
and Saturn both maintain price; Jupiter maintains price while Saturn cuts price;
Jupiter cuts price while Saturn maintains price; and both stations cut price.
We can gather this information in Table 10.1 as follows. Record Jupiter’s alter-
native strategies along the rows, and Saturn’s strategies along the columns. The
columns and rows delineate four cells, each representing one of the four possible
outcomes. In each cell, the first entry is Jupiter’s daily profit and the second entry
is Saturn’s daily profit. For instance, in the cell where both Jupiter and Saturn
maintain price, write “J: 1,000” to show that Jupiter’s profit would be $1,000, and
“S: 1,000” to show that Saturn’s profit would be $1,000. Similarly, in the cell where
Jupiter cuts price and Saturn maintains price, write “J: 1,300” to show that Jupiter’s
profit would be $1,300, and “S: 700” to show that Saturn’s
Game in strategic profit would be $700. Table 10.1 is called a game in strategic
form: Depicts
form. It helps to organize thinking about strategic decisions
one party’s strategies
in rows, other party’s that parties must take simultaneously.
strategies in columns, Let us use the game in strategic form to consider how
and consequences in Jupiter should act. First, look at the situation from Saturn’s
corresponding cells. position. If Jupiter maintains price, then Saturn will earn
$1,000 if it maintains price or $1,300 if it cuts price, so Saturn
prefers to cut price. Now, if Jupiter cuts price, then Saturn will earn $700 if it
maintains price or $800 if it cuts price, so, Saturn prefers to cut price.
Table 10.1 Gasoline stations: price war
Saturn
Maintain price Cut price
J: 1,000 J: 700
Maintain price
S: 1,000 S: 1,300
Jupiter
Cut price J: 1,300 J: 800
S: 700 S: 800
Hence, regardless of Jupiter’s move, Saturn should cut price. For Saturn, the
strategy of maintaining price is dominated by the strategy of cutting price. A strat-
egy is dominated if it generates worse consequences than
some other strategy, regardless of the other party’s choices. Dominated strategy:
Generates worse
It makes no sense to adopt a dominated strategy. consequences than
For Saturn, maintaining price is a dominated strategy. some other strategy,
Accordingly, Jupiter can guess that Saturn will cut price. in all circumstances.
Similarly, by studying the situation from Jupiter’s position,
it is easy to see that, for Jupiter also, maintaining price is a
dominated strategy. Hence, Jupiter also should cut price.
This situation is called the cartel’s dilemma. As Chapter 11 explains, a cartel is an
agreement to restrain competition. Both stations know that, if they maintain price,
then they can increase their profit. The snag, however, is that when each station
acts independently, it will cut price. The final outcome is that both stations lose
profit.
Definition
The pair of strategies – for Jupiter, to cut price, and for Saturn, to cut price – is the
obvious way for the two stations to act. Moreover, this pair of strategies is a stable
situation in the following sense. Even if Saturn knows that Jupiter will cut price,
Saturn’s best action is to cut price, so it will not change its strategy. Likewise, even
if Jupiter knew that Saturn would cut price, Jupiter would still cut price.
In a game in strategic form, a Nash equilibrium is a set
of strategies such that, given that the other parties choose Nash equilibrium:
their Nash equilibrium strategies, each party prefers its own Given that the other
Nash equilibrium strategy. In the cartel’s dilemma, the pair parties choose their Nash
equilibrium strategies,
of strategies in which both Jupiter and Saturn cut price is a
each party prefers its own
Nash equilibrium. Nash equilibrium strategy.
What justifies a Nash equilibrium as a reasonable way for
the relevant parties to act? In many typical strategic situa-
tions such as the cartel’s dilemma, the Nash equilibrium strategies seem the most
reasonable and obvious way to behave. By extension, this provides grounds for
believing that in other, less intuitive settings, the relevant parties should also act
according to Nash equilibrium strategies.
Solving Equilibrium – Formal Method

How should parties solve a game in strategic form for the Nash equilibrium? The
formal solution for a Nash equilibrium is, first, to rule out dominated strategies
and, next, to check all the remaining strategies, one at a time.
The cartel’s dilemma is easy to solve. First, rule out the dominated strategies:
for Jupiter, maintaining price is dominated, and for Saturn, maintaining price is
dominated. Then each station has only one strategy left – to cut price – so that
must be the equilibrium.
To illustrate the solution for Nash equilibrium in another setting, consider the
Battle of the Bismarck Sea in World War II.2 In late February 1943, the Japanese
commander Rear-Admiral Kimura assembled a convoy of 16 transport ships and
destroyers at the port of Rabaul. Kimura’s mission was to bring the convoy to Lae,
on the mainland of New Guinea. US Lieutenant-General Kenney, commander of
Allied Air Forces in the area, sought to intercept and destroy the Japanese convoy.
Kimura had to choose between sailing along a northern route through the
Bismarck Sea and a southern route. Meteorologists forecast that there would
be rain on the northern route, which would reduce visibility. The weather on the
southern route, however, would be fine.
Kenney had to decide the direction in which to concentrate his reconnaissance
aircraft. Once his aircraft spotted the Japanese convoy, Kenney would dispatch
his bombers. The dilemma for Kenney was that his best decision depended on
what he believed Kimura would do.
Table 10.2 represents the Battle of the Bismarck Sea in strategic form. In each
cell, the first entry represents the number of days of bombing that Kenney could
inflict on the Japanese, while the second entry represents the number of days of
bombing that Kimura would suffer (as a negative number).
Recall that the solution for a Nash equilibrium is, first, to rule out dominated
strategies and, next, to check all the remaining strategies, one at a time. First, look
at the situation from Kimura’s position. If Kenney goes north, then Kimura will
Table 10.2 Battle of the Bismarck Sea
Japan (Kimura)
North South
US: 2 US: 2
North
Japan: –2 Japan: –2
US (Kenney)
US: 1 US: 3
South
Japan: –1 Japan: –3
suffer 2 days of bombing whether he goes north or south, so Kimura is indifferent

between the strategies. If Kenney goes south, then Kimura will suffer 1 day of
bombing if he goes north or 3 days of bombing if he goes south, so Kimura pre-
fers to go north. Hence, regardless of Kenney’s strategy, Kimura should go north;
the south strategy is dominated by the north strategy.
Next, look at the situation from Kenney’s position. Knowing that Kimura
would go north, his choice is between north, which yields 2 days of bombing, and
south, which yields only 1 day of bombing, so Kenney should go north.
Indeed, on February 28, Admiral Kimura set sail on the northern route. On
March 2, General Kenney’s planes discovered the Japanese convoy and destroyed
it over two days of bombing.
Solving Equilibrium – Informal Method

A simple, informal method of finding a Nash equilibrium is to draw arrows
between the cells as follows. Suppose that Kenney chooses north, then for Kimura
draw a double-headed arrow between −2 in the top left-hand cell and −2 in the top
right-hand cell (this double-headed arrow represents Kimura being indifferent
between strategies). Next, suppose that Kenny chooses south, then for Kimura
draw an arrow from −3 in the bottom right-hand cell pointing toward −1 in the
bottom left-hand cell (this arrow represents Kimura preferring north).
Now, suppose that Kimura chooses north, then for Kenney draw an arrow
from 1 in the bottom left-hand cell to 2 in the top left-hand cell (this arrow rep-
resents Kenney preferring north). Finally, suppose that Kimura chooses south,
then for Kenney draw an arrow from 2 in the top right-hand cell pointing toward 3
in the bottom right-hand cell (this arrow represents Kenny preferring south).
Using this “arrow” technique, we can easily see if a strategy is dominated.
A strategy is dominated if the row or column corresponding to the strategy has
all the arrows pointing out. From Table 10.2, for Kimura, south is dominated.
The arrow technique also easily identifies an equilibrium. If there is a cell with all
arrows leading in, then the strategies marking that cell are a Nash equilibrium.
By the arrow technique, the Nash equilibrium is for Kenney to fly north and
Kimura to sail north.
Non-equilibrium Strategies
We have explained how to analyze a strategic situation using the concept of a
Nash equilibrium. Given that the other players choose their Nash equilibrium
strategies, each party’s best choice is its own Nash equilibrium strategy. But
what if some party does not follow its Nash equilibrium strategy? Then the
other parties may find it better to deviate from their respective Nash equilibrium
strategies.
For example, from Table 10.2, Kenney’s Nash equilibrium strategy is north,
while Kimura’s Nash equilibrium strategy is north. Suppose, however, that Kimura
decides, for some reason, to go south and that Kenney has this information. Then
Kenney can score three days of bombing and win a bigger victory by flying south.
Accordingly, in the Battle of the Bismarck Sea, if Kimura does not follow his
Nash equilibrium strategy, then it is better for Kenney to choose a strategy that is
not a Nash equilibrium strategy.
However, if the alternative to the Nash equilibrium strategy is a dominated
strategy, then the Nash equilibrium strategy is better, even if the other party
does not follow its Nash equilibrium strategy. For instance, suppose that Kenney
decides to fly south. Then Kimura should still go north. The alternative strategy,
south, is dominated.
PROGRESS CHECK 10B

In Table 10.1, use the arrow technique to identify the Nash equilibrium strategies.
EMBRAER: TO ENTER OR NOT TO ENTER
Boeing forecasts demand for 23,370 new narrow-body jets in the next 20 years.
Given the finite demand, Airbus CEO, Tom Enders, cautioned that there might
not be room for Airbus, Boeing, and four new manufacturers.
Focusing on COMAC and Embraer, suppose that COMAC would have
an advantage in selling to Chinese airlines, while both manufacturers would
compete on an equal basis for sales to other carriers. If both enter the market,
COMAC would lose $1 billion, while Embraer would lose $2 billion. If only
Embraer enters, it would earn $1 billion, while if only COMAC enters, it would
earn $2 billion.
Table 10.3 presents the game in strategic form. Drawing the arrows shows
two equilibria. In one equilibrium, COMAC enters and Embraer does not enter.
In the other equilibrium, Embraer enters and COMAC does not enter. So, the
game in strategic form is ambiguous about the actual outcome.
Table 10.3 Narrow-body jets: new entry
Embraer
Enter Do not enter
C: –1 C: 2
Enter
E: –2 E: 0
COMAC
C: 0 C: 0
Do not enter
E: 1 E: 0
Source: Richard Tortoriello, “Aerospace & defense,” Standard & Poor’s Industry Surveys,
February 10, 2011.
3. Randomized Strategies
When the various parties act strategically, it seems reasonable for them to choose
Nash equilibrium strategies. In some situations, however, there is no Nash equi-
librium of the type that we have been considering. To illustrate, suppose that
Table 10.4 represents the competitive situation between the two gasoline stations.
The change from the original scenario is that Jupiter has a segment of loyal cus-
tomers, so, if Saturn cuts price, Jupiter earns more by maintaining price.
By applying the arrow technique to Table 10.4, we can see that there is no Nash
equilibrium in pure strategies: no cell has all the arrows leading inward. A pure
strategy is one that does not involve randomization. In
Table 10.4, Jupiter has two pure strategies, price high and Pure strategy: Does not
involve randomization.
price low, and Saturn also has two pure strategies, price high
and price low.
Although there is no Nash equilibrium in pure strategies, there is another way for
Jupiter and Saturn to act. Essentially, Jupiter does not want Saturn to know or pre-
dict its price. One way in which Jupiter can keep Saturn in the dark is to randomize
the choice between maintaining and cutting price. If Jupiter randomizes its price,
the station itself will not know its price. Then, of course, Saturn will not know
either. Similarly, Saturn does not want Jupiter to know or predict its price. If Saturn
randomizes its price, Jupiter cannot guess or learn it.
With a randomized strategy, the party specifies a prob-
ability for each of the alternative pure strategies. It then Randomized strategy:
Choose each pure
adopts each pure strategy randomly according to the proba- strategy according to a
bilities. The probabilities must add up to 1. specified probability.
Nash Equilibrium in Randomized Strategies

Suppose that Jupiter adopts the following randomized strategy: maintain price
with probability 1/2 and cut price with probability 1/2. To implement this strategy,
Jupiter’s manager marks a coin “maintain price” on one side and “cut price” on
the other side, then gives the coin to the pump attendant. Jupiter then orders the
pump attendant to toss the coin and fix the price according to which side of the
coin faces up.
Table 10.4 Gasoline stations: price war (modified)
Saturn
Maintain price Cut price
J: 900 J: 1,000
Maintain price
Jupiter S: 900 S: 800
J: 1,300 J: 600
Cut price
S: 500 S: 600
Given that Jupiter has chosen this randomized strategy, how should Saturn
act? Referring to Table 10.4, let us calculate the expected consequence for
Saturn from maintaining price. Saturn’s profit would be $900 if Jupiter main-
tains price, and $500 if Jupiter cuts price. Hence, Saturn’s expected profit from
maintaining price is ($900 × 1/2) + ($500 × 1/2) = $700. Similarly, we can calcu-
late that, if Saturn cuts price, its expected profit would be ($800 × 1/2) + ($600 ×
1/2) = $700.
What should Saturn do? Since Saturn gets the same expected profit from its two
pure strategies, it is indifferent between the two. Accordingly, it would be willing
to randomize between them. Specifically, suppose that Saturn prices high with
probability 1/2.
How will Jupiter act? If Jupiter maintains price, its expected profit would be
($900 × 1/2) + ($1,000 × 1/2) = $950. Similarly, if Jupiter cuts price, its expected
profit would be ($1,300 × 1/2) + ($600 × 1/2) = $950. Therefore, given Saturn’s
strategy, Jupiter is indifferent between maintaining and cutting price.
A Nash equilibrium in randomized strategies is like a Nash equilibrium in
pure strategies: given that the other players choose their Nash equilibrium strate-
gies, each party’s best choice is its own Nash equilibrium strategy. The following
randomized strategies constitute a Nash equilibrium in the (modified) situation
between the gasoline stations: Jupiter prices high with probability 1/2 and Saturn
prices high with probability 1/2. The appendix at the end of this chapter shows
how to calculate the probabilities and thus solve the Nash equilibrium in random-
ized strategies.
Advantages of Randomization
Suppose that Jupiter has adopted the Nash equilibrium strategy of pricing high
with probability 1/2. Suppose further that Saturn has learned of Jupiter’s strat-
egy through a spy. How can Saturn exploit this information? The answer is that
it cannot – as we have calculated earlier, whether Saturn prices high or low, its
expected profit will be $700. Generally, whenever a party adopts a Nash equilib-
rium strategy, the other parties cannot benefit from learning the strategy.
Randomization is useful in competitive contexts. The advantage of randomiza-
tion comes from its being unpredictable. To implement the randomized strategy,
Jupiter must leave the direction of its pricing to the coin toss. Jupiter must not
make any conscious decision on pricing. If it chooses its price in a conscious way,
Saturn may be able to guess or learn Jupiter’s decision and act accordingly.
PROGRESS CHECK 10C

Referring to Table 10.4, suppose that Jupiter maintains price with probability
2/5. Calculate the expected consequences for Saturn if it: (a) maintains price;
and (b) cuts price.
SUPERMARKET PRICING: HIGH OR LOW?
Supermarkets are torn between pricing high to extract buyer surplus from
loyal consumers and pricing low to grab price-sensitive consumers. The
dilemma is compounded in the presence of competition. Cutting price may
not raise profit if a competing supermarket cuts price even lower.
The solution is to randomize the price discounts. Indeed, the dazzling array
of “specials” in a supermarket advertisement does appear like randomized
discounts. By discounting randomly, a supermarket can attract price-sensitive
consumers while preventing competitors from undercutting on price.
4. Competition or Coordination
In the Battle of the Bismarck Sea, described in Table 10.2, if we add the out-
comes for the United States and Japan in each cell, the sum is zero in every cell.
By contrast, in the price war between the gasoline stations, described in Table 10.1,
if we add the outcomes for Jupiter and Saturn in each cell, the sum varies from
1,600 to 2,000.
Strategic situations can be classified by the outcomes as either
Zero-sum game: One
zero-sum games or positive-sum games. A zero-sum game is party can be better off
one where one party can become better off only if another is only if another is worse off.
made worse off. If the outcomes for the parties add up to the
same number (whether negative, zero, or positive) in every cell
of the game in strategic form, then one party can become better Positive-sum game:
One party can be better
off only if another is made worse off. Accordingly, such a stra-
off without another being
tegic situation is a zero-sum game. A positive-sum game is worse off.
one where one party can become better off without another
being made worse off.
A zero-sum game characterizes the extreme of competition: there is no way
for all parties to become better off. By contrast, a positive-sum game involves at
least some element of coordination. For instance, in the gasoline station price war,
described in Table 10.1, both Jupiter and Saturn could agree that both maintain-
ing price is better than both cutting price. However, the challenge for them is to
enforce the agreement. Each station, acting independently, would cut the price.
Some situations involve elements of both competition and coordination. Con-
sider two TV stations, Channel Z and TV Delta, choosing between two time
slots for the evening news, 7:30 pm and 8:00 pm. Market research shows that the
demand for news peaks at 8:00 pm and is lower at 7:30 pm.
Table 10.5 depicts the game strategic form. Each station has two pure strategies:
broadcast at 7:30 pm or at 8:00 pm. Each cell presents the monthly profits of TV
Delta and Channel Z in millions of dollars. The situation is a positive-sum game.
Table 10.5 Scheduling evening news
Channel Z
7:30 pm 8:00 pm
D: 1 D: 3
7:30 pm
Z: 1 Z: 4
TV Delta
D: 4
8:00 pm D: 2.5
Z: 3 Z: 2.5
The total profit of the two TV stations is not a constant. It is largest when the
stations broadcast in different time slots.
While scheduling the evening news involves coordination, it also has an element
of competition. Both stations will be better off if they schedule their news at dif-
ferent times. But one station will benefit relatively more – the station that gets the
8:00 pm slot. Accordingly, there are elements of competition (for the 8:00 pm slot)
as well as coordination (choosing different slots).
Applying the arrows technique, we can identify two Nash equilibria in pure
strategies: Channel Z broadcasts at 8:00 pm and TV Delta at 7:30 pm, or Delta
broadcasts at 8:00 pm and Channel Z at 7:30 pm. There is also an equilibrium in
randomized strategies, where each station chooses 7:30 pm with probability 1/7
and 8:00 pm with probability 6/7.
PROGRESS CHECK 10D

Check whether the following are zero-sum games: (a) entry into the market
for narrow-body jets (Table 10.3); (b) modified gasoline price war (Table 10.4).
5. Sequencing
So far, we have focused on situations where the various parties move simultane-
ously. What if the parties move one at a time? To organize thinking about a stra-
tegic situation in which the parties act in sequence, we use the game in extensive
form. A game in extensive form explicitly depicts the
Game in extensive sequence of moves and the corresponding outcomes. It con-
form: Depicts the sists of nodes and branches: a node represents a point at
sequence of moves and which a party must choose a move, while the branches lead-
corresponding outcomes.
ing from a node represent the possible choices at the node.
Let us apply the game in extensive form to the schedul-
ing of the evening news where Channel Z can schedule its news before Delta. In
Figure 10.1, at the first node, A, Channel Z must choose between 7:30 pm
(the upper branch) and 8:00 pm (the lower branch). TV Delta has the next move.
7:30 pm Z:1, D:1

B
7:30 pm Delta
A 8:00 pm Z:3, D:4

Z
7:30 pm Z:4, D:3

C
8:00 pm
Delta
8:00 pm Z:2.5, D:2.5
FIGURE 10.1 Scheduling the evening news: extensive form.

Note: At node B, Delta will choose 8:00 pm. At node C, Delta will choose 7:30 pm. By backward
induction, at node A, Z will choose 8:00 pm.
Delta’s node depends on Z’s choice. If Z has chosen 7:30 pm, then Delta will be
at node B and must decide between the two branches of 7:30 pm and 8:00 pm. If
Z has chosen 8:00 pm, then Delta will be at node C and must decide between the
two branches of 7:30 pm and 8:00 pm.
The consequences for TV Delta of choosing 7:30 pm or 8:00 pm depend on Z’s
choice. At the end of each branch, we mark the profits to Z and Delta, respectively.
If Z chooses 7:30 pm and Delta also chooses 7:30 pm, they both earn $1 million.
If Z chooses 7:30 pm while Delta chooses 8:00 pm, then Z makes $3 million and
Delta earns $4 million. If Z chooses 8:00 pm while Delta chooses 7:30 pm, then Z
makes $4 million and Delta earns $3 million. Finally, if both stations choose the
peak time of 8:00 pm, they both earn $2.5 million.
Backward Induction
How should the two stations act? We solve the game in exten-
sive form by backward induction, which means looking Backward induction:
forward to the final nodes and reasoning backward toward Look forward to final nodes
and reason backward
the initial node. We can use this procedure to identify the
toward the initial node.
best strategies for the two stations. There are two final nodes:
B and C. At node B, TV Delta can choose 7:30 pm, which yields
$1 million, or 8:00 pm, which yields $4 million. Clearly, at node B, Delta would
choose 8:00 pm. Accordingly, we cancel the 7:30 pm branch. Now consider node C.
Here, Delta must choose between 7:30 pm, which yields $3 million, and 8:00 pm,
which yields $2.5 million. It will choose 7:30 pm, so we cancel the 8:00 pm branch.
Having determined how TV Delta will act at each of its two possible nodes,
B and C, we work back to consider the initial node, A. At node A, if Z chooses
7:30 pm, it can foresee that Delta will choose 8:00 pm, so Z will earn $3 mil-
lion. On the other hand, if Z chooses 8:00 pm, it can foresee that Delta will
choose 7:30 pm, so Z will earn $4 million. Therefore, Z should choose 8:00 pm.
Accordingly, in the scheduling of the evening news, when Z can move first, it will
choose the 8:00 pm slot, while Delta will take the 7:30 pm slot.
Equilibrium Strategy
In a game in extensive form, a party’s equilibrium strategy
Equilibrium strategy: consists of a sequence of its best actions, where each action
The sequence of best
is decided at the corresponding node. In the evening news,
actions, with each
action decided at the when Channel Z can move first, Z’s equilibrium strategy is to
corresponding node. choose the 8:00 pm slot, while TV Delta’s equilibrium strat-
egy is to take the 7:30 pm slot.
We have assumed that Channel Z moves before TV Delta.
If Delta can move first, then the game in extensive form will be like Figure 10.1,
except that Delta would schedule its broadcast time at node A, and Z would sched-
ule its broadcast time at node B or C. Then Delta’s equilibrium strategy would be
to take the 8:00 pm slot, and Z’s equilibrium strategy would be to settle for 7:30 pm.
To decide on a strategy in a situation where the parties move in sequence, the
basic principle is to look forward and anticipate the other party’s responses. So
when TV Delta can set its schedule before its competitor, Delta must look forward
and anticipate how Z will respond to each of Delta’s choices. In this way, Delta
anticipates that, if it chooses 7:30 pm, then Z would choose 8:00 pm, while if it
chooses 8:00 pm, then Z would choose 7:30 pm. By this procedure of backward
induction, Delta and Z can determine their equilibrium strategies.
Practically, what is the difference between the equilibrium strategy in a game in
extensive form and the Nash equilibrium strategy in a game in strategic form? In the
evening news, when the two stations move simultaneously, there are two Nash equi-
libria in pure strategies (plus a Nash equilibrium in randomized strategies). But
when the stations move in sequence, there is only one equilibrium. In the evening
news, the equilibrium in the extensive form is also a Nash equilibrium in the corre-
sponding strategic form. In other situations, however, the equilibrium in the exten-
sive form may not be a Nash equilibrium in the corresponding strategic form.
Accordingly, when analyzing a strategic situation, it is important to consider
carefully the structure of the moves: do the parties move simultaneously or
sequentially? The equilibria with simultaneous moves and with sequential moves
may be different.
First-Mover Advantage
In the evening news, the station that is first to commit its
First-mover advantage: schedule will make the larger profit. The first mover has the
The party moving first advantage. There is first-mover advantage in any strategic
gains an advantage. situation where a party gains advantage by moving before
others. To identify whether a strategic situation involves
first-mover advantage, analyze the game in extensive form.
The first mover has an advantage in situations of coordination and competi-

tion. For instance, in the evening news, both stations can agree to broadcast at
different times. But they cannot agree on who gets the 8:00 pm slot. The power of
the first mover is to determine the equilibrium.
First-mover advantage is a concept that is much emphasized in corporate strategy.
However, first-mover advantage is not a universal rule in strategic situations. Consider,
for instance, the launch of a new product category, such as driverless cars. The pioneer
must invest to develop infrastructure and educate consumers. Other manufacturers
can piggyback on the pioneer’s investment and introduce their products at lower cost.
Uncertain Consequences
In some situations, one party may not be certain about the consequences of the
various actions of the other party. The situation should still be analyzed by back-
ward induction, using all available information. Even if one party does not know
the consequences of the various actions for the other party, it may be able to
assess the probabilities with which the other party will choose between the alter-
native actions. It can apply backward induction using these probabilities.
For example, suppose that, in the battle for the evening news, Channel Z does
not know TV Delta’s profits, but Z can assess the probabilities of Delta’s actions
at each node. Figure 10.2 shows the specific probabilities.
Then Z can calculate as follows. At node B, Delta will choose 7:30 pm with
probability 1/3 and 8:00 pm with probability 2/3, hence Z’s expected profit would
be 1/3 × 1 + 2/3 × 3 = $2.33 million. At node C, Delta will choose 7:30 pm with
probability 1/2 and 8:00 pm with probability 1/2, hence Z’s expected profit would
be 1/2 × 4 + 1/2 × 2.5 = $3.25 million. Thus, at node A, looking forward, Z should
choose 8:00 pm.
probability 1/3
Z:1
7:30 pm
B
7:30 pm Delta
A 8:00 pm Z:3
Z
probability 1/2 Z:4
C 7:30 pm
8:00 pm
Delta
8:00 pm Z:2.5
FIGURE 10.2 Scheduling the evening news: uncertain consequences.

Notes: At node B, Delta will choose 7:30 pm with probability 1/3 and 8:00 pm with probability 2/3.
At node C, Delta will choose 7:30 pm with probability 1/2 and 8:00 pm with probability 1/2. By
backward induction at node A, Z will choose 8:00 pm.
PROGRESS CHECK 10E

In the gasoline station price war (Table 10.1), does the first mover have an
advantage? Why or why not?
COMAC: FIRST-MOVER ADVANTAGE
Bombardier, COMAC, Embraer, and Irkut have entered or are poised to enter
the market for narrow-body jets. Focusing on COMAC and Embraer, Table 10.3
represents the strategies and corresponding profits of the two manufacturers
if they act simultaneously.
What if COMAC moves first? Figure 10.3 shows the situation. If COMAC
enters, then, at node B, Embraer would not enter, and so COMAC would earn
$2 billion. If COMAC does not enter, then, at node C, Embraer would enter,
and so COMAC would earn zero. So, at node A, COMAC would choose to
enter. Likewise, if Embraer moves first, it would choose to enter.
Clearly, this is a situation of first-mover advantage. COMAC has been
marketing aggressively to sign up customers. If it can convince Embraer that
it is committed to production, it might be able to persuade Embraer not to
enter the market. Then COMAC would face one competitor fewer.
The government of China strongly supports COMAC. This government
backing may further help to convince Embraer (as well as Bombardier and Irkut)
that COMAC is committed to production, and persuade them to stay out. This
would leave the field to just three manufacturers – Airbus, Boeing, and COMAC.
COMAC: –$1 billion

Embraer: –$2 billion
B Enter
Embraer
Enter
COMAC: $2 billion
Do not enter
A Embraer: 0
COMAC
COMAC: 0
Embraer: $1 billion
Do not enter
C Enter
Embraer
Do not enter COMAC: 0

Embraer: 0
FIGURE 10.3 Narrow-body jets: new entry.

Note: At node B, Embraer would not enter, while at node C, Embraer would enter. Looking
forward from node A, COMAC would enter.
6. Strategic Move
Scheduling the evening news is a strategic situation of first-mover advantage.

Each station would like to move first and take the 8:00 pm slot, and have the other
station accept the 7:30 pm slot.
Now, suppose that TV Delta outsources the recording studio and other news
production facilities, and contracts to produce the news for broadcast at 8:00 pm.
With the outsourcing, TV Delta is simply not able to broadcast news at 7:30 pm.
Figure 10.4 illustrates the game in extensive form (modified from Figure 10.1).
We analyze by backward induction, beginning with the final nodes. At node B,
Delta will choose the 8:00 pm slot, and similarly, at node C, Delta will also choose
the 8:00 pm slot.
Hence, at node A, looking forward, Channel Z will reason that, whether it
chooses to broadcast at 7:30 pm or 8:00 pm, TV Delta will choose 8:00 pm. So,
Channel Z will earn more by broadcasting at 7:30 pm.
Thus, in equilibrium, Channel Z will broadcast at 7:30 pm, earning $3 million a
month, while TV Delta will broadcast at 8:00 pm, earning $4 million a month. TV
Delta’s outsourcing illustrates a strategic move. By outsourcing the news produc-
tion facilities, TV Delta reduces its profit from the 7:30 pm slot to zero, and
thereby changes the equilibrium of the strategic situation.
A strategic move is an action to influence the beliefs or
Strategic move: An
actions of other parties in a favorable way. In order to influ- action to influence beliefs
ence the other parties, the strategic move must be credible. or actions of other parties
Suppose that, instead of outsourcing the news production in a favorable way. To be
facilities, TV Delta simply announces that it will only broad- effective, a strategic move
cast news at 8:00 pm. Is such an announcement credible? No, must be credible.
because the announcement does not change the strategic
situation.
The situation would still be as depicted in Figure 10.1. In particular, at node C,
TV Delta would still choose 7:30 pm. So, at node A, Channel Z would still choose
8:00 pm.
7:30 pm Z:1, D:0

B
7:30 pm Delta
A
8:00 pm
Z Z:3, D:4
7:30 pm Z:4, D:0
C
8:00 pm
Delta
8:00 pm Z:2.5, D:2.5
FIGURE 10.4 Scheduling the evening news: TV Delta outsources production facilities.
Notes: At node B, Delta will choose 8:00 pm. At node C, Delta will choose 8:00 pm. Looking forward
from node A, Channel Z will choose 7:30 pm.
CSL: FREE TEXT MESSAGES
With over 6.2 million subscribers in a market of 17 million, Hong Kong Telecom
CSL is the largest provider of mobile telecommunications services in Hong
Kong. It skillfully uses discriminatory pricing to consolidate its market share
and lock out competitors.
All CSL mobile service plans include 10,000 units of free on-network text
messages. With its large market share, CSL’s offer of free “on-net” text
messages is a clever strategic move. This offer can draw customers away from
smaller competitors, and help to entrench CSL’s market position. Moreover, on
the cost side, a carrier incurs lower costs for “on-net” communications, as
compared with “off-net” communications that terminate on networks of other
carriers.
Sources: www.hkcsl.com (accessed December 4, 2014); London Economics, “HKT’s proposed
acquisition of CSL,” http://bit.ly/1F1CBEo, April 2014; “Keeping all in the family,” China Daily
Asia, May 16, 2014.
STRATEGIC MOVE IN WAR: DESTROY THE SHIPS
In 207 BC, Xiang Yu, Marquis of Lu, led a relatively small force to attack the Qin
Dynasty’s much larger army at Julu. Upon crossing the River Zhang, Xiang Yu
famously gave the order to sink all the boats and destroy the cooking pots. He
distributed just three days’ rations to the soldiers.
In 1519, the Spanish conquistador, Hernando Cortés, led a small expedition
from Cuba to conquer the Aztec Empire in Mexico. After landing at Veracruz,
some members tried to return to Cuba. Cortes ordered his carpenters to
dismantle the ships and use the timber and metal in the landward expedition.
By cutting off the means of retreat, both Xiang Yu and Cortés persuaded
(or compelled) their soldiers to fight harder. Both generals succeeded despite
their forces being vastly outnumbered. Xiang destroyed the Qin army and
Cortés conquered Mexico.
Sources: “Battle of Julu,” Wikipedia (accessed July 28, 2011); Bernal Díaz del Castillo, The
Discovery and Conquest of Mexico, 1517–21, transl. A.P. Maudslay, New York: Harper &
Brothers, 1928, pp. 168–169.
7. Conditional Strategic Move
Now suppose that, instead of outsourcing the news production facilities, TV Delta
contracts with a fast food chain to broadcast advertisements during the 8:00 pm news.
The advertising contract specifies that, if TV Delta fails to broadcast according
to contract, it must compensate the fast food chain with $2 million a month. This
means that if TV Delta schedules the news at 7:30 pm, it must pay compensation
of $2 million a month.
By working out the game in extensive form, we can see that, with the advertising
contract, the equilibrium is for Channel Z to broadcast at 7:30 pm, while TV Delta
broadcasts at 8:00 pm. TV Delta’s contract with the adver-
tiser is a conditional strategic move. A conditional strategic Conditional strategic
move: An action under
move is an action under specified conditions to influence the specified conditions to
beliefs or actions of other parties in a favorable way. influence the beliefs or
To the extent that a promised or threatened action need not actions of other parties in
actually be carried out, the conditional strategic move has no a favorable way.
cost. In equilibrium, TV Delta broadcasts news at 8:00 pm,
and so it need not pay compensation to the fast food chain.
A more accurate name for an action like TV Delta outsourcing the production
facilities is an unconditional strategic move because the action is not conditioned
on any eventuality. An unconditional strategic move usually involves a cost under
all circumstances. So, conditional strategic moves are more cost-effective than
unconditional strategic moves.
Conditional strategic moves take two forms – promises, which convey benefits;
and threats, which impose costs. TV Delta’s advertising contract with the fast food
chain is a promise. By the contract, TV Delta promises to pay $2 million a month
in compensation if it does not broadcast the advertisements in the 8:00 pm news.
PROGRESS CHECK 10F

Revise Figure 10.1 to illustrate TV Delta’s contract with the fast food chain.
What is the equilibrium?
Promise
A promise conveys benefits, under specified conditions, to
influence the beliefs or actions of other parties in a favorable Promise: Conveys
benefits, under specified
way. TV Delta’s advertising contract is a promise.
conditions, to change the
To take another example, consider deposit insurance. Gov- beliefs or actions of other
ernments are concerned about the stability of their banking parties in favorable way.
systems. Banks take deposits in checking and savings accounts
and lend the funds to short- and long-term borrowers. So,
generally, banks do not have enough cash at hand to repay all their depositors on
short notice and are vulnerable to bank runs.
Consider a typical depositor. Referring to Figure 10.5, at node A, she must
choose between maintaining her $100 savings with the bank or withdrawing the
deposit to get cash. If she maintains the deposit, the outcome depends on the
behavior of other depositors and is uncertain.
Bank solvent
Depositor: $101
(prob. 99%)
B
Maintains
deposit
A
Bank insolvent
Depositor: $0
Depositor (prob. 1%)
Withdraws
Depositor: $100
FIGURE 10.5 Bank deposit – without deposit insurance.

Note: Looking forward from node A, the consumer withdraws her deposit.
Bank solvent
Depositor: $101
(prob. 99%)
B
Maintains
deposit
A Bank insolvent
Depositor: $100
Depositor (prob. 1%)
Withdraws
Depositor: $100
FIGURE 10.6 Bank deposit – with deposit insurance.

Note: Looking forward from node A, the consumer maintains her deposit.
We represent the uncertainty by a circular node. With 99% probability, the bank
will remain solvent, and at the end of the year, the depositor will get $1 interest,
thus a total of $101. With 1% probability, the bank will suffer a run and become
insolvent, and at the end of the year, the depositor will get zero. Her expected
wealth from maintaining the deposit is $101 × 0.99 + 0 = $99.99.
However, at node A, if the depositor withdraws her funds from the bank, she
will get $100 with certainty, which is more than her wealth from maintaining the
deposit. So, she would choose to withdraw.
Now suppose that the government offers deposit insurance. Then, if the bank
cannot repay its customer’s deposits, then the government will pay. Deposit insur-
ance is a promise: the government pays only in the event that the bank cannot repay.
Figure 10.6 illustrates the strategic situation with deposit insurance. If the
depositor maintains her deposit and the bank is insolvent, she would get $100.
So, her expected wealth from maintaining the deposit is $101 × 0.99 + $100 × 0.01 =
$100.99, which is more than she would get by withdrawing. Hence, at node A, the
depositor would choose to maintain the deposit.
Accordingly, with deposit insurance, depositors will not withdraw their funds
when they hear rumors that their bank is in difficulties. Deposit insurance can
effectively prevent bank runs. If the government insures only solvent banks, then
government need never pay out any compensation. That is the beauty of a condi-
tional strategic move.
Threat
A threat imposes costs, under specified conditions, to influ-
ence the beliefs or actions of other parties in a favorable way. Threat: Imposes
costs, under specified
Threats are frequently used in negotiations. In bargaining
conditions, to change the
with employers, unions may threaten to strike if their mem- beliefs or actions of other
bers’ wages are not raised. Employers fear the disruption to parties in a favorable way.
operations resulting from a strike, so the threat of a strike
may persuade an employer to concede on wages.
The crucial issue is whether the union’s threat of a strike is credible. Credibility
is an issue because, during a strike, the workers must forgo part or all of their
wages. Figure 10.7 is a game in extensive form that depicts a union demanding a
$12 million increase in wages. The figure shows gains and losses for the workers
relative to the status quo.
At node B, the union decides whether to strike. If they strike, then the workers
would lose $4 million in current earnings and possibly gain $12 million in future
earnings. If they do not strike, they will get the status quo. The strike will reduce
the employer’s profit by $3 million.
At node A, the employer decides whether to raise wages. If the employer raises
wages, the higher wages would reduce its profit by $12 million. If it refuses to raise
wages, it possibly faces a strike.
However, the union’s threat to strike is only credible if, at node B, the workers
are better off with a strike. This depends on the probability that the employer
will eventually raise wages. Suppose that the probability that the employer will
Employer
refuses
Workers: Lose $4 million
increase
C Employer: Loses $3 million
B Strikes
Refuses
Union Employer Workers: Gain $8 million
increase
A raises Employer: Loses $15 million
wages
Employer Does not
strike Workers: Status quo
Employer: Status quo
Raises
wages Workers: Gain $12 million
Employer: Loses $12 million
FIGURE 10.7 Strike.

Notes: At node B, the union will strike or not depending on the probability that the employer raises
wages at node C. Looking forward from node A, the employer will raise or refuse to raise wages
depending on whether the union will strike.
eventually raise wages is 20%. Then the probability that the employer will not
raise wages is 80%. At node B, if the union strikes, the workers’ expected earnings
would be 8 × 0.2 + (−4) × 0.8 = −1.6, which is a loss of $1.6 million. So, the union’s
threat is not credible, and thus, the employer should refuse to raise wages.
The prevalence of strikes varies. In professional American football, strikes are
rare. The career of a professional American football player is relatively short. If
football players strike, they are trading off the loss of current earnings against
higher earnings in the future. As football players have short careers, they are not
so attracted by higher earnings in the future. Accordingly, football players are
relatively less likely to strike than other players.
PROGRESS CHECK 10G

Referring to Figure 10.7, what is the minimum probability that the employer
will eventually raise wages for the union’s threat of a strike to be credible?
EMBRAER: WORRIED BY AIRBUS AND BOEING
As of the Paris Air Show in June 2011, Embraer had not yet decided to
produce a narrow-body jet to compete with the Airbus A320 and Boeing 737.
The Embraer CEO, Frederico Curado, remarked: “Going up against Boeing
and Airbus in head-to-head competition is really tough . . . . They can have a
very quick response and literally flood the market.”
By contrast with Embraer, the incumbent manufacturers, Airbus and Boeing,
had already established production, benefiting from the experience curve.
Further, they could develop new versions of their existing planes at relatively
low cost, taking advantage of economies of scope from existing models.
Indeed, both Airbus and Boeing raised the rates of production of their existing
models. These increases in production possibly signaled further increases if
new competitors such as Embraer should enter the market. Hence Curado’s
worry about “flood[ing] the market.” If sufficiently credible, Airbus and Boeing
might succeed in scaring off the competitors.
Source: “Airbus-Boeing duopoly holds narrow-body startups at bay at Paris Air Show,”
Bloomberg, June 23, 2011.
CLOROX: POISON PILL
The Clorox Company is famous for its eponymous bleach. In February 2011,
the billionaire Carl Icahn announced that he had bought 12.5 million shares or
9.4% of Clorox’s 133 million shares. Then, on July 15, he offered to buy the
remaining shares for $76.50 each.
Icahn: Lose 24% on existing

Activate
holdings and 2 million
B rights
shares
Buy 2 million Clorox
shares
A Do not Icahn: Continue to
activate accumulate shares and take
Icahn over Clorox
Do not
buy
Icahn: Status quo
FIGURE 10.8 Poison pill.

Notes: At node B, Clorox will activate rights. Looking forward from node A, Icahn will not buy
2 million additional shares.
On July 18, Clorox’s board of directors adopted a “shareholder rights

plan” to ensure “fair and equal treatment” for all shareholders. The company
issued to each holder of a common share one right to buy another share
at half price. The rights could be exercised if any party acquired 10% or
more of Clorox, except that the rights of the acquiring party would be void.
Following the announcement, the price of Clorox’s common shares fell by
2.0% to $73.04.
Figure 10.8 depicts the strategic situation between Icahn and Clorox.
Suppose that, currently, the price of Clorox shares is $70.
At node A, Icahn must decide whether to acquire an additional 2 million
shares. At node B, suppose that Clorox’s board activates all rights except
Icahn’s, and all shareholders (except Icahn) buy additional shares at half
price, or $35 each. With the increase in shares outstanding, the value of each
share would fall to $53.37.3
Thus, the activation of the rights would reduce the value of Icahn’s existing
holding of 12.5 million shares and the additional 2 million shares by a
whopping 24%. It is no wonder that such shareholder rights plans have been
called poison pills. They hugely raise the cost of hostile takeovers. If the plans
succeed in deterring bidders, the rights will never be activated.
Sources: “Icahn bids for Clorox, suggests others step up,” Reuters, July 15, 2011; “The
Clorox Company adopts stockholder rights plan,” Clorox Company press release, July 18,
2011; Yahoo! Finance.
8. Repetition
So far, we have considered strategic interactions that take place only once. Many
strategic interactions, however, are repeated. Generally, the range of possible
strategies is much wider in repeated interactions than in one-shot scenarios. Spe-

cifically, a party may condition its action on the actions of the other party. Then
the parties may be able to achieve better outcomes than in once-only interactions.
Let us see how such strategies can work in the cartel’s dilemma. Referring to
Table 10.1, both Jupiter and Saturn know that, if both maintain price, each would
achieve higher profits than if both cut price. In a once-only situation, however,
each station has an overwhelming incentive to cut price.
What if the situation is repeated? Then Jupiter can adopt a strategy under which
it conditions its price on Saturn’s previous price, and vice versa. One such strategy is
“tit for tat” – I will begin by maintaining price and will continue until you cut price,
in which case, I will cut price for one week and thereafter will follow your price.
A tit-for-tat strategy combines a promise with a threat. The promise is to main-
tain price if the other station maintains price. The threat is to cut price if the other
station cuts price.
Is tit for tat an equilibrium strategy? Referring to Table 10.1, suppose that
both Jupiter and Saturn adopt a tit-for-tat strategy. Each begins by maintaining
price, with Jupiter and Saturn each earning $1,000. Suppose, however, that in
some week, Saturn cuts price, while Jupiter maintains price. Then Saturn will earn
$1,300 instead of $1,000, gaining $300.
In the following week, Jupiter would retaliate and cut price. What should
Saturn do? It could continue cutting price, and then Jupiter would continue
to retaliate. In this case, Saturn would earn $800 indefinitely. Alternatively, Sat-
urn could go back to maintaining the price, and so restore cooperation there-
after. In this case, Saturn would earn $700 for one week and $1,000 afterward.
Tit for tat is an equilibrium strategy for Saturn if the one-time $300 gain is
outweighed by the future losses due to Jupiter’s retaliation. This depends on two
factors. One is how Saturn values money in the future relative to the present,
which is the concept of discounting introduced in Chapter 1. The less Saturn values
money in the future, the more likely it is to cut price.
The other factor affecting whether tit for tat is an equilibrium strategy is the
time horizon. The shorter the time horizon, due, for instance, to expiry of the
leases of the stations and entry of new competitors, the more likely it is that
Saturn would cut price.
Accordingly, under certain conditions, tit for tat is an equilibrium strategy in
the repeated cartel’s dilemma. When competing sellers interact over an extended
period of time, they can maintain profit above the competitive level.
TIT FOR TAT WINS AGAIN

The political scientist, Robert Axelrod, invited various scholars to submit
computer programs specifying strategies for a prisoners’ dilemma (a game
like the cartel’s dilemma) repeated 200 times. Axelrod pitted each program
against the others in a round-robin tournament.
Programs that were “nice” in the sense of not being first to cheat performed
better than those that were not nice. The winning strategy was tit for tat.
Axelrod announced these results and held a second tournament. Despite the
opportunity to devise better strategies, no scholar could beat tit for tat, and
the simple strategy also won the second tournament.
Based on the tournament results, Axelrod proposed four simple rules:
• Do not strike first.

• Reciprocate both good and bad.
• Act simply and clearly.
• Do not be envious.
Source: Robert Axelrod, Evolution of Cooperation, New York: Basic Books, 1984.
KEY TAKEAWAYS
• A situation is strategic if the parties consider interactions with one another in

making decisions.
• Never use a dominated strategy.
• In a situation of simultaneous moves, a Nash equilibrium strategy is stable in
the sense that, if other parties choose their Nash equilibrium strategies, each
party prefers its own Nash equilibrium strategy.
• In competitive situations, it may help to randomize.
• Zero-sum games characterize extreme competition: one party can be better
off only if another is worse off.
• In a situation of sequential moves, plan by looking forward to the final nodes
and reasoning backward toward the initial node.
• Use strategic moves to influence the beliefs or actions of other parties in a
favorable way. To be effective, they must be credible.
• If possible, use conditional strategic moves, both threats and promises, as they
are more cost-effective than unconditional strategic moves.
• In repeated situations, get better outcomes through strategies that condition
actions on the actions of others.
REVIEW QUESTIONS
1. In situations of (a) perfect competition and (b) monopoly, does it matter whether
the seller acts strategically?
2. Explain why you should not use a dominated strategy.
3. Which of the following are reasons to adopt a Nash equilibrium strategy?
(a) I can minimize my expected loss.
(b) I can guarantee a minimum outcome.
(c) Even if the other party knows my strategy, it cannot take advantage of
that information.
4. If others do not act strategically (for instance, they use a non-equilibrium
strategy), should I follow?
5. Explain the meaning of a randomized strategy.
6. Some right-handed boxers also train themselves to box with their left hands.
Which of the following strategies will be more effective? (a) Throw a left-hand
punch after every three right-handers. (b) Box mainly with the right hand and
throw a left-hand punch at random.
7. Explain the associations between: (a) zero-sum game and competition; and
(b) positive-sum game and coordination.
8. In a game in strategic form, the consequences to the two players in every cell
add up to −10. Is this a zero-sum game?
9. In a game in extensive form, what is wrong with planning by reasoning forward
from the initial node?
10. In the Battle of the Bismarck Sea (Table 10.2), does the first mover have an
advantage? Why or why not?
11. Why are conditional strategic moves better than unconditional strategic
moves?
12. Suppose a bank offers deposit insurance to its depositors, with the
compensation to be paid from its own funds. Why is this not credible?
13. In bargaining, a common tactic is to “walk away.” Is this credible?
14. Loan sharks are not allowed to use the legal system to collect debts. Does this
explain why they employ violence?
15. How do strategies in repeated strategic situations differ from those that occur
only once?
1. Susan and Greg stole a car and have been caught by the police. Detective
Lenny Briscoe does not have sufficient evidence to convict them of auto theft.
He ushers Susan and Greg into separate interview rooms and offers each a
deal: “If the other suspect doesn’t confess and you do, we’ll give you a reward
of $1,000.” Each suspect knows that if neither confesses, they will be let off.
If one confesses while the other does not, then the confessing suspect will
receive the $1,000 reward, while the other will be jailed for one year. If both
confess, each will be jailed for one year.
(a) Construct a game in strategic form to analyze the choices of Susan
and Greg between confessing and not confessing.
(b) Identify the equilibrium/equilibria.
(c) Compare this situation to the cartel’s dilemma in Table 10.1.
2. The National Collegiate Athletic Association (NCAA) restricts the amount that
colleges and universities may pay their student athletes. Suppose that there
are just two colleges in the NCAA: Ivy and State. Each must choose between
paying athletes according to NCAA rules and paying more. If both Ivy and
State follow the NCAA salaries, then each would earn $3 million. If one follows
the NCAA salaries and the other pays more than NCAA salaries, then the
college paying more can attract better players and would earn $5 million, while
the college following NCAA would earn just $1 million. If both colleges pay
more than NCAA salaries, they would increase their costs but not get better
players, so both would earn $2 million.
(a) Construct a game in strategic form to analyze the choices of Ivy and
State, and identify the equilibrium/equilibria.
(b) With government backing, the NCAA can punish colleges that pay
more than the NCAA permitted salaries. How would this affect the
equilibrium/equilibria?
(c) Which of the following concepts best describes the NCAA rules on player
salaries: (i) monopoly; (ii) monopsony; (iii) economies of scale; (iv)
economies of scope. Explain your answer.
3. Launched in June 2007, Apple’s iPhone quickly grabbed a substantial share of
the global smart phone market. Competing mobile operating systems include
Nokia’s Symbian, Research-in-Motion’s Blackberry, Google’s Android, and
Microsoft’s Windows Mobile. In September 2010, soon after joining Nokia as
CEO, Stephen Elop reviewed Nokia’s strategy for smart phones.
(a) Consider the strategic situation between Nokia and HTC, a Taiwanese
manufacturer. If they both build phones on Symbian, HTC would
earn $2 billion while Nokia would earn $4 billion. If they both build
phones based on Windows Mobile, both HTC and Nokia would earn
$2 billion. If HTC builds on Symbian while Nokia builds on Windows
Mobile, each would earn zero. If HTC builds on Windows Mobile while
Nokia builds on Symbian, HTC would earn zero and Nokia would earn
$2 billion. Draw the game in strategic form and identify the equilibrium/
equilibria.
(b) Use your answer in (a) to explain why Microsoft would offer to pay
Nokia “billions of dollars” to build phones on Windows Mobile.
(c) Which of the following apply to the game in (a): (i) zero-sum game;
(ii) positive-sum game; (iii) economies of scale; (iv) economies of
scope. Explain your answer.
4. On May 19, 2012, the UEFA Champions League final between Bayern Munich
of Germany and Chelsea of England was decided by a penalty shoot-out.
Manuel Neuer took Bayern’s third kick. Chelsea goalkeeper Petr Čech guessed
correctly that Neuer would kick to the left but the low shot crept past into the
bottom left corner.
(a) Construct a game in strategic form as follows. Neuer must choose
between kicking to the left, center, or right. Čech must choose between
moving to block on the left, center, or right. If Neuer scores, Bayern
wins one point while Chelsea loses one point. If Čech succeeds in
blocking, both teams get zero.
(b) Is this a zero or positive-sum game?
(c) Find the Nash equilibrium/equilibria in pure strategies, if they exist.
(d) Verify that the following is a Nash equilibrium in randomized strategies:

each player plays left/centre/right with probability 1/3.
5. Pluto Limited is financed with $2 billion of debt at an interest rate of 10% and
$8 billion of equity. If management works as usual, Pluto would earn revenues
of $10 billion and incur operating expenses of $9.5 billion, and so operating
income would be $0.5 billion. If management works hard to cut costs, operating
expenses would be $9.0 billion.
(a) Construct a game in extensive form with the following nodes. At the
first node, a private-equity fund chooses between a leveraged buyout
of Pluto and the status quo. At the subsequent node, management
chooses between working as usual and cutting costs.
(b) Suppose that the leveraged buyout recapitalizes Pluto to $8 billion of
debt at an interest rate of 10% and $2 billion of equity. Calculate Pluto’s
profit if management: (i) works as usual; (ii) cuts costs.
(c) Explain how the leveraged buyout serves as a strategic move to cut
costs.
6. A major problem for China’s state-controlled banks is loans to state-owned
enterprises. Historically, many state-owned enterprises were financed with
relatively little equity and large loans from state-controlled banks. On a strictly
commercial basis, these loans should be classified as “non-performing.” Banks
should make provision for non-performing loans and, accordingly, reduce their
profit and the book value of their assets. Banks are run by managers who care
mainly about their job security.
(a) Use a game in extensive form to depict the following scenario. Renmin
Construction has borrowed 1 billion yuan at an interest rate of 5%
from People’s Bank. The loan is about to mature. The management
of People’s Bank must choose between rolling over the loan and
demanding repayment. (i) If the bank rolls over the loan, Renmin
Construction would continue in business and use part of the loan to
pay the interest due to the bank. The loan is classified as performing
and the bank management continues to be employed. (ii) If the bank
demands repayment, Renmin Construction would default and the
bank would have to write off 1 billion yuan. The government would
replace the management.
(b) Identify the equilibrium/equilibria of the game in (a).
(c) Suppose that Renmin Construction is actually bankrupt and has
no chance of ever making a profit. If the bank management aims to
maximize profit, should it roll over the loan or demand repayment?
7. The government of Bulgaria has pegged its currency, the lev, at a rate of
1.95583 levs to one euro. The central bank is committed to exchanging
euros for levs at this rate. In December 2014, the central bank had issued
currency in circulation with face value of 10.17 billion levs, and Bulgaria’s
foreign exchange reserves totaled 35.0 billion levs. Other things being equal,
Bulgarians prefer to hold levs as they are more convenient for daily use than
euros.
(a) Construct a game in extensive form with the following nodes: at the first
node, the typical Bulgarian chooses between redeeming levs for euros
and not redeeming; at the following nodes, the central bank either has
sufficient assets to meet the redemptions and remains solvent, or has
insufficient assets and becomes insolvent.
(b) As of December 2014, would the typical Bulgarian redeem her levs?
(c) Suppose, however, that the central bank increases the amount of
currency in circulation to 40 billion levs. Explain how the typical
Bulgarian’s decision whether to redeem depends on her beliefs about
the decisions of other persons whether to redeem.
8. The German telecommunications provider Deutsche Telekom is rated Aa2
by Moody’s on a scale ranging from the highest grade of Aaa through Aa,
A, Baa, Ba, B, to the lowest grade of Caa. In its June 2000 bond issue worth
$14.5 billion, Deutsche Telekom promised to increase the interest payment by
0.5 percentage points if its credit rating fell below A.
(a) For simplicity, assume that the normal interest on the bonds is
$870 million a year. Suppose that, some years later, Deutsche Telekom
will choose between two investments. For the risky investment, it
must issue new bonds with annual interest of $130 million. The risky
investment will generate a cash flow of $1.5 billion with 50% probability,
and $1 billion otherwise. The safe investment will not require additional
borrowing and will yield $1.07 billion for sure. Using a game in extensive
form, illustrate this choice, while ignoring the promise. At each final
node, show the net cash flow (net of interest on all loans).
(b) Now suppose that the additional borrowings for the risky investment
would reduce Deutsche Telekom’s credit rating below A and trigger
its promise to raise the interest payment by 0.5% × $14.5 billion =
$72.5 million. Suppose that banks will not lend if there is any possibility
of negative net cash flow. How does Deutsche Telekom’s promise
affect the game in extensive form?
9. Bombardier, COMAC, Embraer, and Irkut have entered or are poised to enter
the market for narrow-body jets. Focusing on COMAC and Embraer, Table 10.3
represents the strategies and corresponding profits of the two manufacturers
if they act simultaneously.
(a) Using suitable games in extensive form, show that this is a situation of
first-mover advantage.
(b) Suppose that the government of China guarantees COMAC that it
would reimburse 110% of any losses on developing and producing the
C919. How does that affect the equilibrium/equilibria in (a)?
(c) How much should the Chinese government budget for the
guarantee?

Consider some strategic situation – between your organization and a competitor,
between your organization and a supplier or customer, or even between you and
a rival at work. Apply a game in strategic form or extensive form to plan your
strategy.
Appendix: Solving Nash Equilibrium in Randomized

Strategies
This chapter has introduced the concept of randomization and showed its useful-
ness, especially in competitive contexts. There are two ways to solve for the Nash
equilibrium with randomized strategies – graphical and algebraic. Let us apply
the two methods to the (modified) situation of the gasoline stations described in
Table 10.4.
Graphical
In Figure 10.9, the horizontal axis depicts the probability that Jupiter maintains
price, and the vertical axis shows Saturn’s expected profit. We draw two lines in
Figure 10.9. One line shows Saturn’s profit if it maintains price, as a function of
the probability that Jupiter maintains price. By Table 10.4, if Jupiter maintains
prices with certainty, Saturn’s profit from maintaining price would be $900. How-
ever, if Jupiter maintains price with zero probability (cuts price), Saturn’s profit
from maintaining price would be $500.
$900
Profit ($)
$800
Saturn’s profit if
it maintains price Saturn’s profit if
it cuts price
$600
$500
0 1/2 1
FIGURE 10.9 Solving Nash equilibrium in randomized strategies.

Notes: Saturn gets equal profit from maintaining price and cutting price if Jupiter maintains price
with probability 1/2, and so Saturn is indifferent between maintaining and cutting price. Similarly for
Jupiter. So, the Nash equilibrium in randomized strategies comprises Jupiter maintaining price with
probability 1/2 and Saturn maintaining price with probability 1/2.
The other line shows Saturn’s profit if it cuts price, as a function of the proba-
bility that Jupiter maintains price. By Table 10.4, if Jupiter maintains price with
certainty, Saturn’s profit from cutting price would be $800. However, if Jupiter
maintains prices with zero probability (cuts price), Saturn’s profit from cutting
price would be $600.
The two lines cross at one point. At that point, Saturn’s profit is the same
whether it maintains or cuts price. That point marks Jupiter’s Nash equilibrium
probability, which turns out to be 1/2.
We can use a similar graph to determine Saturn’s equilibrium probability, which
also turns out to 1/2.4
Algebraic
Another way to find the Nash equilibrium probabilities is to use algebra. In
equilibrium, both Jupiter and Saturn must randomize. Suppose that Jupiter
maintains price with probability q. For Saturn to be willing to randomize, it
must be indifferent between the two pure strategies: maintain price and cut
price. This means that Saturn must receive the same expected profit from the
two pure strategies.
To calculate Saturn’s expected profit from maintaining price, refer to Table 10.4.
Saturn’s profit would be $900 if Jupiter maintains price, which occurs with proba-
bility q, while Saturn’s profit would be $500 if Jupiter cuts price, which occurs with
probability 1 − q . Hence, Saturn’s expected profit from maintaining price would
be (900 × q) + (500 × (1 − q)) = 500 + 400q.
Similarly, we can calculate Saturn’s profit from cutting price, which would be
(800 × q) + (600 × (1 − q)) = 600 + 200q. In randomized strategy equilibrium,
Saturn must receive the same expected profit from pricing low and high. This
means that 500 + 400q = 600 + 200q, which implies that q = 1/2.
Likewise, we can determine Saturn’s Nash equilibrium strategy. Suppose that
Saturn maintains price with probability p. For Jupiter to be indifferent between its
alternative pure strategies, it must earn the same expected profit from maintaining
and cutting price. This means that 900p + 1,000(1 − p) = 1,300p + 600 (1 − p), or
p = 1/2.
Notes
1 This discussion is based, in part, on Richard Tortoriello, “Aerospace & defense,” Standard &
Poor’s Industry Surveys, February 10, 2011; “Airbus and Boeing call end to ‘duopoly’,” Financial
Times, June 21, 2011; “Airbus-Boeing duopoly holds narrow-body startups at bay at Paris Air
Show,” Bloomberg, June 23, 2011.
2 This example is based on O. G. Haywood, “Military decision and game theory,” Journal of
the Operations Research Society of America, Vol. 2, No. 4, November 1954, pp. 365–385; and
J. Rohwer and G. Hummelchen, Chronology of the War at Sea, 1939–45, Vol. 2, transl. Derek
Masters, London: Ian Allan, 1974, p. 306.
3 After activation of the rights, the value of the company would be the prior market value,
$70 × 133 million = $9.31 billion, plus the payment for the additional shares, $35 × (133 − 12.5)
million = $4.22 billion, or a total of $13.53 billion. There would be 133 + (133 − 12.5) =
253.5 million shares, and so the value of each share would be $13,530/253.5 = $53.37.
4 This technique is based on Avinash Dixit and Barry Nalebuff, Thinking Strategically: The Com-
petitive Edge in Business, Politics, and Everyday Life, New York: Norton, 1991, Chapter 7.
C H A P T E R
11
Oligopoly1
LEARNING OBJECTIVES
• For sellers competing on price to sell differentiated products, iden-
tify the price that maximizes profit and how to adjust the price.
• Appreciate that prices can be strategic complements.
• For sellers competing on capacity to sell a homogeneous prod-
uct, identify the capacity that maximizes profit and how to adjust
capacity.
• Appreciate that capacities can be strategic substitutes.
• Apply limit pricing to deter entry.
• Apply capacity leadership for first-mover advantage.
• Appreciate how to limit competition.
1. Introduction
Dynamic random-access memory (DRAM) provides high-speed storage and

retrieval of data, and is a key component in computers. NAND flash memory
is essential in devices, such as smart phones, that require a lot of space for data
storage or fast recording. The demand for memory grows in tandem with sales of
computers, smart phones, and other electronic devices. In 2010, global DRAM
sales totaled $39.2 billion; NAND flash memory sales totaled $17.3 billion.2
Competition in the memory industry is intense, with relatively little product
differentiation. However, the industry is not perfectly competitive. Samsung Elec-
tronics dominates the production of both DRAMs and NAND flash memory,
with 38% and 40% shares, respectively (see Table 11.1). In both markets, the top
four manufacturers have about a 90% share.
250 11 Oligopoly
Table 11.1 Memory industry: share of production
Manufacturer DRAM (%) NAND (%) Gross margin,

recent average (%)
Elpida 16 − 5
Hynix 22 10 ~40
Micron 13 10 ~13
Samsung Electronics 38 40 >30
Toshiba − 32 −
A major reason for the high concentration of the memory industry is large
fixed costs. The cost of building a wafer fabrication plant has risen over time, and
presently ranges from $3 billion to $8 billion.
Historically, with improvements in manufacturing technology and expansion
of capacity, the price of memory has fallen by around 20–30% a year. Moreover,
the industry must contend with volatile exchange rates. With prices set in US dol-
lars, manufacturers outside the United States incur exchange rate risks.
Given the huge costs of investment and volatile prices and revenue, the memory
industry is not for the faint-hearted. In 2006, Infineon spun off its DRAM manu-
facturing into Qimonda, which filed for insolvency protection three years later.
Elpida, Japan’s last manufacturer of memory, was sold to Micron in 2013.
With continued volatility in exchange rates, how should Samsung and other
memory manufacturers adjust pricing and capacity to the fluctuation of the Korean
won against the US dollar? How should they respond to the trend of consolidation
as smaller competitors merge or exit completely? How would the consolidation of
the memory industry affect buyers of memory such as manufacturers of comput-
ers and mobile phones?
The global semiconductor industry is obviously an oligopoly.
An oligopoly is a market with a small number of sellers who
Oligopoly: A market with
a small number of sellers
behave strategically. Oligopoly is a market structure that
who behave strategically. lies between the two extremes of perfect competition and
monopoly. Under perfect competition no seller has market
power, while in a monopoly there is only one seller.
To understand the impact of the fluctuation of the Korean won, we must
understand how sellers decide capacity, production, and pricing under conditions
of oligopoly. Since there are few sellers, it is likely that they would actively con-
sider interactions with one another in business decisions. Hence, it is important to
consider their strategic interaction. Accordingly, we apply the techniques of game
theory developed in Chapter 10.
We distinguish between two time horizons – the short and long run. Generally,
production capacity is less flexible than pricing, and hence businesses must decide
on capacity (long run) before pricing (short run).
We first study oligopoly in the short run – how competing sellers set prices and
how their pricing depends on whether the product is homogeneous or differen-
tiated, and how each seller should adjust price in response to competitors’ price
Oligopoly 251
changes. We then study oligopoly in the long run – how competing sellers decide
on production capacity and how each seller should adjust capacity and produc-
tion in response to competitors’ changes in capacity and production.
Applying these models, we can explain how Samsung and other memory man-
ufacturers should adjust capacity and pricing to changes in exchange rates and
other costs. Further, we explain how memory manufacturers should adjust to
consolidation of the industry.
Finally, we study how producers can benefit from restraining competition and
the ways of accomplishing such restraint. The analysis suggests how consolidation
of the memory industry would affect buyers such as computer manufacturers.
Managers in oligopolistic industries can apply the techniques and analysis of
this chapter to plan more effective strategies in the face of competition.
2. Price Competition
Typically, in the short run, the strategic variable for oligopolistic sellers is price.
The outcome of oligopolistic competition on price depends on whether the prod-
uct is homogeneous or differentiated. Let us analyze the outcome in each of these
two settings.
Homogeneous Product
Suppose that, in the market for wireless telecommunications, the market demand
is represented as in Figure 11.1. The marginal cost of service is a constant $30 per
subscriber per month.
100
Price ($ per subscriber per month)
market demand
65
marginal cost
30
marginal
revenue
0 10.5 30
Production (million subscribers per month)
FIGURE 11.1 Monopoly.

Note: The monopoly maximizes profit at a scale of 10.5 million subscribers per month and a price of
$65 per subscriber per month.
252 11 Oligopoly
It will be useful to compare the market outcome in scenarios of monopoly and

oligopoly. Referring to Figure 11.1, the scale at which marginal revenue equals
marginal cost is 10.5 million subscribers and the corresponding price is $65 per
subscriber per month. At that scale and price, the monopoly maximizes profit.
The monopoly profit contribution is ($65 − $30) × 10.5 million = $367.5 million
per month.
For simplicity, we consider a situation of duopoly. Sup-
Duopoly: A market with pose that there are just two sellers in the market, say, Luna
two sellers who behave
strategically.
Cellular and Mercury Wireless, and that they behave strate-
gically. Each seller produces under conditions of constant
marginal cost of $30 per subscriber per month with no
Bertrand oligopoly:
capacity constraint. Luna and Mercury provide identical
Sellers produce at constant
marginal cost with services, so the product is homogeneous. This describes the
unlimited capacity, and Bertrand model of oligopoly, named for the French mathe-
compete on price to sell a matical economist who published it in 1883.
homogeneous product. Under these conditions, the market equilibrium is per-
fectly competitive. Remarkably, even though the industry is a
duopoly, the outcome is the same as with perfect competition. To see this, imagine
that Luna Cellular charges some price above marginal cost, say $32.
Then Mercury has three choices: it can price above, at, or below $32. If Mercury
charges above $32, all consumers would subscribe with Luna, and so Mercury will
sell and earn nothing. If Mercury matches Luna’s price $32, it would get half of the
market demand and earn profit equal to the incremental margin, $32 − $30 = $2,
per sale. However, if Mercury charges below $32, even slightly less, then it would get
the entire market demand.
So, by marginally undercutting Luna’s price of $32, Mercury can almost double
its profit relative to matching the price. The reason is that its incremental margin
is only slightly less than $2, but its sales would double. Indeed, Mercury faces a
demand curve that is infinitely elastic with respect to a price cut below Luna’s price.
The same reasoning applies to Luna. Accordingly, the Nash equilibrium in this
duopoly is for both competitors to set prices equal to the marginal cost of $30. If
either seller sets a price above $30, the other seller would undercut. The strategic
logic is essentially the same as in the situation of the competing gasoline stations
described in Table 10.1.
The result is quite dramatic. Even when there are only two competing sellers,
each faces a demand curve that is so elastic that the market outcome is identical
to that with perfect competition.
The same analysis applies even more strongly if there are more than two com-
petitors. Every seller will prefer to undercut the others – its incremental margin
would be only slightly lower, but its sales would increase in proportion to the
number of competitors it undercuts. (As Chapter 10 shows, competing sellers
can avoid ruinous price competition in repeated interaction. Using “tit-for-tat”
strategies – price high and cut price only if the competitor cuts price – they might
avoid price wars.)
Oligopoly 253
Differentiated Products
What if the competing sellers offer differentiated products? With differentiation,
if one seller undercuts the competitor’s price, it would take away only part of the
competitor’s demand. Hence, the price-cutter’s demand is not infinitely elastic.
So, what is the equilibrium outcome?
Consider the following situation, which applies the Hotelling
model of oligopoly. Suppose that consumers are located Hotelling oligopoly:
uniformly along a street one mile long, with the Luna and Sellers compete on
Mercury dealers at each end of the street. Each consumer price to sell a product
differentiated by distance
is willing to pay up to $100 for the service but must incur a from consumer.
cost of $4 per mile to travel to a dealer to buy the service. So,
a consumer located 0.25 mile from the Luna dealer would
incur $4 × 0.25 = $1 to travel to the Luna dealer, and $4 × 0.75 = $3 to travel to
the Mercury dealer.
In this situation, the differentiation is due to the difference in the consumer’s
distances from the two dealers. Luna and Mercury independently set prices simul-
taneously and incur a constant marginal cost of $30 per subscriber per month to
provide the service.
Since Luna and Mercury set prices simultaneously, it is Residual demand
curve: Demand given
appropriate to apply the concept of Nash equilibrium. For the actions of competing
simplicity, we assume that, in equilibrium, every consumer sellers.
will buy. Let Luna set price pL, and Mercury set price pM. In
the appendix to this chapter, we show that, given Mercury’s
price, Luna’s residual demand curve would be as shown in Figure 11.2. The residual
demand is the demand given the actions of competing sellers.
pM + 4
Luna’s residual
1
2
(pM + 34) demand (given
Mercury’s price, pM)
marginal cost
30
residual
marginal
revenue
0 1
16
(pM −26) 1
8
pM + 1
2
Production (million subscribers per month)
FIGURE 11.2 Price competition with differentiated products: residual demand.

Notes: Given Mercury’s price, pM, Luna’s residual demand slopes downward from quantity 0 at price
pM + 4, to quantity 1/8 pM + 1/2 at price 0. Luna maximizes profit at quantity 1/16(pM − 26) and price
1/2(pM + 34).
254 11 Oligopoly
Given Mercury’s price, pM, and any possible price that it could set, consumers
relatively closer to Luna would buy from Luna and those relatively closer to Mer-
cury would buy from Mercury. The consumers’ choices are relative to the prices
set by Luna and Mercury. With information on the consumers’ choices, Luna can
calculate its sales. By repeating this procedure for every price that it could set,
Luna can construct its residual demand curve.
The residual demand curve slopes downward. If Luna raises its price, some con-
sumers (located relatively far from Luna) would switch to Mercury, and so Luna’s
sales would be lower. With the residual demand curve, Luna can construct its resid-
ual marginal revenue curve. Note that Luna’s residual demand and marginal reve-
nue curves depend on Mercury’s price. Suppose that Mercury raises its price. Then
some consumers (located relatively far from Mercury) would switch to Luna, and so
Luna’s residual demand and marginal revenue curves would be further to the right.
To maximize profit, Luna should produce at the scale where marginal revenue
equals marginal cost. The profit-maximizing scale is 161 ( pM − 26) and the corre-
sponding price is 12 ( pM + 34). Owing to the competition for consumers, Luna’s
profit-maximizing price depends on Mercury’s price, pM.
Figure 11.3 presents a graph of Luna’s profit-maximizing price as a function of
Mercury’s price, pM. It begins at a price of $30, which Luna would set if Mercury
set a price of $26. Luna would never price below the marginal cost of $30. The
graph then increases according to 12 ( pM + 34) .
This graph of Luna’s profit-maximizing price as a func-
Best response function:
A seller’s best action as
tion of Mercury’s price is Luna’s best response function. The
function of competing best response function represents the seller’s best action
sellers’ actions. as a function of the actions of competing sellers. Figure 11.3
Luna’s price, pL ($ per subscriber per month)
Mercury’s best Luna’s best response function

response function with higher marginal cost
Luna’s best
response function
34
30
0 26 34
Mercury’s price, pM ($ per subscriber per month)
FIGURE 11.3 Price competition with differentiated products: best response functions.
Notes: Luna’s best response function shows its profit-maximizing price, pL , as a function of Mercury’s
price, pM. Mercury’s best response function shows its profit-maximizing price, pM, as a function of
Luna’s price, pL.
Oligopoly 255
also depicts Mercury’s best response function, which is its profit-maximizing price
as a function of Luna’s price, pL.
Figure 11.3 is the graphical equivalent of the game in strategic form, where
each party chooses from a continuum of strategies. (Chapter 10 considered strate-
gic situations in which each party chose between two strategies, so could be repre-
sented by a table.) The Nash equilibrium is at the intersection of the best response
functions. For both sellers, the Nash equilibrium strategy is to price at $34. By the
properties of Nash equilibrium, if Mercury believes that Luna will price at $34,
then it is best for Mercury to set price at $34, and vice versa.
In equilibrium, each seller earns an incremental margin of $34 − $30 = $4, and
serves half of the consumers. Thus, in competition on price to sell differentiated
products, the equilibrium price exceeds the marginal cost. Further, the equilibrium
does not depend on fixed costs. Only the sellers’ participation decisions – whether
to produce at all – depend on fixed costs: specifically, each seller will produce only
if the profit contribution exceeds its fixed cost.
Consumer Preferences
In the preceding example, the competing products were differentiated by location.
Actually, the Hotelling model applies to differentiation in terms of any attribute
on which consumers have differing preferences. The attribute could be taste (some
prefer salty, others prefer sweet), design (some like functional, others like aes-
thetic), or membership of a customer loyalty program. Then the “transport cost”
represents the consumer’s disutility from consuming any attribute that differs
from the ideal or most preferred.
What if the consumers’ preferences are relatively stronger? Specifically, in the pre-
ceding example, what if the transport cost is $5 rather than $4 per mile? Referring
to Luna’s residual demand curve in Figure 11.2, the vertical intersection would be
higher and the horizontal intersection would be further to the left. Luna’s residual
demand would be more inelastic, and so its profit-maximizing price would be higher.
Then, by Figure 11.3, Luna’s best response function would be further up. Similarly,
Mercury’s best response function would be further to the right. Thus, the Nash
equilibrium prices would be higher.
So, in the context of oligopoly, sellers can mitigate competition by differenti-
ating their products. The stronger are the consumers’ preferences over the differ-
entiating attribute (the larger is the consumer’s disutility from a less than ideal
attribute), the less price elastic will be each seller’s residual demand. With residual
demand being less price elastic, the equilibrium prices would be higher.

How should oligopolists competing on price respond to changes in demand and
cost? Generally, the response comprises two steps. First, as in a monopoly, each
seller should adjust price until its residual marginal revenue equals marginal cost.
Second, in an oligopoly, the new price is a function of the other sellers’ prices.
256 11 Oligopoly
So, with the new demand or cost, the seller’s best response function would shift
and then establish a new equilibrium.
Consider an increase in market demand. In Figure 11.2, Luna’s residual demand
would shift to the right, and so the residual marginal revenue would shift to the
right. So Luna’s profit-maximizing price would be higher. Hence, in Figure 11.3,
Luna’s best response function would be higher. Similarly, Mercury’s best response
function would shift to the right. Thus, in the new equilibrium, each seller would
set a higher price.
What about an increase in marginal cost? In Figure 11.2, with higher marginal
cost, Luna’s profit-maximizing price would be higher. Then, in Figure 11.3, Luna’s
best response function would be higher. If Mercury’s marginal cost does not change,
its best response function would remain the same. In the new equilibrium, both sell-
ers raise price. However, Luna would raise price relatively more than Mercury.
So, in competition on price to sell differentiated products, if a seller incurs
higher marginal cost, the competitors absorb part of the impact by raising their
prices. The seller with higher cost loses consumers and profit contribution, but the
loss is reduced to the extent that competitors respond with higher prices.
Strategic Complements3
Referring to Luna’s best response function in Figure 11.3, if Mercury were to raise
its price, then Luna should raise its price as well, while if Mercury were to cut its
price, then Luna should also cut its price. Intuitively, if Mercury raises its price,
some consumers would switch to Luna. So the marginal consumer (just indifferent
between Luna and Mercury) would be relatively further from Luna. Hence, Luna’s
demand would become relatively inelastic, and so Luna would raise its price.
Likewise, if Mercury were to cut its price, some consumers would switch to Mer-
cury. So the marginal consumer would be relatively closer to Luna’s product. Hence,
Luna’s demand would become relatively elastic, and so Luna would cut its price.
Similarly, referring to Mercury’ best response function, Mercury should always
adjust its price in the same direction as Luna. Accordingly, in price competition
between oligopolists with unlimited capacity that offer differentiated products
(Hotelling model), prices are strategic complements. Actions
Strategic complements: are strategic complements if an adjustment by one party
An adjustment by one leads other parties to adjust in the same direction. In an
party leads others to oligopoly, if prices are strategic complements, then competi-
adjust in same direction.
tors would maximize profit by following each other’s price
moves in the same direction.
PROGRESS CHECK 11A

Suppose that the transport cost is less than $4 per mile. How would that affect
the equilibrium prices?
Oligopoly 257
VODAFONE: 4G PRICING
In the UK 4G mobile telecommunications market, EE has the largest market

share with over 4 million subscribers, followed by O2 and Three with over
2 million each, and Vodafone with under 1 million.
The SIM-only market segment caters to subscribers with their own handset
and who buy only the service. Among the entry-level plans, Three offers the
cheapest at £7 a month, EE is next lowest at £9.99 a month, and O2’s price is
£11 a month, while Vodafone is the most expensive at £22 a month.
The competing providers differentiate by coverage (EE claims the largest
network), quantity of included voice minutes and data, and prices for additional
usage. Vodafone’s plan includes 1GB of data, which is more than the others’.
Nevertheless, Kester Mann, senior analyst at CCS Insight, remarked that
Vodafone “will continue to lose share if it doesn’t respond to the actions of its
competitors, which have reduced tariffs.”
Vodafone defended its pricing as offering good value especially considering
sports and movies. Vodafone offers Sky Sports, Spotify, and Netflix in a 4G
SIM-only plan for £27 a month.
Source: “Vodafone under pressure to cut its 4G SIM-only prices,” Mobile Magazine, August
13, 2014.
3. Limit Pricing
So far, we have analyzed oligopoly strategy assuming that the competitors act
simultaneously. Accordingly, we have applied the concept of Nash equilibrium.
But what if one seller can act before others? Then it might be able to make strate-
gic moves and exploit first-mover advantage. To develop strategy in such circum-
stances, we apply games in extensive form.
Consider the duopoly between Luna Cellular and Mercury Wireless, and allow
Mercury to set price before Luna. Referring to Figure 11.4, at node A, Mercury
sets its price, and then, at node B, Luna sets its price. Here, we allow a continuous
choice of price, so it is not feasible to represent all of the possible choices in the
extensive form.
In competition on price to sell differentiated products, if production involves
a substantial fixed cost, one possible strategic move is to deter potential compet-
itors from entering the industry. The strategy is to set such a low price that the
potential competitor’s residual demand is so low that the potential competitor
cannot break even.
Referring to Figure 11.5, the lower is Mercury’s price, pM, the further to the left
would be Luna’s residual demand curve. If Mercury’s price is low enough, Luna’s
residual demand curve would everywhere lie below its average cost curve. Note
258 11 Oligopoly
A B
price/ price/
Mercury Luna
capacity capacity
FIGURE 11.4 Strategic move.

Note: Mercury sets its price or capacity to influence Luna’s choice of price or capacity in a favorable way.
1
2
(pM + 34)
average cost
marginal cost
30 Luna’s
residual Luna’s
marginal residual
revenue demand
0 1 (pM − 26)
16
Capacity (million subscribers per month)
FIGURE 11.5 Limit pricing.

Notes: Mercury sets its price so low that Luna’s residual demand falls below Luna’s average cost
curve. So, Luna will choose not to produce at all.
the shape of Luna’s average cost curve: at low capacity, the average cost is high
because of the fixed cost, and then the average cost declines with capacity as the
fixed cost is divided by a larger scale of production.
Since Luna’s residual demand curve is below the average cost curve at all scales
of production, the price would be less than average cost at the profit-maximizing
scale, 161 ( pM − 26). Hence, Luna would incur a loss even at the “profit-maximizing”
scale. Accordingly, its best strategy is not to produce at all.
The strategy of producing so much as to choke off poten-
Limit pricing: The leader tial entry is called limit pricing. It is so called because the
commits to pricing so low leader sets a sufficiently low price that it takes away so much
that a potential competitor
demand that a potential competitor cannot break even. Thus
cannot break even.
no other producer would enter the market.
Oligopoly 259
A strategy of limit pricing makes sense under two conditions. The first condi-
tion is that the leader’s price is credible. Potential competitors must believe that
the leader has committed to the entry-deterring level of price, and will not change
if the potential competitor should enter. This means that, for the leader, it must
be more profitable to produce at the entry-deterring price than to accommodate
entry and produce an equal share with competitors.
The second condition is that production involves a substantial fixed cost. If there
were no fixed cost, a potential competitor could break even at a very small scale
of production. Referring to Figure 11.5, the average cost and marginal cost curves
would be the same. So, at the profit-maximizing capacity, the price would exceed
the average cost. So there is no way to deter the potential competitor from entry.
PROGRESS CHECK 11B

Suppose that Luna’s fixed cost of production is higher. Referring to Figure 11.5,
how would that affect Luna’s cost curves and the price that Mercury must set
to deter Luna from entry?
4. Capacity Competition
Typically, in the long run, the strategic variable for oligopolistic sellers is produc-
tion capacity. We analyze the outcome of oligopolistic competition on capacity
in the setting of a homogeneous product. For simplicity, we focus on a duopoly,
and ignore utilization and inventories, and so capacity, production, and sales are
all equal.
The market is as described in Figure 11.1, except that there are two providers.
Luna and Mercury independently and simultaneously set production capacity,
and then produce at a constant marginal cost of $30 per sub-
scriber per month. The market price equates the demand Cournot oligopoly:
with the total capacity offered by the two providers. This is Sellers compete on
production capacity to sell
the Cournot model of oligopoly, named for the French
a homogeneous product.
economist who published it in 1838.
Since Luna and Mercury set capacities simultaneously, it
is appropriate to apply the concept of Nash equilibrium. Let Luna set capacity
QL, and Mercury set capacity QM. Then, given Mercury’s capacity, Luna’s resid-
ual demand curve is as shown in Figure 11.6.
Recall that the residual demand is the demand given the actions of competing
sellers. Here, the action of the competing seller is Mercury’s capacity, QM. So, at
every possible price, Luna’s residual demand is the market demand less Mercury’s
production, QM.
To maximize profit, Luna should set production capacity at the scale where
residual marginal revenue equals marginal cost. From the residual demand, Luna
260 11 Oligopoly
100
Price ($ per subscriber per month) market demand
100 − 13 QM
Luna’s residual demand (given
Mercury’s capacity, QM)
65 − 16 QM
Luna’s residual
marginal revenue
marginal cost
30
0 10.5 − 12 QM 15 − 12 QM 30−QM 30
Capacity (million subscribers per month)
FIGURE 11.6 Capacity competition: residual demand.

Notes: Given Mercury’s capacity, Luna’s residual demand slopes downward to a quantity 300 − QM
at price 0. Luna maximizes profit at a scale of 10.5 − 1/2QM and price 65 − 1/6QM.
can calculate its residual marginal revenue. Referring to Figure 11.6, Luna’s resid-
ual marginal revenue equals the marginal cost of $30, at the production capacity
of 10.5 − 12 QM . Owing to the competition for consumers, Luna’s profit-maximiz-
ing capacity depends on Mercury’s capacity, QM.
Figure 11.7 presents a graph of Luna’s profit-maximizing capacity as a function
of Mercury’s capacity, QM. This graph is Luna’s best response function. We can
intuitively derive two extreme points on Luna’s best response function. If Mer-
cury sets capacity at zero, then Luna would effectively be a monopoly and should
choose the profit-maximizing capacity of 10.5 million subscribers per month. If
Mercury sets capacity at 21 million, then Luna’s residual demand curve would be
so far to the left that its residual marginal revenue would cross the marginal cost
at a capacity of zero. Hence, Luna would choose zero capacity.
Figure 11.7 also depicts Mercury’s best response function, which is its profit-
maximizing capacity as a function of Luna’s capacity, QL. Figure 11.7 is the
graphical equivalent of the game in strategic form, where each party chooses from
a continuum of strategies.
The Nash equilibrium is at the intersection of the two best response functions. For
both sellers, the Nash equilibrium strategy is to set capacity at 7 million. By the prop-
erties of Nash equilibrium, if Mercury believes that Luna will choose capacity of
7 million, then it is best for Mercury to choose capacity of 7 million, and vice versa.
In equilibrium, the total capacity and production is 7 + 7 = 14 million
subscribers a month. Referring to the market demand, the market price is
10
3
× (30 − 14) = $53.33 per subscriber, so its profit contribution is $(53.33 − 30) ×
7 = $163 million a month. The equilibrium does not depend on fixed costs. Only
the sellers’ participation decisions – whether to produce at all – depend on fixed
Oligopoly 261
Mercury’s best
response function
Luna’s capacity, QL (million
subscribers per month)
Luna’s best response function

10.5 with higher marginal cost
7
Luna’s best
response function
0 7 21
Mercury’s capacity, QM (million subscribers per month)
FIGURE 11.7 Capacity competition: best response functions.

Notes: Luna’s best response function shows its profit-maximizing capacity, QL, as a function of
Mercury’s capacity, QM. Mercury’s best response function shows its profit-maximizing capacity, QM,
as a function of Luna’s capacity, QL.
costs: specifically, each seller will produce only if the profit contribution exceed
its fixed cost.
With a duopoly, the total capacity and production is 14 million subscribers a
month, which exceeds the monopoly profit-maximizing capacity of 10.5 million.
The duopoly price of $53.33 per subscriber is less than the monopoly price of
$65 per subscriber. Finally, the combined profit contribution of the two duopo-
lists, $326 million, is less than the monopoly profit contribution of $367.5 million.

How should oligopolists competing on capacity respond to changes in demand and
cost? Generally, the response comprises two steps. First, as in a monopoly, each
seller should adjust capacity until its residual marginal revenue equals marginal
cost. Second, in an oligopoly, the new capacity is a function of the other sellers’
capacities. So, with the new demand or cost, the seller’s best response function
would shift and then establish a new equilibrium.
Consider an increase in market demand. In Figure 11.6, Luna’s residual demand
would shift to the right, and so the residual marginal revenue would shift to the
right. So Luna’s profit-maximizing capacity would be larger. Hence, in Figure 11.7,
Luna’s best response function would be higher. Similarly, Mercury’s best response
function would shift to the right. Thus, in the new equilibrium, each seller would
choose larger capacity.
What about an increase in marginal cost? In Figure 11.6, with higher marginal
cost, Luna’s profit-maximizing capacity would be smaller. Hence, in Figure 11.7,
262 11 Oligopoly
Luna’s best response function would be lower. If Mercury’s marginal cost does
not change, its best response function would remain the same. In the new equilib-
rium, Luna reduces capacity while Mercury increases capacity.
So, in competition on capacity to sell a homogeneous product, if a seller incurs
higher marginal cost, the competitors take advantage of the weakness to increase
their capacities and grab market share. The seller with higher cost loses in two
ways – by the higher cost itself and by the competitors’ strategic adjustment.
PROGRESS CHECK 11C

In Figures 11.6 and 11.7, illustrate how Luna’s choice of capacity and best
response function would change if its marginal cost is lower.
Strategic Substitutes
Referring to Figure 11.6, if Mercury raises its production capacity, Luna’s resid-
ual demand curve would shift to the left by the amount of Mercury’s increase in
capacity. Then Luna’s profit-maximizing capacity (at which its residual marginal
revenue equals marginal cost) would be lower. So, in Figure 11.7, Luna’s best
response function slopes downward: the larger is Mercury’s capacity, the smaller
the capacity that Luna chooses.
Similarly, Mercury’s best response function slopes downward: the larger is
Luna’s capacity, the smaller the capacity that Mercury chooses. Among sellers
producing at constant marginal cost and competing on production capacity, the
capacity choices are strategic substitutes. Actions are strategic
Strategic substitutes: An
substitutes if an adjustment by one party leads other par-
adjustment by one party ties to adjust in the opposite direction. In an oligopoly, if
leads others to adjust in production capacities are strategic substitutes, then competi-
the opposite direction. tors would maximize profit by adjusting capacity in the
opposite direction to the adjustments of others.
A clear appreciation of whether business choices are strategic complements or
strategic substitutes is very useful in strategic thinking. With this appreciation,
managers can respond to competitor’s actions even without knowing the best
response functions or even the equilibrium.
Generally, whether business choices are strategic complements or strategic substi-
tutes depends on the relevant demand and cost conditions. Among sellers producing
at constant marginal cost with unlimited capacity, and competing on price to sell a
differentiated product (Hotelling model), prices are strategic complements. Among
sellers producing at constant marginal cost and competing on production capacity
to sell a homogeneous product (Cournot model), capacities are strategic substitutes.
What about other business choices such as advertising and R&D expenditure?
Generally, these may be either strategic complements or strategic substitutes
depending on the relevant demand and cost conditions. For instance, increased
R&D spending can have an effect similar to increased capacity. On the other hand,
Oligopoly 263
an increase in one seller’s R&D expenditure may drive competitors to increase R&D
as well, particularly when they compete for patents. So, R&D expenditures might
be strategic complements or strategic substitutes, depending on the circumstances.
PROGRESS CHECK 11D

Suppose that R&D expenditures are strategic substitutes. On a graph like
Figure 11.7, illustrate the best response functions.
MANUFACTURING MEMORY: NOT FOR THE FAINT-HEARTED
The DRAM and NAND flash memory industries are highly concentrated. Samsung
Electronics dominates both industries, with shares of 38% and 40% shares,
respectively. The top four manufacturers control around 90% of both markets.
The markets for memory are global with prices set in US dollars. Any
depreciation of the Korean won against the US dollar would raise the US
dollar cost of Korean manufacturers. In the short run, to the extent of product
differentiation, Korean manufacturers should raise their prices. Since prices
are strategic complements, competitors such as Micron would follow.
Given the high fixed costs and price volatility, the memory industries are
consolidating through mergers and exits. Higher-cost manufacturers are likely
to sell themselves or quit. With the consolidation, lower-cost producers such
as Samsung and Hynix would earn larger margins and can further increase
their production capacity and expand their market share.
Source: Clyde Montevirgen, “Semiconductors,” Standard & Poor’s Industry Surveys, May 12,
2011.
5. Capacity Leadership
So far, we have analyzed oligopolists’ choice of capacity assuming that the com-
petitors act simultaneously. Accordingly, we have applied the concept of Nash
equilibrium. But what if one seller can act before others? Then it might be able
to make strategic moves and gain first-mover advantage. To develop strategy in
such circumstances, we apply games in extensive form, as we did in the analysis
of limit pricing.
Consider the duopoly between Luna Cellular and Mercury Wireless, and allow
Mercury to set capacity before Luna. Referring to Figure 11.4, at node A, Mer-
cury sets capacity, and then, at node B, Luna sets capacity. A possible strategic
move is to set larger capacity and push down the competitor’s residual demand so
that it chooses a smaller capacity. By grabbing a larger market share, the leader
can raise its profit.
264 11 Oligopoly
Luna’s capacity, QL (million

subscribers per month)
10.5
5.25 Luna’s best

response function
0 10.5 21
Mercury’s capacity, QM (million subscribers per month)
FIGURE 11.8 Capacity leadership.

Notes: Mercury sets capacity before Luna. It chooses capacity to maximize its profit, given that Luna
would subsequently set capacity according to its best response function.
Figure 11.8 graphs Luna’s best response function, which shows its choice of
capacity as a function of Mercury’s capacity, QM. Since Mercury sets capacity
before Luna, Mercury can choose the capacity that maximizes
Stackelberg oligopoly: its profit, given that Luna would subsequently set capacity
Both leader and follower according to its best response function. This is the Stackelberg
sell a homogeneous model of oligopoly.
product, and the leader Through trial and error, Mercury can calculate the corre-
commits to some capacity
to grab a larger share.
sponding profit contribution from every level of capacity up
to 21 million subscribers a month. The capacity which max-
imizes Mercury’s profit, given that Luna would subsequently
set capacity according to its best response function, is 10.5 million subscribers per
month. Following its best response function, Luna would choose a capacity of
5.25 million. The market price would be 103 × (30 − 10.5 5.25) = $47.50 per sub-
scriber per month.
As the industry leader, Mercury commits to twice the capacity of Luna, the
industry follower. Since the product is homogeneous, both sell at the same market
price. So Mercury enjoys first-mover advantage, earning double the profit contri-
bution of Luna.
However, there is a downside to Mercury’s commitment to large capacity. Tak-
ing into account Luna’s capacity, the total capacity in the market is larger, and so
the market price is lower. Indeed, in our example, the price with capacity leader-
ship is $47.50 per subscriber per month, as compared with $53.33 if both sellers
set capacity simultaneously.
A strategy of capacity leadership makes sense only if the leader’s capacity is
credible. Potential competitors must believe that the leader has committed to its
Oligopoly 265
capacity, and will not change if the potential competitor should enter. This means
that, for the leader, it must be more profitable to produce at the pre-committed
capacity (which implies a relatively low price) than to accommodate entry and
produce an equal share with the competitor.
PROGRESS CHECK 11E

Referring to Figure 11.8, suppose that Mercury sets capacity at 21 million
subscribers a month. How would Luna respond? What is bad about such a
strategy for Mercury?
LAS VEGAS STRIP: EVOLUTION OF OLIGOPOLY
The Las Vegas Strip actually lies outside the city of Las Vegas, on the stretch
of Las Vegas Boulevard between Sahara Avenue and Russell Road. Over
time, Caesars Entertainment and MGM Resorts have consolidated the hotel
industry on the Strip. In 2010, of a total of 67,000 rooms, Caesars Entertainment
managed 22,880, while MGM managed over 12,000.
In the wake of the Great Recession, two new developments opened on the
Strip. In late 2009, CityCenter, 50% owned and managed by MGM Resorts,
added over 4,000 rooms. Then, in late 2010, Cosmopolitan, acquired by
Deutsche Bank on foreclosure of a construction loan, added 2,000 rooms with
another 1,000 in 2011.
The Moody’s analyst Peggy Holloway described the outcome as a “perfect
storm – visitor volume declined due to the recession at the same time significant
new inventory entered the market.” Between 2008 and 2010, MGM’s hotel
occupancy decreased from 92% to 89%, while its average daily rate fell from
$148 to $108. The Sahara, with 1,720 rooms at the northern end of the Strip,
closed in May 2011.
MGM’s strategy in the Las Vegas Strip is consistent with capacity leadership.
Together with Caesars Entertainment, it dominates the industry. By expanding
capacity, it grabs market share. Weaker hotels like the Sahara shut down.
MGM’s room rates were hit by the Great Recession, but, in the longer term, it
can reap the benefits of the larger share.
Sources: Caesars Entertainment Corp., Annual Report 2010; MGM Resorts, Annual Report
2010; “Sahara’s closure on May 16 will mark ‘the end of an era,’” Las Vegas Sun, March 11, 2011.
6. Restraining Competition
Generally, a monopoly is more profitable than an oligopoly – whether the sellers

compete on price or capacity. Further, an oligopoly is more profitable than a
266 11 Oligopoly
perfectly competitive industry. Hence, a monopoly is the most profitable of all

possible market structures.
Accordingly, rather than compete, sellers can raise profits by restraining compe-
tition among themselves. If sellers restrain competition to a sufficient degree, they
can achieve the profit of a monopoly. Competing sellers can restrain competition
through explicit agreement or by integration.
Cartels
A cartel is an agreement to restrain competition. A seller cartel is an agreement
among sellers to restrain competition in supply, while a buyer cartel is an agree-
ment among buyers to restrain competition in demand.
Typically, a seller cartel sets a maximum sales quota for each participant. By
limiting each participant’s sales, the cartel restricts the quan-
Cartel: An agreement to tity supplied and raises sellers’ profit above the competitive
restrain competition. level. The more effectively the cartel suppresses competition,
the closer the cartel’s profit will be to the monopoly level.
A seller cartel restrains sales to raise the market price above the competitive
level. The higher the price, however, the more attractive it will be for an individual
seller to sell more than its quota. To the extent that any one seller exceeds its quota,
the quantity supplied will increase and the market price will fall. So, to be effective,
a cartel must have some way to compel each participant to abide by its quota.
Further, if a cartel succeeds in raising the price above the competitive level, it
will attract new sellers to enter the market. Hence, another issue for a cartel is how
to keep out new entrants. Therefore, the key to an effective cartel, or, more gen-
erally, effective restriction of competition, is enforcement against existing sellers
exceeding their quotas and against the entry of new competitors.
Enforcement
The laws of most developed countries seek to encourage competition and typ-
ically do not allow cartels except for specific exemptions. Cartels that are not
legal must rely on informal enforcement. Generally, the effectiveness of informal
enforcement depends on five factors.
• Number of sellers. Enforcement is easier when there are fewer sellers to be

monitored. So, a cartel will be more effective in a concentrated than in a
fragmented industry.
• Excess capacity. If all sellers are operating near capacity, then it will be difficult
for them to expand; hence, there will be little incentive to exceed the specified
quotas. By contrast, a seller with substantial excess capacity will have more
incentive to exceed its quota.
• Sunk costs. In the short run, competitive sellers are willing to operate so long
as the price covers avoidable cost. Sellers with substantial sunk costs will be
relatively more willing to cut price and exceed their quotas.
Oligopoly 267
• Barriers to entry and exit. Recall from Chapter 8 that, in a perfectly contestable
market, sellers can enter and exit at no cost. Suppose that all the sellers in
a perfectly contestable market form a cartel. Despite their monopoly, they
cannot raise the price above the long-run average cost, because that would
draw new suppliers into the market, which would drive the market price back
down.
• Product. If the product is homogeneous, then each individual seller faces a
relatively elastic demand, so it can easily sell more than its quota. On the
other hand, if the product is homogeneous, it is also easier for the cartel
to monitor the various sellers. One way in which sellers circumvent a cartel
is to sell more and claim that the additional items sold are not covered by
the cartel agreement. Such cheating is difficult for a homogeneous product;
hence, it is easier to enforce the cartel. Accordingly, whether it is easier
or more difficult to enforce a cartel for a homogeneous or heterogeneous
product is unclear.
Horizontal Integration
Cartels that are illegal must rely on private enforcement to prevent sellers from
exceeding their quotas. However, competing sellers can restrain competition in a
way that does not raise the difficulties of enforcement. The alternative is for the
competing sellers to integrate.
Consider, for instance, a combination of two sellers, each of which has 50% of
a market. The combined business will have a monopoly. While it may be illegal for
two independent competitors to fix prices between themselves, it is certainly legal
for the two parts of the same company to agree on prices. Hence, a combination
that creates a monopoly will certainly be able to set price and
sales at monopoly levels, subject, of course, to the entry of Horizontal integration:
Combination of two
potential competitors.
entities, in the same or
The combination of two entities, in the same or similar similar business, under a
businesses, under a common ownership is called horizontal common ownership.
integration. By contrast, vertical integration is the com-
bination of the assets for two successive stages of production
under a common ownership. Chapter 14 analyzes vertical Vertical integration:
Combination of assets
integration.
for successive stages
The horizontal integration of any two businesses with of production under a
market power will lead to a reduction in the quantity sup- common ownership.
plied, and hence raise the market price and increase profits.
The increase in the market price will benefit competing sell-
ers as they will enjoy higher profits.
PROGRESS CHECK 11F

Explain the difference between horizontal and vertical integration.
268 11 Oligopoly
US MOBILE TELECOMMUNICATIONS: STILL FOUR
Verizon and AT&T dominate the US market for mobile telecommunications,

with Sprint and T-Mobile in distant third and fourth places. With the aim of
competing more effectively, Sprint proposed to merge with T-Mobile.
However, the federal government opposed a merger. The chairman of
the Federal Communications Commission, Tom Wheeler, explained: “Four
national wireless providers are good for American consumers.”
In August 2014, Sprint bowed to the inevitable and withdrew from merger
talks. Within weeks, Sprint and T-Mobile introduced new cheaper pricing plans.
Sprint’s new plan provided unlimited talk, text, and data for $60 a month,
$20 cheaper than T-Mobile, and Sprint offered $350 to subscribers switching
from other carriers. T-Mobile offered unlimited data service for 12 months to
customers introducing new subscribers.
Sources: “No merger of Sprint and T-Mobile US, but plenty of taunts,” New York Times, August 6,
2014; “T-Mobile, Sprint cut prices after merger talks,” Wall Street Journal, August 21, 2014.
KEY TAKEAWAYS
• For sellers competing on price to sell differentiated products, to maximize

profit, set price so that residual marginal revenue equals marginal cost.
• When demand or costs change, adjust price so that: (i) residual marginal revenue
equals marginal cost; and (ii) the best response price is a Nash equilibrium.
• If prices are strategic complements, then adjust price in the same direction as
competitors’ prices.
• For sellers competing on capacity to sell a homogeneous product, to maximize
profit, set capacity so that residual marginal revenue equals marginal cost.
• When demand or costs change, adjust capacity so that: (a) residual marginal
revenue equals marginal cost; and (b) the best response capacity is a Nash
equilibrium.
• If capacities are strategic substitutes, then adjust capacity in the opposite
direction to competitors’ capacities.
• To deter entry through limit pricing, set price so low that potential competitors
cannot break even.
• To apply capacity leadership, choose capacity to maximize profit, given that
competitors would choose capacity according to their best response functions.
• Limit competition through agreements and horizontal integration.
REVIEW QUESTIONS
1. Price will equal marginal cost only in a perfectly competitive industry. True or
false? Explain your answer.
Oligopoly 269
2. Explain the meaning of residual demand in the context of competition on:

(a) price; (b) capacity.
3. Explain the meaning of best response function in the context of competition
on: (a) price; (b) capacity.
4. If advertising expenditures are strategic complements, and your competitor
raises advertising, how should you respond?
5. In an oligopoly, how does differentiation raise profit?
6. How is Figure 11.3 related to a game in strategic form?
7. In competition between sellers on capacity to sell a homogeneous product,
how would an increase in fixed cost affect the equilibrium?
8. Explain the meaning of strategic substitutes.
9. Suppose that advertising expenditures are strategic substitutes. On a graph
like Figure 11.7, illustrate the best response functions.
10. Explain why limit pricing does not make sense in an industry where production
involves no fixed cost.
11. Suppose that your cost of capacity is sunk, once incurred. Does this help or
hinder a strategy of limit pricing?
12. Suppose that you can commit to production capacity before other sellers
set their capacity. Should you set a relatively larger or smaller capacity as
compared to the situation where you commit to capacity simultaneously with
other sellers?
13. In committing to production capacity before other sellers, why is it important
to look forward and consider how the other sellers would set capacity?
14. What are the five factors that influence the effectiveness of a cartel?
15. Is a cartel easier or more difficult to enforce in a market with less heterogeneous
products than in a market with more heterogeneous products?
1. Major US carriers, including American and United Airlines, operate on a hub-

and-spoke system, offering many frequent connections through their respective
hubs. Travelers may prefer one airline to another depending on which hub is
more convenient.
(a) Suppose that American and United compete on price to sell differenti-
ated products. If American cut fares on flights at its Dallas/Fort
Worth airport hub, should United raise or reduce fares at that
airport?
(b) Compare competition between American and United at Chicago’s
O’Hare International Airport, which is a hub for both airlines, and at
Dallas/Fort Worth Airport, which is a hub only for American. At which
hub would price competition be more intense?
(c) Suppose that the strength of traveler preferences over alternative hubs
were to diminish. How should airlines adjust their prices?
2. Among the three major national newspapers in the USA, the Wall Street
Journal, with average daily (Monday to Friday) circulation of 2.12 million, is
270 11 Oligopoly
reputed to be conservative. By contrast, the New York Times is reputed to be

liberal. Its average daily circulation is 916,900, of which 46% is sold in greater
New York City. (Sources: “Wall Street Journal still first in daily circulation,”
mediadecoder.blogs.nytimes.com, May 3, 2011; New York Times Company,
2009 Annual Report.)
(a) Explain how a media industry analyst could apply the Hotelling model
to competition between the New York Times and the Wall Street
Journal.
(b) Suppose that the marginal cost of producing the New York Times
increases by 20 cents. Using relevant diagrams, explain: (i) how the
New York Times should adjust its price; and (ii) how the Wall Street
Journal should respond.
(c) Suppose that people with more intense political preferences switch
from reading newspapers to listening to radio talk shows, while a new
generation of readers is more politically moderate. How should the
New York Times and Wall Street Journal adjust their prices to these
changes?
3. In the wake of expanding demand for their cars, Japanese manufacturers,
Toyota and Honda, have expanded production at existing factories and
established new plants in North America. In June 2006, Honda announced a new
$550 million factory, creating 2,000 jobs, at Greensburg, Indiana. Meanwhile,
US manufacturers, General Motors and Ford, facing declining demand, have
offered incentives for workers to quit, so as to reduce production capacity.
(a) Suppose that, in the short run, Ford and General Motors compete on
price to sell differentiated products. If GM increases price discounts,
how should Ford adjust prices?
(b) Use the Cournot model to relate the expansion of capacity by Toyota
and Honda to the contraction in capacity by Ford and General
Motors.
(c) Suppose that Toyota exercises capacity leadership. How would that
affect your explanation in (b)?
4. In February 2015, broadband and fixed line carrier BT (British Telecom)
re-entered the UK mobile market by buying EE. BT will be able to sell bundles of
service combining mobile, fixed line, broadband, and television. It will challenge
Vodafone, which sells only mobile services. (Source: “BT unveils £1bn share
placing to help fund EE takeover”, Telegraph, February 12, 2015.)
(a) How will BT’s purchase of EE affect the concentration of the UK mobile
services market?
(b) Suppose that the competing mobile carriers set capacity by the
Cournot model. If BT’s purchase does not affect the marginal cost of
EE, how would BT’s entry affect the equilibrium capacities?
(c) Using a suitable figure, explain how BT’s entry would affect the residual
demand for Vodafone and how Vodafone should adjust its prices.
5. Emirates and other Middle East carriers have expanded aggressively into the
“kangaroo” route between Britain and Australia. In 2013, Qantas and Emirates
Oligopoly 271
agreed to coordinate schedules, advertising, and pricing, and share revenues.

Qantas changed its kangaroo stopover from Singapore to Dubai. Subsequently,
Virgin Atlantic withdrew from the kangaroo market. (Source: “Qantas-Emirates
alliance receives full (draft) approval from Australia’s competition authority,”
Centre for Aviation, December 20, 2012.)
(a) Suppose that airlines set capacity according to the Cournot
model. If Emirates and Qantas reduce their marginal cost through
the code-share agreement, how should other airlines adjust their
capacity?
(b) Suppose that Emirates-Qantas exercises capacity leadership. How
would other airlines adjust their capacity as compared with (a)?
(c) How will the agreement between Qantas and Emirates to share
revenues affect their incentive to cut price?
(d) If Qantas and Emirates raise prices, how should other airlines
respond?
6. In 2010, of a total of 67,000 rooms on the Las Vegas Strip, Caesars
Entertainment managed 22,880, while MGM Resorts managed over 12,000.
However, owing to the Great Recession and new hotel openings, between
2008 and 2010, MGM’s hotel occupancy decreased from 92% to 89%, while
its average daily room rate fell from $148 to $108. Meanwhile, CityCenter,
managed by MGM Resorts, and the Cosmopolitan opened with 4,000 and
3,000 rooms respectively, and the 1,720-room Sahara closed. (Sources:
Caesars Entertainment Corp., Annual Report 2010; MGM Resorts, Annual
Report 2010; “Sahara’s closure on May 16 will mark ‘the end of an era,’” Las
Vegas Sun, March 11, 2011.)
(a) Using a suitable figure, explain how the opening of CityCenter and the
Cosmopolitan affects the residual demand for an existing hotel and
how the existing hotel should adjust prices.
(b) If MGM Resorts had not reduced its room rates, what would have been
the effect on occupancy?
(c) Use the Cournot model to explain MGM Resorts’ opening a new hotel
and Sahara’s closing.
7. The OPEC cartel sets production quotas for each member country. As of May
2006, OPEC produced 33.33 million barrels per day (mbd), with an excess
capacity of between 1.3 and 1.8 mbd. Virtually all of the excess capacity was in
Saudi Arabia, while several OPEC members exceeded their production quotas.
(Sources: US Energy Information Administration, Short Term Energy Outlook,
June 6, 2006; Energy Economist, May 12, 2004; BP Statistical Review of World
Energy, June 2006.)
(a) Explain why OPEC members individually would wish to produce more
than their production quota. How would that affect the price of oil and
producer profits?
(b) Suppose that Saudi Arabia enforces the cartel by threatening to
increase production if any OPEC member exceeds its quota. Why is it
important that Saudi Arabia has large excess capacity?
272 11 Oligopoly
(c) Consider the production of oil by non-OPEC members. Suppose that

they have no short-run excess capacity. How do they affect OPEC’s
ability to raise prices in the short and long run?
8. DRAM is used to store data in PCs and other consumer and industrial electronic
products. DRAM manufacturing is a capital-intensive industry. In 2004, Elpida,
Hynix, Infineon, Micron, and Samsung produced 80% of the world DRAM
supply. In the early 2000s, executives of the five companies agreed to fix the
prices of DRAMs for sale to computer manufacturers. They held meetings to
discuss prices, and exchanged information on sales to particular customers.
(a) How easy would it be to enforce the DRAM cartel?
(b) Why did the DRAM cartel members provide sales information to each
other?
(c) Supposing that the DRAM cartel members wished to limit the exchange
of information, which customers should they select to monitor?
9. Gesamtmetall is the federation of employers and IG Metall is the labor union
in Germany’s engineering industry. Under the German system of national
collective bargaining, Gesamtmetall and IG Metall negotiate pay and working
conditions for the entire industry.
(a) With respect to labor, does Gesamtmetall serve as a buyer cartel or
seller cartel? What about IG Metall?
(b) Consider large employers such as Daimler Benz and Robert Bosch. Why
might they prefer to negotiate separate deals with their own workers
rather than comply with the national collective agreement?
(c) If all large employers negotiate separate deals, how will this affect the
wages and conditions that small companies must offer?

Suppose that your major competitor has increased production capacity. Write
a memorandum to the board of directors explaining how your organization
should respond.
Appendix: Solving Equilibria
Here, we derive the major results of this chapter with a linear demand curve,
Q = a − bp, and constant marginal cost, c. To get the numerical results presented
in the chapter, substitute a = 30, b = 3/10, and c = 30.
Monopoly
A monopolist facing a linear demand curve will produce at the scale where mar-
ginal revenue equals marginal cost. First, rewrite the demand curve for price as a
function of quantity (the inverse demand function)
Oligopoly 273
a 1
p= − Q. (11.1)
b b
Then profit is total revenue minus total cost, or
⎛a 1 ⎞ a 1
Π( ) = ⎜ − Q Q − cQ = Q − Q 2 cQ. (11.2)
⎝b b ⎠ b b
To identify the profit-maximizing scale, differentiate (11.2) with respect to pro-

duction, Q, and set the derivative equal to zero. After simplifying, we have
1
Q∗ ( a − bc ).
2
Cournot Model: Capacity Competition with

Homogeneous Product
In the Cournot model, the market price is the price at which the quantity
demanded equals the capacity of the two sellers. By (11.1), with two sellers, Luna
and Mercury, the inverse market demand is
a 1
p= − (Q
(QL QM ), (11.3)
b b
where QL represents Luna’s capacity and QM represents Mercury’s capacity.

Each seller’s profit depends on the other seller’s capacity. Luna’s profit is
⎡a 1 ⎤
Π L = p QL c QL = − (QL QM )⎥ QL cQL ,
⎣b b ⎦
after substituting from (11.3). To identify the profit-maximizing capacity, differ-

entiate the profit with respect to Luna’s choice of capacity, QL, and set the deriva-
tive equal to zero. After simplifying, we have Luna’s best response function,
1 (11.4)
QL ( a bc QM ).
2
Similarly, we can derive Mercury’s best response function,
1 (11.5)
QM ( a bc QL ).
2
274 11 Oligopoly
Figure 11.7 graphs the two best response functions.

In the Nash equilibrium, strategies QL and QM would satisfy both (11.4) and
(11.5). Solving these two equations simultaneously, we have the equilibrium
capacities,
1
QL* *
QM = ( a bc ).
3
Differing Costs
If the sellers have different marginal costs, cL and cM, then we can use the same
procedure as above to solve for the Nash equilibrium,
1
QL* [aa b cL − cM )]
3
and
* 1
QM [aa b cM cL )].
3
Hotelling Model: Price Competition with

Differentiated Products
The setting of the Hotelling model differs from the setting of the monopoly and
Cournot duopoly. Here, we solve the equilibrium of a duopoly, where Luna and
Mercury each produce at constant marginal cost, c, and consumers are willing to
pay $100 for the service and incur a “transport cost” of t per mile to buy.
The key to calculating the demand for each seller is to identify the marginal
consumer, located at some point, x*, who is just indifferent between buying from
the two sellers. Generally, for a consumer at location x, her expected utility from
Luna’s product would be
100 − pL − t × x. (11.6)
That consumer would lie at a distance 1 − x from Mercury, so her expected

utility from Mercury’s product would be
100 − pM − t × (1 − x ). (11.7)
Hence, for the marginal consumer, the two expected utilities must be equal.
Equating (11.6) and (11.7),
100 − pL − tx 100 − pM − t(1

tx* =100 ( − x*).
Oligopoly 275
After simplifying, the location of the marginal consumer is
1
x∗ ( pL , pM ) = ( pM − pL + t ). (11.8)
2t
Consumers located to the left of x* would buy from Luna, while those to the
right would buy from Mercury. So, x* is the market share of Luna and 1− x* is
the market share of Mercury.
Consider Luna. Its profit (as a function of Mercury’s price) is
1
Π L = ( pL − c)
c ) x* = ( pL c) ( pM pL t ),
2t
after substituting from (11.8). To identify the profit-maximizing price, differenti-

ate the profit with respect to Luna’s price, pL, and set the derivative equal to zero.
After simplifying, we have Luna’s best response function,
1 (11.9)
pL = ( pM t c ).
2
Similarly, we can derive Mercury’s best response function,
1 (11.10)
pM = ( pL t c ).
2
Figure 11.3 graphs the two best response functions.

In the Nash equilibrium, strategies, pL and pM, would satisfy both (11.9) and
(11.10). Solving these two equations simultaneously, we have the equilibrium
prices,
p *L p *M = c t.
Notes
1 This chapter is more advanced. All but the section on capacity leadership (Section 5) may be
omitted without loss of continuity. The section on capacity leadership provides background
relevant to Chapter 15 on regulation.
2 The following discussion is based, in part, on Clyde Montevirgen, “Standard & Poor’s Industry
Surveys: Semiconductors,” May 12, 2011; Elpida Memory, Inc., FY2010 Financial Review &
Business Updates, May 12, 2011; “Qimonda,” Wikipedia (accessed August 12, 2011).
3 This section is based on Jeremy Bulow, John Geanakoplos, and Paul Klemperer, “Multimarket
oligopoly: strategic substitutes and complements,” Journal of Political Economy, Vol. 93, No. 3,
June 1985, pp. 488–511.
PA R T
III
Imperfect Markets
C H A P T E R
12
Externalities
LEARNING OBJECTIVES
• Understand positive and negative externalities.
• Appreciate economic efficiency as a benchmark and the opportu-
nity to profit by resolving an externality.
• Appreciate how to resolve externalities and the corresponding
barriers.
• Understand network effects and appreciate how to manage
demand with network effects.
• Understand public goods.
• Appreciate how to provide public goods commercially.
1. Introduction
With over 1 million square feet of net leasable area, VivoCity is Singapore’s larg-
est mall. It is located on the western edge of the city, at the intersection of two
subway lines, and attracts 40 million visitors annually.1
In 2010, 98.3% of VivoCity was leased to tenants, generating S$116.3 million in
gross rental income. The mall’s retail tenants pay a three-part rental: 86% is fixed,
13% is based on the retailer’s sales revenue, while the remainder comprises service
charges and contributions to advertising and promotion.
VivoCity’s two largest retail tenants, renting 22.4% of the mall, are VivoMart, a
combination of a hypermarket, upscale supermarket, and drug store, and Tangs
Department Store. In 2010, VivoMart and Tangs paid an average rent of S$56 per
square foot per year. By contrast, fashion stores paid almost three times as much,
an average rent of S$157 per square foot per year.
280 12 Externalities
Why does VivoCity charge systematically lower rentals to supermarkets and

department stores than to specialty retailers? Why must VivoCity tenants contrib-
ute to a fund for advertising and promotion? Why do real estate developers build
malls near transportation facilities such as subways?
We can address these questions through the concepts of externality and public
good. An externality arises when one party directly conveys a benefit or cost to
others (not through a market). An anchor store generates a
Externality: A benefit or positive externality for other stores. Department stores and
cost directly conveyed to cineplexes can serve as anchors. By drawing additional shop-
others.
pers to the mall, they directly convey benefits to specialty
stores.
Similarly, transport facilities generate externalities for retailers by moving con-
sumers in large numbers. This explains why developers choose to locate shopping
malls near subways and other transport facilities.
Here, we introduce the concept of an externality, characterize its economically
efficient level (which maximizes value added), and then discuss how to achieve that
level. Economic efficiency can be achieved through collective action of the parties
involved or by merging the source and beneficiary of the externality. For instance,
VivoCity charges lower rents to supermarkets and department stores. They attract
shoppers who benefit fashion stores, which pay relatively higher rents to VivoCity.
A particular class of externalities is network externalities. These arise when a
party directly conveys benefit or cost to everyone else in the same network. Net-
work externalities underlie the rapid growth of the Internet. The same concept
applies more widely to communications, media, and high-technology industries.
Finally, we introduce the concept of a public good, charac-
Public good: One terize its economically efficient level, and discuss how to achieve
person’s increase in
that level. An item is a public good if one person’s increase in
consumption does not
reduce the quantity consumption does not reduce the quantity available to others.
available to others. Broadcast TV nicely illustrates a public good: one hour of
TV can serve 100,000 or 1 million viewers with no difference
in cost. In the provision of a public good, there is an extreme economy of scale
with respect to the number of customers served (not with respect to volume of
production). For tenants of a shopping mall, expenditures by the mall on adver-
tising and promotion that attract more shoppers are a public good. The addi-
tional shoppers benefit all the stores in the mall.
In any market with externalities and public goods, managers can increase value
added (and profit) by bringing provision closer to the economically efficient level.
This chapter explains how to do so in a wide range of contexts.
2. Benchmark: Economic Efficiency
Externalities can be positive or negative. A positive externality arises when one

party directly conveys a benefit to others. The additional customers that an anchor
Externalities 281
store generates for nearby specialty retailers constitute a pos-

Positive externality:
itive externality. By contrast, a negative externality arises A benefit directly
when one party directly imposes a cost on others. A betting conveyed to others.
shop imposes a negative externality on nearby toy stores by
discouraging families with children from browsing nearby. Negative externality:
A cost directly imposed
on others.
Positive Externality
To explain the benchmark level of a positive externality, consider the example of a
department store on Main Street. It must decide how much to spend on advertising
to attract shoppers. To attract more shoppers, it must spend more on advertising.
The additional shoppers would increase the store’s sales and profit contribution.
Table 12.1 presents the department store’s revenues and costs at the various
levels of customer traffic on a monthly basis. For instance, with 50,000 customers
a month, the store’s sales revenue would be $1.9 million, while the variable cost
would be $1.0 million, and so the profit contribution would be $900,000. The cost
of advertising to attract the 50,000 customers is $200,000. Accordingly, the over-
all profit would be $700,000.
The marginal profit contribution is the increase in profit contribution associated
with the increase in customers. At 50,000 customers a month, the marginal profit
contribution is $(900 − 800)/(50 − 40) = $10 per customer. The marginal customer
cost is the increase in advertising expenditure associated with the increase in cus-
tomers. At 50,000 customers a month, the marginal customer cost is $(200 − 160)/
(50 − 40) = $4 per customer.
What level of customer traffic maximizes profit? Referring to Table 12.1, the
department store maximizes profit at 60,000 customers a month. Its maximum
Table 12.1 Customer traffic and profit
Customers Revenue Variable Profit Marginal Advertising Marginal Profit

(’000) ($ ’000) cost contri- profit expenditure customer ($ ’000)
($ ’000) bution contribu- ($ ’000) cost ($)
($ ’000) tion ($)
0 0 0 0 0 0
10 460 200 260 26 40 4 220
20 880 400 480 22 80 4 400
30 1,260 600 660 18 120 4 540
40 1,600 800 800 14 160 4 640
50 1,900 1,000 900 10 200 4 700
60 2,160 1,200 960 4 240 4 720
70 2,380 1,400 980 2 280 4 700
80 2,560 1,600 960 −2 320 4 640
90 2,700 1,800 900 −6 360 4 540
100 2,800 2,000 800 −10 400 4 400
110 2,860 2,200 660 −14 440 4 220
37.20
Marginal benefit/cost ($ per customer per month)
Combined marginal
profit contribution
26
Possible increase in
Dept store marginal combined profit
profit contribution contribution
11.20
Restaurant marginal
profit contribution Dept store
marginal cost of
4
3.20 advertising
0 60 69 80
Customer traffic at department store (’000 customers a month)
FIGURE 12.1 Positive externality.

Notes: The combined marginal benefit (profit contribution) is the vertical sum of the individual marginal
benefits (profit contributions). The economically efficient level of customers is 69,000, where the
combined marginal benefits equal the marginal cost. If the department store considers only its own
benefit and cost, it would attract only 60,000 customers. By increasing to 69,000 customers, the
department store and restaurant could gain the shaded area in additional profit.
profit is $720,000. At 60,000 customers a month, the marginal profit contribution

and the marginal customer cost are both $4 per customer.
Figure 12.1 represents the same analysis graphically. The level of customer traf-
fic that maximizes the department store’s profit is that where the marginal profit
contribution equals the marginal customer cost of $4 per customer.
The customers that the department store attracts would also shop at nearby
specialty retailers selling fashion, lifestyle items, and food and beverages. How-
ever, there is no market through which specialty retailers pay the department store
for the additional business. By definition, a positive externality conveys a benefit
directly rather than through a market. Accordingly, the customers are an externality
from the department store to the specialty retailers.
Consider a restaurant situated next to the department store. Table 12.2 pres-
ents the restaurant’s revenues and costs at the various levels of the department
store’s customer traffic on a monthly basis. For instance, with 60,000 customers a
month, the restaurant’s sales revenue would be $1.392 million, while the variable
cost would be $960,000, and so profit contribution and profit would be $432,000.
The restaurant does not spend anything on advertising, so its profit contribution
is the same as its profit.
Externalities 283
Table 12.2 Customer traffic and externality
Customers Revenue Variable cost Profit Marginal

(’000) ($ ’000) ($ ’000) ($ ’000) profit ($)
10 272 160 112 11.20

20 528 320 208 9.60
30 768 480 288 8.00
40 992 640 352 6.40
50 1,200 800 400 4.80
60 1,392 960 432 3.20
70 1,568 1,120 448 1.60
80 1,728 1,280 448 0.00
90 1,872 1,440 432 −1.60
100 2,000 1,600 400 −3.20
110 2,112 1,760 352 −4.80
By the nature of an externality, the source of the externality considers only

the benefits and costs to itself, while ignoring the benefits and costs to others. So,
acting independently, the department store chooses to attract 60,000 customers a
month. Referring to Figure 12.1, at that level of customer traffic, the restaurant’s
marginal profit is $3.20 per customer.
If the department store were to attract a 60,001st customer, that customer
would raise the department store’s profit contribution by a little less than $4, raise
the department store’s cost of advertising by $4, and raise the restaurant’s profit
by $3.20. So, the combination of the department store and restaurant would be
better off by slightly less than $3.20.
Indeed, the department store could continue to raise the combined profit con-
tribution of the department store and restaurant by increasing customer traffic
up to the economically efficient level. Recall from Chapter 6 that, at the econom-
ically efficient level, the marginal benefit equals the marginal cost. In the present
context, the department store’s marginal benefit from customers is its marginal
profit contribution, while the restaurant’s marginal benefit from customers is its
marginal profit contribution.
So, for economic efficiency in customer traffic, the number of customers should be
such that the combined marginal benefit (department store’s plus restaurant’s mar-
ginal profit contribution) equals the marginal cost of attracting customers. Referring
to Figure 12.1, the combined marginal benefit is the vertical sum of the department
store’s marginal profit contribution and the restaurant’s marginal profit contribution.
For instance, at 60,000 customers, the combined marginal benefit is $4 + $3.20 = $7.20.
The economically efficient level of customer traffic is 69,000 customers. By
increasing the number of customers from 60,000 to 69,000, the department store
and restaurant can collectively gain the shaded area in additional profit contribu-
tion. So, if the department store could collect a fee from the restaurant for gener-
ating customer traffic, it would maximize total profit (from sales to customers and
fee from restaurant) at 69,000 customers.
Another perspective is to suppose that the department store owns the restau-
rant. Then the department store would choose customer traffic to maximize the
combined profit contribution of the store and restaurant. It would increase the num-
ber of customers up to the level that the combined marginal profit contribution
equals the marginal cost of attracting customers, at 69,000 customers a month.
PROGRESS CHECK 12A

Suppose that the restaurant’s marginal profit contribution is lower by $3.20
at all levels of customer traffic. Revise Figure 12.1 to show the economically
efficient level of customer traffic.
Negative Externality
To explain the benchmark level of a negative externality, consider the example of
a betting shop, located on Main Street next to the restaurant. The betting shop
must decide how much to spend on advertising to attract gamblers.
Figure 12.2 represents the betting shop’s marginal profit contribution and mar-
ginal cost of customers. The level of customer traffic that maximizes the betting
Marginal benefit/cost ($ per customer per month)
Possible increase in combined

profit contribution
Betting shop marginal

profit contribution
Combined marginal cost
Restaurant marginal cost
4
Betting shop
marginal cost of
2 advertising
0 7 10
Customer traffic at betting shop (’000 customers a month)
FIGURE 12.2 Negative externality.

Notes: The combined marginal cost is the vertical sum of the individual marginal costs. The
economically efficient level of customers is 7,000, where the marginal benefit equals the combined
marginal costs. If the betting shop considers only its own benefit and cost, it would attract 10,000
customers. By reducing to 7,000 customers, the betting shop and restaurant could gain the shaded
area in additional profit.
Externalities 285
shop’s profit contribution is that where the marginal profit contribution equals
the marginal customer cost of $2 per customer. That level is 10,000 customers a
month.
However, the betting shop’s customers discourage families with children from
patronizing the next-door restaurant. So, the betting shop imposes a cost on the
restaurant. Since it imposes the cost directly and not through a market, the cost
is a negative externality.
Suppose that the greater the number of customers that the betting shop attracts,
the higher is the restaurant’s marginal cost, as measured by the restaurant’s mar-
ginal loss of profit contribution. Figure 12.2 also shows the restaurant’s marginal
cost, as a function of the betting shop’s customer traffic.
If the betting shop advertises to get 10,000 customers, the 10,000th customer
imposes a marginal cost of $4 on the restaurant. If the betting shop were to reduce
its customers by one, its profit contribution would be $2 lower, its advertising cost
would be $2 lower, and the restaurant’s cost would be $4 lower. So, the combina-
tion of the betting shop and restaurant would be better off by $4.
So, the reduction in customers would raise the combined profit contribution of
the betting shop and restaurant. Indeed, the betting shop could increase the com-
bined profit contribution by reducing customer traffic down to the economically
efficient level. The economically efficient level is where the betting shop’s marginal
benefit (marginal profit contribution) equals the combined (betting shop’s plus
restaurant’s) marginal cost of customers.
Referring to Figure 12.2, the combined marginal cost is the vertical sum of
the betting shop’s marginal cost of advertising and the restaurant’s marginal cost
(loss of profit contribution). For instance, at 10,000 customers a month, the vertical
sum is $2 + $4 = $6.
The economically efficient level of customer traffic is 7,000 customers. By
reducing the number of customers from 10,000 to 7,000, the betting shop and
restaurant can collectively gain the shaded area in additional profit contribution.
So if the betting shop could collect a fee from the restaurant for reducing cus-
tomer traffic, it would maximize total profit (from sales to customers and fee from
restaurant) at 7,000 customers.
Another perspective is to suppose that the betting shop owns the restaurant.
Then the betting shop would choose customer traffic to maximize the combined
profit contribution of the betting shop and restaurant. It would reduce the num-
ber of customers up to the level that the combined marginal profit contribution
equals the marginal cost of attracting customers, at 7,000 customers a month.
PROGRESS CHECK 12B

Referring to Figure 12.2, suppose that the negative externality on the restau-
rant is stronger. How will this affect (a) the restaurant’s marginal cost curve,
and (b) the number of customers that maximizes the group profit contribution?
EXTERNALITIES IN TALENT: SILICON VALLEY
There is an exceptional concentration of information technology related

businesses in the peninsula between the San Francisco Bay and the Pacific
Ocean. The area, aptly nicknamed Silicon Valley, is home to high-technology
leaders such as Apple, Hewlett-Packard, Intel, Google, and Facebook.
Stanford University and the Xerox Palo Alto Research Center (PARC) played
key roles in the development of Silicon Valley. Basic and applied research at
the two institutions has provided the foundation for many successful high-
tech companies.
A local area network links separate personal computers over short
distances. Robert Metcalfe and David Boggs invented the Ethernet local area
network at Xerox PARC. In 1979, Metcalfe left Xerox PARC to found 3Com
(the three “coms” being computer, communication, and compatibility), which
commercialized the Ethernet technology.
Routers are devices that link computer networks that use different protocols.
Sandy Lerner and Len Bosack started Cisco Systems while working at
Stanford. In 1986, they left the university to run Cisco full-time. The company
became the world’s largest manufacturer of data-networking systems.
In 1996, two Stanford computer science graduate students, Sergey Brin and
Lawrence Page, developed an algorithm for Internet search. Their research
blossomed into Google, now a multi-billion-dollar company.
Sources: Douglas K. Smith and Robert C. Alexander, Fumbling the Future, New York: William
Morrow, 1988, pp. 95–103; letter from Robert Metcalfe, July 16, 1996; letter from Cisco
Systems, May 23, 1996; www.cisco.com and www.3com.com (accessed May 22, 1996), www.
google.com (accessed August 9, 2011).
General Benchmark
We have separately discussed positive and negative externalities. Generally, in the
presence of both positive and negative externalities, the affected parties maxi-
mize combined profit at the following benchmark: where the
Externality resolved: combined marginal benefits equal the combined marginal
Combined marginal costs, which is the economically efficient level of the external-
benefits equal combined
ity. An externality is resolved when the combined marginal
marginal costs.
benefits equal the combined marginal costs.
When the source of an externality considers only its own
benefit and cost, it overlooks an opportunity for increasing profit. The source of
a positive externality can raise profit by collecting a fee from the beneficiaries and
increasing the level of externality. By contrast, the source of a negative external-
ity can raise profit by collecting a fee from the sufferers and reducing the level of
externality. There will be an opportunity for such a profit whenever the combined
marginal benefits differ from the combined marginal cost.
Externalities 287
The benchmark of economic efficiency applies whether or not the externali-

ties are monetary. To take account of non-monetary externalities, just measure
the benefits and costs in terms of money, and then apply the economic efficiency
analysis.
ACADEMIC EXTERNALITIES: A FREE CAMPUS
In May 1995, the regents of the University of California met to choose the
location for a new campus to become the tenth branch of the university.
The regents considered two alternatives: a 2,000-acre site at Lake Yosemite
offered free of charge by the Virginia Smith Trust and another site near the
city of Fresno that had to be purchased. Swayed in part by the free land, the
regents voted for Lake Yosemite.
The Virginia Smith Trust owned 7,000 acres of land by Lake Yosemite,
used to graze cattle. The Trust projected that, with establishment of the
new campus, it could earn $350 million from development of the remaining
5,000 acres of real estate. The positive externalities were sufficient to justify
the “gift” to the University of California. UC Merced opened in 2005, and
presently has over 6,000 students.
Source: “Lake Yosemite selected for proposed campus,” UC Focus (Office of the President,
University of California), Vol. 9, No. 5, June–July 1995, pp. 1 and 7.
3. Resolving Externalities
The benchmark for an externality is the economically efficient level, where the
combined marginal benefits equal the combined marginal costs. Chapter 6 showed
that, in a perfectly competitive market, the invisible hand ensures economic effi-
ciency. By definition, however, externalities do not pass through markets. Hence,
externalities will be resolved only through deliberate action.
Generally, externalities can be resolved in two ways – through common owner-
ship of the source and recipient, or by agreement among the source and recipient.
However, there are two obstacles to the resolution of externalities – the assign-
ment of rights and possible free riding.
Common Ownership
In the example of the department store and restaurant, if the department store
owned the restaurant, then the store would consider all the benefits and costs
of additional customers. Hence, it would increase advertising to attract the
economically efficient number of customers. Similarly, in the example of the bet-

ting shop and restaurant, if the betting shop owned the restaurant, then it would
reduce advertising to attract the economically efficient number of customers.
These examples suggest one way to resolve an externality: combine the source
and recipient of the externality under common ownership. The common owner
would choose the economically efficient level of the externality, whether positive
or negative.
Agreement
Besides common ownership, another way to resolve externalities is by agree-
ment. The source and recipient of the externality could negotiate and agree
on the level of the externality. For instance, the restaurant could offer to pay
part of the department store’s advertising cost. This would encourage the store
to increase advertising. Similarly, the restaurant could pay the betting shop to
reduce advertising.
An agreement to resolve an externality involves two steps. First, the affected
parties must agree on how to resolve the externality. This step involves collecting
information about the benefits and costs to the various parties, and then agreeing
on the level of the externality.
The second step is to enforce compliance with the agreement. Enforcement
includes monitoring the source of the externality and applying incentives to
ensure that the source complies with the agreed level of the externality.
Assignment of Rights
One hurdle to resolution of externalities is unclear assignment of rights. In
the example of the betting shop and restaurant, the externality might actually
be resolved in two possible ways. The betting shop could reduce its advertising.
Another way is for the restaurant to move away.
So does the betting shop have the right to impose an externality on the restau-
rant? Or does the restaurant have the right not to suffer an externality? If rights
are not clearly assigned, then the two parties would have difficulty agreeing on
common ownership or the level of the externality.
The assignment of rights is an acute issue in externalities from a common
resource. Consider several oil producers drilling from the same oil field, multiple
trawlers fishing the same stock, or various lumber producers cutting from the
same forest. In these situations, without a clear assignment of property rights, the
various competitors rush to grab production, ignoring the long-term impact on
the resource.
Free Riding
The other hurdle is possible free riding. In the department store example, sup-
pose that the department store customers also patronize nearby retailers that sell
Externalities 289
fashion, lifestyle items, and food and beverages. So, the department store provides
positive externalities to all of these specialty retailers.
We have shown that the source and recipients of the externalities can resolve the
externalities through common ownership or agreement. Now consider a florist
that benefits from the department store’s positive externalities. The florist might
adopt a strategy of not paying the department store for advertising. The florist
knows that the department store would still advertise, while other specialty retail-
ers might still pay to increase the advertising. The department store and other
specialty retailers cannot prevent their customers from patronizing the florist. So,
the florist can benefit from the customers without paying anything.
In this strategy, the florist is taking a free ride on the advertising by the depart-
ment store and fees paid by the other specialty retailers. Generally, a free rider
pays less than its marginal benefit to resolution of the exter-
nality. At the extreme, the free rider pays nothing at all. Free rider: Contributes
less than marginal benefit
Free riding arises whenever it is costly to exclude those to resolution of externality.
who do not pay from benefiting from the externality. When it
is easy to exclude those who do not pay, then all beneficiaries
of the externality will pay for resolution, otherwise they would not benefit. Free
riding is more serious when the externality affects many recipients. When an exter-
nality affects many recipients, the amount that any particular recipient must pay
is relatively small. Hence, the other recipients may pay to resolve the externality
even if some recipients take a free ride.
PROGRESS CHECK 12C

In the example of the betting shop and restaurant, what is the possible hurdle
to resolving the externality?
VIVOCITY: RESOLVING EXTERNALITIES
With over 1 million square feet of net leasable area, VivoCity is Singapore’s largest
mall, attracting 40 million visitors a year. VivoCity’s two largest retail tenants are
VivoMart, a combination of a hypermarket, upscale supermarket, and drug store,
and Tangs Department Store, which collectively occupy 22.4% of the mall.
VivoMart and Tangs, together with a cineplex and several medium-sized
stores, what VivoCity management calls “mini anchors,” pull customers into
the mall. The customer traffic benefits smaller specialty stores, which occupy
33% of the mall.
In 2010, on a per square foot basis, the major and mini anchors achieved
average sales of S$678 a year while paying rent of S$78 a year. By contrast,
the specialty retailers achieved average sales of S$950 a year and paid an
average rent of S$196 a year. So, the specialty retailers paid higher rent in
both dollar terms and as a proportion of sales.
VivoCity resolves the externality between anchor tenants and other retailers
by charging relatively lower rents to anchors and higher rents to other retailers.
This is an established practice in the management of shopping malls. Through
common management, retailers in a shopping mall can resolve externalities
more effectively than retailers distributed on a public street.
Source: Mapletree Commercial Trust, Prospectus, April 18, 2011.
MICKEY MOUSE EXTERNALITIES
The Walt Disney Company owns and manages Disneyland in Anaheim,

California. This theme park is surrounded by hundreds of businesses, including
motels, restaurants, souvenir stores, and transportation services. Disneyland
visitors are the major source of income for these neighboring businesses.
In the late 1980s, Disney decided to embark on a large program of invest-
ments to upgrade Disneyland. Before commencing construction, the company
secretly bought much of the property around the theme park, including the
Disneyland Hotel for $200 million. By purchasing the surrounding property,
Disney ensured that it would capture the benefits from new attractions.
Consequently, the company had a greater incentive to make the economically
efficient level of investment in new attractions at the theme park.
Disney applied the same principle when developing its theme park on
Lantau Island, Hong Kong. As a condition of the investment, the Hong Kong
government awarded Disney a 20-year option to purchase an adjoining site.
The government further agreed to restrict economic activity in the vicinity.
It limited the height of nearby buildings, banned aircraft from flying over the
park, and excluded ships from the seafront.
Hong Kong Tourism Commissioner Mike Rowse explained: “It is an essential
element of a Disney theme park that people outside the park not be able to
see in, and those inside not be able to see the ‘real world’ outside.”
Sources: Gary Wilson, Chief Financial Officer, Disney Company, speech at the Anderson
School, UCLA, March 15, 1989; “Disney given controls over area around park,” South China
Morning Post, November 20, 1999.
4. Network Effects and Externalities
A network effect is a benefit or cost that increases with

Network effect: A the size of the network. The adjective network emphasizes
benefit or cost that that the benefit or cost is generated by the entire network of
increases with the size of users. An instant messaging service nicely illustrates a net-
the network. work effect. The benefit that one user derives from an instant
messaging service increases with the number of other users.
Externalities 291
The marginal benefit and demand from an item that exhibits network effects
increases with the number of other users. For instance, when one more person
subscribes to an instant messaging service, the marginal benefit and demand
curves of all other users will shift up.
Related to the concept of a network effect is the network externality. A network
externality is a benefit or cost directly conveyed to others
that increases with the size of the network. Accordingly,
Network externality:
a network externality is a network effect that is conveyed A benefit or cost directly
directly, and not through a market. As with externalities conveyed to others that
in general, the benchmark for a network externality is eco- increases with the size of
nomic efficiency: the combined marginal benefits equal the the network.
combined marginal costs.
In markets with network effects, the character of demand and competition dif-
fers from that in conventional markets. We discuss each of these differences below.
Critical Mass
In a market with network effects, the demand may be zero unless the number of
users exceeds critical mass. The critical mass is the number of users at which the
demand becomes positive. Consider, for instance, instant
messaging service. There is no demand if the number of Critical mass: The
number of users at
users falls below some level. Who would want instant mes-
which demand becomes
saging if there is no one to communicate with? positive.
Suppose that demand for instant messaging service is zero
if there are fewer than 10,000 users. Then the critical mass
is 10,000 users. The demand will be positive only if the price or other factors are
sufficient to attract 10,000 users.
Technical standards play a key role in markets with network effects. Consider
two incompatible instant messaging services. If each can attract at most 5,000
users, then neither will achieve critical mass. However, if both services conform to
a common technical standard and interconnect, then they can collectively achieve
the critical mass of 10,000.
The demand for some items depends on the presence of complementary hard-
ware. For instance, the demand for web-based social media
depends on consumer access to computers and smart phones. Installed base:
Similarly, the demand for Android apps depends on the The quantity of
complementary hardware
number of Android smart phones and tablets. In this con- in service.
text, the installed base is the stock of the complementary
hardware in service.
Expectations
In a market with network effects, user expectations are important. Consider a
situation where users simultaneously decide whether to subscribe to a new instant
messaging service. There are two possible equilibria. In the good equilibrium,
every potential user expects the others to subscribe, and accordingly subscribes.
Then demand exceeds critical mass and the service succeeds as expected. By
contrast, in the bad equilibrium, potential subscribers are pessimistic. Each one
expects less than the critical mass to subscribe, and so does not subscribe. Then
demand fails to reach critical mass and the service flops as expected.
Thus user expectations are very important. In equilibrium, expectations –
whether pessimistic or optimistic – can be self-fulfilling. How to influence the
expectations of potential users? One way is through commitments to ensure that
demand will exceed critical mass. For instance, sellers of items with network effects
may give away large quantities in order to establish a sufficient installed base.
Another way of influencing expectations is hype. For instance, a grand launch
attended by famous athletes and movie stars may generate the self-fulfilling
prophecy that demand for the item will attain critical mass.
Tipping
The demand in markets with network effects may be extremely sensitive to small
differences among competitors. Suppose there are several competing products,
each of whose demand is close to critical mass. Then a small
Tipping: The tendency increase in the user base of one product can tip the market
for demand to shift demand toward that product. Tipping is the tendency for
toward a product with a
small initial lead.
demand to shift toward a product that has gained a small
initial lead.
To illustrate tipping, consider two providers competing to
introduce new instant messaging services. Each can guarantee 9,000 users. Then
the market would tip toward the service which can get the additional 1,000 users
and cross the threshold for critical mass. The other service would fail.
Accordingly, a market with network effects may “tip” toward one product. By
contrast, in a conventional market, several competitors of similar size may con-
tinue profitably for a long time. Even if one seller gains an advantage in pricing or
product quality, the entire market demand will not tip in its favor.
Price Elasticity
Chapter 3 introduced the concept of (own) price elasticity of demand, which is
the percentage by which the quantity demanded changes if the price of the item
rises by 1%. The presence of network effects influences the price elasticity in dif-
ferent ways, depending on whether the demand has reached critical mass.
For a product with network effects in demand, when the demand is below
critical mass, the demand will be zero, and hence extremely price inelastic. Price
reductions, however large, will not increase the demand at all.
The demand will be sensitive to price only when it exceeds critical mass. Then the
network effect causes the market demand to be relatively more price elastic. Con-
sider, for instance, a price increase. This would reduce the demand. The reduction
Externalities 293
would feed back through the network effect to further reduce the demand. The
network effect tends to amplify the effect of a price increase on demand. Simi-
larly, the network effect would amplify the effect of a price reduction.
Around the critical mass, the demand would be very price elastic. A slight price
cut would pull demand above critical mass and further increase demand through
the network effect. By contrast, a slight price increase would push demand below
critical mass and then demand would collapse to zero.
PROGRESS CHECK 12D

How do conventional markets differ from markets with network effects?
HARNESSING NETWORK EXTERNALITIES: GOOGLE
Google harnesses network externalities in a very subtle way. Whenever a

user submits a search to Google, the service uses its proprietary page-rank
algorithm to display links to webpages in order of predicted relevance. Google
tracks the links that people click on and updates its algorithm accordingly.
Every search helps to refine Google’s algorithm, and so, the more people
search on Google, the better the quality of its service.
5. Public Goods
Open-air fireworks nicely illustrate the concept of a public good. An item is a pub-
lic good if one person’s increase in consumption does not reduce the quantity avail-
able to others. Equivalently, a public good provides non-rival
consumption. Consumption is non-rival if one person’s Non-rival consumption:
One person’s increase
increase in consumption does not reduce the quantity avail- in consumption does
able to others. If Mary watches a show of fireworks, she does not reduce the quantity
not reduce the quantity that others can watch. Hence, fire- available to others.
works provide non-rival consumption and are a public good.
Another way of appreciating non-rival consumption is
through the concept of economies of scale. Given that open-air fireworks are
being provided to one viewer, the marginal cost of providing the same display
to additional viewers is zero. There is an extreme economy of scale in providing a
public good with respect to the number of consumers. The cost of provision is fixed
with respect to the number of consumers and the marginal cost of serving an
additional consumer is zero.
The economy of scale with respect to the number of consumers differs from an
economy of scale with respect to the scale of provision. Increasing the length of a
fireworks show does involve more costs and need not exhibit economies of scale.
Private good Public good

Food Congestible Scientific formula
Clothing Air Musical composition
Computers Internet access TV broadcast
Electricity Roads and bridges Fireworks
Restaurant meal
FIGURE 12.3 Rivalness.
Given the extreme economy of scale (with respect to the number of consumers),
every additional customer brings in pure profit contribution. The business impli-
cation is to sell the public good to more customers (not quite to maximize the
number of customers, but to maximize the revenue).
For instance, the content of movie is a public good. Thus movie producers
increase their profit by selling the content through multiple formats – theaters,
DVDs, cable TV, and free-to-air TV.
Rivalness
Public goods lie at one end of a spectrum of rivalness, with private goods at the other
extreme (Figure 12.3). An item is a private good if one person’s increase in consump-
tion reduces the total available to others by the same quantity.
Rival consumption:
Equivalently, a private good provides rival consumption,
One person’s increase which means that one person’s increase in consumption reduces
in consumption reduces the quantity available to others by the same amount.
the quantity available Scientific formulas, musical compositions, TV broadcasts,
to others by the same and fireworks provide non-rival consumption. If one more
amount.
person uses them, they do not reduce the quantity available
for others.
By contrast, clothing and restaurant meals are private goods. If you wear a new
polo shirt, no one else can wear it at the same time. If you eat an eight-ounce
steak, then there will be eight ounces less for others.
Some items are neither public nor private, but provide
Congestible congestible consumption. This means that one person’s
consumption: One increase in consumption partially reduces the total quantity
person’s increase in available to others. Congestible items are public goods when
consumption partially consumption is low but are private goods when consump-
reduces the total available
tion is high.
to others.
The Internet is congestible. At off-peak times, one person’s
increase in usage will not reduce the quality or speed of ser-
vice to others. At peak times, however, the more people use the Internet, especially
if they are downloading large files such as movies, the slower the service will be
for others. The air around us is also congestible. The quality of the air is fine until
the discharge of pollutants by drivers and factories is too high, whereupon the
quality degrades.
Externalities 295
10
8.9
Marginal benefit/cost ($ per minute)
Vertical sum of marginal benefits
Marginal
5.6
5 cost
4.5 Alan
4
3.6
Mary
1 Peter
0.8
0 1 4 5 10
Minutes of fireworks
FIGURE 12.4 Economically efficient provision of public good.

Notes: At the economically efficient quantity of public good, the sum of the individual marginal
benefits equals the marginal cost. Every individual marginal benefit curve lies below the marginal cost
of $5.60; hence, no individual person would be willing to buy even one minute of fireworks. If each
individual tries to get a free ride, it might not be possible to provide even one minute of fireworks on
a commercial basis.
Benchmark: Economic Efficiency

We showed earlier that the economically efficient level of an externality is where
the combined marginal benefits equal the combined marginal costs, and at that
level there are no further opportunities to profit from adjusting the externality.
Let us now show the same for a public good. Suppose that there are three viewers
of open-air fireworks – Alan, Mary, and Peter – with marginal benefits as shown
in Figure 12.4. The graph also shows the marginal cost of open-air fireworks, a
constant $5.60 per minute.
Consider the provision of one minute of fireworks. Since fireworks are a public
good, there will be one minute for each of the three viewers. Accordingly, each
would be willing to pay their marginal benefit for that minute. By Figure 12.4, the
sum of the marginal benefits to the three viewers is $0.80 + $3.60 + $4.50 = $8.90.
The cost of one minute is $5.60, hence there is an opportunity to make a profit
by providing one minute of fireworks. This same argument applies for additional
quantities up to four minutes.
At four minutes of fireworks, the sum of the individual marginal benefits
equals the marginal cost. Can someone make money by increasing provision to
five minutes? By Figure 12.4, the fifth minute provides an additional benefit to
Alan and Mary only. The sum of their marginal benefits is $2.00 + $2.50 = $4.50.
Since the cost of each minute is $5.60, providing the fifth minute would result in
a loss.
So the opportunities for profit are exhausted at the point where the combined
marginal benefits equal the marginal cost, which is the economically efficient quan-
tity of the public good. Accordingly, at the economically efficient quantity, there
are no further opportunities to profit from adjusting the provision of the public
good.
Referring to Figure 12.4, notice that each of the individual marginal benefit
curves lies below the marginal cost of $5.60. Hence, no individual person would
be willing to buy even one minute of fireworks. The combined marginal benefit
curve, however, lies above the marginal cost curve at quantities of between zero
and four minutes.
Since a public good provides non-rival consumption, the three persons’ will-
ingness to pay is given by the combined marginal benefit, which is the vertical
sum of the individual marginal benefit curves. While no single person would buy
even one minute of fireworks, they collectively would be willing to pay for four
minutes.
PROGRESS CHECK 12E

In Figure 12.4, include another marginal benefit curve and show the impact on
the economically efficient quantity of fireworks.
VIVOCITY: ADVERTISING AND PROMOTION
With over 1 million square feet of net leasable area, VivoCity is Singapore’s
largest mall, attracting 40 million visitors a year. About 1% of the gross
rental that VivoCity collects from tenants comprises service charges and
contributions to advertising and promotion.
Advertising and promotion by the mall attract more customers to the mall.
The additional customers could visit any of the retailers in the mall. Hence, the
benefit of the advertising and promotion is non-rival among all the retailers.
So, VivoCity finances the advertising and promotion from a charge to all retail
tenants.
Source: Mapletree Commercial Trust, Prospectus, April 18, 2011.
6. Excludability
Having discussed the nature of public goods and their economically efficient
quantity, we are now ready to discuss how public goods can actually be provided.
Externalities 297
There is a fundamental condition for commercial provision of any product that is

easy to overlook. That condition is that, in selling a product, the seller must be
able to exclude those who do not pay. The condition is crucial in the commercial
provision of public goods.
Consumption is excludable if the provider can exclude
particular consumers. Typically, commercial provision of an Excludable
item depends on excludability. consumption: The
When consumption of an item is not excludable, commer- provider can exclude
particular consumers.
cial provision will be difficult. To illustrate, let us suppose
that Neptune Entertainment wants to provide one minute of
open-air fireworks to Alan, Mary, and Peter. Referring to Figure 12.4, Neptune
knows that, at that quantity, the individual marginal benefits of Alan, Mary, and
Peter are $4.50, $3.60, and $0.80, respectively. If Neptune charges each person a
price equal to their marginal benefit, then Neptune will collect $8.90 from these
three people. Neptune’s cost of providing one minute of fireworks is $5.60; hence,
it will realize a profit of $3.30.
Peter, however, might reason that if he refuses to pay, while Alan and Mary do
pay, then Neptune would collect $4.50 + $3.60 = $8.10, which would be enough to
cover the cost of $5.60. Then Neptune would provide the one minute of fireworks
and Peter would enjoy a free show. Alan and Mary, however, might also think in
similar ways. If everyone tries to get a free show, Neptune will incur a loss. Hence,
in the extreme, the free-rider problem prevents the provision of even one minute
of fireworks on a commercial basis.
The basic problem is that, whenever consumption of some item is non-excludable,
individual consumers will have an incentive to free-ride. Free riding will cut into
the seller’s revenues and thus reduce the seller’s profit and hamper commercial
provision of the item.
To understand the scope for commercial provision of public goods, we must
first appreciate the distinction between content and delivery, and then the two
factors that affect excludability, law and technology.
Content and Delivery

Television programming can be broadcast over the air or by cable. Regardless of
the method of delivery, the content of broadcast TV is non-rival. If Nancy
switches on her TV to watch the evening news, she does not affect the quantity
of news available to other people. This is true whether the signal comes over the
air or by cable.
The method of delivery, however, may be a public or pri-
vate good. Delivery by free-to-air transmission is a public A public good can be
good. The same signal can serve any number of TV sets delivered in the form of a
within the range of transmission. Delivery by cable, how- private good.
ever, is a private good. Each cable serves only one TV set.
Scientific knowledge also illustrates the distinction between content and deliv-
ery. The formula for the drug atorvastatin is a public good. If one more man-
ufacturer uses the formula, it does not reduce the quantity available to other
manufacturers. However, the drug is delivered in the form of manufactured tab-
lets. The tablet is a private good: if one person consumes a tablet, that tablet is
unavailable to others.
Since private goods are excludable, a key way to provide public goods on a com-
mercial basis is to deliver them in the form of private goods.
Law
The use of scientific formulas, musical compositions, and computer algorithms is
non-rival. However, through the legal concept of intellectual property, the law has
made their use legally excludable. The objective of intellectual property rights is
to encourage innovation. The policy intent is that the profits from exclusive rights
would encourage innovators to invest more in creative activities.
Two particular forms of intellectual property rights are
Patent: A legal exclusive
right to a product or
patents and copyright. A patent is a legal exclusive right to a
process. product or process. It is illegal to manufacture a product or
use a process covered by a patent without the permission of
the patent owner. A copyright is a legal exclusive right to an
Copyright: A legal artistic, literary, or musical expression. It is illegal to repro-
exclusive right to artistic,
duce copyrighted material without the permission of the
literary, or musical
expression. copyright owner.
Technology
In addition to the law, the other factor affecting excludability is technology.
The content of TV programming is a public good. But the medium of deliv-
ery can be excludable. Free-to-air TV is not excludable. The station cannot
prevent particular individuals from watching a broadcast. However, by using
scrambling technology, the broadcast can be made excludable. Scrambling
technology transforms the medium of delivery from being non-excludable to
excludable.
Software provides another example. The algorithms and code underlying com-
puter software are public goods. Users might copy software from one another.
However, if the publisher includes in the software an activation code that it pro-
vides only to licensed users, then it can exclude those who do not pay.
PROGRESS CHECK 12F

Distinguish the content and delivery of a public good. How does this matter
for commercial provision of a public good?
Externalities 299
UNCLE SAM – PATENT INFRINGER?
“Cheap imitations are killing our business, and destroying thousands of good
jobs.” Who is complaining? Microsoft? Universal Studios? Bertelsmann?
None of the above. It is Rosoboronexport, the Russian arms exporter.
During World War II, while being treated in hospital, Mikhail Timofeevich
Kalashnikov conceived of a new assault rifle. In 1942, he was discharged from
hospital and developed his concept. In 1949, following rigorous competitive
trials, the Soviet Army adopted the new rifle as the AK-47. Kalashnikov was
awarded the Stalin Prize First Class and Russian arms manufacturer Izhmash
began production.
Fast forward 50 years. The US government is a leading buyer of AK-47s
for supply to Afghanistan and Iraq. But not from Izhmash, which owns the
relevant patents. Rather, the United States buys AK-47s from unlicensed
manufacturers in Bulgaria, Hungary, and Romania. Somewhat helplessly,
Rosoboronexport official Igor Sevastyanov declared: “We would like to inform
everybody in the world that many countries, including the United States, have
unfortunately violated recognized norms.”
Source: “Russians take aim at AK-47 imitators,” International Herald Tribune, July 26, 2004.
FREE-TO-AIR TV: COMMERCIAL PROVISION

WITHOUT EXCLUSION?
Typically, commercial provision of a public good depends on excludability.

Free-to-air TV is non-excludable. In America, however, four major networks
provide free-to-air TV on a commercial basis. Far from limping along, the
business has flourished. What is the networks’ secret?
The primary source of revenues of free-to-air TV networks is advertising.
While viewers do not directly pay for TV, they do pay indirectly. They pay by
increasing their purchases of the advertised products and services. In the
year 2013, 20th Century Fox’s TV businesses earned operating income of
$855 million on revenues of $4.86 billion.
Source: 20th Century Fox, Annual Report 2013.
KEY TAKEAWAYS
• Externalities are benefits or costs directly conveyed to others.

• The economically efficient level of an externality is where the combined marginal
benefits equal the combined marginal costs.
• There is an opportunity to make a profit by resolving an externality up to the

economically efficient level.
• Resolve an externality through merger or agreement.
• The two hurdles to resolution of an externality are unclear assignment of rights
and possible free riding.
• Network effects are benefits or costs that increase with size of network.
• In the presence of network effects, manage user expectations to achieve
critical mass.
• A public good provides non-rival consumption: One person’s increase in
consumption does not reduce the quantity available to others.
• Provide public goods on a commercial basis by excluding non-payers through
law or technology.
• Increase profit by selling the public good to more consumers.
REVIEW QUESTIONS
1. What is the difference between a negative and a positive externality?

2. Luna and Neptune have both launched a new consumer electronic device.
Luna’s advertising is raising Neptune’s demand. What is the economically
efficient level of Luna’s advertising?
3. Explain the relation between the combined marginal benefit curve from additional
customers for all retailers in a mall and the retailers’ individual marginal benefit
curves.
4. When the sum of the marginal costs from a negative externality exceeds the
marginal benefit to the source, what is the profit opportunity? Explain.
5. Explain how the following action will help to resolve the externalities generated
by a new subway line: the subway system buys the property around the new
stations.
6. What is the possible free-rider problem in the following context? Saturn City
lies two miles off a busy highway. The city’s major businesses have proposed
to build an exit from the highway to draw traffic into the city.
7. Explain the possible network effects in the demand for spreadsheet
software.
8. Give an example of a network externality, and in that context explain the
concepts of critical mass and tipping.
9. Does the presence of network effects cause demand to be more or less price
elastic?
10. Where are technical standards relatively more important: in markets with
network effects or those without? Explain your answer.
11. Give an example of a public good. Explain how the use of the good is
non-rival.
12. Which of the following are public goods? (a) Dental treatment at a public
hospital. (b) Welfare payments to unemployed people. (c) National defense.
13. For a public good, why is the combined marginal benefit equal to the sum of
the individual marginal benefits?
14. In what way does excludability depend on law and technology?
15. Why does it make sense to sell a public good to more consumers?
Externalities 301
1. VivoCity is Singapore’s largest mall, attracting 40 million visitors a year. Its two
largest retail tenants, occupying 22.4% of the mall, are VivoMart, a combination
of a hypermarket, upscale supermarket, and drug store, and Tangs Department
Store. In 2010, on a per square foot basis, VivoCity’s major and mini anchors
achieved average sales of S$678 a year while paying rent of S$78 a year.
Specialty retailers achieved average sales of S$950 a year and paid an average
rent of S$196 a year.
(a) With a relevant diagram, illustrate a department store’s marginal benefit
from and marginal cost of attracting customers. Show the externality
on a specialty retailer and identify the economically efficient number of
customers.
(b) Compare the ratio of rent to sales for major and mini anchors and for
specialty retailers.
(c) Explain why VivoCity charges lower rents to anchors than to specialty
retailers.
(d) Are externalities from a department store to nearby retailers more
likely to be resolved for a store on a public street or in a shopping mall?
Explain your answer.
2. One of the world’s most abundant anchovy fisheries lies off the Pacific coast
of Peru. Historically, Peruvian fisheries operated on an “Olympic system.” The
Managing Director of the fish producer Pacific Andes, Ng Joo Siang, remarked:
“Nobody has time to preserve the fish because they want to catch as quickly as
possible.” Properly caught and processed for human consumption, anchovies
could sell for 40–50 US cents per kilogram. The alternative is animal feed worth
10 cents per kilogram. (Source: “Fishery player Pacific Andes bets on regulatory
changes, positions vessels in Peru,” The Edge Singapore, August 18, 2008, p. 18.)
(a) Explain the externalities under the “Olympic system.” Compare
production relative to the economically efficient level.
(b) In 2008, the government of Peru changed the system and awarded
each fish producer an individual quota of fish to catch. What would be
the impact on economic efficiency?
(c) The quotas were transferable. How would this help to achieve economic
efficiency?
3. Choice International, a business-format franchisor of hotels, launched Cambria
Suites to serve an upscale market segment. Choice International is concerned
that franchisees maintain a consistent standard. It has contracted with a quality
assurance specialist, LRA Worldwide, to conduct anonymous inspections of
each Cambria Suites hotel twice yearly.
(a) Using a relevant graph with ‘effort’ on the horizontal axis, illustrate a
franchisee’s marginal benefit and marginal cost of effort and explain its
choice of effort.
(b) How does the franchisee’s effort benefit other franchisees? Explain in
terms of an externality. On your graph for (a), identify the economically
efficient effort.
(c) Why would one Cambria Suites franchisee want Choice International
to inspect other franchisees?
(d) Consider the deviation between the franchisee’s choice of effort and
the economically efficient effort. Would this be smaller or larger in a
chain of hotels that are all in a single city than in an international chain?
4. BAA Airports Limited operates London’s Heathrow Airport. Between 2008
and 2009, with the Great Recession, Heathrow passenger traffic fell by 1.5%.
However, the proportion of long-haul traffic increased from 52.2% to 52.9%
and the proportion of transfer passengers increased from 35.9% to 37.4%.
(Source: BAA Airports Limited, Results for the year ended 31 December
2009.)
(a) Explain how passenger traffic affects the demand for retail services
such as restaurants and car rental at the airport.
(b) Using relevant demand curves, compare airport expenditures on
promoting passenger traffic at: (i) airports with retail facilities; and
(ii) airports without retail facilities. Which airports would spend more
on promoting passenger traffic? Assume that the airport maximizes
the net benefit to all parts of the airport (including the retailers, if any).
(c) BAA’s financial report remarked that “strong in-terminal shopping
performance reflected a higher proportion of intraterminal transfer
passengers, providing longer departure lounge dwell times for such
passengers.” Would you advise BAA to extend the transfer times
between flights, so that passengers must spend more time at the
terminal?
5. In 2007, the computer manufacturer Dell spent $943 million on advertising. Many
of Dell’s advertisements feature “Intel inside.” These advertisements boost
the demand for other computer manufacturers using Intel microprocessors.
Between 2007 and 2008, Dell’s advertising expenditure fell from $943 to
$811 million. The major reason was an increase in receipts for cooperative
advertising. (Source: Dell Inc., Form 10K, 2010.)
(a) With a relevant diagram, illustrate Dell’s marginal benefit from and
marginal cost of advertising. Show the externality on other computer
manufacturers and identify the economically efficient level of
advertising.
(b) Explain why Intel makes cooperative advertising payments to Dell.
(c) AMD competes with Intel to supply microprocessors to computer
manufacturers. How should Intel take account of this competition in
deciding payments for cooperative advertising?
6. With the growth of international trade and travel, more people are learning
English. Major European companies have adopted English as a common
language. English is also the standard language for communication among
aircraft pilots and air-traffic controllers. Linguists estimate that, of 7,000
languages in use worldwide, nearly half will disappear within this century.
(a) Compare the benefit of speaking a common language among:
(i) truckers; (ii) pilots.
Externalities 303
(b) Does the growth of English generate a positive or negative externality

for: (i) people who are already fluent in English; (ii) people who do not
speak English?
(c) Can the externalities in (b) be described as network externalities?
Explain your answer.
(d) How do critical mass and expectations apply to the survival of a
language?
7. Microsoft’s Excel dominates the market for spreadsheet software. Presently,
its main competitor is Calc, a product of OpenOffice. OpenOffice emphasizes
that Calc “looks and feels familiar and is instantly usable by anyone who has
used a competitive product.” (Source: OpenOffice.org.)
(a) Explain the network effects in the demand for spreadsheet software.
(b) Use the concept of critical mass and tipping to explain the demise of
Lotus 1-2-3.
(c) Why does OpenOffice stress that Calc is so similar to Excel?
(d) Should Microsoft cooperate with OpenOffice in the further development
of Calc?
8. Football, cricket, and other sports entertain thousands through live attendance
and millions through TV. TV subscribers pay substantial fees to watch live
sports. Others watch free but “pay” by consuming advertisements.
(a) Alan, Mary, and Peter are sports fans. Using a suitable figure with
quantity of sports (in hours per month) on the horizontal axis, explain
the economically efficient production of sports for the three fans.
(b) Use the figure to identify the marginal benefits of the three fans at the
economically efficient level of production.
(c) If a sports league were to broadcast only through subscription channels
that charge a uniform price, can it achieve the efficient level of
production?
(d) What pricing policy are sports leagues applying? Does it help to
achieve economic efficiency?
(e) Some websites broadcast live sport without license from the sports
league. How does this affect economically efficient production of
sports?
9. Mikhail Kalashnikov conceived the AK-47 assault rifle while recuperating from
wounds during World War II. Within 50 years, over 100 million AK-47s had been
produced. Russian arms manufacturer Izhmash owns patents to the AK-47.
However, the US government buys AK-47s from unlicensed manufacturers in
Bulgaria, Hungary, and Romania for supply to its allies in Afghanistan and Iraq.
(Sources: “Russians take aim at AK-47 imitators,” International Herald Tribune,
July 26, 2004; “AK-47’s inventor peacefully retired,” Guardian, October 26,
2003.)
(a) Use the AK-47 to explain how engineering design is a public good.
(b) Patents are only effective within the country in which they are granted.
Explain how manufacturers in Bulgaria, Hungary, and Romania can
free-ride on Kalashnikov’s design.
(c) Maxim Piadiyshev, editor of Arms Export Review, explained the success
of the rifle: “Compared to other automatic rifles at the time . . . it was
very simple in production, use and maintenance, with eight moving
parts.” From Izhmash’s viewpoint, does it help or hurt that the AK-47’s
design is simple yet effective?

Identify any externalities or network effects in your organization’s activities.
How should your organization take advantage of the profit opportunities?
Note
1 This discussion is based, in part, on Mapletree Commercial Trust, Prospectus, April 18, 2011.
C H A P T E R
13
Asymmetric Information
LEARNING OBJECTIVES
• Understand imperfect information and risk.
• Appreciate asymmetry of information and its consequences.
• Appreciate and apply appraisal to resolve information asymmetry.
• Appreciate and apply screening to resolve information asymmetry.
• Understand how to design and bid in auctions.
• Appreciate and apply signaling to resolve information asymmetry.
• Appreciate and apply contingent contracts to resolve information
asymmetry.
1. Introduction
In 2005, Countrywide Financial Corporation, the largest US mortgage lender,

originated $490 billion of residential mortgages. Countrywide bundled the mort-
gages into pools and sold them completely or in part through mortgage-backed
securities. Mortgage sales contributed $451.6 million, or 10.9% of the company’s
overall pre-tax earnings.1
Historically, Countrywide mostly lent through “prime conforming” mortgages,
a standard set by the US federal government sponsored mortgage corporations,
Fannie Mae and Freddie Mac. From 2004, faced with declining demand, Coun-
trywide increased lending through prime non-conforming mortgages, subprime
mortgages, and home equity loans from 40.2% to 64.1% of all mortgages.
Countrywide masked substantial risks in classifying loans. Banking regulators
recommended that only loans to borrowers with a FICO (a credit score based on
306 13 Asymmetric Information
a system created by the Fair Isaac Company) score exceeding 620 be classified as
“prime.” However, Countrywide set no minimum FICO score for prime loans.
Countrywide’s “prime non-conforming” category included “Alt-A” mortgage
loans with reduced or no documentation, no proof of income, or with loan to
value ratios of 95% or more.
Other high-risk mortgages that Countrywide originated included 80/20 subprime
loans and pay-option adjustable rate mortgages (ARMs). The 80/20 loan combined
a first trust deed mortgage for 80% of the value of the home, and a second trust
deed mortgage for the remaining 20% to provide a 100% loan. Pay-option ARMs
did not require verification of the borrower’s income.
Angelo Mozilo, Chairman and CEO of Countrywide, worried about the pay-
option ARMs. In an email dated September 26, 2006, he directed the President
and Chief Operating Officer, David Sambol, and the Chief Financial Officer, Eric
Sieracki, to “sell all newly originated pay options and begin rolling off the bank
balance sheet.”
HSBC had purchased pools of 80/20 mortgages from Countrywide subject to
the right to require Countrywide to buy back mortgages that did not meet specific
conditions. In early 2006, HSBC exercised its right to require Countrywide to buy
back the specific mortgages.
Countrywide’s business and the market for mortgages present several questions.
Why are mortgage borrowers subject to FICO scoring? Why did HSBC condition
its purchase on the buyback right? Why did Mozilo ask senior management to sell
the pay-option ARMs?
The relationship between borrowers and Countrywide, and that between
Countrywide and investors such as HSBC, illustrate asymmetric information. In a
situation of asymmetric information, one party has better
Asymmetric information than another. Borrowers had better information
information: One party about their income, assets, and willingness to repay (or
has better information default) than Countrywide.
than another.
In turn, Countrywide had better information about its
mortgages than HSBC and other buyers of its mortgages
and mortgage-backed securities. Mozilo’s direction to sell all pay-option ARMs
perfectly illustrates this asymmetry of information. Being particularly worried
about the creditworthiness of the pay-option ARM borrowers, he decided to sell
the mortgages to other investors.
Managers of financial institutions and investors are aware of the possible risks
due to asymmetric information. Under such circumstances, they may refuse to
engage in transactions and relationships which would increase value added unless
the information asymmetry is resolved.
This chapter introduces four ways to resolve information asymmetry – appraisal,
screening, signaling, and contingent contracts. FICO scores illustrate the technique
of appraisal: originators of mortgages and buyers of mortgages use the scoring
system to appraise the creditworthiness of borrowers. Through FICO scores, lend-
ers try to gain more information about borrowers.
Asymmetric Information 307
Less-informed parties use screening to discern the information of the better-

informed parties. By limiting loans to borrowers who agree to credit checks (as
Countrywide did not), lenders screen borrowers for their creditworthiness. In
contrast, better-informed parties use signaling to communicate their information
to less-informed parties. A contingent contract specifies actions under particular
conditions. HSBC’s right to require Countrywide to buy back specific mortgages
illustrates a contingent contract.
The concept of information asymmetry and techniques to resolve asymmetries
apply very broadly beyond finance, to commercial, non-commercial, and personal
settings. Managers can use these methods to resolve information asymmetry and
thus realize transactions and relationships to increase value added and profit.
2. Imperfect Information
Before analyzing situations of asymmetric information, we should understand

the concept of imperfect information. To have imperfect
information about something means not having certain Imperfect information:
knowledge about it. Most people have imperfect informa- The absence of certain
tion about future events such as next Monday’s Standard & knowledge.
Poor’s 500 Index, the severity of the coming winter, and next
year’s growth in employment. It is also possible to have imperfect information
about things in the present or past. For instance, do you know precisely the height
of K2 or the distance between Sydney and Tokyo?
Imperfect and Asymmetric Information

A single person can have imperfect information. By contrast, asymmetric infor-
mation involves two or more parties, one of whom has better information than
the other or others. Asymmetric information will always be associated with imper-
fect information, because the party with poorer information definitely will have
imperfect information. For instance, if the seller knows the quality of a wine, but
a potential buyer does not, then the buyer has imperfect information. The wine
could be good or lousy, but the buyer does not know which for sure.
Although the concepts of asymmetric and imperfect information are related,
it is important to appreciate the distinction. The reason is that a market can be
perfectly competitive even when buyers and sellers have imperfect information, so
long as their information is symmetric. Under perfect competition, the forces of
demand and supply channel resources into economically efficient uses; hence, no
further profitable transactions are possible.
For instance, the current demand for heating oil depends on expectations about
temperatures in the coming winter. Buyers and sellers have equal access to meteo-
rological forecasts. Based on these forecasts, each buyer determines its demand for
heating oil. In a market equilibrium, the quantity demanded equals the quantity
supplied and the marginal benefit equals the marginal cost. Hence, any further
sales would be unprofitable.
By contrast, a market where information is asymmetric cannot be perfectly com-
petitive. This means that, if buyers and sellers can resolve the information asym-
metries, they can increase their benefits by more than their costs, and so add value.
Risk
When information is imperfect, there is risk. To understand the meaning of risk, let us
consider the following example. Alice knows that, with probability 1.5%, someone
will steal her $20,000 car within the next 12 months. If that were to happen, Alice
would lose $20,000. If Alice’s car is not stolen, however, Alice will lose nothing.
The probability that her car will not be stolen is 100 − 1.5 = 98.5%.
Alice has imperfect information about her future losses, because she does not
know for sure whether her car will be stolen. Alice bears a risk: either she will
lose $20,000 with probability 1.5% or she will lose nothing
Risk: Uncertainty about with probability 98.5%. Risk is uncertainty about benefits
benefits or costs. or costs and arises whenever there is imperfect information
about something that affects benefits or costs.
If Alice knew for sure that her car would not be stolen within the next 12 months,
then she would not bear any risk. Similarly, if she knew for sure that her car would
be stolen, she would also not bear any risk. It is because her information is imper-
fect that she faces risk.
To explain the distinction between risk and imperfect information, consider Bob,
who is unrelated to Alice. Bob also has imperfect information about whether Alice’s
car will be stolen. But the fate of Alice’s car does not affect Bob’s benefits or costs.
Hence, Bob does not bear any risk with regard to Alice’s car.
Risk Aversion
How a person responds to situations involving risk depends
Risk averse: A person on the extent to which he or she is risk averse. A person is
prefers a certain amount
risk averse if she or he prefers a certain amount to risky
to risky amounts with the
same expected value. amounts with the same expected value. A person is risk
neutral if she or he is indifferent between a certain amount
and risky amounts with the same expected value.
Risk neutral: A person Given Alice’s possible losses and the probabilities, her
is indifferent between a expected loss is ($20,000 × 0.015) + ($0 × 0.985) = $300. If
certain amount and risky Alice is risk averse, she will prefer to lose $300 for certain
amounts with the same
expected value.
than to lose $20,000 with probability 1.5% or lose nothing
with probability 98.5%. If Alice is risk neutral, she would be
indifferent between the two scenarios.
Risk-averse persons will pay to avoid risk. Insurance is the business of taking
certain payments in exchange for eliminating risk. Suppose that an insurer offers
Alice an insurance policy that pays her $20,000 if her car is stolen but pays noth-
ing if her car is not stolen.
If Alice has the policy and her car is stolen, she loses the car but receives $20,000,
so on balance, she gains and loses nothing. If her car is not stolen, the insurer will
not pay her anything, so she gains and loses nothing. Thus, the insurance policy
eliminates the risk that Alice must otherwise bear. Recall that, without insurance,
Alice’s expected loss is $300. Hence, if she is risk averse, she would pay at least
$300 for the insurance policy.
How much risk-averse persons are willing to pay for insurance depends on their
degree of risk aversion. A more risk-averse person will be willing to pay a larger
amount to avoid risk. By contrast, a risk-neutral person will not pay more to
avoid risk. For instance, suppose that Emily faces the same situation as Alice. If
Emily is risk neutral, she would pay no more than $300 for the insurance policy.
It is important to understand the meaning of risk and risk aversion because,
whenever information is asymmetric, the less-informed party has imperfect infor-
mation. To the extent that this means uncertainty about benefits or costs, the
less-informed party faces risk.
PROGRESS CHECK 13A

Suppose that Alice buys an insurance policy that pays her $20,000 if her car is
stolen and nothing if her car is not stolen. Who of the following has imperfect
information and who faces risk: (a) Alice; (b) the insurer?
3. Adverse Selection
The quality of wine depends on multiple factors, including the grapes used, location
of the vineyard, weather, method of production, and age. Producers of wine have
better information about these factors and the quality of the wine than consumers.
So information is asymmetric.
Let us use the example of wine to analyze the nature of equilibrium and the
effect of price changes on the quantity demanded and supplied in a market with
asymmetric information. Consider a very basic setting where wine producers have
no reputation and produce wine in plain bottles without labels.
Demand and Supply

Initially, suppose that producers offer only high-quality wine. Figure 13.1 shows
the supply and demand for high-quality wine. In equilibrium at point b, the price
would be $50 per bottle and production would be 300,000 bottles a month.
Now suppose that producers also produce 100,000 bottles of low-quality wine,
at marginal cost ranging from zero to $1. In principle, two markets could exist: one
Supply
Demand
Price ($ per bottle)
b
50
1
0 3
Production of wine (hundred thousand bottles a month)
FIGURE 13.1 Market with symmetric information.
High-quality
Demand supply
(marginal
benefit) for Combined
high quality supply of low
70 and high quality
Expected demand
30 a (marginal benefit)
1
0 1 2
FIGURE 13.2 Market with adverse selection.

Notes: The combined supply of low- and high-quality wine begins with the supply of low-quality
wine up to 100,000 bottles at $1 per bottle, then runs parallel to the supply of high-quality wine. The
expected demand, which reflects the probability of getting wine of low quality, is the demand curve
for wine of high quality shifted down by the proportion 1/Q. The expected demand and combined
supply cross at point a, where the price is $30 per bottle and production is 200,000 bottles a month.
for low-quality wine and another for high-quality wine. Suppose, however, that
potential consumers cannot distinguish high from low quality. Then there can be only
one market. In that single market, low-quality wine trades alongside high-quality
wine, so the supply of high- and low-quality wine is combined.
Figure 13.2 illustrates the market with both low- and high-quality wine. The
combined supply of low- and high-quality wine begins with the supply of
low-quality wine from zero to 100,000 bottles at a marginal cost of $1 per bottle.
Then the combined supply curve runs parallel to the supply of high-quality
wine.2
What about the demand side of the market? Consumers do not know whether
the wine is of low or high quality. Assume that they are risk neutral. Each consumer
has a marginal benefit curve for high-quality wine. Suppose that their marginal
benefit from low-quality wine is zero. So, each consumer’s expected marginal benefit
is the marginal benefit from high quality multiplied by the probability of getting
high quality.
Suppose that consumers purchase a total of Q hundred thousand bottles, of
which 100,000 are of low quality. Then each consumer has probability 1/Q of get-
ting low-quality wine and probability (Q − 1)/Q of getting high-quality wine. So,
the consumer’s expected marginal benefit is (Q − 1)/Q of her marginal benefit
for high-quality wine. The consumer’s expected marginal benefit curve, which
reflects the probability of getting low-quality wine, is her marginal benefit curve
for high-quality wine, shifted down by the proportion 1/Q at every quantity.
Accordingly, the expected demand for wine, which reflects the probability of
getting low-quality wine, is the demand curve for high-quality wine shifted down
by the proportion 1/Q. Equivalently, at every possible quantity, the consumers’
actual willingness to pay is only a fraction (Q − 1)/Q of their willingness to pay for
the same of quantity of wine that is definitely of high quality.
Market Equilibrium
Having laid out the demand and the supply, we can identify the equilibrium. The
expected demand and the combined supply cross at point a, where the price is
$30 per bottle and the quantity Q is 200,000 units a month. Hence, the probability
that a consumer gets low-quality wine is 100,000/200,000 = 50%. and the probabil-
ity of getting high-quality wine is 50%.
Consumers cannot distinguish good from bad wines, and
Adverse selection: The
so they get an adverse selection – a mixture of bad and good less-informed party draws
wines. Adverse selection arises in situations of asymmetric a selection with relatively
information: the less-informed party draws a selection with bad characteristics.
relatively bad characteristics.
What if the production of low-quality wine is lower or
higher? If, for instance, there are 50,000 bottles of low-quality wine, then the
expected demand would be higher and the combined supply of low- and high-quality
wine would be further to the left. Then the equilibrium will have a higher price.
The equilibrium production, however, might be lower or higher: the demand is
higher, but supply is further to the left.
By contrast, if the production of low-quality wine is larger, then the expected
demand would be lower and the combined supply would be further to the right.
Hence, the market price would be lower. Again, production might be lower or
higher, depending on the balance between demand and supply.
PROGRESS CHECK 13B

Using Figure 13.2, illustrate the market equilibrium with production of 50,000
bottles of low-quality wine at a marginal cost of up to $1 per bottle.
Economic Inefficiency
Referring to Figure 13.2, in the market equilibrium, the price is $30 per bottle and
production is 200,000 bottles a month. Production comprises 100,000 bottles of
low-quality wine and 100,000 bottles of high-quality wine. By the supply curve
of high-quality wine, the marginal cost of producing the 100,000th bottle is $30.
Consumers buy up to the point that the expected marginal benefit (adjusted for
the probability of getting low quality) equals the market price. Low-quality wine
provides no marginal benefit, so, in equilibrium, the marginal benefit of consum-
ers who get low-quality wine is less than the marginal cost.
By contrast, referring to the marginal benefit curve for high-quality wine, at
100,000 bottles a month, the marginal benefit is $70 per bottle. This exceeds the
marginal cost of producing the high-quality wine, $30 per bottle. Accordingly, the
equilibrium is not economically efficient.
If, somehow, another bottle of high-quality wine could be produced and sold,
then there would be a consumer willing to pay a little less than $70, and it would
cost a little more than $30 to produce. Accordingly, there is a potential value added
of almost $40.
Essentially, sellers of low-quality wine impose a negative externality on con-
sumers and producers of high-quality wine. By Chapter 12, we know that the
negative externality would exceed the economically efficient level. This means that
a profit can be made by resolving the externality, which, in this setting, means
resolving the information asymmetry.
Market Failure
Before considering how to resolve an asymmetry of information, let us look at an
extreme possibility. Suppose that wine producers produce F hundred thousand
bottles of low-quality wine. Referring to Figure 13.3, the combined supply of low
and high-quality wine has a kink at point c.
What if the quantity of low-quality wine is so large that the expected demand
crosses the combined supply at some point d below the kink? In this case, there will
be no supply of high-quality wine and the entire supply will be low-quality wine.
Then a consumer’s probability of getting high-quality wine would be zero. This
means that the expected marginal benefit and the expected demand curve must
coincide with the horizontal axis. Thus, the initial supposition that the expected
demand crosses the combined supply at some point d is not valid.
So there cannot be a market equilibrium with the expected demand curve crossing
the combined supply below the kink. Indeed, the same logic also shows that there
Demand
(marginal High-quality
benefit) for supply
high quality
Combined
supply of low
and high quality
Expected demand
(marginal benefit)
c
1 d
0 F
FIGURE 13.3 Market failure.

Notes: Supposing that the expected demand crosses the combined supply at point d, then the entire
production will be of low quality. Then, the expected demand curve would coincide with the horizontal
axis, and cannot cross the combined supply at point d. Generally, there cannot be an equilibrium with
the expected demand curve crossing the combined supply above but close to the kink at point c.
cannot be an equilibrium with the expected demand curve crossing the combined
supply above but close to the kink. If there cannot be an equilibrium, then con-
sumers and sellers cannot trade. This means that the intrusion of low-quality wine
has caused the entire market to fail!
Let us view the market failure from another perspective. A change in price has
very different effects in a perfectly competitive market as compared to a market
subject to adverse selection. Suppose that, in a perfectly competitive market, the
quantity supplied exceeds the quantity demanded. Then a price reduction will
reduce the quantity supplied and raise the quantity demanded. A sufficient reduc-
tion of the market price will restore equilibrium.
By contrast, suppose that the market for wine is out of equilibrium, with the
quantity supplied exceeding the quantity demanded. Consider a reduction in the
market price. The marginal cost of producing high-quality wine slopes upward
from a minimum of $1 per bottle. So, when the market price is lower, high-quality
wine producers produce less.
However, a reduction in the market price (so long as it remains above $1) does
not affect the production of low-quality wine. Hence, the lower market price
raises the proportion of low-quality wine, leaving consumers with a worse adverse
selection. So their willingness to pay and the expected demand curve would fall.
Thus, in a market with adverse selection, a cut in price reduces both the quan-
tity demanded and the quantity supplied, so it need not reduce the excess supply
and restore equilibrium. At the extreme, if the price is cut very low, the expected
demand curve would fall to zero and the market would fail completely.
PROGRESS CHECK 13C

Referring to Figure 13.3, suppose that the expected demand curve crosses the
combined supply curve at point c. What would be the market equilibrium price
and production?
LIFE INSURANCE: BETTING ON DEATH
Insurance is a market where buyers have better information (about themselves)

than sellers. Life insurance is insurance against the event of death. It would
be more accurate, but much less appealing, to call it death insurance. The
probability of dying within a given period of time depends on the person’s
health and lifestyle.
The price that an insurer charges for an insurance policy is called a premium.
Life insurers face an adverse selection problem: if an insurer charges a high
premium, it is likely to draw applicants in relatively poor health or who maintain
risky lifestyles.
Insurers collect information about applicants for life insurance – whether
they smoke, their age and their medical history, and their employment, sports,
and other activities. However, it is difficult for an insurer to obtain all the
relevant information about an applicant’s state of health and lifestyle. Hence,
there remains asymmetric information between insurers and applicants for
insurance.
VOLUNTARY RETIREMENT: DOWNSIZING THE HUMANE WAY
Given a mandate to reduce budgets and headcount, one appealing way is

to encourage employees to quit or retire voluntarily. In the mid-2000s, under
pressure to reduce middle management, the International Monetary Fund
offered a generous voluntary separation program to all managers, regardless
of age.
Among the professionals who left the IMF was Shang-jin Wei, Assistant
Director in the Research Department. He took voluntary separation and
became a professor at Columbia University.
Voluntary separation seems much more humane than forced retrenchment.
But does it leave behind an adverse selection?
Sources: “Staff in black: IMF faces structural adjustment,” Bretton Woods Project, January 31,
2008; Shang-jin Wei, curriculum vitae, October 12, 2010.
4. Appraisal
When information is asymmetric, the market outcome will not be economically

efficient. Resolving the information asymmetry would raise benefits by more than
costs, and so there is an opportunity to add value and make a profit.
The most obvious way to overcome asymmetric information is to obtain the
information directly. Consumers and producers can engage experts to appraise
the wine. Referring to Figure 13.2, in the equilibrium at point a, the marginal
benefit of high-quality wine is $70, while the market price is $30. For the marginal
consumer, there is a potential gain of almost $40 by identifying a high-quality
wine. Similarly for the marginal producer of high-quality wine.
To the extent that appraisals are available, there will be a separate market for
good wine. In that market, consumers and producers would have equal informa-
tion, and so perfect competition will lead to economic efficiency.
An information asymmetry can be resolved by appraisal
under two conditions. One is that the characteristic about Appraisal: Works if the
which information is asymmetric must be objectively characteristic about which
verifiable. If an expert cannot objectively distinguish information is asymmetric
high-quality from low-quality wine, then appraisals would is objectively verifiable and
if it is not too costly.
not help. The appraisal must be objective: if different
appraisers give different opinions, then information will
still be asymmetric.
The other condition is that the potential gain from resolving the asymmetry
must exceed the cost of appraisal. This, in turn, depends on two factors. One is
the proportion of low quality relative to high quality. The other is the difference
between the marginal benefit and marginal cost.
Who should procure the expert appraisal? A wine producer can obtain one
appraisal and present it to many potential buyers. The appraisal is a public good:
any number of potential buyers can use the same information. Hence, it is less
costly for the seller to get an appraisal than for every potential buyer to procure
an appraisal of the same item.
By contrast, if the buyer is the better-informed party and dealing with multiple
sellers, it would be less costly for the buyer to procure the appraisal. Commer-
cial borrowers are buyers of finance, while lenders and investors are suppliers of
finance. Commercial borrowers typically pay credit rating agencies to appraise
their creditworthiness and provide the credit rating to potential lenders and
investors.
PROGRESS CHECK 13D

Will appraisals be more common in the market for cheap or more expensive wines?
COUNTRYWIDE: PAY-OPTION ARMS
In Countrywide’s quest to generate income, pay-option ARM loans accounted

for 17–21% of the mortgages that it originated. Typically, banks guard against
adverse selection by appraising applicants for loans, and in particular, check
the borrower’s income.
Countrywide changed the lending model for the pay-option ARM by not
requiring verification of the borrower’s income. In June 2006, an audit revealed
that over half of the borrowers had overstated their income by more than 10%,
while over one-third had overstated by more than 50%.
By 2007, with interest rates rising and house prices falling, Countrywide
began to incur large losses. The next year, Countrywide sold itself to the Bank
of America.
Source: Securities and Exchange Commission v. Angelo Mozilo, David Sambol, and Eric Sieracki,
US District Court for Central District of California, case no. CV09-03994, filed June 4, 2009.
5. Screening
Information asymmetries can be directly resolved through appraisal. However,

appraisal is costly and not always feasible. So it is important to consider indirect
alternatives. There are three ways to indirectly resolve asym-
Screening: An initiative metric information. Here, we present screening, and discuss
of a less-informed party
the two other methods – signaling and contingent contracts –
to indirectly elicit the
better-informed party’s in Sections 7 and 8. Screening is an initiative of a less-
characteristics. informed party to indirectly elicit the better-informed party’s
characteristics.
Self-Selection
Screening exploits the sensitivity of the better-informed party to some variable
that the less-informed party can control. The strategy is to design choices around
that variable to induce self-selection, meaning that parties
Self-selection: with different characteristics choose different alternatives.
Parties with different Consider the market where consumers cannot distinguish
characteristics choose low- from high-quality wine. The unknown characteristic is
different alternatives.
the true quality of the wine. Suppose that consumers insist
that they get the first bottle of wine free and pay double for
the second bottle. Can this deal effectively screen low- from high-quality wine?
Suppose that consumers who get high-quality wine would buy the second bot-
tle, while those who get low-quality wine would not buy the second bottle. So
producers of low-quality wine would not get any revenue. They would not accept
the deal. The deal makes sense only for producers of high-quality wine. Thus, it
credibly screens the two types of producers.
Screening is possible only if the less-informed party can control some vari-
able to which the better-informed parties are differentially sensitive. In the wine
example, low-quality producers are more sensitive to the deal than high-quality
producers.
By contrast with appraisal, screening reveals the information of the better-
informed party indirectly. Screening is an indirect way of resolving an information
asymmetry. Consumers do not directly determine whether wine is of high quality.
Rather, they require producers to make a choice that indirectly communicates
their characteristics.
A key business application of screening is to pricing. Chapter 9 on pricing pre-
sented indirect segment discrimination. Such discrimination applies screening to
induce self-selection among buyers with different price elasticities of demand.
Differentiating Variables
In some instances, the less-informed party can choose between several differen-
tiating variables. Ideally, the less-informed party should structure the choice that
drives the biggest possible wedge between the better-informed parties with the
different characteristics.
For instance, in pricing, airlines could use clothing as a differentiating variable,
offering a lower fare to travelers wearing casual clothing and a higher fare to
travelers in business attire. Business travelers could easily circumvent this differen-
tiating variable; hence, it is not effective. By contrast, airlines have found advance
booking and penalties for changes to be effective differentiating variables.
The most effective screening may involve a combination of the differentiating
variables. For instance, airlines use a combination of advance booking, penalty
for changes, and frequent flyer benefits to screen leisure from business travelers.
PROGRESS CHECK 13E

Collateral is property that a borrower provides to a lender as a security for a
loan. If the borrower defaults on the loan, the lender can seize the collateral.
Explain how a lender can use requirements for collateral to screen among bor-
rowers with different willingness to repay.
SCREENING FOR LOVE: THE PRENUP
Human relationships are rife with information asymmetry. A particular concern

among rich men and women entering into marriage is that their loving spouse
might be marrying for the money. But they can sleep more easily with a prenuptial
agreement. Typically, the prenuptial agreement specifies a minimum period of
marriage before which the spouse can share a limited portion of the wealth and
income. It is an effective way to screen gold-diggers from true life partners.
In October 2004, Tiger Woods married Elin Nordegren at the Sandy Lane
resort, Barbados. Their prenuptial agreement specified that Ms Nordegren
would receive $20 million after 10 years of marriage.
Fast forward five years. After Woods mysteriously crashed his car, a string
of women came forward to reveal that they had had sex with him during his
marriage. Woods reportedly renegotiated the prenuptial to pay Ms Nordegren
$5 million immediately, shorten her vesting period to 7 years, and raise her
maximum payout to $75 million.
Sources: “New details on Tiger’s prenup,” Daily Beast, December 3, 2009; “Tiger Woods
mistress list rises to 11,” go.com, December 9, 2009.
6. Auctions3
Sellers can apply indirect segment discrimination to screen among buyers who
differ in their price sensitivity. Auctions are a particular form of indirect segment
discrimination that exploits strategic interaction among the bidders.
Suppose that a wine grower wants to sell a vineyard. There are many potential
buyers, and the seller wants to get the highest possible price. An auction applies
competitive pressure to the participating bidders. Each bidder must act strategi-
cally since its best bid depends on the competing bids: if the other bids are low,
then a bidder can win with a relatively low bid, while if the other bids are high,
then a bidder must bid relatively high to win.
Each bidder faces a fundamental tradeoff. By bidding more aggressively, it will
improve its chances of winning the auction. On the other hand, if it bids more
aggressively, it will get a smaller profit from winning the auction.
The differentiating variable in an auction is the probability of winning. A bidder
that values the item more will gain relatively more from winning the auction, and
hence will bid higher. Thus, the auction induces self-selection among the partici-
pants according to their valuation of the item.
Auction Methods
Auctions may be conducted in various ways. The bidding can be open or sealed.
In an auction for multiple items, each winning bidder can be required to pay the
price that she bid, or the price bid by the marginal winning
Reserve price: The price bidder. (The marginal winning bidder is the last bidder (in
below which the seller will
not sell the item.
descending order of bids) to get an item.) The seller may
specify a reserve price, below which it will not sell the item.
• Open/sealed bidding. Auction houses such as Christie’s and Sotheby’s use open
auctions. The auctioneer calls out prices in an ascending sequence, and the
bidders indicate whether or not they wish to continue participating. By contrast,

developers choosing a building contractor usually conduct sealed bid auctions.
• Reserve price. The seller may specify a reserve price, below which it will not
sell the item. The reserve price forces bidders to bid higher. However, all the
bids may fall below the reserve price and then the seller will get no revenue.
So, in setting a reserve price, the seller must balance the increased revenue
from a sale against the probability of no sale.
Discriminatory auction:
• Prices for multiple units. In a discriminatory auction, Each winning bidder pays
each winning bidder pays their bid. By contrast, in a their bid.
non-discriminatory auction, each winning bidder pays
the bid of the marginal winning bidder. Bidders at non- Non-discriminatory
discriminatory auctions should bid relatively higher than auction: Each winning
at discriminatory auctions. Whether a seller gets higher bidder pays the bid of the
revenue from a non-discriminatory auction depends on marginal winning bidder.
two factors. One is that bidders make relatively higher
bids, but the other is that the seller collects only the bid of the marginal bidder
for every item sold.
Winner’s Curse
In addition to not knowing their competitors’ strategies, the bidders participat-
ing in an auction may also be uncertain about the value of the item for sale. For
instance, in the auction of the vineyard, the bidders may be unsure about the quan-
tity of grapes that the land will yield. How will this uncertainty affect the bidding?
Consider the bidder who wins the auction of the vineyard. Then, by the fact of
winning, the winning bidder can infer that it probably had the highest estimate of
the yield. On the basis of all of the bidders’ estimates, however, it may have over-
estimated the true yield and actually would incur a loss on the purchase.
The vineyard auction illustrates the winner’s curse. The
winner’s curse arises in an auction where the various bid- Winner’s curse:
The winning bidder
ders are uncertain about some common element in the value overestimates the true
of the item for sale. A bidder whose estimate of that com- value of the item.
mon element is high is more likely to win. Hence, on average,
the winning bidder is the one most likely to have overesti-
mated the true value of the item.
A bidder should take account of the possibility of the winner’s curse by bidding
more conservatively, that is, aiming for a larger margin between its estimate of the
value of the item and its bid. By bidding conservatively, the bidder can reduce the
likelihood of overbidding for the item.
There are three circumstances in which the winner’s curse is more severe and
bidders should bid more conservatively:
• If there are more bidders. With 20 bidders, the winner will probably be one
whose estimate is higher than 19 other estimates. By contrast, with four
bidders, the winner’s estimate is probably higher than three other estimates.
An estimate that is higher than 19 others is more likely to exceed the true
value than an estimate that is higher than three others.
• If the true value of the item is more uncertain. Consider, for instance, an
auction for gold. There is no uncertainty about the true value of an ounce of
gold. Every bidder would know the market price of gold, and hence all their
estimates and bids would be the same. The winner’s curse arises only when
there is uncertainty about the true value of the item. If the true value of the
item is more uncertain, then the probability that the highest estimate exceeds
the true value will be higher, and so the extent of the winner’s curse would
be greater.
• In a sealed-bid compared with an open auction. In an open auction, bidders
with relatively low values for the item will drop out progressively as the price
ascends. Since the record of bidding is open, the remaining bidders can see
the prices at which others drop out. The remaining bidders can then infer the
dropping bidders’ estimates of the true value. They can use this additional
information to refine their estimate of the true value. Hence, open bidding
mitigates the winner’s curse.
PROGRESS CHECK 13F

The seller of a vineyard has procured an independent appraisal and provided it
to bidders. How would this affect the extent of the winner’s curse?
WINNER’S CURSE: TUAS SEWERS
Governments use auctions to procure goods and services. The principles

of an auction to buy are exactly the same as for an auction to sell. In an
auction where the bidders compete to supply an item with uncertain cost, the
winner’s curse is to overestimate the true cost.
In the construction industry, the cost of underground works is more
difficult to estimate than above-ground works. In July 2014, the Public
Utilities Board of Singapore invited tenders by sealed bid for a contract
to build sewers in the Tuas area. Eight contractors submitted bids.
Aik Leong Plumbing Construction bid the lowest, S$22.3 million, while
Building Construction bid the highest, S$48.8 million. The average bid was
S$32.4 million.
The Public Utilities Board awarded the contract to Aik Leong. It will be
interesting to see whether Aik Leong underestimated the cost of construction.
Source: Government Electronic Business, www.gebiz.gov.sg (accessed December 9,
2014).
7. Signaling
Screening, an initiative of the less-informed party to elicit the characteristics of

the better-informed party, is one indirect way to resolve
asymmetric information. Another is signaling, an initiative Signaling: An initiative
of the better-informed
of the better-informed party to communicate its characteris- party to communicate its
tics to the less-informed party. The key is that the communi- characteristics to the less-
cation must be credible, that is, the parties with different informed party.
characteristics choose different signaling policies.
Recall the wine market where consumers cannot distinguish low- from high-quality
wine. Suppose that a high-quality producer offers to give a full refund to any con-
sumer who returns a partly-consumed bottle. Does the refund offer credibly signal
that the wine is indeed of high quality?
Consider a producer of low-quality wine. If it offers the refund, all consumers
would return their purchases for refund. So the low-quality producer will not offer
the refund. Thus, only high-quality producers offer the refund, and the refund
does induce self-selection and is a credible signal of high quality.
Signaling indirectly communicates the characteristics of the better-informed party
and thereby resolves the information asymmetry. The producer of high-quality
wine does not directly declare the product quality. Rather, the producer takes an
action that indirectly and credibly communicates the high quality.
To be credible, a signal must induce self-selection among the better-informed
parties. Specifically, the cost of the signal must be sufficiently lower for parties
with superior characteristics than for parties with inferior characteristics. Then
only those with superior characteristics will offer the signal.
Suppose that a wine producer labels its product “high quality.” Would this be a
credible signal? The answer depends on whether the action is relatively more costly
for a low-quality producer. The cost of the “high quality” label is the same for
producers of high and low-quality wine. Hence, such a label alone will not induce
self-selection among producers and cannot be a credible signal.
PROGRESS CHECK 13G

Explain the difference between screening and signaling.
TOSHIBA: LEADING INNOVATION – SIGNALING THROUGH

ADVERTISING AND REPUTATION
If labeling products as “high quality” is not a credible signal, why does Toshiba
advertise “Leading Innovation”? Investment in advertising and building
reputation can be a credible signal only if it induces self-selection – it must
pay off for sellers of high quality and not pay off for sellers of low quality. Such
signaling depends on three conditions:
• The investment must be large and sunk. If the advertising expenditure is

reversible, then a seller of low quality can also make the same investment,
pass on inferior products, and get its money back. A reversible investment
will not be a credible signal.
• Buyers must be able to detect poor quality fairly quickly. If a seller can
fool buyers for a long time, then even a seller offering poor quality can
afford the sunk investment.
• Information about the poor quality must cut the seller’s future sales and
quickly. A one-time seller can afford to pass off poor quality because it
will never face the punishment of losing repeat business.
8. Contingent Contracts
We have discussed two indirect ways to resolve an information asymmetry: screen-

ing and signaling. Another indirect approach is a contingent
Contingent contract: contract. A contingent contract specifies actions under
Specifies actions under particular conditions. Bets are contingent contracts: you get
particular conditions.
a dollar from me if the coin turns up heads, while I will get a
dollar from you if it turns up tails. In this bet, the contin-
gency is the side of the coin that faces up after the toss.
To see how contingent contracts can resolve an information asymmetry, con-
sider a wine producer selling a vineyard. Suppose that the seller knows that the
vines will yield an exceptional quantity of grapes, say, 6,000 tons a year. However,
it has no independent appraisal or other way to directly convince potential buyers.
What if the seller were to specify that it would sell the vineyard for a share
of the grape production rather than a straight cash payment? By asking for a
share, the seller is taking a payment that is contingent on the yield from the vines.
If the yield is high, then the seller will get a larger payment, while if yield is low,
then the seller will receive less.
By selling the vineyard for a share of the production, the seller can credibly convey
its information to potential buyers. Sellers of average or relatively low-yielding vines
would prefer to sell for cash. Hence, those selling relatively better vines can distin-
guish themselves by selling for a share of the production. The result is self-selection.
For a workplace example, suppose that an employer offers production-line
workers a choice between a fixed wage and piece-rate wage. The piece-rate wage is
essentially a contingent contract, with the wage contingent on the worker’s pro-
ductivity. The choice between fixed wage and piece-rate screens between workers
of differing productivity.
RESOLVING INFORMATION ASYMMETRY: BUYBACK
As the originator of the mortgages, Countrywide had better information about

the mortgages that it pooled for sale than buyers such as HSBC. Obviously
aware of this asymmetry of information, HSBC bought 80/20 mortgages subject
to a provision that it could require Countrywide to buy back mortgages that
did not meet specific conditions.
The right to require buyback is a contingent contract. Indeed, in early 2006,
HSBC exercised its right, and Countrywide had to repurchase the specified
mortgages and incurred substantial losses.
Source: Securities and Exchange Commission v. Angelo Mozilo, David Sambol, and Eric
Sieracki, US District Court for Central District of California, case no. CV09-03994, filed June 4,
2009.
HP: BUYING AUTONOMY WITHOUT CONTINGENT CONTRACT
In mergers and acquisitions, the buyer has less information about the value of
the acquisition than the seller. One way to resolve the information asymmetry
is a contingent contract that conditions payments on sales or profit targets.
Another way is to pay in shares, so that the value of the payment depends on
the value of the acquisition.
In August 2011, Hewlett-Packard (HP) announced that it would acquire British
software producer Autonomy plc for £25.50 per share in cash without any
conditions on future sales or profits. Analysts roundly criticized the $11.6 billion
acquisition. However, HP CEO Leo Apotheker defended the deal, saying: “We
have a pretty rigorous process inside H.P. that we follow for all our acquisitions,
which is a D.C.F.-based model . . . . And we try to take a very conservative view.”
Within a month, HP dismissed Mr Apotheker. Just over a year later, in November
2012, HP wrote down the acquisition by $8.8 billion, citing a “willful effort on
behalf of certain former Autonomy employees to inflate the underlying financial
metrics of the company in order to mislead investors and potential buyers”.
Source: “From H.P., a blunder that seems to beat all,” New York Times, November 30, 2012.
KEY TAKEAWAYS
• Risk is uncertainty about benefits or costs.

• A risk-averse person will pay to avoid risk.
• If information is asymmetric, the allocation of resources is not economically
efficient, and there is an opportunity to add value and profit by resolving the
asymmetry.
• If adverse selection is severe, the market will fail.

• Use appraisal if the characteristic about which information is asymmetric is
objectively verifiable and if it is not too costly.
• Use screening to indirectly elicit the characteristics of the better-informed
party if the screening induces self-selection.
• Adjust for the winner’s curse in auction bidding.
• A better-informed party can use signaling to indirectly communicate its
characteristics if the signaling induces self-selection.
• Use a contingent contract to screen or signal.
REVIEW QUESTIONS
1. Explain the difference between imperfect information and risk.

2. In the following situations, explain the asymmetry of information, if any:
(a) Investors do not know the next day’s Standard & Poor’s 500 Index.
(b) Acquirer is planning a takeover bid for Target at $50 a share, which is 25%
above the current market price of $40, and is secretly buying shares of Target.
3. A bank has just rejected Ming’s application for a car loan. Ming approaches the
loan officer and offers to pay a higher interest rate. Why does the loan officer
laugh?
4. If a borrower defaults on a secured loan, the creditor can seize and sell the item
against which the loan is secured. Explain why the interest rate on secured
loans is lower than that on unsecured loans.
5. A manufacturer of women’s clothing pays production workers a piece rate. The
human resources manager has proposed offering workers the alternative of a
fixed salary. Explain the possible adverse selection.
6. Jill is about to buy a secondhand car at a below-market price. The seller
assured Jill that the car is in perfect condition. Why should Jill get an expert
to evaluate the car?
7. Young people do not have much driving history. Does this explain why insurers
are reluctant to insure young drivers?
8. Give an example of screening. Explain: (a) the asymmetry of information; and
(b) self-selection.
9. An automobile insurance policy with a $2,000 deductible only covers loss in
excess of $2,000. Typically, automobile insurers offer policies with a choice
between higher deductibles and higher premiums. Explain how this choice can
screen among drivers with different probability of accident.
10. During peak hours, a tunnel is congested. From the standpoint of economic
efficiency, the tunnel service should be allocated to the drivers who value time
most highly. Explain how a toll can achieve economic efficiency.
11. This question relies on the auctions section. The seller of the rights to oil in a
particular area has undertaken a geological study. Will the winner’s curse be
more serious if the seller (a) provides the study to all bidders, or (b) keeps the
study secret?
12. How should a bidder adjust for the winner’s curse in (a) an auction to sell, and
(b) an auction to buy?
Table 13.1 Southwest Airlines
2010 2009 2008
Operating revenues ($ million) 12,104 10,350 11,023

Operating expenses ($ million) 11,116 10,088 10,574
Operating income ($ million) 988 262 449
Net income ($ million) 459 99 178
Available seat miles (millions) 98,437 98,001 103,271
Average fuel cost ($ per gallon) 2.51 2.12 2.44
Fuel consumed (million gallons) 1,437 1,428 1,511
13. Give an example of signaling. Explain: (a) the asymmetry of information; and
(b) self-selection.
14. A producer of financial management software offers full refunds to any
dissatisfied purchaser. Is the refund policy a credible signal of product quality?
15. How does the following contract help to resolve asymmetric information?
Acquirer buys all of Target’s shares and pays partly in cash and partly in Acquirer
shares.
1. Southwest Airlines pioneered the concept of the low-cost airline, operating from
secondary airports with short hops and quick turnarounds. In 2010, Southwest
increased net income by 4.6 times to $459 million. In 2008–2009, as the price
of oil fell sharply, Southwest decided to reduce its hedging, and so realized
hedging-related losses that reduced net income. Table 13.1 reports selected
financial and operating information.
(a) Referring to Southwest’s fuel consumption in 2010, explain how a
10-cent increase in the price of jet fuel would affect Southwest’s costs
and income.
(b) Southwest is averse to risk. Explain why it buys crude oil derivatives to
hedge the price of jet fuel.
(c) Suppose that the actual price of crude oil turns out to be lower than
the price at which Southwest hedged. Was Southwest wrong to have
hedged?
2. Some Chinese dairy producers add melamine to raw milk, which raises the
apparent protein content in testing. In 2008, melamine contamination caused six
babies to die and harmed hundreds of thousands of others. As Chinese parents
flocked to buy baby formula in Hong Kong, supermarkets had to ration sales.
(a) Using a relevant diagram, illustrate the equilibrium price and production
in the following market. The supply curve of contaminated milk is a
straight line from zero production at zero marginal cost to production of
1 million kilograms a month at 10 yuan per kilogram. The supply curve of
pure milk is a straight line from zero production at 10 yuan per kilogram
to 3 million kilograms a month at 25 yuan per kilogram. Consumers
cannot distinguish pure and contaminated milk. Their demand for pure
milk is zero at 50 yuan per kilogram and 10 million kilograms a month
at zero price. The demand for contaminated milk is zero at all prices.
(b) Using your diagram in (a), illustrate how the market can fail if farmers
increase the production of contaminated milk.
3. Hong Kong University of Science and Technology has contracted with HSBC
Insurance to provide medical insurance to all university employees. Medical
insurance covers the cost of the treatment and prescriptions necessary for the
insured party to recover from an illness or accident.
(a) Explain the asymmetry of information between the insurer and the
University employees.
(b) How does a group policy, covering all employees of an organization,
mitigate adverse selection for the insurer?
(c) Unlike falling sick or meeting with accidents, in many cases, women
voluntarily enter into pregnancy. Explain why HSBC covers pregnancy
as part of the basic coverage, rather than as an option.
4. Corporations borrow by issuing commercial paper and bonds. They may
commission ratings by credit rating agencies and provide the ratings to investors.
Moody’s ratings correlate well with actual defaults. Between 2005 and 2010,
the average five-year default rate was 0.81% among investment-grade (Baa
and higher) issuers, and 22.38% among speculative-grade (Caa and lower)
issuers. (Source: Moody’s Investors Service, “The performance of Moody’s
corporate debt ratings,” March 2010 quarterly update.)
(a) Explain the asymmetry of information between issuers of securities
(borrowers) and potential investors (lenders).
(b) Why do issuers of securities rather than potential investors commission
credit ratings?
(c) Why is it important that ratings have correlated well with actual
defaults?
(d) By issuing more debt, a corporation may raise the probability of default.
Should ratings be fixed or adjusted over time?
5. Historically, credit rating agencies rated issuers, i.e., companies and govern-
ments. A new and fast-growing business is to rate structured products such
as mortgage-backed securities. A single investment bank may issue tens or
hundreds of structured products. Ray McDaniel, CEO of Moody’s, acknowledged
that issuers negotiate and redesign structured products to achieve particular
credit ratings, and that issuers shop around the competing credit rating agencies
for the best rating. (Source: “Ray McDaniel, Moody’s,” FT View from the Top,
video, October 11, 2007.)
(a) Explain the asymmetry of information between issuers of structured
products and potential investors.
(b) Compare the degree of asymmetry of information about issuers vis-à-
vis structured products.
(c) In light of McDaniel’s disclosures, how effective is appraisal in resolving
asymmetries of information over structured products?
(d) Legislators have proposed to require issuers of mortgage-backed

securities to retain a minimum fraction of the securities that they issue.
How would this help to resolve the asymmetry of information?
6. In October 2004, Tiger Woods married Elin Nordegren at the Sandy Lane
resort, Barbados. Their prenuptial agreement specified that Ms Nordegren
would receive $20 million after 10 years of marriage. (Source: “New details on
Tiger’s prenup,” Daily Beast, December 3, 2009.)
(a) Explain the asymmetry of information in a marriage between a wealthy
man or woman and a relatively poorer spouse.
(b) Explain the prenuptial agreement in terms of a contingent contract.
(c) How can the prenuptial agreement screen gold-diggers from true life
partners?
(d) Can the relatively poorer person use the prenuptial agreement to signal
his or her true quality?
7. This question relies on the auctions section. In 2001, Singapore’s Land Transport
Authority shortlisted three contractors to build a section of the Marina Line
subway: Nishimatsu-Lum Chang Joint Venture (NLC), S$275 million; Impregilo
SPA-Hua Kok Realty Joint Venture, S$343 million; and Samsung Corporation
Engineering and Construction, S$345 million. Three years later, in May 2004,
the lowest three bids for contract 855 of the Circle Line were much closer:
NLC, S$376–389 million; Woh Hup-Shanghai Tunnel Engineering Co.–Alpine
Mayreder consortium, S$390–398 million; and Obayashi, S$400.3 million.
(a) Explain the winner’s curse in the context of the bidding for the Marina
Line.
(b) Explain how a contractor’s degree of risk aversion would affect the
amount that it bids.
(c) What experience from the Marina Line bidding did NLC apply in
bidding for contract 855?
(d) Explain why the Authority should provide all available information
about the soil conditions to the bidders.
(e) In tunneling projects, should the Authority offer to share part of the
contractor’s cost overrun? How would such sharing affect the
contractors’ bids?
8. Tesla Motors, pioneer of high-performance electric automobiles, launched a
sports car, the Roadster, in 2008, followed by the luxury Model S, in 2012.
In early 2014, Tesla announced an eight-year warranty on the battery pack
and drive unit with no mileage limit. CEO Elon Musk emphasized: “If we truly
believe that electric motors are fundamentally more reliable than gasoline
engines, with far fewer moving parts and no oily residue or combustion
by-products to gum up the works, then our warranty policy should reflect
that.” (Source: “Do warranty extensions make sense?” The Star (Toronto),
November 21, 2014.)
(a) Explain the asymmetry of information between Tesla and its potential
customers.
(b) Explain the extended warranty in terms of a contingent contract.
(c) Under what conditions will an extended warranty be a credible signal

of superior quality?
(d) How does Tesla’s financial condition affect the credibility of the warranty
as a signal?
9. Genzyme manufactures “orphan drugs” that target rare genetic diseases. In
July 2010, Sanofi-aventis offered to acquire Genzyme at $69 a share. Genzyme’s
board rejected the offer and sought $75 a share. Sanofi disagreed with Genzyme
about the future sales of its new multiple sclerosis drug, Lemtrada. Eventually,
in February 2011, Genzyme accepted a revised offer of $74 per share in cash
plus a contingent value right (CVR) worth up to $14 per share depending on the
sales of Lemtrada and other milestones. (Sources: “Sanofi steps up Genzyme
pursuit,” Financial Times, August 29, 2010; Genzyme Corporation media
release, “Sanofi-aventis to acquire Genzyme for $74.00 in cash per share plus
contingent value right,” February 16, 2011.)
(a) Explain the asymmetry of information between Sanofi and Genzyme.
(b) How could the CVR resolve the asymmetry of information? Compare
Sanofi’s valuation of the CVR with Genzyme’s valuation.
(c) Could the asymmetry be resolved directly by Sanofi inspecting
Genzyme’s accounts and facilities?

Identify any asymmetry of information in your organization’s activities. Write a
memo to the board of directors explaining how your organization could exploit
the profit opportunities.
Notes
1 This discussion is based, in part, on Securities and Exchange Commission v. Angelo Mozilo, David
Sambol, and Eric Sieracki, US District Court for Central District of California, case no. CV09-
03994, filed June 4, 2009.
2 This model is adapted from B. Peter Pashigan, Price Theory and Applications, New York:
McGraw-Hill, 1995, pp. 520–526.
3 This section is more advanced. It may be omitted without loss of continuity.
C H A P T E R
14
Incentives and Organization
LEARNING OBJECTIVES
• Appreciate moral hazard.
• Apply monitoring and incentives to resolve moral hazard.
• Appreciate that incentives create risk and affect performance on
multiple responsibilities.
• Appreciate holdup and the application of detailed contracts.
• Understand ownership and its consequences.
• Appreciate vertical integration.
• Understand and apply organizational architecture.
1. Introduction
In April 2004, Boeing launched development of the 787 Dreamliner with an order
of 50 planes from All Nippon Airways. The Boeing 787 is a twin-engine medium-
to long-range wide-body jet with capacity of 200–300 passengers. With extensive
use of composite materials in construction, the 787 is designed to consume 20%
less fuel than the similar-sized Boeing 767.1
Traditionally, Boeing designed and built the major parts, and then assembled
the planes at its Everett and Renton, Washington, factories. In a major departure,
the then Chairman and CEO, Harry Stonecipher, and the then head of Commer-
cial Airplanes, Alan Mulally, decided to subcontract the design and construction
of major parts of the 787, including the wing, fuselage, and tail. Boeing would
then need only a relatively small workforce at the Everett factory to carry out final
assembly.
330 14 Incentives and Organization
By radically outsourcing the production work, Boeing aimed to substantially

reduce the assets required to support production, and so boost the return on net
assets. Mr Stonecipher had applied the same outsourcing strategy as President
and CEO of military contractor McDonnell Douglas, which merged with Boeing
in 1997.
Boeing had planned the first flight of the 787 Dreamliner for August 2007 and
deliveries to All Nippon Airways by late 2008. However, by mid-2007, construc-
tion was behind schedule. The Chairman and CEO, James McNerney (successor
to Stonecipher, who was dismissed in March 2005 for a “personal relationship”
with another employee) acknowledged that subcontractors had not delivered
according to specification: “We were surprised on the physical reality of some of
the things that we received from suppliers versus the documentation.”
After repeated delays, the first flight actually took place only in December 2009.
Meanwhile, Boeing was forced to buy a factory in Charleston, South Carolina,
from Vought Aircraft Industries at a cost of $1 billion. Vought had contracted to
design and build the 787’s rear fuselage. However, facing technical and financial
challenges, Vought’s owner, the private-equity investor Carlyle Group, would not
increase its investment and instead sought to sell the business. The Vought CEO,
Elmer Doty, explained that “the financial demands of this program are clearly
growing beyond what a company our size can support.”
How was Boeing’s outsourcing related to supplier delays and its surprise at the
quality of supplier performance? Why did the outsourcing lead Boeing to buy the
Charleston factory from Vought? How would Boeing’s purchase affect the perfor-
mance of the Charleston factory?
Organizational architecture comprises the distribution
Organizational of ownership, incentive schemes, and monitoring systems.
architecture: The Boeing’s experience with production of the Dreamliner
distribution of ownership, clearly demonstrates the importance of organizational archi-
incentive schemes, and tecture. Far from increasing the company’s return on net
monitoring systems.
assets, the outsourcing strategy inflated development costs by
an estimated 50–80% over the original plan of $8–10 billion.
This chapter presents a framework for analyzing organizational architecture. An
efficient organizational architecture resolves four issues of internal management –
holdup, moral hazard, monopoly power, and economies of scale and scope. Holdup
is an action to exploit the dependence of another party. For instance, Vought Air-
craft Industries refused to increase investment despite challenges in producing the
rear fuselage sections. Vought’s holdup forced Boeing to buy the Charleston plant.
Moral hazard arises when one party’s actions affect but are not observed by
another party. It results from asymmetry of information about actions. The Boeing
CEO James McNerney expressed surprise at the subcontractors’ quality of work.
Evidently, Boeing suffered from the moral hazard of its subcontractors. The moral
hazard was further manifest in the systematic delays to the Dreamliner program.
The architecture of an organization, particularly vertical integration, affects
the degree of moral hazard and potential for holdup. Vertical integration is the
Incentives and Organization 331
opposite of outsourcing. When an organization outsources a particular input,

it is vertically disintegrating out of the production of that input. By acquiring
the Charleston factory, Boeing reversed its outsourcing strategy, and vertically
integrated into the manufacturing of the rear fuselage and so avoided further
holdup.
Besides moral hazard and holdup, the architecture of an organization also influ-
ences the extent of internal monopoly power and economies of scale and scope.
Together, all of these determine the economic efficiency of the organization. To
the extent that there is any economic inefficiency, there will be an opportunity for
managers to add value and increase the profit.
2. Moral Hazard
Marie is a salesperson. By the very nature of her job, a salesperson operates inde-
pendently. Hence, it is difficult for Marie’s employer to monitor her work. Marie
alone decides how many customers to visit and how much effort to spend on sell-
ing. Her employer wants her to exert the maximum effort – to be patient and
persuasive, yet persistent.
In this example, the salesperson is subject to moral hazard
relative to her employer. A party is subject to moral hazard Moral hazard: One
if its actions affect but are not observed by another party. party’s actions affect
but are not observed by
another party.
Asymmetric Information about Actions
If the employer can freely monitor the salesperson at all times, then it can directly
specify her effort – how many customers to visit, how to persuade them, etc.
Then the salesperson would not be subject to moral hazard. Moral hazard arises
because information is asymmetric. The employer depends on the salesperson’s
effort but cannot observe it.
Chapter 13 discussed asymmetric information in markets such as life insurance
where the applicant has better information about her health, and lending where
the borrower has better information about her ability to repay. In those situations,
the asymmetry of information concerned some characteristic(s) of the better-
informed party.
By contrast, in the case of the salesperson, Marie has better information about
her effort than her employer. In this case, the asymmetry of information concerns
some action of the better-informed party. This information asymmetry is a neces-
sary condition for moral hazard.
Economic Inefficiency
The salesperson acts independently. She chooses her effort to maximize her per-
sonal net benefit. The salesperson’s net benefit is her compensation less her cost of
effort. The level of effort that maximizes her net benefit is that where her marginal
compensation equals her marginal cost. Referring to Figure 14.1, the salesperson
will choose 120 units of effort. At that level of effort, the salesperson’s marginal
compensation and marginal cost are $10 per unit of effort.
What if the salesperson were to increase her effort by one unit? Figure 14.1
also shows the employer’s marginal profit contribution. The salesperson’s effort
increases her employer’s profit contribution. When the salesperson exerts 120
units of effort, the employer’s marginal profit contribution is $25. So, the salesper-
son’s additional unit of effort would increase the employer’s profit contribution
by a little less than $25 (a little less because the marginal profit contribution curve
slopes downward).
The cost to the salesperson of the additional unit would be a little more than
$10 (a little more because the marginal cost curve slopes upward). So, the addi-
tional unit of effort would add value by increasing the employer’s marginal profit
contribution by more than the salesperson’s marginal cost. Indeed, the salesper-
son could continue adding value by raising effort up to 250 units. The additional
value (the excess of profit contribution over cost) to the employer and salesperson
would be the shaded area abc.
Essentially, the salesperson is generating a positive externality for her employer.
So, as with any positive externality, if it is not resolved, then the situation is not
economically efficient. So there is an opportunity to add value and raise profit by
resolving the externality.
In the sales context, resolving the externality means resolving the moral hazard
between the salesperson and her employer. The shaded area abc represents the
Marginal benefit/cost ($ per unit effort)
b Salesperson’s
25 marginal cost
c
Employer’s
a marginal profit
10 contribution
Salesperson’s
marginal
compensation
0 120 250
Quantity of effort (units a month)
FIGURE 14.1 Economically efficient effort.

Notes: The salesperson chooses 120 units of effort, which balances her marginal compensation
with her marginal cost. The economically efficient level of effort, 250 units, balances the employer’s
marginal profit contribution with the salesperson’s marginal cost.
amount of profit that can be earned by resolving the moral hazard. The challenge
then is how to do so.
Degree of Moral Hazard

Suppose that the salesperson’s marginal compensation coincides exactly with the
employer’s marginal profit contribution. Then the salesperson would choose the
economically efficient level of effort. Under these conditions, there will be no
moral hazard.
Referring to Figure 14.1, the lower the salesperson’s marginal compensation is
relative to the employer’s marginal profit contribution, the lower will be the effort
that the salesperson chooses relative to the economically efficient level.
We can measure the degree of moral hazard by the difference between the eco-
nomically efficient action and the action chosen by the party subject to moral
hazard. The larger this difference is, the greater are the degree of moral hazard
and the added value that can be realized by resolving the moral hazard.
PROGRESS CHECK 14A

Suppose that the salesperson’s marginal cost of effort in Figure 14.1 is higher.
Draw the new marginal cost curve. How does this affect: (a) the economically
efficient level of effort; (b) the effort that the salesperson actually chooses?
MORAL HAZARD AT THE TOP
Top executives in large publicly listed corporations are subject to moral hazard
relative to shareholders. Publicly listed companies have many diverse share-
holders ranging from pension funds with million-dollar holdings to individuals
with several hundred shares. It is not worthwhile for small shareholders to
monitor the managers of the company. They can instead free-ride on the
monitoring efforts of other shareholders.
Shareholders are primarily concerned about the value of their shares. By
contrast, CEOs and other senior managers may have different objectives.
On April 20, 2010, British Petroleum’s oil rig, Deepwater Horizon, located in
the Gulf of Mexico, exploded. BP’s CEO, Tony Hayward, downplayed the incident,
describing the Gulf as a “big ocean” and claiming that the environmental
impact would be “very very modest.”
During the ensuing crisis, Hayward took time off to sail his yacht. Facing
mounting criticism, he decided to resign. The price of BP shares rose by
4.85%, increasing the market value of the company by $5.6 billion.
Source: “BP chief Hayward ‘negotiating exit deal,’” BBC News, July 25, 2010; “BP oil spill: the
rise and fall of Tony Hayward,” Telegraph, July 27, 2010.
BOEING 787: A NEW WAY TO BUILD PLANES
Traditionally, Boeing designed and built the major parts, and then assembled
planes at its Everett and Renton, Washington, factories. In a major departure, the
then Chairman and CEO, Harry Stonecipher, and the then head of Commercial
Airplanes, Alan Mulally, decided to subcontract design and construction of
major parts of the 787, including the wing, fuselage, and tail. Boeing would then
need only a relatively small workforce at the Everett factory to carry out final
assembly. By extensive outsourcing, Boeing hoped to reduce its production
assets, and so boost its return on net assets.
Boeing had planned the first flight of the 787 Dreamliner for August 2007
and deliveries to All Nippon Airways by late 2008. However, by mid-2007,
construction was behind schedule. The Chairman and CEO, James McNerney
(successor to Stonecipher), acknowledged that subcontractors had not delivered
according to specification: “We were surprised on the physical reality of some
of the things that we received from suppliers versus the documentation.”
The outsourcing strategy had placed Boeing in a position of asymmetric
information relative to its subcontractors. The result was severe moral hazard,
to the extent that Boeing had to despatch hundreds of manufacturing engineers
and procurement staff to assist major subcontractors. Boeing had to provide
almost $1 billion for just the first six-month delay in production.
Source: “787 delay could wind up costing Boeing $1 billion,” Seattle Times, October 25, 2007.
3. Incentives
We have identified the potential gains from resolving moral hazard. Generally,
there are two complementary approaches. One is to invest in monitoring, surveil-
lance, and other methods of collecting information about the actions of the party
subject to moral hazard. The other approach is to align the incentives of the party
subject to moral hazard with those of the less-informed party.
Monitoring systems and incentive schemes are two elements of organizational
architecture. They are complementary because all incentives must be based on
actions that can be observed, so the better the available information, the wider
will be the choice of incentive schemes. Ideally, the relevant parties would like to
completely resolve the moral hazard, so that the better-informed party will make
the economically efficient choice. Let us now discuss how to resolve the moral
hazard of a salesperson relative to her employer.
Monitoring
The simplest monitoring system focuses on objective measures of performance
such as hours on the job. However, hours on the job and effort are not the same
thing. A salesperson can start at 8:00 am and finish at 5:00 pm, but do nothing
during that time. Accordingly, employers need monitoring systems that provide
more than basic objective information.
One method that employers frequently use to collect information is supervi-
sion. It is not cost-effective for supervisors to monitor subordinates all the time,
however. So supervisors make random checks. Chapter 10 discussed the advan-
tages of randomization. The same principle applies to supervision: the supervisor
should check staff at random, rather than according to some regular pattern.
Employers can also enlist customers to monitor the performance of employees.
Customers have a natural advantage in monitoring salespersons, as salespersons
spend more time with customers than at the employer’s office. Employers can
encourage customers to report salespersons’ performance.
Performance Pay
The counterpart to monitoring systems is incentive schemes. Incentive schemes
resolve moral hazard by linking compensation to some measure of performance.
The schemes depend on a link between the unobservable action and some observ-
able measure of performance. Generally, the scope of incentive schemes depends
on what indicators of the unobservable action are available.
An employer can use the information provided by monitoring systems to structure
incentives for its workers. While Marie’s employer cannot observe her effort, it can
observe the value of her sales. So, it can base incentives on Marie’s sales revenue.
One common incentive scheme is performance pay,
which bases pay on some measure of performance. Let us
Performance pay: An
consider performance pay for the salesperson. Suppose first
incentive scheme that
that Marie receives a fixed wage of $4,000 a month and that bases pay on some
her employer does not monitor her at all, not even requiring measure of performance.
her to keep a diary of sales calls.
With a fixed wage and no monitoring, Marie cannot affect
her earnings in any way, however hard she works. Her marginal compensation
from effort will be zero. In Figure 14.2, the salesperson’s marginal compensation
with a fixed wage is the horizontal axis. This is lower than the salesperson’s mar-
ginal cost at all levels of effort. Hence, she would choose zero effort.
Now suppose that the employer pays Marie a 10% commission on sales reve-
nue. This is an example of payment based on performance. The more Marie sells,
the more she will earn. With this incentive scheme, her marginal compensation
from effort will be positive. The height and slope of her marginal compensation
curve depend on how her effort affects sales revenue.
Figure 14.2 also shows the marginal compensation with a 10% commission.
This crosses her marginal cost at an effort of 30 units. With a 10% commission,
the salesperson chooses 30 units of effort. Accordingly, the commission partly
resolves the salesperson’s moral hazard.
An incentive scheme is relatively stronger if it provides a higher marginal com-
pensation for effort. Suppose that the employer strengthens the incentive scheme
by raising the commission to 15%. Then the marginal compensation curve would
Salesperson’s
Marginal benefit/cost
marginal cost
($ per unit effort)

c Employer’s
marginal profit
contribution
e
Salesperson’s Salesperson’s marginal
marginal compensation with 10%
compensation commission
with fixed wage
0 30 250
FIGURE 14.2 Performance pay.

Notes: With a fixed wage, the salesperson’s marginal compensation is the horizontal axis, and she
chooses zero effort. With a 10% commission, she chooses 30 units of effort.
be higher, and it would cross the marginal cost curve at a higher level of effort. This
shows that a stronger incentive scheme induces the salesperson to increase effort.
PROGRESS CHECK 14B

Using Figure 14.2, draw a marginal compensation curve such that the salesper-
son would choose the economically efficient level of effort.
Performance Quota
If the employer pays a 100% commission, Marie’s marginal compensation
would align with the employer’s marginal revenue and she would choose the
economically efficient 250 units of effort. But then her employer would make no
profit.
Another way to induce the salesperson to choose the
Performance quota: economically efficient level of effort is a performance quota.
A minimum standard A performance quota is a minimum standard of perfor-
of performance, below mance, below which penalties apply. The penalties could
which penalties apply.
include deferral of promotion, reduction in pay, or even
dismissal.
To apply a performance quota, the employer must identify the sales revenue
that would result if the salesperson chose the economically efficient 250 units
of effort. Suppose that 250 units of effort would generate sales revenue of
$30,000. Then the employer should set the performance quota at $30,000 a
month. Figure 14.3 shows the salesperson’s marginal compensation curve with
such a quota.
Marginal benefit/cost ($ per unit effort)
Salesperson’s
marginal cost
c Employer’s
marginal profit
contribution
Salesperson’s marginal
compensation with quota
0 250
FIGURE 14.3 Performance quota.

Note: With a performance quota, the salesperson chooses 250 units of effort, which balances her
marginal compensation with her marginal cost.
The salesperson’s marginal compensation curve has three parts. Recall that the
employer pays the salesperson a fixed monthly wage provided that she meets the
quota; otherwise the salesperson will be dismissed. At 249 units of effort and
below, the salesperson will be dismissed. Additional effort does not affect her
earnings; hence, her marginal compensation is zero.
The incentive scheme pays the salesperson no extra for additional effort above
250 units. Accordingly, at 251 units of effort and above, the salesperson’s marginal
compensation also is zero. The marginal compensation, however, is very high at
exactly 250 units of effort. An increase in effort from 249 to 250 units is just
enough to satisfy the quota and hence allows the salesperson to retain her job.
Thus, the salesperson’s marginal compensation curve follows the horizontal
axis from 0 to 249 units of effort, spikes up at 250 units, and then follows the
horizontal axis again at 251 units of effort and above. Therefore, the marginal
compensation curve crosses the marginal cost at 250 units of effort. Accordingly,
the salesperson chooses 250 units of effort.
A performance quota is a cost-effective way of inducing the salesperson to
choose the economically efficient level of effort. It is cost-effective because it does
not reward effort below or above the economically efficient level. It focuses the
incentive at the economically efficient level of effort.
PROGRESS CHECK 14C

Suppose that 200 units of effort would generate $25,000 worth of sales. Using
Figure 14.3, illustrate the salesperson’s marginal compensation if the employer
specifies a sales quota of $25,000 a month.
MONITORING TRUCK DRIVERS: ONBOARD COMPUTERS
Truck drivers operate independently and may travel far from base. The
management of a trucking business may equip trucks with onboard computers
to monitor drivers.
A study of the US trucking industry found significant patterns in the adoption
of onboard computers. Only 7% of owner-operated trucks were equipped
with onboard computers, as compared to 19% of employee-operated trucks.
An owner-operator would not need to monitor himself, hence derives less
benefit from an onboard computer.
Only 6% of trucks that operated at less than 50 miles from base were
equipped with an onboard computer. By contrast, 19% of trucks that operated
at distances of 100–200 miles had onboard computers. Truckers who
operate at greater distance will be more difficult to monitor through personal
supervision. Hence, management derives a greater benefit from installing an
onboard computer to monitor such drivers.
Source: Thomas N. Hubbard, “The demand for monitoring technologies: the case of trucking,”
Quarterly Journal of Economics, Vol. 115, No. 2, May 2000, pp. 533–560.
WHEN INCENTIVES ARE TOO WEAK: REAL ESTATE AGENTS
Real estate agents are paid a percentage commission on the sales price. So
both the agent and home seller would like to get a higher price. However, the
agent incurs effort in selling, which the seller might not observe. So the agent
is subject to moral hazard.
Consider, for instance, a $300,000 house. Assume that the brokerage fee
is 6%, split equally between the seller’s and buyer’s agents. The seller’s agent
must split half of her fee with her agency, so she actual receives 1.5%.
Suppose that the agent could sell the house for $310,000 with more effort.
The seller would get 94% of the additional $10,000, or $9,400. The agent
would get 1.5% of the additional $10,000, or $150. Would the agent consider
the extra effort worth $150?
Professor Steven Levitt studied the sales of 100,000 Chicago houses,
some of which were owned and sold by the agents themselves. On average,
the agent-owned homes were marketed for 10 days longer and achieved a
3% higher price. Not surprising, since an agent selling her own house would
not be subject to moral hazard.
Source: Steven D. Levitt and Stephen J. Dubner, Freakonomics, London: HarperCollins,
2005, pp. 8–9.
4. Risk and Multiple Responsibilities
The combination of incentives and monitoring can help to resolve moral hazard.
Here, we expand on the analysis of incentives to consider two serious side-effects of
incentives. One is risk and the other is worse performance on other responsibilities.
Risk
Incentive schemes resolve moral hazard by linking compensation to some observ-
able measure of the unobservable action. But what if the measure is affected by
factors other than the unobservable action? Then the payments will depend on
these other factors. A party who is subject to moral hazard and has imperfect
information about these factors will face risk.
Consider, for instance, the salesperson’s commission on monthly sales revenue.
Besides the salesperson’s effort, the actual sales may depend on the general state
of the economy, competition and other factors. The salesperson may be uncertain
about these other factors. So the commission scheme imposes risk on her. Risk will
arise if the party subject to moral hazard is uncertain about her compensation.
To achieve economic efficiency, the incentive scheme must balance the incentive
for effort with the cost of risk. The cost of risk depends on three factors:
• Impact of extraneous factors. The extent of risk depends on the degree to

which the extraneous factors affect the measure on which incentives are
based. If the measure is sensitive to these extraneous factors and the factors
are subject to wide swings, then the risk would be relatively large.
• Risk aversion. If the party subject to moral hazard is risk neutral, then the risk
imposes no cost. The cost of risk increases with the degree of risk aversion.
• Strength of incentive scheme. Stronger incentives would induce the salesperson
to increase her effort. However, stronger incentives would also impose a heavier
burden of risk. For instance, with a higher commission, a larger part of the
salesperson’s income would depend on sales and vary with uncertain factors.
Generally, the incentive scheme should be stronger if the Cost of risk depends on:
extraneous factors are weaker and the party subject to moral • impact of extraneous
hazard is relatively less risk averse. Conversely, the incentives factors;
should be weaker if the extraneous factors have a stronger • risk aversion;
influence and risk aversion is higher. • strength of incentive
scheme.
Relative Performance Incentives

One way to resolve moral hazard without imposing risk is to use relative perfor-
mance incentives. In the sales context, Marie’s employer could pay each salesper-
son a fixed monthly wage plus a commission for sales revenue in excess of the
average for all salespersons.
This incentive scheme would not penalize salespersons for extraneous factors
like a bad economy. If the economy weakens, this would affect all the sales of all
salespersons. If Marie exerts relatively more effort, she will still achieve higher
sales than the average and hence will earn a larger commission.
By gauging performance on a relative basis, the incentive scheme cancels out the
effect of extraneous factors to the extent that they affect all salespersons equally.
This reduces the risk due to extraneous factors. Relative incentive schemes are
most useful where common extraneous factors are important.
Multiple Responsibilities
In many situations, the party subject to moral hazard has multiple responsibilities.
For instance, a sales representative may be responsible for initial sales as well as
providing post-sales service. In a factory, production supervisors may be respon-
sible for meeting output targets as well as maintaining quality.
Ideally, in these situations, an incentive scheme should resolve a balance of
the multiple responsibilities. This means that there should be some investment in
monitoring each of the unobservable actions and incentives based on the corre-
sponding indicators.
Balancing multiple responsibilities becomes harder when it is more difficult to
measure performance on some responsibilities than others. An incentive scheme
may focus on a particular responsibility because that dimension is relatively easier
to monitor. Recall that the scope of an incentive scheme depends on the available
indicators of the unobservable action.
Suppose that the incentive scheme focuses on just one responsibility. Then
it will induce better performance on that dimension but have the side-effect of
aggravating the moral hazard with regard to the other responsibilities.
For instance, a factory wants its production supervisors to meet output targets
as well as maintain product quality. Output is easy to measure. However, prod-
uct quality may be difficult to measure. If the factory adopts a strong incentive
scheme for output, the supervisors would tend to focus on output, and quality
would fall.
Incentive schemes focus on actions for which there are reliable measures of per-
formance. If, however, there are important responsibilities for which it is difficult
to measure performance, then it may be better to adopt relatively weak perfor-
mance incentives in general. A deliberate use of weak incentives is a way to achieve
a balance among multiple responsibilities.
PROGRESS CHECK 14D

Suppose that a department store has switched its sales clerks from a fixed sal-
ary to a salary plus a commission on sales. How will this affect the sales clerks’
incentive to process returns?
EVALUATING MANAGERIAL PERFORMANCE: MEDTRONIC
Headquartered in Minneapolis, Medtronic manufactures medical devices for

a wide range of conditions, including cardiac rhythm disease, coronary artery
and peripheral vascular disease, diabetes, and neurological illnesses. In
2010–2011, Medtronic earned net income of $3.10 billion (or $2.86 per share)
on revenue of $15.93 billion.
One measure of the performance of a company is total shareholder return,
which includes dividends and appreciation in the share price. Between 2006
and 2011, Medtronic yielded total shareholder return of −9.4%. If that were not
bad enough, during the same period, the total shareholder return on the S&P
500 Health Care Equipment Index was 15%.
This example suggests that a reasonable way of evaluating companies is
to measure their performance against that of other companies in the same
industry. Relative performance evaluation cancels out background factors
over which managers have no control and provides a more accurate measure
of management’s performance.
Indeed, Medtronic links incentive pay for senior managers to three measures:
growth of earnings per share, and return on capital and growth of revenue
relative to a peer group of manufacturers of pharmaceuticals and medical
instruments.
Sources: Medtronic, Inc., proxy statement, July 15, 2011; Form 10K, 2011.
5. Holdup
Moral hazard arises when one party’s actions affect but are not observed by
another party. A related managerial issue is holdup. To explain holdup, suppose that
Luna Supermarket engaged Mercury Logistics to deliver grocery orders. Their
contract specified that Mercury should make two rounds of deliveries a day, at
12:00 noon and 4:00 pm. One day, however, Luna received so many orders that it
asked Mercury to make a third round of deliveries. Mercury took the opportunity
to demand twice the usual fee.
In this example, Mercury took advantage of Luna’s special need to hold up the
supermarket. A holdup is an action to exploit another par-
ty’s dependence. Holdup is distinct from moral hazard in Holdup: An action to
that it does not require asymmetric information. For exam- exploit another party’s
dependence.
ple, when Luna requested the extra delivery, Mercury openly
demanded a higher fee. Luna could clearly observe Mercu-
ry’s action – there was no asymmetry of information.
Mercury’s holdup has implications beyond the exceptional fee that Luna paid
on one occasion. The potential for holdup in the future would lead Luna to take
precautions. For instance, Luna might limit the number of delivery orders that
it accepts, or it might even establish its own delivery service. These precautions
would either reduce Luna’s revenues or increase its costs.
Generally, whenever there is the potential for holdup, other parties will take
precautions to avoid dependence. These precautions either reduce benefits or
increase costs. So, the potential for holdup reduces overall value and economic
efficiency. This means that there is an opportunity to add value and increase profit
by resolving the holdup.
Specific Investments
One particular type of precaution against holdup is worth emphasizing. It is to
reduce specific investments. Suppose that, to optimize delivery time and fuel,
Mercury has installed a computerized route planning system. Clients must pre-
pare delivery orders using specialized software.
In this example, acquiring, installing, and learning to use
Specificity: The the delivery system is an investment by Luna that is spe-
percentage of an
cific to its relationship with Mercury. The specificity of an
investment that will be lost
if the asset is switched to investment in an asset is the percentage that will be lost if the
another use. asset is switched to another use.
For instance, suppose that Luna must spend $4,500 to
acquire the delivery system, comprising software and a computer, and $500 on
training. If Luna switches to another logistics provider, it can reuse only the com-
puter, worth $1,000. It would have to write off the software and training. So, the
specificity of the investment in the delivery system is (4,500 + 500 − 1,000)/(4,500 +
500) = 80%.
Any asset, whether physical or human, can be specific. Learning to use an orga-
nization’s systems is a specific human investment by the employee. The training
that organizations provide to employees is a specific investment. If the employees
quit, the organization will have to write off the investment.
The potential for holdup will deter all forms of specific investments. If holdup
could be prevented, the relevant parties would increase specific investments and
so add value and increase profit.
Incomplete Contracts
Suppose that the contract between Luna and Mercury had specified conditions under
which Luna could request an additional delivery and the corresponding fee. Then
Mercury would not have been able to hold up Luna. Generally, the scope for a holdup
depends on the extent to which the contract between the parties is incomplete.
A complete contract specifies the actions of all parties
Complete contract: under every possible contingency. By contrast, a contract is
Specifies the actions of incomplete if it does not specify the actions of the parties
all parties under every in some contingency. The actions specified by the contract
possible contingency.
possibly include payments.
It would be extremely costly for Luna and Mercury to agree and prepare a
complete contract. A huge number of contingencies must be covered: the need for
a fourth delivery, the possibility that Mercury’s truck may break down, and the
possibility of an earthquake are just a few. Rather than consider every such detail,
the two parties will probably agree on an incomplete contract.
So, in practice, all contracts are incomplete, and deliberately so. The issue is
then how incomplete the contract should be. Generally, the appropriate degree of
incompleteness depends on two factors:
• Potential benefits and costs. The larger the stakes, the more the parties should
invest in preparing the contract. Luna Supermarket sells dairy products and small
stationery items. Dairy products account for a large part of Luna’s sales and are
much more important than small stationery items. Accord-
ingly, Luna should write a more detailed contract with its A contract should be
supplier of dairy products than its supplier of stationery. more detailed if:
• Possible contingencies. Dairy products are perishable and • potential benefits and
sold in high volume, hence Luna needs frequent supply. costs are larger;
Moreover, the supply of dairy products is relatively more • possible contingencies
vulnerable to disruptions by bad weather, transportation are more serious.
problems, and labor disputes. So, Luna needs more assurance
about the supply of dairy products, and so will write a more detailed contract.
PROGRESS CHECK 14E

What factors should the parties consider in deciding how detailed a contract
to prepare?
AVOIDING SPECIFICITY IN ELECTRIC POWER INVESTMENTS:

FLOATING POWER PLANTS
Plagued by violence and instability, Iraq faces many challenges, one of which
is to provide a reliable supply of electricity. Many Iraqi businesses and families
have bought generators to provide a necessary but costly backup supply.
In 2007, the Turkish energy group, Karadeniz Holding AŞ, initiated the
“Power of Friendship” project to supply 615 megawatts (MW) of electricity
to Iraq. Karadeniz bought secondhand generating sets in China and Dubai. It
then arranged to overhaul the generators and install them on four vessels at
shipyards in Turkey and Singapore.
The Doğan Bey, with a capacity of 126.4 MW, and Kaya Bey, with a capacity
of 216.4 MW, are moored at the ports of Um Qasr and Al Zubayr, and supply
electricity to the Basra region in south-eastern Iraq.
Source: “Wärtsilä receives O&M agreement for Iraqi floating power plants,” Wärtsilä
Corporation, trade and technical press release, May 12, 2010.
6. Ownership
Holdup can be resolved by changing the ownership of the

Ownership: Rights to
residual control, which are
relevant assets. Ownership means the rights to residual con-
those rights that have not trol, which are those rights that have not been contracted
been contracted away. away.
To explain the meaning of residual control, suppose that
Saturn Properties borrowed $5 million from a bank to develop a supermarket,
which it has rented to Luna on a five-year lease. The bank has a mortgage against
the building. This means that, if Saturn fails to make the loan payments on time,
the bank will have the legal right to take possession of the building. This is a right
that Saturn contracted away to get the loan.
Saturn has also entered into a five-year lease with Luna Supermarket. Through
the lease, Luna has the right to use the property for five years. This is another right
that Saturn has contracted away.
As owner, Saturn has residual control. This means that it has all rights except
those contracted away. For instance, it may have the right to enter into a second
mortgage on the building, and it has the right to use the building after the expira-
tion of Luna’s lease.
A transfer of ownership means shifting the rights of residual control to another
party. Suppose that Luna buys ownership of the supermarket building from Saturn.
Then Luna would have all the rights that previously belonged to Saturn.
Residual Income
One dimension of residual control is particularly worth empha-
Residual income:
sizing. The owner of an asset also has the right to receive
Income remaining after
payment of all other the residual income from the asset, which is the income
claims. remaining after payment of all other claims.
To illustrate, suppose that Saturn collects $100,000 a
month in rent from Luna. Saturn’s expenses include $50,000 in interest and prin-
cipal to the bank as well as $20,000 in taxes and other expenses. As owner of the
building, Saturn receives the residual income of $100,000 − $50,0000 − $20,000 =
$30,000.
The important implication is that, as the recipient of the residual income, the
owner gets the full benefit of changes in income and costs. For instance, if Saturn
can raise the rent by $5,000 to $105,000, then its profit would increase by $5,000 to
$35,000. Similarly, if Saturn can reduce expenses by $2,000, then its profit would
increase by $2,000 to $32,000.
So the owner has the full incentive to maximize the value of the asset. Other
parties would not have the same incentive. If information about their actions is
asymmetric, they would be subject to moral hazard relative to the owner. Even in
the absence of asymmetric information, they may hold up the owner and exploit
the owner’s dependence.
Vertical Integration
As introduced in Chapter 11, vertical integration is the
Vertical integration: The
combination of the assets for two successive stages of
combination of assets for
production under a common ownership. With common two successive stages
ownership, the owner would have full incentive to maxi- of production under
mize the value of the combined assets. By contrast, with common ownership.
separate ownership, the owner of each asset would only
maximize the value of its asset, and possibly at the expense of the owner of the
other asset.
Vertical integration is downstream or upstream, depending on whether it
involves the acquisition of assets for a stage of production nearer to or further
from the final consumer. Suppose, for instance, that a food manufacturer acquires
a supermarket. Since the supermarket operates a stage of production nearer to
the final consumer, this is an example of downstream vertical integration. By con-
trast, if the food manufacturer were to acquire a dairy farm, it would be vertically
integrating upstream.
The decision to vertically integrate upstream is often characterized as the choice
of whether to “make or buy.” The food manufacturer can either buy milk from
farmers or establish its own farm to make the input for its production.
Similarly, the decision to vertically integrate downstream may be characterized
as the choice of whether to “sell or use.” The dairy farmer can either sell its prod-
ucts to food manufacturers or establish its own factory and use its milk as input
into production of processed foods.
Vertical integration or disintegration changes the ownership of assets, and
hence alters the distribution of the rights to residual control and residual income.
As we explain in the next section, these in turn affect the degree of moral hazard
and the potential for holdup.
PROGRESS CHECK 14F

Explain the difference between upstream and downstream vertical integration.
CONTRACT AND OWNERSHIP: THE NORTH EUROPEAN

GAS PIPELINE
Gazprom, one of the world’s largest energy producers, initiated the North
European Gas Pipeline to transport gas from central Russia under the Baltic
Sea to western Europe. The total investment is €4.6 billion. Of course, once
the pipeline is completed, the costs of construction are sunk. So, Gazprom is
vulnerable to holdup by the customers for its gas.
Before construction, in December 2005, Gazprom established the North

European Gas Pipeline Company with two German companies, Wintershall,
a unit of BASF, and E.ON Ruhrgas, as minority shareholders to build the
pipeline. Subsequently, the company was renamed Nord Stream, and the
German shareholders sold part of their shares to the leading gas suppliers,
Gasunie of the Netherlands and GDF/Suez of France.
Soon afterward, Gazprom locked in the key German customers with long-
term contracts. In January 2006, Wintershall signed a 20-year contract to buy 8
billion cubic meters of gas a year from 2010 to 2030. Then, in August 2006, E.ON
signed a 15-year contract to buy 4 billion cubic meters a year from 2020 to 2035.
Sources: Nord Stream AG; “Like it or not, many countries are locked in to Gazprom,” New
York Times, January 5, 2006.
7. Organizational Architecture
Chapter 1 discussed the vertical and horizontal boundaries of an organization. An

oil company that produces crude, refines it, and markets gasoline is more vertically
integrated than one that only produces crude. A media organization that publishes
newspapers and provides cable television and broadband service has wider hori-
zontal boundaries than one that specializes in cable television service.
Vertical and horizontal boundaries are just two aspects of the organizational
architecture, which comprises the distribution of ownership, incentive schemes,
and monitoring systems. From the viewpoint of managerial economics, the design
of organizational architecture depends on a balance among four issues:
• holdup,
Organizational
• moral hazard,
architecture depends on
a balance among: • internal market power, and
• economies of scale, scope, and experience,
• holdup;
• moral hazard;
• internal market power; and the mechanisms by which these issues may be resolved.
• economies of scale, We have already discussed specific ways of resolving
scope, and experience. holdup and moral hazard. Here, we consider how ownership
will affect holdup, moral hazard, internal market power, and
economies of scale, scope, and experience.
Holdup
How would changing the ownership of the relevant assets resolve holdup?
Recall that when Luna requested an additional delivery at short notice, Mercury
extracted a higher fee. What if Luna had an in-house delivery service? To make
the additional delivery, it might have to order a driver to work overtime. The
driver could demand a special overtime payment. By doing so, however, the driver
runs the risk that Luna may replace her with another worker.
If the driver strikes, the cost imposed on Luna is the cost of replacing the driver
at short notice. This is less costly than replacing the truck and driver which would
be necessary if an outsourced delivery service should withhold its services.
As this example suggests, even an employee can engage in holdup. However, an
employee is less likely than an external contractor to engage in holdup and would
impose lower costs if she did so. The reason is that the external contractor owns
the assets necessary to provide the service. Residual control of an asset includes
the right to withhold its services. Hence, an external contractor has the power to
withhold the services of its assets.
By contrast, an employee has no such power since the assets on which she works
belong to the employer. With less power, the employee is less likely to behave
opportunistically and engage in holdup and would impose lower cost if she did so.
Accordingly, vertical integration can mitigate the potential for holdup.
Moral Hazard
Changes in ownership also affect the degree of moral hazard. If Luna Supermar-
ket vertically integrates into the delivery business, it must engage drivers. Then it
would change from dealing with an owner who supplies a service to dealing with
an employee. As we showed above, employees are subject to moral hazard.
Generally, the employer’s marginal profit contribution exceeds the worker’s
marginal compensation. Since the worker chooses effort according to her mar-
ginal compensation, she chooses less than the economically efficient level of effort.
By contrast, suppose that the employee owns the business. Then the worker receives
the residual income. Hence, if the worker exerts an additional unit of effort, she will
receive the entire marginal profit contribution. If she reduces effort by one unit, she
will suffer the entire reduction in the profit contribution. Thus, when balancing her
marginal benefit with the marginal cost, the worker will choose the economically
efficient level of effort. Giving ownership to the worker will resolve the moral hazard.
This explains why many businesses pay senior managers through shares and
stock options. To the extent that these managers have a share of ownership, their
interests will be more closely aligned with those of the business, which reduces the
degree of the moral hazard.
Vertical integration changes ownership. Since an employee is subject to rel-
atively greater moral hazard than an owner, vertical integration increases the
degree of moral hazard.
Internal Market Power

Changes in ownership will affect the monopoly power of internal sources of
inputs and monopsony power of internal users of outputs. If Luna establishes
its own delivery service, it will prefer using its own delivery service to engaging an
external contractor. This is a reasonable policy to the extent that some of the costs
of the in-house delivery service are sunk.
The preference in favor of an internal provider, however, means creating an
internal monopoly. Chapter 8 showed that a seller with market power would
restrict production and raise price. An internal provider may also use its market
power to raise its price. Then the organization as a whole will find that the cost of
internal provision may rise above the price charged by external contractors. This
higher cost must be borne by the organization.
One way of resolving an internal monopoly is a policy to outsource whenever
the internal provider’s cost exceeds that of external sources. Outsourcing is the
purchase of services or supplies from external sources. It subjects internal provid-
ers to the discipline of market competition, and so limits the extent to which the
cost of internal provision diverges from the competitive level.
Similarly, if a business produces an input that it uses internally for downstream
production, the downstream user may acquire monopsony power. Chapter 8
showed that a buyer with market power would restrict purchases to depress the
price. A policy to sell externally whenever the external price is higher than the
internal transfer price can resolve the internal monopsony power.
Economies of Scale, Scope, and Experience

Finally, changes in ownership affect the extent of economies of scale, scope, and
experience. Recall from Chapter 7 that, if there are economies of scale, then the
average cost of provision will be lower with a larger scale. Typically, an internal
provider will operate at a smaller scale than an external contractor. It then is
necessary to consider how the average cost varies with the scale of production.
For instance, Luna Supermarket’s deliveries may occupy a truck and driver for
only four hours a day. If Luna were to set up its own delivery service, its utilization
of the assets and human resources would be relatively low. By contrast, an exter-
nal delivery contractor may get 10 or 12 hours of work per day from its equipment
and personnel. The external contractor would have better capacity utilization and
hence a lower average cost. To this extent, it would be less costly to purchase the
service from the external contractor.
We can also illustrate the significance of scale economies by comparing a super-
market’s need for delivery service with that for armored truck service to convey
cash and checks. While some supermarkets own a delivery service, they are not
likely to own an armored truck service. It would hardly be efficient for a super-
market to buy an armored truck that makes one daily trip to the bank.
The analysis of ownership and the experience curve are similar to those for
economies of scale. If the business expects a relatively low cumulative volume,
it would be high up on the experience curve. By contrast, a specialized provider
might consolidate sufficient orders from multiple users that it can push further
down the experience curve and realize lower average cost.
From Chapter 7, if there are economies of scope across two products, then
the total cost of production will be lower when the two products are produced
together than when they are produced separately. Economies of scope are a
key reason for the growth of supermarkets at the expense of traditional stores.
A supermarket can supply items at a relatively lower cost than more specialized
stores such as bakeries, groceries, and newsstands.
Economies of scope are the major factor in favor of wide horizontal orga-
nizational boundaries. However, for an individual organization, economies of
scope have countervailing effects. If it already produces one item, then it can
reduce total cost by producing the other one as well. However, if it does not
already produce either item, then economies of scope imply that it should out-
source both.
Balance
The appropriate organizational architecture depends on a balance among the four
issues – the scope for holdup, the degree of moral hazard, internal market power,
and the extent of economies of scale, scope, and experience – and on other ways
to resolve these issues.
Specifically, holdup can be resolved through more detailed contracts, moral
hazard can be resolved through incentives and monitoring, and internal market
power can be resolved through outsourcing and external sales. Generally, the eco-
nomically efficient solution will involve a mix of all policies.
Let us illustrate the application of this framework with two examples. One
concerns the vertical boundaries of the organization, while the other concerns
its horizontal boundaries. The two examples show that the same framework can
be applied to address both vertical and horizontal dimensions of organizational
architecture.
A function that many organizations consider whether to “make or buy” is
information technology services. As Figure 14.4 shows, this decision resolves into
a balance among four issues. Two factors favor the decision to “make.” One is
Vertically integrate
• Holdup
• Economies of scope Outsource
• Moral hazard
• Internal market power
• Economies of scale,
scope, and experience
FIGURE 14.4 Vertical integration.

Notes: In favor of vertical integration: holdup and economies of scope. In favor of outsourcing: moral
hazard, internal market power, and economies of scale, scope, and experience.
the extent to which the potential for holdup cannot be cost-effectively reduced
through more detailed contracts. The other is the extent to which the internal
provider enjoys economies of scope in information technology services.
There are three factors in favor of deciding to “buy.” One is the extent to
which the moral hazard of the internal information technology group cannot
be cost-effectively resolved through incentive schemes and monitoring systems.
Another is the extent to which internal monopoly power cannot be cost-effectively
redressed through outsourcing. A third is the extent to which the internal unit will
lack economies of scale, scope, or experience in providing information technology
services.
Decisions whether to “make or buy” concern the vertical boundaries of the
organization. An organization must also consider its horizontal boundaries. For
instance, should a bus operator also be in the trucking business? As the busi-
nesses are not vertically related, holdup and internal market power are not major
issues. The key factor in favor of owning both businesses is economies of scope,
while the factor against is the extent to which moral hazard in the trucking unit
cannot be cost-effectively resolved through incentive schemes and monitoring
systems.
PROGRESS CHECK 14G

What are the four issues to consider in the decision whether to “make or buy”?
NISSAN: SECOND-SOURCING BATTERIES
Projecting huge growth in the demand for electric cars, Nissan set up plants
at Zama (Japan), Sunderland (Britain), and Smyrna (Tennessee) with capacity
to manufacture up to 500,000 batteries a year.
In late 2014, Nissan reportedly reviewed its strategy for batteries. The
demand for electric cars had grown much less than projected. The previous
year, Nissan and strategic partner, Renault, sold only 67,000 electric cars.
Another problem was Nissan’s cost of producing batteries. A company
executive remarked: “We set out to be a leader in battery manufacturing but
it turned out to be less competitive than we’d wanted . . . . We’re still between
six months and a year behind LG in price-performance terms.”
In the strategic review, one option was for Nissan to concentrate production
in Japan and sell the British and American facilities to LG Chem (South Korea).
Renault pressed Nissan to adopt dual sourcing and buy batteries from LG
Chem.
Source: “Nissan faces battery plant cuts as electric car hopes fade,” Reuters, September
15, 2014.
BOEING’S CHARLESTON PURCHASE: NOT ACCORDING TO PLAN
In a major departure from previous strategy, the then Chairman and CEO,
Harry Stonecipher, and the then head of Commercial Airplanes, Alan Mulally,
decided to radically outsource design and construction of the 787. Boeing
would focus on final assembly. The new strategy would reduce the assets
required to support production, and so boost Boeing’s return on net assets.
Boeing contracted with Vought Aircraft Industries to design and build two
sections of the 787’s rear fuselage at Vought’s factory in Charleston, South
Carolina. It also contracted with an adjacent joint venture between Vought
and Alenia of Italy to join the various fuselage sections – the two rear sections
built by Vought, two mid-sections built in Italy by Alenia, and a section built by
Kawasaki in Japan. The joint venture would then send the assembled fuselage
to Everett for final assembly.
However, Vought faced technical and financial difficulties. Vought’s owner,
the private-equity investor Carlyle Group, refused to increase its investment
and instead sought to sell the company. Vought’s CEO, Elmer Doty, explained
that “the financial demands of this program are clearly growing beyond what
a company our size can support.”
In 2008, Boeing bought Vought’s share of the joint venture with Alenia. Finally,
in July 2009, to resolve the holdup, Boeing again resorted to vertical integration
and bought the Charleston factory from Vought at a cost of $1 billion.
Source: “Boeing’s buy of 787 plant will cost $1B,” Seattle Times, July 7, 2009.
KEY TAKEAWAYS
• Moral hazard arises if one party’s actions affect but are not observed by
another party.
• Resolve moral hazard through monitoring and incentives.
• Incentives create risk, the cost of which increases with the impact of the
extraneous factors, risk aversion, and strength of the incentives.
• For better performance on multiple responsibilities which vary in their ease of
measurement, use weaker incentives.
• Holdup is action to exploit another party’s dependence.
• Resolve holdup through detailed contracts, with more detail if potential benefits
and costs are larger, and if contingencies have larger impact.
• Ownership is the rights to residual control, which are those rights that have not
been contracted away.
• Vertical integration is the combination of the assets for two successive stages
of production under a common ownership.
• Apply organizational architecture to balance holdup, moral hazard, internal
market power, and economies of scale, scope, and experience.
REVIEW QUESTIONS
1. In the context of your business or organization, explain the meaning of

organizational architecture.
2. Explain the moral hazard in the following situation. Leah has just bought
comprehensive insurance on her car that covers loss and damage for any
reason, including theft. Her insurer is concerned that she may take fewer
precautions against theft.
3. By considering benefits and costs to the various parties, explain why resolving
moral hazard can be profitable.
4. Why is it better to pay a real estate broker by commission, whereby she receives
a percentage of the selling price of the property, rather than an hourly rate?
5. Your compensation package includes shares of the company, which vest after
three years. Explain the risk that you bear.
6. Explain how a taxi company can structure incentives based on relative
performance to motivate its drivers to maintain their taxis carefully and avoid
breakdowns.
7. A secretary’s job includes typing letters and other responsibilities. Comment
on a proposal to pay a secretary according to the number of letters that he
types.
8. Maria is a pilot. Which of the following investments by her is relatively more
specific to the airline that she works for? (a) An executive MBA program. (b)
Training on the airline’s flight management system.
9. Why do businesses enter into contracts that are deliberately incomplete?
10. In the context of an incorporated business, explain the meaning of: (a) residual
control; (b) residual income.
11. Give an example of a holdup. Explain how this will induce the affected parties
to avoid specific investments.
12. How does the potential for holdup reduce the value of transactions and
relationships?
13. A car manufacturer has experienced holdup by its supplier of electronics.
Should the car manufacturer produce the electronics internally?
14. How can outsourcing resolve the monopoly power of an internal supplier?
15. Explain the roles of economies of scope in decisions about horizontal
integration.
1. The National University of Singapore provides outpatient medical insurance to

faculty and staff. The plan covers the entire bill for treatment and medicine at
an approved general practitioner subject to a copayment of S$5.
(a) Construct a diagram with quantity of healthcare (including treatment
and medicines) on the horizontal axis and marginal benefit and marginal
cost of healthcare on the vertical axis. Draw the patient’s marginal
benefit and marginal cost, and the insurer’s marginal cost.
(b) Compare the level of healthcare that the patient would choose with the
level that maximizes the patient’s benefit less the insurer’s cost.
(c) How does the S$5 copayment affect the patient’s choice between (i) a
generic drug that costs $5 and (ii) a branded drug that costs $50?
(d) Suppose that the University replaces the S$5 copayment with a 10%
copayment. Use your diagram to illustrate how that would affect the
patient’s choice of healthcare.
2. Mapletree Commercial Trust owns Singapore’s largest mall, VivoCity. The Trust
has contracted with Mapletree Investments to manage the mall. The manager
is responsible for the carpark, as well as cleaning, security, and other common
services. The mall’s retail tenants pay a three-part rental: in 2010, 86% was
fixed, and 13% was based on the retailer’s sales revenue, while the remainder
comprised service charges and contributions to advertising and promotion.
(Source: Mapletree Commercial Trust, Prospectus, April 18, 2011.)
(a) How is the manager of VivoCity subject to moral hazard relative to its
tenants?
(b) How does the variable payment align the interests of landlord and tenant?
(c) Ideally, should the variable payment be based on the tenant’s gross
revenue or net revenue (gross revenue less cost of goods sold)?
(d) Mapletree Investments is a developer of commercial and industrial real
estate. It sells completed projects to Mapletree Commercial Trust.
Discuss the asymmetry of information between the two entities.
3. Pension funds and trusts may be limited to investing in securities of particular
credit rating, e.g., AAA. To appeal to such investors, investment banks that
issue structured products need to meet the required credit rating. The credit
rating agencies, Fitch Ratings, Moody’s Investor Services, and Standard &
Poor’s Ratings Services, will rate a security and, after issuance, continue to
update the rating over the life of the security. The agency charges the issuer a
fee for rating as well as a fee for updating.
(a) Consider the managers of an investment bank which plans to issue a
structured product. Would they be more concerned about the initial
rating or updating?
(b) How does your answer to (a) depend on the turnover of investment
bank managers?
(c) What is the incentive for the rating agency to rate a structured product
as AAA?
(d) Who would be more affected by inaccuracy in updating: the issuer or
the rating agency?
4. Mayo Clinic, a non-profit provider of healthcare, whose motto is “the needs of
the patient come first,” pays doctors a salary rather than a fee for service. Dr
John C. Lewin, chief executive of the American College of Cardiology, argues
that salaried doctors can provide better healthcare and lamented that most US
healthcare is “divided between small practices and community hospitals that
aren’t linked together with incentives to coordinate care.” (Source: “A new way
to pay physicians,” NYTimes.com, September 23, 2009.)
(a) Consider the conventional model of healthcare which pays doctors

according to the amount of treatment that they provide. Using a relevant
figure (with hypothetical demand and marginal cost), explain how the
doctor will over-treat the patient.
(b) If doctors are paid a fixed salary, how would that affect their incentive
to over-treat?
(c) Dr Lewin believes that a vertically integrated system provides better
healthcare. Suppose that a hospital integrates with a group of heart
specialists. Explain the moral hazard of the doctors relative to the hospital.
(d) Following the integration in (c), how should the hospital motivate the
doctors?
5. Alibaba, which operates Taobao and Tmall, is the world’s largest online
marketplace. In the 15 months before its initial public offering, it awarded
employees 68.5 million units of restricted shares and options for 21.8 million
shares, with a total value exceeding $5 billion. (Source: “Alibaba and the
kingdom of sumptuous stock grants,” Fortune, September 12, 2014.)
(a) Using a relevant diagram, illustrate the company’s marginal profit
contribution, and a senior executive’s marginal compensation and
marginal cost as a function of the senior executive’s effort.
(b) Use the diagram to explain how the stock options are intended to
motivate senior executives of Alibaba.
(c) The Alibaba executives also receive a salary from the company. How
do the stock options affect the risk that they bear?
(d) Explain why Alibaba should condition incentive compensation on
Alibaba’s performance relative to competitors.
6. Boeing contracted with Vought Aircraft Industries to design and build two
sections of the 787’s rear fuselage at Charleston, South Carolina. Boeing also
contracted with an adjacent joint venture between Vought and Alenia of Italy to
assemble fuselage sections. Facing technical and financial difficulties, Vought’s
owner, the Carlyle Group, refused to increase its investment and instead sought
to sell the company. In 2008, Boeing bought Vought’s share of the joint venture
with Alenia, and then, in July 2009, Boeing bought the Charleston factory from
Vought at a cost of $1 billion. (Source: “Boeing’s buy of 787 plant will cost $1B,”
Seattle Times, July 7, 2009.)
(a) Explain the mutual dependence between Boeing and Vought.
(b) Explain Carlyle Group’s refusal to increase investment in terms of holdup.
(c) The potential for holdup by a subcontractor may lead a manufacturer to
contract with two subcontractors, a practice called “second-sourcing.”
Explain why second-sourcing may be inefficient in terms of the experience
curve.
(d) Suppose that the Carlyle Group had a higher discount rate (shorter
time horizon) than Alenia. Would that explain why the Carlyle Group
but not Alenia tried to exit their Boeing contracts?
(e) The Carlyle Group approached various potential buyers. Why would
Boeing be willing to pay more than other potential buyers?
7. The €4.6 billion North European Gas Pipeline transports natural gas from
central Russia to Germany. Gazprom first conceived of the project in 1997. In
January 2006, Gazprom signed a 20-year contract with Wintershall, a unit of
BASF, to sell 8 billion cubic meters of gas a year from 2010 to 2030. Then, in
August 2006, Gazprom signed a 15-year contract with E.ON to sell 4 billion
cubic meters a year from 2020 to 2035.
(a) Why was it important for Gazprom to sign these contracts before
commencing construction of the pipeline?
(b) Would it be just as good for Gazprom to sign an initial five-year contract
with the German customers and then renew the contract later on?
(c) Should the contracts specify the price of the gas or leave it for later
negotiation?
(d) The German companies also bought shares in the pipeline. How would
this help to resolve the potential for holdup?
8. Wärtsilä, with headquarters in Helsinki, Finland, advertises the operating flexibility
of its electric power generating plants: “If you need to relocate, modular
construction will make the move easier. Or you can choose a Wärtsilä floating
power plant or a Wärtsilä Power Module and become truly mobile.”
(a) How is the specificity of an investment related to sunk costs?
(b) Explain the concept of specific investments in relation to building an
electric power generation facility.
(c) Is the problem of holdup more or less serious for investments with a
longer payback period?
(d) Do you expect the demand among investors for floating power
plants to be greater or less in countries with high political risk?
Explain why.
9. Nissan and NEC jointly established Automotive Energy Supply Corporation
(AESC) to manufacture high-performance lithium-ion batteries. AESC invested
over $1 billion in new plants at Zama (Japan), Sunderland (Britain), and Smyrna
(Tennessee) with capacity to manufacture up to 500,000 batteries a year. The
Sunderland and Smyrna plants are co-located with Nissan car manufacturing.
Nissan contracted with NEC to buy electrodes for 220,000 units of 24 kWh
batteries a year. However, the demand for electric cars grew less than projected.
In 2013, Nissan and strategic partner, Renault, sold just 67,000 electric cars.
Further, AESC’s cost of production was higher than that of external competitor
LG Chem of South Korea. (Source: “Nissan faces battery plant cuts as electric
car hopes fade,” Reuters, September 15, 2014.)
(a) Evaluate the specificity of AESC’s investment in the Sunderland and
Smyrna plants.
(b) What holdup problem does NEC face?
(c) Why would Nissan contract to buy a minimum number of electrodes
each year from NEC?
(d) How do (i) economies of scale; and (ii) the experience curve explain
why AESC cannot produce as cheaply as LG Chem?
(e) In light of (d), explain the costs and benefits of vertical integration.

Consider the architecture of your organization. How can your organization
add value by:
(a) outsourcing/insourcing,
(b) revising incentive schemes,
(c) adjusting ownership?
Notes
1 The following discussion is based, in part, on “The affair that grounded Stonecipher,” Business
Week, March 8, 2005; “787 delay could wind up costing Boeing $1 billion,” Seattle Times, Octo-
ber 25, 2007; “Boeing’s buy of 787 plant will cost $1B,” Seattle Times, July 7, 2009; “A ‘prescient’
warning to Boeing on 787 trouble,” Seattle Times, February 5, 2011; “Boeing 787 Dreamliner,”
Wikipedia (accessed August 21, 2011).
C H A P T E R
15
Regulation
LEARNING OBJECTIVES
• Appreciate the conditions for a natural monopoly and how govern-
ments should regulate a monopoly.
• Appreciate the conditions for a potentially competitive market and
how governments should foster competition.
• Appreciate how governments should regulate markets with asym-
metric information.
• Appreciate how governments should regulate externalities.
1. Introduction
North Delhi Power Limited (NDPL), a joint venture of the Tata Power Company
(51%) and Government of Delhi (49%), has the monopoly franchise to distribute
electricity in the north and northwest areas of India’s capital. It serves 1.2 million
customers in a population of 5 million. In 2010, NDPL earned net profit after tax
of 3.5 billion Indian rupees on revenue of 34.0 billion rupees.1
Between 2002 and 2011, NDPL reduced losses of electricity due to theft and
technical reasons from 74% to just 14%. Under the terms of its franchise, NDPL
splits the additional revenue from reduction in losses between its shareholders and
customers.
Historically, NDPL specialized in distribution of electricity. It buys over 1,100
megawatts (MW) of electricity on long-term contracts from Pragati Power Cor-
poration and other generators of electricity. In a new initiative toward vertical inte-
gration, NDPL is building a 108 MW gas-based combined cycle power generation
358 15 Regulation
plant at Rithala in North Delhi. NDPL justified the new plant as “obviat[ing] the
need for an equivalent amount of expensive bilateral power.”
The Delhi Electricity Regulatory Commission regulates the electricity industry
and limits profits to a maximum return on equity of 14% or 16%. The maximum
allowed return differs by the provider’s function (generation, distribution, or
transmission) and has varied over time. In 2007, Pragati Power petitioned the
Commission to increase its approved equity. However, NDPL challenged the
proposed increase and the Commission rejected the petition.
Why did the Delhi government award NDPL a monopoly franchise to distribute
electricity? Should the government allow NDPL’s move toward vertical integra-
tion? Why does the Delhi Electricity Regulatory Commission limit the return on
equity of electricity providers? Why did NDPL challenge Pragati Power’s petition
to increase its approved equity?
This chapter addresses the role of the government in situations where buyers
and sellers, acting independently and selfishly, do not equalize marginal benefit
and marginal cost. So, the invisible hand fails – the allocation of resources is not
economically efficient.
If private action fails to resolve the economic inefficiency, then there may be
a role for government regulation. If the government can resolve the divergence
between marginal benefit and marginal cost, then it will increase net benefit for
society.
To understand the role of government regulation, we consider the various pos-
sible sources of economic inefficiency: market power, externalities, and asym-
metric information. For each source of economic inefficiency, we analyze the
conditions under which the government should intervene and the appropriate
form of regulation.
In businesses with significant economies of scale or scope relative to demand,
it may be economically efficient to award a monopoly franchise. Then the
government must regulate the monopoly’s exercise of market power. The Delhi
government chose to regulate NDPL’s profits according to a maximum return
on equity.
Under regulation of return on equity, the regulated business would seek to increase
its approved equity. The larger its approved equity, then the higher will be its allowed
profit. This explains why Pragati Power petitioned to increase its approved equity,
and why NDPL, which buys electricity from Pragati, opposed the petition.
Besides market power, the invisible hand may also fail because of informa-
tion asymmetries and externalities. This chapter also discusses how the gov-
ernment can resolve asymmetric information through regulation of disclosure,
conduct, and business structure, and resolve externalities through standards
and user fees.
Regulated industries serve both businesses and consumers. Hence, as NDPL’s
opposition to Pragati’s petition illustrates, it is important for businesses as well as
consumers to understand the basis and methods of government regulation.
Regulation 359
2. Natural Monopoly
A market is a natural monopoly if the average cost of production is minimized

with a single supplier. Essentially, economies of scale or scope
are large relative to the market demand. For instance, electric- Natural monopoly:
A market where
ity is distributed through a network of cables. Consider a the average cost of
town with two competing electricity distributors with separate production is minimized
infrastructure. Then two sets of cables would run into every with a single supplier.
home, office, and factory. In the electricity distribution indus-
try, allowing competition would result in wasteful duplication.
Other examples of natural monopolies include broadband service, distribution
of gas and water, and sewage collection. In all of these markets, the economies of
scale may be large relative to the market demand. So, the average cost of produc-
tion is lowest when there is only one supplier.
If a market is a natural monopoly, the government should prohibit competition
and allow only a single supplier. This would establish the conditions for produc-
tion at the lowest average cost.
The monopoly, however, might exploit its exclusive right to raise its price at
the expense of its customers. The increase in the price would force the marginal
benefit above the marginal cost. Accordingly, to ensure economic efficiency, the
government must control the monopoly. The government can do this in two ways:
• The government itself can own the business, and operate at the economically
efficient level.
• Award a monopoly franchise to a commercial enterprise and subject the
monopoly to regulation.
Government Ownership
In principle, government ownership and operation is the
simplest and most direct way to ensure economic efficiency. Controlling monopoly:
However, in practice, government-owned enterprises tend to • government ownership
be relatively inefficient, and so government ownership and and operation; or
operation fails to achieve economic efficiency. • regulate commercial
enterprise.
One source of inefficiency is that government-owned enter-
prises are prone to be coopted by employees, so that the enter-
prise serves its employees rather than its customers. Some symptoms of employee
control are high wages and overstaffing, both of which inflate the cost of production.
Another source of inefficiency is that government-owned enterprises depend on
the government for investment funds. The government budget must finance every-
thing from social welfare to military equipment. A government-owned enterprise
must compete with other priorities for an allocation from the budget and may not
be able to secure the economically efficient level of investment.
360 15 Regulation
Owing to the limitations of government ownership and operation, a worldwide

trend has been to privatize government-owned enterprises.
Privatization: Transfer Privatization means transferring ownership from the gov-
of ownership from
government to private
ernment to the private sector. It does not necessarily mean
sector. allowing competition. Indeed, many privatized enterprises
are monopolies in their markets.
Price Regulation
Recall from Chapter 6 that the provision of a good or service will be economically
efficient at the level where marginal benefit equals marginal cost. Suppose that the
government awards an exclusive franchise for electricity distribution to Jupiter
Power. Jupiter’s costs include a fixed cost and a constant marginal cost of $40 per
megawatt-hour (MWh). Figure 15.1 shows the cost of distribution and the demand
for electricity.
Suppose that the government requires the provider to set its price equal to its
marginal cost and to meet the quantity demanded. Then, referring to Figure 15.1,
at every possible quantity, the price will be the marginal cost.
Marginal cost pricing: In effect, the government’s policy forces the provider to
The provider must set
the price equal to the
behave like a perfectly competitive supplier. This policy is
marginal cost and supply called marginal cost pricing.
the quantity demanded. The demand curve crosses the marginal cost curve at point
a. If Jupiter sets a price of $40 per megawatt-hour, the market
would demand a quantity of 10,000 MWh. Under marginal cost pricing regulation,
Jupiter must supply the quantity demanded; hence, it must produce 10,000 MWh.
Now, each customer buys the quantity that balances its marginal benefit with the
Marginal benefit/cost ($ per MWh)
d
55
50 Average cost
40 a Marginal cost
c
Demand
b
0 7 10
Quantity (thousand MWh a month)
FIGURE 15.1 Price regulation.

Notes: The demand curve crosses the marginal cost curve at a. Under marginal cost pricing, the
provider sets a price of $40 per MWh and the market quantity demanded is 10,000 MWh a month.
The demand curve crosses the average cost curve at d. Under average cost pricing, the provider sets
a price of per $55 MWh and the market quantity demanded is 7,000 MWh.
Regulation 361
price. Hence, marginal benefit equals marginal cost, which is the condition for
economic efficiency.
Price regulation presents two challenges. One is that, under marginal cost pric-
ing, the provider may incur a loss. Jupiter’s revenue is represented by the area
0bac, which is $40 × 10,000 = $400,000 a month. The average cost at 10,000 MWh
is $50 per MWh, which means that the total cost is $50 × 10,000 = $500,000. With
marginal cost pricing, the provider would incur a loss of $500,000 − $400,000 =
$100,000 a month.
Accordingly, the government must provide a subsidy of $100,000 a month to
ensure that the provider is financially viable. The subsidy is necessary to achieve
economic efficiency. However, the regulator must then raise funds to provide the
subsidy.
How should the government regulate Jupiter if it does not wish to provide a
subsidy? The policy that most closely approaches economic efficiency while allow-
ing the provider to break even is average cost pricing. Under
average cost pricing, the provider must set the price equal Average cost pricing:
The provider must set the
to average cost and supply the quantity demanded. With price equal to the average
average cost pricing, the provider would exactly cover its cost and supply the
costs. Applying average cost pricing to Jupiter, the price quantity demanded.
would be $55 per megawatt-hour and the quantity demanded
would be 7,000 MWh a month.
The other challenge in price regulation is to acquire information about the pro-
vider’s costs. The franchise holder has a strong incentive to overreport its costs.
Then the regulator will allow a higher price, which would enable the provider to
increase its profit. Indeed, the situation of the regulator and franchise holder is
one of asymmetric information. As Chapter 13 shows, the information asymme-
try, if not resolved, would result in economic inefficiency.
PROGRESS CHECK 15A

Referring to Figure 15.1, suppose that the demand is higher. How would that
affect the optimal price?
Rate-of-Return Regulation
Under price regulation, the franchise holder may exaggerate its reported costs and
so inflate the regulated price. The alternative to price regulation is rate-of-return
regulation. This avoids the issue of costs by focusing on the franchise holder’s
profit. The regulator allows the franchise holder to set prices freely, provided that
it does not exceed the maximum allowed profit.
Under rate-of-return regulation, the regulator stipulates Rate base: Assets or
equity on which the
the franchise holder’s maximum allowed profit in terms of
franchise holder may earn
a maximum rate of return on the value of the rate base. the allowed rate of return.
The rate base may be specified by assets or equity. Whenever
362 15 Regulation
the franchise holder’s rate of return exceeds the specified maximum, it will be
required to reduce its prices.
Suppose, for instance, that Jupiter Power is regulated to a maximum 12% rate of
return on allowed assets of $5 million. Then the maximum allowed profit would
be 0.12 × $5 million = $600,000 a year.
Rate-of-return regulation presents three challenges in implementation:
• Rate of return. One challenge is to set the allowed rate of return. Typically,
the rate base is large, so a small difference in the allowed rate of return will
translate into a large sum of money. Since the franchise holder is a monopoly,
there would be few comparable businesses, so it would be difficult to determine
the appropriate rate of return.
• Rate base. Another challenge is to determine what assets are needed to
provide the regulated service and, hence, should be counted in the rate base.
The franchise holder will seek the widest possible definition to increase its
profit.
• Overinvestment. Finally, the franchise holder has an incentive to invest beyond
the economically efficient level. By enlarging the rate base, the allowed rate
of return will be applied to a larger base, and so the franchise holder can
increase its profit.
AUSTRALIA: PRICE REGULATION OF TRUNK

TELECOMMUNICATIONS
The Australian Competition and Consumer Commission (ACCC) subjects

telecommunications with bandwidth of 2 megabits per second or greater
to regulation. The ACCC regulates price according to total service long-run
incremental cost (TSLRIC), which is essentially the marginal cost of service
based on commercially available efficient technology, and taking account of
economies of scale and scope.
The ACCC has justified regulating by TSLRIC for four reasons:
• If the service provider faced effective competition, its prices would

approximate TSLRIC.
• TSLRIC provides a risk-adjusted return on investment, and so encourages
efficient investment in infrastructure.
• TSLRIC encourages the efficient use of existing infrastructure.
• TSLRIC allows the service provider to fully recover its efficient costs of
providing service.
Source: Australian Competition and Consumer Commission, Domestic Transmission

Capacity Service: Final Report, March 2009, pp. 36–37.
Regulation 363
3. Potentially Competitive Market
By contrast with a natural monopoly, a potentially com-

Potentially competitive
petitive market is one where economies of scale and scope market: Economies
are small relative to market demand. So having two or more of scale and scope are
competing suppliers would not raise average costs. In a small relative to market
potentially competitive market, with perfect competition, the demand.
invisible hand will ensure economic efficiency. Hence, in any
potentially competitive market, the government should promote competition.
Competition Law
The basic way in which governments promote competition is through competition
law (also called “antimonopoly” or “antitrust” law). However, industries that are
subject to specific regulation may also be required to comply with competition law
specific to the industry.
Table 15.1 lists the key laws regarding competition in various jurisdictions.
In addition to these laws, individual countries within the European Union and
Table 15.1 Competition laws
Jurisdiction Law Enforcement agency
Australia Competition and Consumer Australian Competition and

Act 2010 Consumer Commission
Canada Competition Act 1985 Competition Bureau
China Anti-Monopoly Law 2007 National Development and Reform
Commission; State Administration
for Industry and Commerce;
Ministry of Commerce
European Union Treaty on the Functioning of Directorate General for Competition,
the European Union, 2009: European Commission
Articles 101–106
Japan Anti-Monopoly Act 1947 Fair Trade Commission
Korea Monopoly Regulation and Fair Trade Commission
Fair Trade Act 1980
New Zealand Commerce Act 1986 Commerce Commission
Taiwan Fair Trade Act 1991 Fair Trade Commission
United Kingdom Competition Act 1998; Competition and Markets Authority
Enterprise Act 2002
United States of Sherman Act 1890; Clayton Department of Justice; Federal
America Act 1914; Federal Trade Trade Commission
Commission Act 1914;
Robinson–Patman Act 1938;
Hart–Scott Rodino Antitrust
Improvement Act 1976
364 15 Regulation
individual US states may have their own competition laws. Table 15.1 also lists the
agencies responsible for enforcing competition laws in the various jurisdictions.
Generally, competition laws prohibit the following:
• competitors from colluding on price or other aspects of purchases or sales;

• businesses with market power from abusing their market power;
• mergers or acquisitions that would create substantial market power.
Competition laws In addition, competition laws may prohibit or restrict spe-

prohibit: cific business practices such as control over resale prices and
• collusion;
exclusive agreements. What exactly is prohibited varies from
• abuse of market power; one jurisdiction to another.
• harmful mergers. The mission of the competition agency is to enforce com-
petition laws. One role is to prosecute competitors that
engage in collusion and businesses that abuse market power. The other role is
to review proposals for mergers and acquisitions. The competition agency must
ensure that any proposed merger or acquisition complies with the law. The agency
may approve the proposed combination subject to specific conditions such as the
divestment of particular businesses to mitigate the anticompetitive impact of the
combination.
Besides government enforcement, the competition laws may also provide for
persons harmed by anticompetitive behavior to take legal action in civil law.
Under US federal antitrust laws, for instance, plaintiffs in civil actions can recover
damages of three times the harm that they suffered. Further, civil plaintiffs can
petition courts for an injunction against anticompetitive conduct.
Structural Regulation
The fact that one market is a natural monopoly does not necessarily mean that
related upstream or downstream markets are also natural monopolies. In electric-
ity, for instance, production may be potentially competitive even while transmis-
sion and distribution are natural monopolies. Likewise, distribution of water may
be a natural monopoly, while production is potentially competitive.
Under such circumstances, the government must consider how to preserve the
benefits of monopoly in one market while fostering competition in the other.
A special challenge to the regulator arises when the monopoly franchise holder
also participates in the potentially competitive market.
To illustrate the issues, suppose that the government awards a monopoly fran-
chise for distribution of electricity, while allowing competition in generation.
Jupiter Power has the monopoly franchise over distribution. So, it also has a
monopsony over the purchase of electricity from generators. Hence, the govern-
ment must regulate Jupiter’s monopoly over distribution of electricity as well as
its monopsony over purchases of electricity.
Regulation 365
Now suppose that Jupiter has vertically integrated upstream into the generation
of electricity. Then Jupiter may have an incentive to favor its internal genera-
tor of electricity and discriminate against competing generators. The regulator
should intervene to ensure fair competition. However, Jupiter might exploit its
superior information about technical and other issues to confound the regulator.
So competing generators of electricity may be at a disadvantage in supplying the
distribution monopoly.
One solution to this challenge is structural regulation to separate the natural
monopoly from the potentially competitive market. Under
structural regulation, the regulator stipulates the condi- Structural regulation:
Stipulating conditions for
tions under which a business may produce vertically related a business to produce
goods and services. vertically related goods
In the electricity case, the regulator could require Jupiter and services.
to separate its electricity distribution and generation busi-
nesses. In the extreme, it could simply disallow Jupiter from generating electricity.
With Jupiter purely distributing electricity, it would have no incentive to discrim-
inate among generators of electricity.
PROGRESS CHECK 15B

Explain the meaning of structural regulation to promote competition.
EXECUTING A GLOBAL MERGER
In May 2011, the German automobile group, Volkswagen, owned 72% of

Swedish truck maker, Scania, and 30% of German truck maker, MAN.
Ferdinand Piech, the Chairman of Volkswagen’s Supervisory Board and also
Chairman of MAN, aimed to strengthen the collaboration between Scania
and MAN.
Volkswagen commenced a tender offer for MAN at €95 per ordinary
share and €59.90 per preference share. However, Volkswagen anticipated
that “[a]ntitrust restrictions [would] pose high hurdles for a more in-depth
co-operation, as MAN on the one hand and Scania and Volkswagen on the
other hand are regarded as competitors.”
To complete the merger with MAN, Volkswagen had to seek the approval of
the European Commission as well as competition and regulatory authorities
in Australia, Brazil, Canada, China, France, Germany, Italy, Japan, Mexico,
Turkey, and the USA.
Sources: Volkswagen AG, offer document, May 30, 2011; “Volkswagen submits merger
control application to EU Commission,” Press Release 282/2011, August 23, 2011.
366 15 Regulation
ELECTRICITY IN NORTH DELHI
In the electricity industry, transmission and distribution exhibit strong

economies of scale, while generation does not. In 2001, the government of
India’s capital, Delhi, restructured and privatized its electricity agency into
seven companies – two for generation, one for transmission, and four for
distribution.
The Delhi government awarded each of four companies a monopoly franchise
for distribution of electricity in designated areas. NDPL holds the franchise for
the north and northwest areas. It buys over 1,100 MW of electricity on long-
term contracts from generators of electricity.
In a new initiative toward vertical integration, NDPL is building a 108 MW
gas-based combined cycle power generation plant at Rithala in North Delhi.
NDPL justified the new plant as “obviat[ing] the need for an equivalent amount
of expensive bilateral power.” But what does that portend for competition in
the market for generation of electricity?
Source: North Delhi Power Limited, Annual Report, 2009–10.
4. Asymmetric Information
Another situation in which the invisible hand may fail is where information
about some characteristic or future action is asymmetric. Chapters 13 and 14
show that, if the information asymmetry is not resolved, the marginal benefit will
diverge from marginal cost, and the allocation of resources will not be economi-
cally efficient.
Consider, for instance, the market for medical services. Patients rely on sur-
geons for advice as well as treatment. Owing to the asymmetry of information
between surgeon and patient, the surgeons are subject to moral hazard. Surgeons
may overprescribe treatment to increase their incomes.
Figure 15.2 illustrates the market equilibrium. The true demand is the patients’
marginal benefit if they had the same information as their surgeons. The inflated
demand results from asymmetry of information and exceeds the true demand to
the extent that surgeons induce patients to get excessive treatment.
The inflated demand crosses the supply of medical services at point a. In the
market equilibrium, the price is $140 per hour and the quantity of treatment is
210,000 hours a month. At that quantity, the true marginal benefit of medical
services is $50, which is the height of the true demand curve. The marginal cost of
medical services is $140, the height of the supply curve at the equilibrium quan-
tity. In equilibrium, the marginal cost exceeds the true marginal benefit by $90.
This economic inefficiency results from the asymmetry of information between
surgeons and patients.
Regulation 367
Supply
Price ($ per hour)
a
140
b
100 Inflated demand
50 c
True demand
0 200 210
Quantity (thousand hours a month)
FIGURE 15.2 Moral hazard in medical services.

Notes: The inflated demand is higher than the true demand. The market equilibrium lies at a, where
the inflated demand crosses the supply. The price is $140 per hour and the quantity is 210,000 hours
a month. If the moral hazard can be resolved, the inflated demand would shift down to the true
demand and the equilibrium would be at b, where the true demand (marginal benefit) equals the
supply (marginal cost).
In situations of asymmetric information, the regulator might possibly resolve

the asymmetry by regulating the better-informed party’s
Regulate information
• disclosure of information, asymmetry through:
• conduct, and • information disclosure;
• business structure. • conduct;
• business structure.
In medical services, to the extent that the regulation is
effective, the inflated demand in Figure 15.2 would shift down toward the true
demand. Then the equilibrium would be closer to point b, where the true mar-
ginal benefit equals the marginal cost.
Disclosure
The most obvious way to resolve asymmetric information is to require the better-
informed party to disclose its information truthfully. However, disclosure will
resolve the asymmetry only if the information can be objec-
tively verified. In the case of surgery, the patient’s need is a Disclosure resolves
asymmetry if information
matter of professional judgment, hence disclosure may not can be objectively verified.
resolve the information asymmetry.
Consider another situation of asymmetric information:
financial advisors hard-selling risky investments. The regulator can require finan-
cial advisors to disclose the riskiness of investments. This might help to resolve
the information asymmetry if the client understands the disclosure.
368 15 Regulation
Conduct Regulation
Instead of directly resolving an information asymmetry, an alternative is to reg-
ulate the conduct of the better-informed party and so limit the extent to which it
can exploit informational advantage. If parties with better information cannot
exploit their advantage, then the outcome would be closer to the economically
efficient level.
For example, the regulator of financial services can stipulate that financial
advisors must evaluate their clients’ risk profile before marketing any investment
products. The regulator can also stipulate a minimum “cooling-off period” during
which an investor may withdraw from the purchase of any investment without
penalty. Such regulations of conduct restrict the scope for a financial advisor to
pressure investors into buying unsuitable investments.
Structural Regulation
Another way to limit the extent to which a better-informed party can exploit an
informational advantage is to regulate the structure of the industry. By enforcing
separation of different businesses, a regulator may reduce the opportunities for
exploiting superior information.
The market for medical services illustrates structural regulation. In some coun-
tries, doctors are limited to providing advice and treatment, and are prohibited
from selling medicines and medical supplies. This regulation effectively dissuades
doctors from excessive prescription of medicines.
The market for financial services also illustrates structural regulation. Follow-
ing the subprime financial crisis, various regulators have proposed that commer-
cial banks, which take deposits from retail customers, should not be allowed to
engage in trading of derivatives and other securities. This structural regulation
limits the exposure of commercial banks to risk.
PROGRESS CHECK 15C

How can the government regulate asymmetries of information?
STRUCTURAL REGULATION IN REAL ESTATE BROKERAGE
In real estate brokerage, a dual agent represents both the buyer and seller.
A dual agent is inherently conflicted. The agent may tell the seller how much the
buyer is willing to pay, or tell the buyer how much the seller is willing to accept.
Does dual agency favor the buyer or seller? The empirical evidence is
mixed. In a study of over 21,000 residential transactions in central Virginia
between 2000 and 2009, dual agency was associated with 1.7% lower prices.
Regulation 369
By contrast, among over 18,000 residential transactions in Johnson County,

Indiana, between 2000 and 2010, dual agency was associated with prices that
were 4.8% higher for property owned by the agent. However, there was no
significant difference in prices for properties owned by unrelated individuals.
The government of Singapore prohibits dual agency and requires that the
buyer and seller be represented by different agents.
Sources: Raymond T. Brastow and Bennie D. Waller, “Dual agency representation: incentive
conflicts or efficiencies,” Journal of Real Estate Research, Vol. 35, No. 2, 2013; Ken Johnson,
Zhenguo Lin, and Jia Xie, “Dual agent distortions in real estate transactions,” Real Estate
Economics, 29 March 2015, DOI: 10.1111/1540-6229.12073.
5. Externalities
The invisible hand may fail in situations of externalities, that is, when some ben-
efit or cost passes directly from source to recipient and not through a market. In
the absence of a market, the invisible hand cannot work. There are no markets
for factory emissions or discharge of contaminated water. Furthermore, pollution
affects so many entities that private action would probably not resolve the exter-
nality. Government regulation may be the only solution.
Chapter 12 showed that, for economic efficiency, the level of an externality
should be such that marginal benefit equals marginal cost. The benefit of emis-
sions is in allowing the sources to avoid the cost of clean disposal. The cost of
emissions is the harm to the health of the victims.
As depicted in Figure 15.3, the economically efficient rate of emissions is 800,000
tons a year, where the marginal benefit equals the marginal
cost to society. How can this be achieved? Generally, there are Regulation of
two ways of regulating externalities: externalities:
• user fees or taxes;
• user fees or taxes; • standards or quotas.
• standards or quotas.
User Fee/Tax
One method of regulation aims to mimic Adam Smith’s invisible hand: allow all
sources to emit as much as they like provided that they pay the appropriate user
fee or tax. Referring to Figure 15.3, at the economically efficient rate of emissions,
the marginal cost to society is $35 per ton. Suppose that the regulator sets a user
fee for emissions of $35 per ton and allows every source (user) to buy as large an
amount of emissions as it would like at that price.
Consider an oil refinery that emits pollutants. To maximize profits, the refin-
ery should buy emissions up to the rate where the marginal benefit of emissions
370 15 Regulation
Marginal benefit/cost ($ per ton) Marginal cost to

society
a
35
Marginal benefit
to society
0 800
Emissions (thousand tons a year)
FIGURE 15.3 Economic efficiency in emissions.

Note: The economically efficient rate of emissions is 800,000 tons a year, where the marginal benefit
and the marginal cost to society are both equal to $35 per ton.
Marginal benefit/cost ($ per ton)
45
35 user fee
25
marginal benefit
to refinery
0 40 50 60
Quantity (thousand tons of emissions a year)
FIGURE 15.4 User fee.

Notes: The refinery’s marginal benefit equals the $35 fee, at an emissions rate of 50,000 tons a year.
If the refinery emitted 40,000 tons, the marginal benefit would be $45, which exceeds the $35 fee, so
by increasing emissions, the refinery could increase profits. If the refinery emitted 60,000 tons, then
the marginal benefit would be $25, which is less than the $35 fee, so the refinery should cut back on
emissions.
balances the $35 fee. Suppose that, as shown in Figure 15.4, the refinery’s mar-
ginal benefit balances the $35 fee at an emissions rate of 50,000 tons a year.
To see why 50,000 tons maximizes profit, consider emissions of less than 50,000
tons a year – say, 40,000 tons. Then the refinery’s marginal benefit from more
emissions would be $45, which exceeds the $35 fee. So it should increase emissions
and raise profit. If it emits more than 50,000 tons a year, then its marginal benefit
would be less than the $35 fee. So it should reduce emissions.
Regulation 371
All sources would emit up to the level that their marginal benefit equals $35.
Since the regulator charges the same $35 fee to all sources, the marginal benefits
of all sources would be equal.
Furthermore, the regulator sets the fee according to the social marginal cost
of emissions, $35. Thus, the marginal benefits equal the user fee which equals the
social marginal cost of emissions. So, the user fee implements the economically
efficient rate of emissions.
Quota/Standard
The other method of regulation is directly through standards or quotas. From
Figure 15.3, the economically efficient rate of emissions is 800,000 tons a year.
So the regulator could simply stipulate a maximum quota of 800,000 tons a year.
However, there are many sources of emissions. How should the regulator allocate
the quota among the various sources?
One way is to issue licenses for 800,000 tons of emissions a year and sell the
licenses through public auction. Each source would demand licenses according
to its marginal benefit from emissions. The market demand would be the social
marginal benefit curve in Figure 15.3.
The supply of licenses would be perfectly inelastic at 800,000 tons a year. By
Figure 15.3, the marginal benefit of the 800,000th ton of emissions is $35. Thus,
the equilibrium price of a license – where quantity demanded equals quantity
supplied – would be $35.
Each source would buy licenses up to the level that its marginal benefit equals
the price of $35. Hence, the quantity of emissions would be economically effi-
cient. By selling emissions licenses, the regulator is effectively charging a user fee
that is determined by a competitive market.
PROGRESS CHECK 15D

Referring to Figure 15.3, compare the social benefit and cost with user fees of
$25 per ton and $45 per ton.
Congestion
The demand for facilities such as bridges and tunnels varies with time. Outside
peak hours, the facilities provide non-rival use. By contrast, during peak hours,
each additional user generates congestion, the costs of which include delays and
more accidents. Congestion is a negative externality.
Consider a tunnel that can smoothly convey up to 30 vehicles a minute. From
the standpoint of economic efficiency, traffic through the tunnel should be man-
aged so that the marginal benefit of each user balances the marginal cost. When
there are fewer than 30 vehicles a minute, the marginal cost of a crossing is noth-
ing. So the tunnel should not exclude any driver.
372 15 Regulation
When, however, 30 vehicles per minute are in the tunnel, the marginal cost
becomes positive. For economic efficiency, the tunnel should set a toll equal to the
marginal cost. This will ensure that the only drivers who enter the tunnel are those
whose benefit exceeds the marginal cost.
What if the tunnel does not charge a toll? Consider the decision of a mar-
ginal driver (the 31st) whether to enter the tunnel or wait until there is less traffic. If
the driver enters immediately, she will save some time. In making the decision, that
driver compares her private benefit from crossing at that time with her private cost,
and ignores the additional costs on other drivers. Owing to this negative externality,
too many drivers enter the tunnel relative to the economically efficient number.
Generally, for economic efficiency, congestible facilities should levy a user fee
equal to the marginal cost of use, where the cost includes the externalities imposed
on other users. As the marginal cost varies with the time of day, so should the
price. This pricing would ensure economically efficient usage of facilities such as
bridges, tunnels, roads, and subways.
Congestion illustrates an externality that varies over time. A similar principle
applies to externalities that vary geographically. With differences in marginal ben-
efit and marginal cost, the economically efficient level of the externality would
vary. And so would the appropriate user fee or tax.
Accidents
Accidents are a specific class of externalities which are probabilistic. For instance,
the possibility of automobile accidents means that every driver imposes a possi-
ble negative externality on others. The likelihood and severity of the externality
depend on the care with which she drives.
As Figure 15.5 illustrates, the driver chooses the level of care where her mar-
ginal benefit (in terms of the reduced expected harm from accidents) equals
marginal cost. However, the economically efficient degree of care balances the
marginal benefit to society with the driver’s marginal cost of care. The marginal
benefit to society includes the reduction in expected harm to other drivers, so
exceeds the marginal benefit to the driver herself.
Accidents possibly involve so many people that private action will not resolve the
externality. Then government intervention is necessary. Typically, however, the gov-
ernment does not intervene directly. Rather, the government assigns rights which then
establish the basis for the relevant parties to resolve the exter-
Liability: The conditions nality. Specifically, the law specifies the liability of the parties
under which one party
to an accident. Liability is the set of conditions under which
must pay compensation for
causing harm to another. one party must pay compensation for causing harm to another.
In effect, by specifying liability and compensation, the law
implicitly sets a price for causing an accident or, equivalently,
a price for failing to take care. Unlike prices in conventional markets, the price for
an accident is paid only after the event.
So, given the law, each driver chooses the level of care to balance her marginal
benefit of care (in terms of the reduced expected liability for compensation and
Regulation 373
Marginal benefit/cost ($ per unit of care)
Marginal cost to
driver
s Marginal benefit
30 to society
Marginal benefit
to driver
0 50
Quantity (units of care)
FIGURE 15.5 Accidents.

Notes: The driver would choose the level of care where her marginal benefit equals marginal cost. This
is lower than the economically efficient level of care, where the marginal benefit to society equals the
marginal cost of care to the driver.
reduced harm from accidents) against her marginal cost. With the appropriate
liability and compensation, the driver would choose the economically efficient
level of care. Referring to Figure 15.5, this will happen if the marginal benefit of
care to the driver balances the marginal cost at the economically efficient level.
PROGRESS CHECK 15E

Suppose that, relative to the situation in Figure 15.5, the courts double the
compensation that injurers must pay accident victims. How will this affect a
driver’s choice of care?
AUSTRALIA: REDUCING CARBON
Australia generates about 80% of its electricity from coal. Among economically
advanced countries, it emits the greatest amount of greenhouse gases per
capita. Prime Minister Kevin Rudd was overthrown by his deputy, Julia Gillard,
after he failed to pass legislation to reduce carbon emissions.
In July 2011, the new government of Prime Minister Julia Gillard committed
to reducing Australia’s carbon emissions by 80% from 2000 levels by 2050.
From July 2012, the government would impose a tax of A$23 per ton of carbon
emissions on the top 500 polluters. From 2015, it would replace the tax with a
market-based emissions trading scheme. Meanwhile, the government would
provide financial incentives to close the country’s most polluting power stations.
Source: “Breaching the brick wall,” The Economist, July 11, 2011.
374 15 Regulation
KEY TAKEAWAYS
• A market is a natural monopoly if the average cost of production is minimized

with a single supplier.
• Governments can regulate monopolies by either rate of return or price.
• In a potentially competitive market, economies of scale and scope are small
relative to market demand.
• Governments can foster competition through competition law and structural
regulation.
• Governments can regulate markets with asymmetric information through
disclosure, conduct, and structure.
• Governments can regulate externalities through user fees/taxes or quotas/
standards.
• Governments can regulate accidents by establishing legal rights.
REVIEW QUESTIONS
1. Explain the concept of a natural monopoly.

2. What are the challenges for the efficiency of businesses owned and operated
by government?
3. Explain marginal cost pricing.
4. What are the problems with price regulation?
5. Explain rate-of-return regulation.
6. What are the problems with rate-of-return regulation?
7. Does privatization necessarily increase competition? Explain why or why not.
8. Generally, what does competition law seek to prohibit?
9. Mandatory disclosure is always the best way to resolve asymmetric information.
True or false?
10. Explain how to resolve asymmetric information by regulating conduct and
business structure.
11. Explain how to regulate construction site noise with a user fee.
12. Explain how to regulate automobile emissions by a standard.
13. Why should the government charge a toll for use of a tunnel?
14. Why should the charge for using a bridge vary by the time of day?
15. How does the government regulate accidents?
1. The Delhi electricity industry is vertically separated into generation, transmis-

sion, and distribution. The government awarded monopoly franchises for trans-
mission and distribution to various companies. The Delhi Electricity Regulatory
Commission regulates the monopolies to a maximum 14% rate of return on
equity. Between 2005 and 2011, the Commission refused increases in prices,
although two of the distribution companies incurred losses. (Source: “Delhiites to
Regulation 375
pay 22% more for power as regulator clears hike,” Financial Express, August 27,
2011.)
(a) Why did the Delhi government separate the industry into generation,
transmission, and distribution? Why did it award monopolies for
transmission and distribution?
(b) What challenges would the Commission face in administering rate-of-
return regulation?
(c) Explain the Commission’s refusal to allow price increases in terms of
holdup.
(d) Three of the distribution companies are joint ventures between the
private sector and the Delhi government. How does the government
shareholding affect the potential for holdup by the government?
2. The Independent Pricing and Regulatory Tribunal regulates the prices set
by the Office of Water, the monopoly supplier of water in New South Wales,
Australia. In 2010, the Office of Water sought higher prices to cover an A$186
million increase in costs. The Tribunal approved only 41% of the cost increase.
It also implemented a two-part tariff for users with meters. The fixed part of
the tariff would yield 70% of revenue, with the remaining 30% coming from the
usage charge. (Source: Independent Pricing and Regulatory Tribunal, Review
of Prices for the Water Administration Ministerial Corporation, October 2010.)
(a) Explain natural monopoly in the context of water distribution.
(b) Using a relevant diagram, explain average cost pricing of water.
(c) If the monopoly’s costs increase, use your diagram in (b) to explain
how the regulator should adjust price.
(d) Why would the Office of Water inflate its reported costs?
(e) Explain how the two-part tariff would encourage users to economize
on water.
3. In Hong Kong, the supply of electric power is monopolized by two vertically
integrated companies. Hong Kong Electric supplies Hong Kong Island, while
China Light and Power supplies Kowloon and the New Territories. The Hong
Kong government regulates each company to a maximum 9.99% rate of return
on net fixed assets. Although the two companies’ networks are just a few
kilometers apart, they do not interconnect.
(a) Discuss whether vertical integration of the Hong Kong electricity
industry is economically efficient.
(b) Given differences in the demand for electricity between Hong Kong
Island and Kowloon and the New Territories, how would interconnection
increase economic efficiency?
(c) Why do the two electricity companies oppose interconnection?
4. In the United States, all stockbrokers must insure customer accounts up to
$500,000 with the Securities Investor Protection Corporation (SIPC). SIPC
insurance covers customers for losses on cash, stocks, and bonds due to
default by the broker. The SIPC does not cover losses on other investments
such as commodity futures.
(a) Can you explain the government requirement for SIPC insurance in
terms of asymmetric information between brokers and investors?
376 15 Regulation
(b) Some brokers purchase private insurance to cover losses that exceed
the minimum SIPC insurance coverage. Explain this practice as a way
in which brokers can signal their financial reliability.
(c) Commodity futures are more complex than stocks and bonds. Why is
it reasonable for SIPC not to insure losses on futures?
5. With a fast-growing population, India’s economy would boom if young people
were sufficiently educated. Many parents appreciate the value of education and
send their children to school. However, some private schools fail to educate
their pupils, and illiterate parents do not know that they are being cheated.
(a) Using a relevant figure, explain the effect of asymmetry of information
in the market for education.
(b) Discuss how regulation of (i) disclosure, (ii) conduct, and (iii) structure
might help to resolve the asymmetry of information.
(c) How can charities and social enterprise help?
6. In India, many poor parents prefer their children to work rather than attend
school. In 2014, Akshaya Patra provided free lunches to 1.4 million children in
over 10,600 government and government-aided schools across ten states to
encourage attendance at school.
(a) Suppose that some parents have a very high discount rate. How would
that affect their choice between sending their children to school and
work?
(b) Does education of children generate an externality? Does that justify a
subsidy to education?
(c) To what extent does your analysis in (b) apply to tertiary as compared
to primary and secondary education?
(d) Generally, economists argue that giving cash is more efficient than
gifts in kind. Discuss whether Akshaya Patra should give students an
equivalent cash subsidy instead of the free lunch.
7. The US Federal Aviation Administration (FAA) classifies all civil transport
aircraft according to noise. The categories range from Stage 1, which is the
noisiest, to Stage 4, which is the quietest category. In 2013, the FAA prohibited
the operation of Stage 2 jet aircraft weighing less than 75,000 pounds. Some
of the jets can be retro-fitted to meet Stage 3 standards. (Source: “FAA Stage 2
noise ban rule bars older jets in U.S.,” ainonline.com, August 4, 2013.)
(a) Who should regulate aircraft noise: the federal government or local
airport authorities?
(b) Consider an airport that has specified standards for aircraft noise.
Should the airport create permits for aircraft noise and allow airlines to
trade these permits?
(c) Using a relevant diagram, explain how the airport should determine the
level of noise to allow. Should the level vary with the time of day?
8. Among economically advanced countries, Australia emits the largest amount
of greenhouse gases per capita. In July 2011, the government of Prime Minister
Julia Gillard committed to reducing Australia’s carbon emissions. From July
2012, the government would impose a tax of A$23 per ton of carbon emissions
Regulation 377
on the top 500 polluters. From 2015, it would replace the tax with a market-
based emissions trading scheme.
(a) Using a relevant diagram, explain how the government should determine
the tax on carbon emissions.
(b) Using the same diagram, explain how the government should determine
the quota for total emissions.
(c) How would emissions trading establish a price for emissions?
9. With rental rates exceeding $3,000 per square foot per year, retail space in
Causeway Bay, Hong Kong, is among the world’s most expensive. In 2011,
the Town Planning Board limited the height of future buildings in the area to
between 130 and 200 meters. Real estate developer Hysan Group, which owns
nine properties in the area, asked the Board to relax the limits, but the Board
denied the application. (Source: “Long hours didn’t cloud our judgment on
Causeway Bay building restrictions, town planners say,” South China Morning
Post, December 1, 2014.)
(a) What externality does the height restrictions resolve?
(b) How would the Hysan Group benefit from relaxing the height restrictions?
(c) Using a relevant diagram, explain how the Town Planning Board should
determine the restrictions on height.
(d) Should the restrictions on height be the same throughout Hong Kong?

How does government regulation affect your organization? Write a memorandum
to the government explaining how to improve the administration of regulation.
Note
1 This discussion is based, in part, on Delhi Electricity Regulatory Commission, Petition for
Approval of Aggregate Revenue Requirement and Multi Year Generation Tariff for Pragati
Power Corporation Limited for the FY 2007–08 to FY 2010–11, Petition No. 39/2007; North
Delhi Power Limited, Annual Report, 2009–10.
16 C H A P T E R
Answers to Selected
Progress Check and
Review Questions
Chapter 1. Introduction to Managerial Economics

1A. Value added = Buyer benefit − Seller cost = Buyer surplus + Seller
economic profit.
1B. 0, 0, … , 0, 300.
1C. Max’s NPV from the MBA would be
50,000 95,000 95,000 95,000

−$50,000 − + + + = $119,
1 319.
1.10 1.102 103
11.10 1.104
He should continue in his current job, which gives a higher NPV.

1D. [Omitted.]
1E. The three branches are: (i) competitive markets; (ii) market power; and
(iii) imperfect markets. Competitive markets have large numbers of buy-
ers and sellers, none of which can influence market conditions. By con-
trast, a buyer or seller with market power can influence market conditions.
A market is imperfect if one party directly conveys benefits or costs to
others, or if one party has better information than another.
1. Value added is the difference between buyer benefit and seller cost. Eco-
nomic profit is the difference between seller revenue and seller cost.
2. Value added is the difference between buyer benefit and seller cost.
So, even though the charity receives no revenue, it need not be
destroying value. For the free mosquito nets to create value, the buy-
er’s benefit must exceed the seller’s cost.
3. [Omitted.]
4. [Omitted.]
5. People act with bounded rationality because they have limited cogni-
tive ability and lack self-control.
Answers to Selected Progress Check and Review Questions 379
6. The employer could invest the $1 million today, which would yield a positive
return, and so the pension fund would have more than required to pay out
$1 million in the future.
7. If the inflows exceed the outflows after accounting for the timing of the
inflows and outflows, then the NPV will be positive.
8. With regard to vertical boundaries, a local cable TV provider that produces
TV series is more vertically integrated than a local cable TV provider that
buys TV series from others.
9. A university merging with a hospital is expanding its horizontal boundaries.
A university shutting some faculty is shrinking its horizontal boundaries.
10. When Apple outsources the manufacturing of iPhones to a contractor in
China, it is vertically disintegrating.
11. (a) The electricity market includes buyers and sellers. (b) The electricity
industry consists of sellers only.
12. (a) False. (b) False.
13. Demand and supply model.
14. A manufacturer with market power can influence conditions of demand
and/or supply.
15. (b).
Chapter 2. Demand
2A. The theater must cut its price by $3 from $11 to $8.
2B. (a) It slopes downward because of diminishing marginal benefit. (b) Assum-
ing that all-in-one stereos are an inferior product, the drop in the consumer’s
income will cause the demand curve to shift to the right.
2C. Online movies are a substitute for seeing movies in the theater. A fall in the
price of online movies will cause the demand curve for theater movies to
shift to the left.
2D. If the price of movies is $8, Joy’s buyer surplus would be the area under her
demand curve above the horizontal line at the price of $8.
1. [Omitted.]
2. [Omitted.]
3. [Omitted.]
4. [Omitted.]
5. Introduction of the new birth-control device will: (a) reduce the demand for
male condoms; (b) reduce the demand for birth-control pills.
6. Pepsi advertising will: (a) increase the demand for Pepsi; and (b) decrease the
demand for Coca-Cola.
7. Changes in consumer incomes would affect consumer demand for Apple
iPhones, and so affect Apple’s demand for microprocessors.
8. ATMs are a substitute for bank clerks. The higher labor costs will increase
the demand for ATMs.
380 16 Answers to Selected Progress Check and Review Questions
9. Buyer surplus is the difference between buyer’s total benefit from consump-
tion and the buyer’s actual expenditure.
10. False. Whether the buyer is a consumer or business, buyer surplus is the dif-
ference between buyer’s total benefit from consumption and the buyer’s
actual expenditure.
11. Each person has a different marginal benefit for the flight. The market
demand curve shows the marginal benefits of the various consumers,
arranged from the consumer with highest marginal benefit to the consumer
with lowest marginal benefit.
12. Draw any straight line which includes the point at 200 minutes and 10 cents.
Antonella’s buyer surplus is the area between the demand curve and the
10 cents line.
13. Not true. Each additional item provides lower marginal benefit. Actual sav-
ings are the difference between total benefit and the price paid.
14. A package deal provides a fixed level of consumption at a lump-sum price.
A broadband service provider can design a package deal providing a fixed level
of consumption at a price just a little less than the consumer’s total benefit.
15. Two part pricing is a pricing scheme comprising a fixed payment and a charge
based on usage. A broadband service provider can design a two-part price
such that the fixed charge extracts the consumers’ buyer surplus.
Chapter 3. Elasticity
3A. The proportionate change in quantity demanded is (1.5 − 1.44)/1.44 = 0.06/
1.44 = 0.042. The proportionate change in price is (1 − 1.10)/1.10 = −0.10/
1.10 = −0.091. The own-price elasticity is 0.042/(−0.091) = −0.46.
3B. The intuitive factors that influence the own-price elasticity of demand are:
(a) availability of direct or indirect substitutes; (b) the buyer’s prior commit-
ments; (c) the benefits/costs of economizing.
3C. The proportionate change in quantity demanded would be −2.5 × 7% = −17.5%.
The proportionate change in expenditure would be 7% − 17.5% = −10.5%.
3D. The own-price elasticity is always a negative number. But the income elastic-
ity could be negative or positive, depending on whether the item is a normal
or inferior good.
3E. For a non-durable good, the longer the time that buyers have to adjust, the
bigger will be the response to a price change.
1. [Omitted.]
2. The demand curve slopes downward, hence the own-price elasticity is negative.
The proportionate changes in quantity demanded and price of an item are
pure numbers. Hence, the own-price elasticity is a pure number with no units.
3. If a 1% price increase causes less than a 1% drop in quantity demanded, then
the demand is price inelastic. If a 1% price increase causes more than a 1%
drop in quantity demanded, then the demand is price elastic.
4. The demand of executives traveling at the expense of their employers will be

less elastic. They incur the costs of economizing but their employers receive
the benefit.
5. Rise. The proportionate change in expenditure would be 10% × (1 − 0.7) = 3%.
6. The volume of sales would change by −1.5 × (− 5%) = 7.5%.
7. [Omitted.]
8. True.
9. Complements.
10. Advertising by one product brand will draw customers from customers of
other brands as well as increase the demand for beer in general. Advertising
of beer in general can only increase the market demand.
11. The change in quantity demanded would be 1.3 × 5% = 6.5%.
12. More elastic in the long run.
13. The increase in quantity demanded would be 2.85% in the short run and
3.91% in the long run.
14. If consumers are influenced by sunk costs, they will be less responsive to
changes in price, and so their demand will be less elastic.
15. (a) More elastic. (b) Less elastic.
Chapter 4. Supply
4A. If the fixed cost were higher, the variable cost curve would not be affected,
while the total cost curve would be shifted up by the extent of the increase in
fixed cost.
4B. If the fixed cost were higher, the average variable cost and marginal cost
curves would not be affected, while the average cost curve would be
shifted up.
4C. Luna should produce at the rate where marginal cost is $7.50 per sheet.
4D. It should produce 6,000 sheets a week.
4E. The new seller surplus would be the area between the new price line of $9 per
sheet and Luna’s supply curve. The seller surplus would increase by the area
between the $9 and $7 price lines.
4F. Supply would be inelastic if production capacity is fully used and the adjust-
ment time is short.
1. The short run is a time horizon within which a seller cannot adjust at least
one input. By contrast, the long run is a time horizon long enough that the
seller can adjust all inputs. Assuming that all fixed costs are also sunk, while
all variable costs are not sunk, then there are fixed costs only in the short
run, while all costs are variable in the long run.
2. The average cost of $5 includes the average fixed cost. But, in the short run,
the fixed cost is unavoidable. In the short run, the factory should accept
orders so long as the price of a shirt covers the average variable cost.
3. Nothing. Marginal costs are not necessarily related to fixed costs.
4. The marginal product of advertising for a manufacturer of shampoos is the

increase in sales arising from an additional dollar of advertising. With more
of the advertising inputs, there will be a diminishing marginal product.
5. If the seller is so small that it can sell as much as it would like at the market price.
6. Since the price is less than the marginal cost, the producer can raise profit by
reducing production.
7. The analysis underestimates the increase in profit. It considers only the
increase in profit on the existing production, and ignores the increase in
profit resulting from an increase in production.
8. Price is revenue divided by sales (or production), while average variable cost
is variable cost divided by sales (or production).
9. Not necessarily. In the short run, it depends on whether or not the revenue
covers the variable cost. If not, then they should shut down. In the long run,
it depends on whether or not the revenue covers average cost.
10. Price is revenue divided by sales (or production), while average cost is total
cost divided by sales (or production).
11. In the short run, the business should continue in production so long as its
revenue covers the variable cost. In the long run, the business should con-
tinue in production so long as its revenue covers the total cost.
12. (a) The supply of furniture will be lower (shift to the left). (b) The supply of
furniture will be higher (shift to the right).
13. For a given increase in price, if the supply curve is more elastic, the increase
in seller surplus will be larger.
14. True.
15. The supply will be relatively more elastic in the long run.
Chapter 5. Market Equilibrium

5A. The conditions for a market to be in perfect competition are as follows.
(a) The product is homogeneous. (b) There are many buyers, none of whom
has market power. (c) There are many sellers, none of whom has market
power. (d) New buyers and sellers can enter freely, and existing buyers and
sellers can exit freely. (e) All buyers and all sellers have symmetric informa-
tion about market conditions.
5B. See Figure 5.1. If price drops to $160 per ton-mile, the quantity demanded
will exceed the quantity supplied.
5C. (a) False. (b) True.
5D. In Figure 5.4, the entire demand curve would shift to the left by 2 billion ton-
miles. The supply curve would not shift. The equilibrium price would fall below
$200 per ton-mile and production would fall below 10 billion ton-miles a year.
5E. The short-run price would be higher and the long-run price would be lower.
Hence, the difference between the short- and long-run prices would be greater.
1. (a) The service is close to homogeneous. (b) There are many consumers in
the market, each of whom buys a small quantity. (c) There may be few dry
cleaners, depending on the density of population. (d) Entry and exit are
relatively free. (e) Consumers and dry cleaners probably have symmetric
information about market conditions.
2. (a) The good is close to homogeneous. (b) There are many buyers in the
market, each of whom buys a small quantity. (c) There may be few manufac-
turers. (d) Entry and exit are relatively free. (e) Buyers and manufacturers
probably have symmetric information about market conditions.
3. This regulation presents a barrier to entry, and hence reduces the degree of
competition.
4. The price will tend to fall.
5. The price will tend to rise.
6. The increase in consumer incomes would shift the demand for clothing upward
(consumers are willing to pay more) and to the right (at every price, consumers
want to buy more). The effect on the market price depends on the price elas-
ticity of supply. If supply is completely inelastic, the price would increase the
most. If supply is completely elastic, the price would not increase at all.
7. The effect of any upward shift in demand depends on the price elasticities of
demand and supply. The impact on price will be greater if demand is more
elastic, and supply is more inelastic.
8. The effect of any upward shift in demand depends on the price elasticities of
demand and supply. The impact on quantity consumed will be greater if
demand is more elastic, and supply is more elastic.
9. If demand is completely inelastic, consumers would pay a lot for additional
cars, and production would increase the most. However, if demand is com-
pletely elastic, consumers would not be willing to pay more for additional
cars, and production would not increase at all.
10. The effect of any shift in supply depends on the price elasticities of demand
and supply. The impact on the price will be greater if demand is more inelas-
tic, and supply is more inelastic.
11. The effect of any shift in supply depends on the price elasticities of demand
and supply. The impact on the quantity produced will be greater if demand
is more elastic, and supply is more inelastic.
12. An increase in wages would shift upward the supply of restaurant meals.
The effect of any change in supply depends, in part, on the price elasticity of
demand. If the demand is very elastic, a reduction in supply would not affect
the price. By contrast, if demand is very inelastic, a reduction in supply
would reduce the price.
13. The supply curve is more elastic in the long run than in the short run.
Accordingly, the price increases more in the short run than in the long run.
14. The supply curve is more elastic in the long run than in the short run. Accord-
ingly, the production increases more in the long run than in the short run.
15. The effects of an increase in demand depend on the elasticities of demand
and supply. Hence, the difference in short- and long-run effects of an increase
in demand depends on the difference between short- and long-run price
elasticities.
Chapter 6. Economic Efficiency

6A. The concept of economic efficiency extends beyond technical efficiency. For
economic efficiency, the production of the item must be such that the mar-
ginal benefit equals the marginal cost. Technical efficiency means providing
an item at the minimum possible cost. It does not imply that scarce resources
are being well used.
6B. The price in a market system has two roles. First, the price communicates
necessary information. It tells buyers how much to purchase and tells sellers
how much to supply. Second, the price provides a concrete incentive for each
buyer to buy that quantity and maximize net benefit, and for each seller to
supply that quantity and maximize its profit.
6C. With decentralization, Jupiter should set the transfer price equal to the
market price of semiconductors. Then the electronics division will use
semiconductors up to the point where marginal benefit equals the transfer
price. This will ensure economic efficiency.
6D. It does not affect the difference between freight-inclusive and free on board
pricing.
6E. The buyer’s price increases relatively more. Hence, the incidence of the tax
on travelers will be relatively higher.
1. Children were using bread (in sport) up to the point that the marginal ben-
efit equaled the very low price. This price was less than the marginal cost.
Hence, the marginal benefit of use was less than the marginal cost and not
economically efficient.
2. The condition that all users receive the same marginal benefit.
3. Each prisoner of war received the parcel for free. Prisoners vary in their
preferences for cigarettes and chocolate. If each prisoner gets the same
quantity of each item, then their marginal benefits will be different.
4. The self-employed workers would get a higher wage (because they need not
pay tax), so they would work up to a level of hours where their marginal cost
would be higher than that of the employed workers. Hence, the allocation of
labor is not economically efficient.
5. In a competitive finance market, all buyers (investors) purchase up to the
quantity where the marginal benefit equals the market price of the funds, and
all sellers (banks) supply up to the quantity where the marginal cost equals
the market price of the funds. Buyers and sellers face the same market price;
hence, the allocation of investment funds is economically efficient.
6. Suppose that the price control forces the price to below the marginal cost.
Consumers would buy to the quantity such that their marginal benefit equals
the price. This would be less than the marginal cost, causing economic inef-
ficiency in the allocation of rice.
7. Outsourcing is the purchase of services or supplies from external sources.
The opposite of outsourcing is vertical integration.
8. The transfer price is the price charged for the sale of crude oil from the crude
oil division to the refining division.
9. To ensure that the department store uses the economically efficient quantity
of retail space.
10. A price that includes freight is called cost including freight. A free on board
price does not include the freight cost.
11. Payment refers to who conveys the money to the agent. Incidence refers to
the impact in terms of price – buyer’s price and seller’s price.
12. With free shipping, the retail supply curve would be lower and the demand
curve would be higher. If vendors charge for shipping, the retail supply curve
would be higher and the demand curve would be lower. The total price (good
plus shipping) to consumers would be the same.
13. If the demand is perfectly elastic, the tax will be incident completely on the
supply side and there would no incidence on consumers.
14. Both demand and supply increase. The incidence on airlines and consumers
depends on the price elasticities of supply and demand.
15. If the demand is inelastic relative to supply, the tax will be incident mostly
on the demand side. The difference between the pre-tax prices at the airport
and in the city would be larger.
Chapter 7. Costs
7A. The opportunity cost is $8 million. So, if Luna commences R&D, its profit
would be the profit contribution minus the R&D expense minus the opportu-
nity cost, 20 − 10 − 8 = $2 million. So, the correct decision is to commence R&D.
7B. (a) Set the transfer price equal to the market price. (b) Set the transfer price
equal to the opportunity cost of the input (which is the consuming division’s
marginal benefit from that input).
7C. The avoidable cost of R&D is 10 − 1 = $9 million. Hence, if Luna continues
R&D, the cost would be $9 million, and the profit would be the profit con-
tribution minus the avoidable cost of R&D, 8 − 9 = −$1 million, that is, a loss
of $1 million. So, the correct decision is to cancel the R&D.
7D. [Omitted.]
7E. Total cost = $26,900 + $3,500 = $30,400. There are neither economies nor
diseconomies of scope.
7F. The new experience curve is higher at every level of cumulative production
than the experience curve with 20% cost reduction.
1. The client could have used the time spent with the salesman on other activi-
ties. In addition, the salesman buys lunch for his client with the objective of
increasing sales.
2. Value added = Buyer benefit − Seller cost = Buyer surplus + Economic
profit. The social enterprise does not receive any revenue. However, it does
create value to the extent that the benefit (of education) exceeds the cost.
Since the enterprise does not earn any revenue, its EBITDA is the negative
of its costs.
3. The residential development group should pay the market price of lumber to
the building materials division.
4. Luna Biotech’s manufacturing division should set its price equal to the
opportunity cost of the process.
5. Jupiter (loan-financed) looks like it is less profitable than Mercury (an iden-
tical business, but equity-financed) because conventional accounting requires
interest payments to be expensed, but not expected dividends. In terms of
EBITDA, Mercury must account for the cost of equity, so the performance
of the two companies would be more similar.
6. No. The pensions of retired employees are sunk costs. Once the business com-
mits to hire the employees, it must pay its retired employees their pensions.
The pensions are already committed and unavoidable. In forward-looking
business decisions, managers should only consider avoidable costs.
7. Sunk costs are more significant in situation (a). The business depends on
permanent staff rather than part-time workers. The employment of permanent
staff is a long-term commitment.
8. No. Although the average fixed costs will fall with the scale of the produc-
tion, the variable cost will most probably increase with the scale of production
due to nurses working inefficiently (for example, too many nurses will get
in each other’s way). Since the fixed costs are not substantial, the increase in
average variable cost will outweigh the decrease in average fixed cost.
9. Fixed costs can be either sunk or avoidable. Fixed costs that are sunk are
those fixed costs that are committed and can never be recovered. Fixed costs
that are avoidable are those fixed costs that can be avoided with a change of
plan. Sunk costs need not be fixed in the sense of supporting any scale of
production.
10. In an industry with economies of scale, businesses operating on a larger scale
will be able to achieve a lower average cost. Therefore, businesses should aim
to produce on a large scale. This means mass marketing and low prices.
11. Economies of scale pertain to a single product. Economies of scope pertain
to multiple products. With economies of scale, increasing the scale of pro-
duction would reduce average costs. With economies of scope, increasing
the scope of production would reduce average costs.
12. When there are economies of scope for two products, it is cheaper to pro-
duce them together. Therefore, companies should produce both products to
achieve lower cost than competitors which specialize in producing one of
the products.
13. An experience curve with a learning percentage of 100% is horizontal (no
reduction in average cost with increase in cumulative production).
14. With an experience curve, average cost falls with cumulative production. So,
it is essential to forecast cumulative production accurately. This forecast is
crucial for planning investments and setting prices.
15. Production of a newspaper involves a substantial fixed cost. Decision-makers

subject to the fixed-cost fallacy include part of the fixed cost in variable cost.
Hence, they overestimate the average variable (and marginal) cost.
Chapter 8. Monopoly
8A. Barriers to competition and the elasticity of demand or supply.
8B. If the demand is very elastic, then the marginal revenue will be close to the
price.
8C. It should raise price, so reducing sales to the production where its marginal
cost equals its marginal revenue.
8D. The new marginal cost will cross the (unchanged) marginal revenue at a
smaller quantity of production. Hence, Venus should set a higher price to
induce that quantity of sales.
8E. Advertising expenditure equals (135 − 71) × 0.14 × 1.3 = $11.65 million.
8F. With a higher price and lower marginal cost, the incremental margin will be
higher, and so R&D expenditure should be higher.
8G. Venus’s total expenditure is represented by either the area u0vx under the
marginal expenditure curve from a quantity of 0 to 6,000 tons or the rectan-
gle t0vz.
1. [Omitted.]
2. Patents, copyright, trademarks, and trade secrets.
3. With economies of scale, a producer that produces on a larger scale will have
a cost advantage over other producers. So, it can price lower and dominate
its market.
4. To sell additional units, a seller must reduce its price. So, when increasing
sales by one unit, the seller will gain the price of the marginal unit but lose
revenue on the inframarginal units. Hence, the marginal revenue is less than
or equal to the price.
5. True.
6. The publisher should reduce its price and sell additional units. For these
units, the marginal revenue is greater than marginal cost, thereby increasing
total profit. It should reduce price until the marginal revenue equals the mar-
ginal cost.
7. The profit-maximizing quantity is such that the marginal revenue equals the
marginal cost. Hence, after a change in costs, the seller should look for the
quantity where the marginal revenue equals the new marginal cost. So it
must consider both the marginal revenue and the marginal cost.
8. The profit-maximizing quantity is such that the marginal revenue equals the
marginal cost. Hence, after a change in demand, the seller should look for
the quantity where the new marginal revenue equals the marginal cost. So it
must consider both the marginal revenue and the marginal cost.
9. Advertising expenditure equals (100 − 40) × 0.01 × 500,000 = $0.3 million.
10. Reduce advertising expenditure until the advertising–sales ratio equals the
incremental margin multiplied by the advertising elasticity of demand.
11. The R&D elasticity of demand depends on two factors: one is the effec-
tiveness of R&D in generating new products and enhancing existing
products; and the other is the effect of new and enhanced products on
demand.
12. Raise R&D expenditure until the R&D–sales ratio equals the incremental
margin multiplied by the R&D elasticity of demand. Intuitively, the higher
is the incremental margin and the more sensitive is demand to R&D expen-
diture, the more the business should spend on R&D.
13. False.
14. True.
15. In a perfectly competitive market, every buyer purchases at a scale where its
marginal expenditure equals the market price; hence, its incremental margin
percentage is zero. By contrast, a monopsony restricts purchases to get a
lower price and increase its net benefit above the competitive level. The more
inelastic is the market expenditure, the more the buyer can reduce price
below its marginal expenditure.
Chapter 9. Pricing
9A. ( p − 70)/p = 1/2, so price equals $140.
9B. It does not extract the entire buyer surplus, and it does not provide the eco-
nomically efficient quantity.
9C. Area adb = 0.5 × 2,500 × (400 − 240) = $200,000. Area bec = 0.5 × 2,500 ×
(240 − 80) = $200,000.
9D. Adults: With the marginal cost of $100, the marginal cost is equal to marginal
revenue at a smaller quantity. At the new quantity, the price would be higher.
Seniors: With the marginal cost of $100, the marginal cost is equal to marginal
revenue at a smaller quantity. At the new quantity, the price would be higher.
9E. The CF price in Japan is
p − 30,000 1
=− ,
p −2.5
or p = ¥50,000. The domestic price is $350 = ¥35,000. So the difference is ¥15,000.

9F. Buyer surplus of business traveler: unrestricted = $501 − $500 = $1; restricted =
$401 − $180 = $221. Buyer surplus of leisure traveler: unrestricted = $201 −
$500 = −$299; restricted = $101 − $180 = −$79. The leisure traveler will not buy
either the unrestricted fare or the restricted fare. So indirect segment discrimi-
nation will not work.
9G. Indirect segment discrimination is less profitable than direct segment dis-
crimination for two reasons. (a) Buyer benefit: to induce buyers with differ-
ent attributes to choose different products, sellers may have to resort to
designing low-end products in a less appealing way. (b) Cost: implementing

indirect discrimination may involve relatively higher costs.
1. With uniform pricing, the price of a product depends on both the price elasticity
of demand and marginal cost. Whether to set a higher price for the own-label
merchandise depends on two factors. (a) Since the demand for the own-label
good is less elastic than for the branded good, the incremental margin percent-
age should be higher for the own-label good. (b) However, the marginal cost of
the own-label good would be lower than that of the branded good.
2. By the formula for uniform pricing,
p−2 1
=− ,
p −1.25
so the price is $10.

3. No. It should adjust price so that the incremental margin percentage,
( p − 40)/p, equals the absolute value of the reciprocal of the price elasticity
of demand.
4. The pricing method used by the book publishers is cost-plus pricing. It
ignores the price elasticity of demand.
5. Complete price discrimination increases profit by resolving two shortcom-
ings of uniform pricing. Complete price discrimination extracts the entire
buyer surplus by pricing each unit at the buyer’s benefit. In addition, it pro-
vides the economically efficient quantity, where marginal benefit equals the
marginal cost.
6. The two conditions that are necessary to implement complete price discrim-
ination are as follows: (a) the seller must know each potential buyer’s indi-
vidual demand curve; (b) the seller must be able to prevent customers from
buying at a low price and reselling to others at a higher price (arbitrage).
7. [Omitted.]
8. FOB pricing considers only the difference in the cost of delivery to the dif-
ferent markets, and it ignores the difference in the price elasticity of demand.
CF pricing takes into consideration both the different price elasticities and
the different marginal costs.
9. Newspapers. To discriminate by the buyer’s location, buyers must not adjust
location to take advantage of the price differences. Buyers will not shift loca-
tion to save on the price of a newspaper. Also, the value of newspaper con-
tent depreciates fast, so no buyer would wait for resale of a newspaper from
a cheaper location.
10. [Omitted.]
11. Those who rent at the expense of others will be less sensitive to the higher
gasoline prices at the car rental agencies and will not make the additional
effort to look for cheaper gasoline. The person who makes the buying deci-
sion does not pay the bill.
12. Coupons can be used to discriminate on price because it takes time and
effort to redeem a coupon. To the extent that consumers whose demand is
more price elastic are also those whose time is less valuable, they will redeem
the coupon and get a lower price.
13. From highest profit to lowest: complete price discrimination, direct segment
discrimination, indirect segment discrimination, uniform pricing.
14. Information technology both assists and hinders the seller’s ability to dis-
criminate on price. Through the Internet, marketers can collect detailed
information on consumer preferences. Furthermore, technology helps to
reduce the cost of storing, analyzing and applying this information, enabling
sellers to offer products that better suit customer demand. However, tech-
nology has enabled the growth of search services. So, consumers can more
easily compare products and prices.
15. Cannibalization occurs when high-benefit segments buy the item aimed
at low-benefit segments. First, sellers can make the differences between
the high-margin item and low-marginal item more salient by upgrading
the high-margin item and degrading the low-marginal item. Second, prod-
ucts can be designed with multiple discriminating variables. Lastly, canni-
balization can be mitigated by controlling the availability of the low-margin
item.
Chapter 10. Strategic Thinking

10A. In perfect competition, there are so many producers that decisions made by
a single producer do not affect others. The increase in the production of
nails (made by your company) will not affect the market price, so this deci-
sion is not strategic.
10B. The Nash equilibrium is for both Jupiter and Saturn to cut price.
10C. (a) If Saturn maintains price, the expected consequence is
2 3
× 900 + × 500 = 660.
5 5
(b) If Saturn cuts price, the expected consequence is
2 3
× 800 + × 600 = 680.
5 5
10D. (a) Positive-sum game. The sums of the outcomes for the parties are differ-
ent (between −3 and 2) in the various cells of the game in strategic form.
(b) Positive-sum game. The sums of the outcomes for the parties are different
(between 1,200 and 1,800) in the various cells of the game in strategic form.
10E. There is no first-mover advantage because the equilibrium is the same – both
cut price – whether Jupiter or Saturn moves first.
10F. The new game in extensive form begins at a node where TV Delta chooses
between contracting and not contracting with the fast food chain. Each pos-
sible action leads to nodes where Channel Z chooses between 7:30 pm and
8:00 pm. In turn, each of Channel Z’s actions leads to nodes where TV Delta
chooses between 7:30 pm and 8:00 pm. In equilibrium, TV Delta chooses to
advertise, Channel Z chooses 7:30 pm, and TV Delta chooses 8:00 pm.
10G. Let the minimum probability that the employer will eventually raise wages
be R. Then the expected increase in earnings from a strike would be (−4 + 12) ×
R + (−4) × (1 − R). For the threat of strike to be credible, the expected increase
in earnings from a strike must be at least zero, (−4 + 12) × R + (−4) × (1 − R) ≥ 0,
or 8R −4 + 4R ≥ 0, or R ≥ 1/3. So, the minimum probability is 1/3.
1. (a) In perfect competition, there are so many sellers that decisions made by
a single player do not affect others. So, it does not matter whether the seller
acts strategically. (b) In a monopoly, there is only one seller. However, it
must consider potential entrants to its market. Therefore, it has to act
strategically.
2. A dominated strategy generates worse consequences than some other strat-
egy, regardless of the other parties’ choices. Therefore, you should not adopt
a dominated strategy.
3. Only (c).
4. If others do not act strategically, I should choose the best strategy, even if it
is not a Nash equilibrium strategy. But, I should never use a dominated
strategy.
5. In a randomized strategy, the party specifies a probability for each of the
alternative pure strategies. Then it adopts each pure strategy randomly
according to the probabilities. The probabilities must add up to 1.
6. Strategy (b) is more effective. It is random and unpredictable, and so the
boxer’s opponent will not be able to anticipate the move and counter it.
7. In a zero-sum game, the outcomes for the parties add up to the same num-
ber in every cell of the game in strategic form. Therefore, one party can
become better off only if another is made worse off – which is the essence
of competition. By contrast, in a positive-sum game, the outcomes for the
parties add up to different numbers in the various cells. Hence, one party
can become better off without another being made worse off. By coordi-
nating, one party can become better off without another being made
worse off.
8. Yes. The outcomes for the parties add up to the same number in every cell of
the game in strategic form.
9. By reasoning forward from the initial node, the party will not be able to
anticipate the other parties’ responses and adjust its own response to maxi-
mize its expected outcome.
10. The equilibrium is the same – both US and Japan head north – whether the
US moves first or second. So there is no first-mover advantage.
11. If the conditional strategic move (promised or threatened action) need not
actually be carried out, the conditional strategic move has no cost. On the
other hand, an unconditional strategic move will involve a cost under all
circumstances. Therefore, conditional strategic moves are more cost-effective
than unconditional strategic moves.
12. Deposit insurance pays compensation to depositors if the bank becomes
insolvent. The deposit insurance funded by the bank itself may not be cred-
ible because, if the bank becomes insolvent, it may not have funds to pay the
compensation.
13. The “walk away” tactic is credible only if the person has a better offer else-
where and is better off by walking away. So the credibility of “walking
away” depends on the probability that the person has a better offer
elsewhere.
14. Both the legal system and violence are threats to persuade borrowers to
repay their loans. Since loan sharks are not allowed to use the legal system
to collect debts, they employ violent threats (costs) to persuade borrowers to
repay.
15. There are many more possible strategies in repeated interactions than in
once-only scenarios. In particular, a party can choose a strategy in which it
conditions its action on the actions of the other party.
Chapter 11. Oligopoly

11A. Luna’s residual demand curve would be more elastic, so its profit-maximizing
price would be lower. Then its best response function would be further to the
left. Similarly, Mercury’s best response function would be lower. The equi-
librium prices would be lower.
11B. The average cost would be higher (the fixed cost does not affect the marginal
cost). The limit price (the lowest price that Mercury can set and deter Luna
from entry) would be higher.
11C. In Figure 11.6, if Luna’s marginal cost is lower, its profit-maximizing price
would be lower and production would be higher. So, in Figure 11.7, Luna’s
best response function would be further to the right.
11D. The graph would be like Figure 11.7, with the downward-sloping best
response functions.
11E. If Mercury sets capacity at 21 million subscribers a month, Luna’s best
response is capacity of zero. This strategy is bad for Mercury because it
would depress the market price and reduce its profit.
11F. Horizontal integration is the combination of two entities, in the same or
similar business, under a common ownership. Vertical integration is the
combination of assets for two successive stages of production under a com-
mon ownership.
1. False. In a Bertrand oligopoly, price equals marginal cost.

2. (a) The residual demand curve of one seller is the demand given the prices
of competing sellers. (b) The residual demand curve of one seller is the
demand given the capacities of competing sellers.
3. (a) The best response function of one seller is the seller’s profit-maximizing
price as a function of the prices of competing sellers. (b) The best response
function of one seller is the seller’s profit-maximizing capacity as a function
of the capacities of competing sellers.
4. You should adjust in the same direction as your competitor, which is to raise
your advertising.
5. With differentiation, if one seller undercuts the competitor’s price, it would
take away only part (and not all) of the competitor’s demand. So, differenti-
ation softens competition.
6. A strategic form shows the strategies of the various parties and the corre-
sponding consequences.
7. The increase in fixed cost does not affect the equilibrium (provided that it is
not so large as to cause the producer to cease production).
8. Actions are strategic substitutes if one party’s adjustment of action leads
others to adjust in the opposite direction.
9. The graph would be like Figure 11.7, with the downward-sloping best
response functions.
10. If the fixed cost of production is zero, the average cost and marginal
cost curves would be the same and a seller can break even at small scale.
So the price leader must set a price below its own average and marginal
cost (and incur a loss) to deter the potential competitor from entering the
market.
11. Sunk cost of capacity helps to implement a strategy of limit pricing. Looking
forward, unavoidable costs would be lower, and so a strategy of maintaining
a low price would be more credible.
12. Relatively larger. Your commitment to a larger capacity would push your
competitors to smaller capacity.
13. Because the production committed will affect the capacity set by other sellers.
This will in turn affect the price and profits.
14. Number of sellers, excess capacity, sunk costs, barriers to entry and exit, and
product homogeneity.
15. It depends. If the product is more homogeneous (less heterogeneous), each
seller will have a relatively elastic demand, therefore, it easily can sell more
than its quota. However, having a homogeneous product enables the cartel
to monitor the various sellers more easily.
Chapter 12. Externalities

12A. The new combined marginal profit contribution is lower by $3.20. The new
economically efficient level of customer traffic would be 60,000 customers a
month.
12B. (a) The restaurant’s marginal cost curve will be higher. (b) The number of
customers that maximizes the group profit contribution will be lower.
12C. The possible hurdle to resolving the externality is the assignment of rights.
Does the betting shop have the right to attract customers or does the restaurant
have the right not to be bothered by gamblers? (This situation involves only two
parties, betting shop and restaurant, so free riding is not a problem.)
12D. There are four differences: (a) Critical mass: in markets with network effects,
the demand may be zero unless the number of users exceeds critical mass.
(b) Expectations: in markets with network effects, user expectations are
important in demand. (c) Tipping: in markets with network effects, the
demand may be extremely sensitive to small differences among competitors.
(d) Price elasticity: in markets with network effects, the price elasticity may
differ according to whether the demand has reached critical mass.
12E. An additional marginal benefit curve would increase the vertical sum of
marginal benefits, and so possibly increase the economically efficient quan-
tity of fireworks.
12F. Consider, for instance, a musical composition. It is a public good. However,
it can be delivered in the form of a private good through copyright. The
ability to deliver a public good in the form of a private good enables the
commercial provision of the public good.
1. A positive externality directly conveys benefits to others. A negative exter-

nality directly imposes costs on others.
2. Luna’s economically efficient level of advertising is where the combined mar-
ginal profit contribution equals the combined marginal cost of advertising.
3. The combined marginal benefit curve from additional customers for all
retailers is the vertical sum of the retailers’ individual marginal benefit curves.
4. When the sum of the marginal costs from a negative externality exceeds the
marginal benefit to the source, there is an opportunity for profit by reducing
the externality. The amount that the victims would be willing to pay to
reduce the externality would exceed the amount that the source would
require in compensation to reduce the externality.
5. The new subway will generate positive externalities for the property around
the new stations as people benefit from convenient transportation. There-
fore, by purchasing the properties around the new stations, the subway com-
pany will capture the benefits that the new stations will generate.
6. Once the highway exit is built, it would be difficult to exclude any particular
business from benefiting from the additional traffic. The free-rider problem
is that each individual business would try to avoid contributing to the cost
of building the highway exit (knowing that it can still enjoy the benefit).
7. The benefit that one person obtains from using a particular spreadsheet pro-
gram increases with the number of other users. The more widespread is the
usage of that spreadsheet software, the more convenient it will be for users
to share their data and work.
8. [Omitted.]
9. It depends on whether the demand is below or above critical mass. If demand
is below critical mass, the demand will be zero and extremely price inelastic.
On the other hand, if demand is above critical mass, if there is a price reduc-
tion, this will increase the demand, so the demand would be elastic. If
demand far exceeds critical mass, any increase in demand will feed back
through the network effect to further increase the demand. Then the demand
would be inelastic to price.
10. Technical standards are relatively more important in markets with network
effects. Having common technical standards for different suppliers helps
those suppliers to jointly achieve critical mass in demand.
11. [Omitted.]
12. Only (c) is a public good.
13. The quantity of the public good available to each user does not decrease
with additional users.
14. Laws such as patents and copyright enable owners of scientific formulas or
artistic, literary, or musical expression (examples of public goods) to exclude
those who do not pay for using them. Technology may enable the delivery of
the public good to be excludable.
15. If the public good is being provided to one person, the marginal cost of
providing the same public good to additional users is zero. The cost of pro-
vision is fixed with respect to the number of consumers, and the marginal
cost of serving additional users is zero.
Chapter 13. Asymmetric Information

13A. Alice has imperfect information but does not face risk. The insurer has
imperfect information and faces risk.
13B. Relative to the market with 100,000 bottles of low-quality wine, the com-
bined supply would be closer to the supply curve of high-quality wine, and
the market price and quantity produced would be higher.
13C. If the expected demand curve crosses the combined supply curve at point c,
in the market equilibrium, all wine would be low quality. Hence, the expected
demand would drop to zero, and the market would fail.
13D. Appraisals will be more common in the market for more expensive
wine. The price of the wine would be more likely to cover the cost of
appraisal.
13E. Borrowers who are more willing to repay will be relatively more likely to post
collateral.
13F. This information would reduce every bidder’s uncertainty about the true
value of the vineyard. Hence, it would reduce the extent of the winner’s
curse.
13G. Screening is an initiative of the less-informed party, while signaling is an
initiative of the better-informed party.
1. Imperfect information is the absence of certain knowledge. Risk is uncer-

tainty about benefits or costs. A person can have imperfect information
about some possibility, but if that possibility does not affect her/his benefits
or costs, it does not impose any risk on her/him.
2. (a) No asymmetry. (b) The directors of Acquirer have better information
than the general investor.
3. Ming’s offer to pay a higher interest rate indicates that he is less likely to
repay the loan.
4. Two reasons: (a) Less risk. If a borrower defaults on a secured loan, the
creditor can seize and sell the item against which the loan is secured. (b) Less
adverse selection. The security serves to screen good from bad borrowers.
5. The fixed salary will draw relatively less hardworking workers than a piece
rate.
6. It is in the seller’s interest to conceal negative information about the car.
7. With less information about the younger drivers, the insurers would be less
able to screen between good and bad drivers among them. So insurers are
more vulnerable to adverse selection.
8. [Omitted.]
9. Drivers with a lower probability of accident will prefer a higher deductible
as they are less likely to make a claim.
10. Drivers whose value of time is less than the toll will not pay, and hence
will be screened away in favor of those whose value of time exceeds
the toll.
11. The winner’s curse would be more serious in (b). By contrast, in (a), the
study would reduce the uncertainty of every bidder, and so reduce the win-
ner’s curse.
12. In all auctions, the bidder should bid more conservatively. (a) In an auction
to sell, this means bidding higher. (b) In an auction to buy, this means bid-
ding lower.
13. [Omitted.]
14. Yes. It is relatively more costly for producers of poorer-quality software to
offer the full refund, because their customers are more likely to demand the
refund.
15. By accepting payment in Acquirer’s shares, the actual amount that target
receives will depend on Target’s true value (which Target knows but Acquirer
does not).
Chapter 14. Incentives and Organization

14A. (a) The economically efficient effort will be lower. (b) The effort that the
worker actually chooses will be lower at e.
14B. Draw any salesperson’s marginal compensation curve that crosses the sales-
person’s marginal cost curve at 250 units of effort a month.
14C. The salesperson’s marginal compensation curve is zero up to just below 200
units of effort a month, vertical at 200 units, and zero above 200 units.
14D. The commission on sales will reduce the sales clerks’ incentive to process
returns.
14E. Potential benefits and costs, and possible contingencies.
14F. Upstream vertical integration is the acquisition of assets for a stage of produc-
tion further from the final consumer. Downstream vertical integration is the
acquisition of assets for a stage of production closer to the final consumer.
14G. The potential to reduce holdup through detailed contracting; the potential to
resolve moral hazard through incentives and monitoring; the potential for
outsourcing to reduce internal monopoly power; and the extent of economies
of scale, scope, and experience.
1. [Omitted.]
2. It is costly for the insurer to monitor Leah’s precautions, so information is
asymmetric. Leah bears the cost of precautions but receives only part of the
benefit.
3. The party subject to moral hazard will choose the activity at the level where
their own marginal benefit equals own marginal cost. At that level, the sum
of marginal benefits to all parties differs from the sum of marginal costs to
all parties. So a profit can be made by adjusting the activity to the level where
the sum of marginal benefits equals the sum of marginal costs.
4. It can help to resolve moral hazard by linking compensation to some measure
of performance.
5. Your compensation will be affected by factors other than your performance,
such as the general state of the economy and competition.
6. Reward the drivers for repair and breakdown expenses below the average for
all drivers, and penalize them for expenses above the average.
7. The scheme will reduce the secretary’s incentive for effort in the other
tasks.
8. (b) Training on the airline’s flight management system is more specific.
9. Because the additional cost of preparing a complete contract outweighs the
potential benefit in avoidance of holdup.
10. (a) Shareholders have residual control (rights that have not been contracted
away). For instance, they may dismiss the current board of directors and man-
agement. (b) Shareholders also have the rights to residual income (income
remaining after the payment of all other claims). They receive dividends only
after all other claims, such as interest and debts, have been paid.
11. [Omitted.]
12. The potential for holdup in the future would lead one party to take precau-
tions such as writing more detailed contracts and avoiding specific invest-
ments. The precautions would increase its costs and reduce its revenues.
13. It should consider the disadvantages of internal production – moral hazard,
internal monopoly, and inefficient scale, scope and experience.
14. Outsourcing constrains the monopoly power of the internal supplier.
15. Economies of scope are one reason for horizontal integration: the organization
can produce a variety of products at lower cost than if each item were
produced separately. However, if the organization produces none of the items

for which there are economies of scope, then it should outsource all of them
rather than producing a subset.
Chapter 15. Regulation

15A. The optimal price would be the same (equal to the marginal cost). However,
the consumption would be larger.
15B. Under structural regulation, the regulator stipulates the conditions
under which a business may produce vertically related goods and services
(to separate the natural monopoly from the potentially competitive
market).
15C. In situations of asymmetric information, the regulator may be able to resolve
the asymmetry by regulating the disclosure of information, conduct, and
business structure of the better-informed party.
15D. If the user fee is $25 per ton, polluters would emit to a level where the
marginal benefit of emissions is less than the marginal cost. If the user fee is
$45 per ton, polluters would emit to a level where the marginal benefit of
emissions exceeds the marginal cost.
15E. The driver would choose a new level of care where her marginal benefit
equals marginal cost. It would be higher than the original level.
1. A market is a natural monopoly if the average cost of production is mini-

mized with a single supplier.
2. One source of inefficiency is that government-owned enterprises are prone
to be coopted by employees, so that the enterprise serves its employees rather
than its customers. Another source of inefficiency is that government-owned
enterprises may not get the optimal quantity of investment funds.
3. Marginal cost pricing means that a provider must set its price equal to
marginal cost and supply the quantity demanded.
4. Price regulation gives the regulated entity an incentive to exaggerate its
reported costs and so increase its actual profit.
5. Under rate-of-return regulation, the regulator stipulates the franchise holder’s
maximum allowed profit in terms of a maximum rate of return on the value
of the rate base.
6. Under rate-of-return regulation, the regulated entity has little incentive to
minimize costs and has an incentive to inflate the rate base.
7. Privatization means transferring ownership from the public to the private
sector. Allowing competition means removing an exclusive (monopoly)
right. A government-owned monopoly may be privatized without allowing
competition.
8. To increase economic efficiency, competition law prohibits: (a) collusion;
(b) abuse of monopoly power; and (c) mergers that create substantial market
power.
9. False. If the information cannot be objectively verified, disclosure will not

resolve the asymmetry.
10. Regulation of conduct and structure may prevent the party with superior
information from exploiting that advantage, and so resolve the information
asymmetry.
11. (a) The regulator could charge a user fee for noise generated by the construc-
tion equipment. (b) The regulator could set a standard and make it illegal
for noise from construction equipment to exceed the standard.
12. Set a standard for engines in terms of emissions (quantity per hour).
13. The toll can induce the user to balance the marginal cost of congestion
against her/his marginal benefit, achieving economic efficiency and resolv-
ing congestion.
14. Because the marginal cost of congestion varies by the time of day.
15. The government specifies the liability and compensation for accidents
through law. The potential source of harm will take care to reduce liability.
Index
Abbreviated New Drug Application airlines: advance booking and penalties

(ANDA) 168 317; frequent flyer programs 49;
academic externalities 287 pricing 317
accidents 372–3 airport congestion 119
accounting statements: opportunity costs AK-47 rifle, patent infringement 299
and 141; sunk costs and 146 Alaska Marine Highway System 78
accuracy, own-price elasticity of Albaugh, Jim 1, 2, 218
demand 47 Alenia 351
adjustment time 87–8, 108–13; demand All Nippon Airways 329, 330
increase 109–10; demand reduction Amazon 13; Prime service 43–4, 47
110–12; elasticity of demand 56–8; American football 238
long-run equilibrium 109; short and anchoring 7–8, 59
long run 112–13; short-run equilibrium antimonopoly see competition law
108–9 antitrust law see competition law
adverse selection 309–14; definition 311; Apotheker, Leo 323
demand and supply 309–11; economic Apple 286; product differentiation and
inefficiency 312; market equilibrium innovation 171
311–12; market failure 312–14 appraisal 315
advertising 179–80; advertising–sales ratio approved equity, regulation and 358
179–80; factor on demand 30; free-to-air Arkes, Hal 7
TV 299; positive externality 281; prod- assignment of rights 288
uct differentiation 170; profit- asymmetric information 305–28, 366–8;
maximizing 179; rule for expenditure about actions 331; adverse selection
180; spending on 182 309–14; appraisal 315; auctions 318–20;
advertising elasticity 54–5 conduct regulation 368; contingent
advertising elasticity of demand 179 contracts 322; definition 306; disclosure
advertising–revenue ratio see 367; imperfect information 307–9; not
advertising–sales ratio required for holdups 341; regulation and
advertising–sales ratio 179–80 franchise holders 361; screening 316–17;
agreement, resolving externalities 288 signaling 321; structural regulation 368
Aik Leong Plumbing Construction AT&T 268
320 atorvastatin 167–9, 170, 180, 298
air travel market 122–3; fare differences auctions 318–20; auction methods 318–19;
193–4 winner’s curse 319–20
air travel, taxes 130–2 Australia: price regulation 362; reducing
Airbus 1–3, 218, 238; A320neo 139–40, carbon 373
141; increasing production 156 Australian Competition and Consumer
airline landing fees 118–19 Commission (ACCC) 362
Index 401
automobiles see cars break even analysis 76–8, 82–3, 175–6

Autonomy plc 323 bridges, demand for 371
available production capacity 87 Brin, Sergy 286
average cost 70–2, 151–2 British Petroleum (BP) 333
average cost pricing 361 broadband service: economies of scale
average revenue per user (ARPU), China 154; natural monopoly 359; scope
Mobile 22, 28 economies 157
average value 5 broadcast TV, non-rival 297
Aviation Industry Corporation of Buffet, Warren 143
China 219 Building Construction (Singapore) 320
avoidable costs 140 business demand 31–2
Axelrod, Robert 240–1 business travelers, airline pricing 317
butter, demand for 106
backward induction 229–30 buyer cartels 266
balance 349–50 buyer commitments, own-price elasticity
banking 124–5; bank runs 235–7; 48–9
diseconomies of scope 158 buyer surplus 32–6; benefit and expendi-
Barker, R.P. 156 ture 32–3; definition 33; package deals
Barnes and Noble 13 34–5; price changes 33–4; total benefit
barriers to competition, market power 169 32; two-part pricing 35
barriers to entry and exit 184, 267 buyer’s price 130–2
Barry, P. 149 buyouts 219
batteries 350
Battle of the Bismarck Sea 222–3, 227 cable television market 99; scope
behavioral biases 130 economies 157–8
benchmark, economic efficiency 120–2; Caesars Entertainment 265
conditions for economic efficiency cannibalization 212–13
120–1; consumer sovereignty 121; capacity competition 259–63; demand and
technical efficiency 121 cost changes 261–2; strategic substitutes
benefit/cost economizing, own-price 262–3
elasticity 49 capacity leadership 263–5
Bertrand model of oligopoly 252 capacity utilization 198
best response function 254–5, carbon emissions 373
260, 261 Carlyle Group 330, 351
betting shops 284–5, 288 cars: repair shops 198
Big Pharma versus generics 176 cartel’s dilemma 221–2, 240
Bill and Melinda Gates Foundation 187 cartels 266–7
Blumer, Catherine 7 cement market 127–9
Boeing 1–3, 139–40, 218, 238, 329–31, 334, central planning, banks 125
351; Dreamliner 330–1, 334 charitable monopsony 187
Boggs, David 286 China Mobile 21–2; “Worldwide Connect”
Bombardier 2–3, 218, 232; CSeries 37, service 35; average revenue per user
139–40, 160–1 (ARPU) 22, 28; prepaid mobile
Bosack, Len 286 service 28
bounded rationality 7–8, 161; elasticity of Christie’s 318
demand 59; systematic biases 7–8, 59 Cisco Systems 286
branches, sequencing 228 Citigroup 158
brand extension 157–8 CityCenter 265
402 Index
Clorox 238–9 Countrywide Financial Corporation 305,

Coca-Cola 198; secrecy 172 316, 323
COMAC see Commercial Aircraft Cournot model of oligopoly 259, 262,
Corporation of China 273–4
Commercial Aircraft Corporation of credibility 237
China (COMAC) 2, 139, 218, 219, 224; credible signal 321
first-mover advantage 232 creditworthiness, appraisal and 315
commercial borrowers, appraisal 315 critical mass 291
commitments 148, 149 cross-price elasticity 54
common ownership 287–8 CSL 234
competition law 363–4 cumulative production 141, 159–60
competition, zero-sum game 227 Curado, Frederico 218, 238
competitive market 122–3 customer relationship management
complementary hardware 291 (CRM) 202
complete contract 342–3
complete price discrimination 198–202; Dampier–Bunbury pipeline 149
definition 200; information and resale decentralized management 124–6; internal
201; price discrimination 199–202; market 125–6; outsourcing 126
shortcomings of uniform pricing 198–9 decision-making 5–8; buyer and seller
compliments and substitutes, factors on biases 130; participation and extent 5–6
demand 28–30 decreasing returns to scale see disecono-
compliments, cross-price elasticity 54 mies of scale
conditional strategic move 234–8; Deepwater Horizon 333
promise 235–7; threat 237–8 Delhi Electricity Regulatory Commission
conduct regulation 368 358
congestible consumption 294 demand 21–42, 292–3; advertising 30;
congestion 371–2 business demand 31–2; butter 106;
consumer choice, China Mobile 22 buyer surplus 32–6; compliments and
consumer preferences 255 substitutes 28–30; demand curve 22–4;
consumer sovereignty 121 excess 102; for French products 106; and
content and delivery 297–8 income 25–8; increase 109–10, 177; indi-
contingent contracts 322 vidual 22–5; market demand 36–7; nor-
contract and ownership 345–6 mal and inferior products 27–8; price
coordination 227 elasticity of 104–5; reduction 110–12
copyright 170, 298 demand and cost changes 177–9, 196–7,
Cortés, Hernando 234 255–6, 261–2; demand change 177;
Cosmopolitan 265 fixed-cost change 178; marginal cost
cost accounting 143 change 177–8
cost including freight (CF) pricing demand and supply 101–2, 309–11; on
206–7 Valentine’s Day 107
cost-plus pricing 197 demand change 177
costs 139–66, 174; bounded rationality demand curve 22–4
161; economies of scale 150–6; econo- demand shift 106–7
mies of scope 156–8; experience curve demand–supply framework 97–8
159–61; opportunity 141–4; sunk 146–9; demand–supply model 12–13
transfer pricing 144–6; see also demand deposit insurance 235–7
and cost changes design, product differentiation 169
cotton market 12–13, 99 differentiated products 253–5, 262
Index 403
differentiating variables 317, 318 average costs 151–2; public good and
diminishing marginal benefit 24 280; strategic implications 153–4; sunk
diminishing marginal product 71 and fixed costs 154–5; tanker construc-
direct segment discrimination 194, 203–6; tion 155
heterogeneous segments 203–4; economies of scope 141, 156–8, 170–1,
homogeneous segments 203; 348–9; diseconomies of scope 158; joint
implementation 205 cost 156–7; strategic implications 157–8
disclosure 367 EE 257
discounting 8–9 elastic/inelastic demand 48–9; intuitive
discriminatory auction 319 factors 48–9
diseconomies of scale 153 elasticity 43–64; adjustment time 56–8;
diseconomies of scope 158 advertising 54–5; bounded rationality
Disneyland 290 59; cross-price 54; of demand or supply
distribution, product differentiation 170 169; elastic/inelastic demand 48–9;
Diversified Utility and Energy Trust forecasting 50–3; income 53–4; multiple
(DUET) 149 factors 55; own-price 45–7
dominated strategy 221 elasticity of supply 66, 86–8; intuitive
Doty, Elmer 330, 351 factors 87–8; price elasticity 86–7
downsizing 69, 314 electricity distribution 359
dry cleaning market 99 electricity industry: North Delhi 366;
dual agent 368–9 power investments 343
duopoly 252; Airbus and Boeing 218 Elpida 250
durable goods, elasticity of demand Embraer 2–3, 218–19, 224, 232, 238
57–8 employers, strategic thinking 237–8
Durham Furniture 65, 66, 84 enforcement, cartels 266–7
dynamic models 8 entry and exit 99–100
dynamic random-access memory (DRAM) Epic Energy 149
249–50 equal marginal benefit 120
equal marginal cost 120–1
e-commerce 13, 130 equilibria, solving 272–5; see also Nash
E.ON Ruhrgas 346 equilibrium
EBITDA 144 equilibrium change 103–4
economic efficiency 118–36, 201, 280–7, equilibrium strategy 230
295–6, 331; Adam Smith’s invisible hand equity capital 144
122–4; benchmark 120–2; decentralized estate brokerage 368–9
management 124–6; economizing 49; estimation, own-price elasticity of demand
general benchmark 286–7; incidence 45–6
129–30; intermediation 127–9; natural Ethernet, technology 286
monopoly 359; negative externality European Commission 365
284–5; positive externality 281–4; taxes excess capacity 87, 266
130–2 exchange rates 250–1
economic inefficiency 312, 331–3 excludability 296–9; content and delivery
economic profit 3–4, 77–8, 140 297–8; law 298; technology 298
economies of experience 170–1, 348–9 excludable consumption 297
economies of scale 150–6, 170–1, 293, expectations 291–2
348–9; diseconomies of scale 153; fixed expenditure: own-price elasticity
and variable costs in the long run 150–1; of demand 51; total benefit and
intuitive factors 152–3; marginal and 32–3
404 Index
experience curve 159–61, 238, 348; strate- frequent flyer programs 49

gic implications 160 Friis, Janus 11
extensive form, game theory 219 Frontline 97–8, 113
externalities 279–304, 369–73; accidents Fuchs, Victor 202
372–3; congestion 371–2; economic full capacity 145–6
efficiency 280–7; excludability 296–9; furniture industry 65–7, 84
network effects and 290–3; public goods furniture supply, United States 89
293–6; quota/standard 371; resolving
286, 287–90; in talent 286; user fee/tax gambling duopoly 185
369–71 game in extensive form 228–9, 237–8, 257,
extraneous factors, risk 339 263
game in strategic form 219, 220, 255
Facebook 286 game theory 219, 250
FAME see fatty acid methyl ester (FAME) Garber, Alan M. 202
Fannie Mae 305 gasoline prices 25, 30, 56, 58
fatty acid methyl ester (FAME) 73 gasoline stations, price war 225–6
Federal Aviation Administration 119, 124 Gasunie 346
Federal Aviation Administration Gazprom 345–6
(FAA) 119 GDF/Suez 346
Federal Communications Commission 268 Gillard, Julia 373
FICO score 305–6 Global Alliance for Vaccines and
fireworks 293, 297 Immunizations (Gavi) 187
first-mover advantage 219, 230–1, 257, 264 global mergers 365
fixed and variable costs 67–9, 140–1; in Gmail 122
long run 150–1 gold 98
fixed cost fallacy 161 Google, 122, 170, 286, 293
fixed costs: compared to variable 67–9; government ownership 359–60
definition 67; limit pricing and 259; and government subsidy 361
sunk costs 154–5 Great Recession 96, 106, 265
fixed wage 322 GRS Wood Products 66
fixed-cost change 178
flash memory industry 263
Foie Gras, demand for 106 Hatch-Waxman Act 168
forecasting, own-price elasticity of Hayward, Tony 333
demand 50–3 Heineken beer 198
franchises, price regulation and 361 Heinz Ketchup 205
Freddie Mac 305 Hermitage Museum 213
Fredriksen, John 97 heterogeneous segments 203–4
free entry and exit 99–100 Hewlett-Packard 286, 323
free lunch 143 Hockney, David 143
free on board (FOB) price 127, 128–9, holdup 330, 341–3, 346–7; incomplete
206–7 contracts 342–3; resolving by ownership
free online storage 122 344–6; specific investments 342
free riding 288–9; problem 297 Holloway, Peggy 265
free shipping, online retail 129 homogeneous product 98, 251–2
free to air TV 298, 299 homogeneous segments 203
freight-inclusive pricing 127–8 horizontal boundaries 10–11, 346
French products, demand for 106 horizontal integration 267
Index 405
Hotelling model 253, 255, 262, 274–5 installed base 291

HSBC 306, 307, 323 instant messaging service 290–1, 292
Hynix 263 insurance, risk and 308–9
Intel 286
intellectual property 170, 298
Icahn 239 interest payments 144
imperfect information 307–9; and asym- intermediation 119, 127–9; free on board
metric 307–8; risk 308; risk aversion pricing 128–9; freight-inclusive pricing
308–9 127–8
imperfect markets 14 internal market 125–6
incentive scheme, strength of 339 internal market power 347–8
incentives 334–8; monitoring 334–5; internal monopoly 348
performance pay 335–6; performance International Monetary Fund (IMF) 314
quota 336–7 internet retailing 13
incentives and organization 329–56; Internet, network externalities 280
holdup 341–3; incentives 334–8; moral intuitive factors: own-price elasticity 48–9
hazard 331–3; organizational architec- intuitive factors, supply elasticity: adjust-
ture 346–50; ownership 344–6; risk and ment time 87–8; available production
multiple responsibilities 339–40 capacity 87
incidence 129–30 invisible hand 122–3, 358, 369
income changes, effect on demand 25–7 Irkut 2, 139, 218, 232
income elasticity 53–4
income statements see accounting Jetstar Airlines 193
statements joint cost 156–7
incomplete contracts 342–3
increasing returns to scale see economies Kahneman, Daniel 7
of scale Karadeniz Holding AS 343
incremental margin 176 Kawasaki 351
incremental margin percentage 176, 179, Knetsch, Jack 7
184–5, 196
indirect segment discrimination 194, labor demand 79
208–10, 317, 318; definition 209; imple- Las Vegas Strip: evolution of
mentation 209–10; structured choice oligopoly 265
208–9 law 298
individual demand 22–5; constructing learning curve see experience curve
curve 22–4; marginal benefit 24; learning percentage 159–60
preferences 25 Lerner, Sandy 286
individual supply curve 78–9, 83 LG Chem 350
industry 12; structure 153–4 liability, accidents 372
inefficiency 119; see also economic life insurance 314
efficiency limit pricing 257–9
inferior product 27–8, 54 location: FOB and CF pricing 206–7;
Infineon 250 parallel imports 207
information and resale 201 long-run 67, 84, 88–9, 95, 112–13, 250;
information technology services 349–50 planning horizons 148
inframarginal units 173 long-run break-even 83
input demand 79 long-run costs 80–1
inputs, business demand 31, 32 long-run equilibrium 109
406 Index
long-run individual supply 80–4; break- medical services 366, 368

even analysis 82–3; individual supply Medtronic 341
curve 83; long-run costs 80–1; produc- memory industry 249–50, 263
tion rate 81–2; short and long run 84 Merck 167
long-run, elasticity of demand 56–8 mergers 365
Metcalf, Robert 286
MAN 365 method of delivery 297–8
managerial economics 1–3; definition 3 MGM Resorts 265
managerial performance, evaluating 341 Mickey Mouse externalities 290
marginal benefit 24 Micron 250, 263
marginal benefit equals marginal cost 121, Microsoft 11, 122, 170
283, 286, 360, 369 migrant workers 132
marginal benefit/cost 6 mineral water 98
marginal cost 70, 151–2, 197, 256 mini-anchors, VivoCity 289
marginal cost change 177–8 mini-bars 197–8
marginal cost pricing 360 mobile telecommunications market
marginal expenditure 186 257, 268
marginal product 71 monitoring 334–5
marginal revenue 75 monopoly 167–92, 272–3; advertising
marginal value 5–6 179–80, 182; definition 168; demand and
market demand curve 36–7, 41–2, 99 cost changes 177–9; market structure
market equilibrium 96–117, 311–12; 182–5; monopsony 185–7; more profit-
adjustment time 108–13; definition able than oligopoly 265–6; profit maxi-
100–1; demand and supply 101–2; mum 172–6; research and development
demand shift 106–7; excess demand 102; 181; sources of market power 169–72
excess supply 102; market equilibrium monopoly franchise 358, 364
100–2; perfect competition 98–100; monopoly production scale 174–5
supply shift 103–5 monopoly revenue 173–4
market failure 312–14 monopsony 168, 185–7, 364; benefit and
market power 13–14, 168; economies expenditure 186; charitable 187;
of scale, scope and experience 170–1; maximizing net benefit 187
intellectual property 170; measuring moral hazard 330, 331–3, 334, 347; asym-
184–5; product differentiation 169–70; metric information about actions 331;
regulation 171; small buyers and sellers degree of 333; economic inefficiency
99; sources of 169–72 331–3; resolving 334, 339; surgeons
market price system 123–4 subject to 366
market structure 182–5; effects of compe- mortgages 305–7
tition 182–3; measuring market power Mozilo, Angelo 306
184–5; potential competition 183–4 Mulally, Alan 329, 334, 351
market supply 88–9; constructing 94–5; multiple factors, effect on demand 55
properties 89; short and long run 88–9 multiple responsibilities 340
market supply curve 88, 99
markets 12–14; competitive 12–13; defini- NAND flash memory 249–50
tion 12; e-commerce 13; imperfect 14; narrow body jets 224
market power 13–14 Nash equilibrium 220–4, 230, 252, 253,
maximum return on equity 358 255, 257, 259, 260; definition 221–2;
McDonnell Douglas 330 non-equilibrium strategies 223–4; in
McNerney, James 330, 334 randomized strategies 225–6; solving
Index 407
equilibrium 222–3, 246–7; solving in organizational architecture 346–50; bal-

randomized strategies 246–7 ance 349–50; definition 330; economies
National University Hospital 206 of scale, scope and experience 348–9;
natural monopoly 359–62, 364; govern- holdup 346–7; internal market power
ment ownership 359–60; price regulation 347–8; moral hazard 347
360–1; rate-of-return regulation 361–2 organizational boundaries 10–11
negative externality 284–5, 312; congestion outsourcing 11, 126, 330–1, 348
371–2 overinvestment 362
net benefit maximum 187 overstaffing 359
net present value 6, 9 own-price elasticity 45–7; accuracy 47;
network effects and externalities 290–3; Amazon “Prime” 44, 47; definition 45;
critical mass 291; expectations 291–2; estimation 45–6; properties 46
Google 293; price elasticity 292–3; ownership 344–5; residual income 344;
tipping 292 vertical integration 345
network externalities 280; harnessing 293
New York Times 52 package deals, buyer surplus and 34–5
newspaper production 150–2 Page, Lawrence 286
niche marketing 154 parallel imports 207
Nissan 350 patents 170, 298, infringement 299
nodes 228 pay-option ARM loans 316
non-discriminatory auction 319 perfect competition 98–100, 182, 307; free
non-durable goods, elasticity of demand entry and exit 99–100; homogeneous
56–7 product 98; many small buyers 99;
non-monetary externalities 287 many small sellers 99; symmetric infor-
non-rival consumption 293, 294 mation 100
Nord Stream 346 perfectly competitive market 145
normal product 27–8, 54 perfectly contestable market 184
North Delhi Power Limited (NDPL) performance pay 335–6
357–8 performance quota 336–7
North European Gas Pipeline 345–6 Pfizer 167–9, 170, 180–1
piece-rate wage 322
O2 257 poison pills 238–9
observable measure of performance 335 pollution 369–71
off-net communications 234 Port Authority of New York 118–19
oil production 80, 96 positive externality 280, 281–4
oligopoly 249–76, 249–75; Bertrand model positive-sum game 227
252; capacity competition 259–63; potential competition 183–4
capacity leadership 263–5; definition potentially competitive market 363–5;
250; Hotelling model 253; limit pric- competition law 363–4; structural regu-
ing 257–9; price competition 251–6; lation 364–5
restraining competition 265–8; Stackel- Pragati Power Corporation 357–8
berg model 264 prenuptial agreements 317–18
on-net communications 234 price changes, buyer surplus and 33–4
online retail, free shipping 129 price competition 251–7; consumer pref-
open/sealed bidding 318–19 erences 255; demand and cost changes
opportunity cost 141–4; alternative courses 255–6; differentiated products 253–5;
of action 141–2; of capital 144; identify- homogeneous product 251–2; strategic
ing 142–3 complements 256; Vodafone 257
408 Index
price discrimination 50, 199–201, 202 randomized strategies 225–7; advantages

price elasticity: of demand 104–5, 195–6, of randomization 226; Nash equilibrium
292–3; of supply 86–7, 105; who is the in 225–6; supermarket pricing 227
customer 198 rate base 361, 362
price regulation 360–1, 362 rate-of-return regulation 361–2
price, elastic/inelastic 48 real estate agents 338
pricing 193–217; complete price dis- regulation 171, 357–77; asymmetric
crimination 198–201; direct segment information 366–8; externalities 369–73;
discrimination 203–5; indirect segment natural monopoly 359–62; potentially
discrimination 208–10; location 206–7; competitive market 363–5
select pricing policy 211–13; uniform relative performance incentives 339–40
194–8 Renault 350
pricing policy, selection of 211–13; repetition 239–40
cannibalization 212–13; technology research and development (R&D) 3, 176,
211–12 181, 182, 262, 263
pricing strategy, own-price elasticity and reserve price 318, 319
51–2 residential mortgages 14
privatization 360 residual control 344
product differentiation 169–70 residual demand 253–4, 255
production rate 81–2 residual income 344
production technology 72–3 resolving externalities 287–90; agreement
profit contribution 176 288; assignment of rights 288; common
profit maximizing 74–5, 82, 83 ownership 287–8; free riding 288–9;
profit maximizing price 195–6 VivoCity 289–90
profit maximizing scale 174–5 restraining competition 265–8; cartels 266;
profit maximizing transfer price 144–5 enforcement 266–7; horizontal integra-
profit maximum 172–6; break-even analy- tion 267
sis 175–6; costs 174; profit measures risk 308; aversion 308–9, 339; cost of 339;
176; profit-maximizing scale 174–5; and multiple responsibilities 339–40;
revenue 173 relative performance incentives
profit measures 176 339–40
promise 235–6 rival consumption 294
properties: of market supply curve 89; Roche 99
own-price elasticity of demand 46 Rosoboronexport 299
public goods 280, 293–6, 298; economic Rudd, Kevin 373
efficiency 295–6; non-rival consumption
293; rivalness 294–5 Salesforce.com 202
pure strategy 225 Samsung Electronics 249, 263
Scania 365
Qantas Group 193–4, 210 scientific knowledge 298
Qimonda 250 screening 316–17; differentiating variables
quantity demanded 50 317; self-selection 316–17
quota/standard 371 segment see direct segment discrimination
self-selection 316–17
R&D see research and development seller cartels 266
Ranbaxy Laboratories 168, 180–1 seller surplus 66, 84–6
randomization 225, 246, 335; advantages seller’s price, air travel tax 130–2
of 226 semiconductor industry 250
Index 409
sequencing 228–31; backward induction repetition 239–40; sequencing 228–31;

229–30; equilibrium strategy 230; strategic move 233; tit-for-tat 240–1
first-mover advantage 230–1; uncertain strategy 219
consequences 231 strikes 237
sewage collection, natural monopoly structural regulation 364–5, 368; in real
359 estate brokerage 368–9
shareholders 333 substitutes: cross-price elasticity 54;
short-run 84, 88–9, 94–5, 112–13, 250; own-price elasticity 48
planning horizons 148 sunk-cost fallacy 7, 59, 161
short-run break even 77 sunk-costs 76–7, 112, 130, 140, 146–9,
short-run costs 67–73; average cost 70–2; 266; alternative courses of action 146–7;
fixed and variable costs 69; marginal commitments 148; and fixed costs
cost 70; production technology 72–3; 154–5; identifying 147–8; ignoring 146;
total cost 69 planning horizon 148; strategic implica-
short-run equilibrium 108–9 tions 148–9
short-run individual supply 73–80; supermarket pricing 227
break-even analysis 76–8; individual supervision 335
supply curve 78–9; input demand 79; supply 65–95; elasticity of supply 86–8;
production rate 73–6 excess 102; long-run individual 80–4;
signaling 321 market 88–9; price elasticity of 105;
Silicon Valley 286 seller surplus 84–6; short-run costs
Sinden, Jack 7 67–73; short-run individual 73–80
Singapore, tax on foreign workers 132 supply curve, concept of 66
Sinopec 127 supply shift 103–5; equilibrium change
Skype 11 103–4; price elasticity of demand 104–5;
small buyers 99 price elasticity of supply 105
small scale production 154 symmetric information 100
Smith, Adam: invisible hand 122–3; systematic biases 7–8
invisible hand disabled 124
Sotheby’s 318 T-Mobile 268
specific investments 342 takeovers 219
sport-utility vehicles (SUVs) 30, 56 Tamiflu 99
Sprint 268 Tangs Department Store 279, 289
Stackelberg model of oligopoly 264 tanker industry 96–8, 101, 103; economies
Stanford University 286 of scale in construction 155; and interest
star anise, market for 99 rates
static models 8 Tata Power Company 357–8
statins 167–9, 180–1 Tattersall’s and Tabcorp 185
status quo bias 7, 161 taxes 130–2; buyer’s and seller’s price
Stonecipher, Harry 329–30, 334, 351 130–2; tax incidence 132
strategic complements 256–9, 262 technical efficiency 121
strategic move 233; in war 234 technical standards 291
strategic situation 219 technology, excludability and 298
strategic substitutes 262–3 text messages 234
strategic thinking 218–48; competition threat 237–8
or coordination 227–8; conditional Three 257
strategic move 234–8; Nash equilibrium timing 8–10
220–4; randomized strategies 225–7; tipping 292
410 Index
tit-for-tat strategy 240–1, 252 user fee/tax 369–71

Toshiba 321–2
total benefit, buyer surplus 32 Valentine’s Day 107
total cost 69 value-added 3–4, 119, 120
total service long-run incremental cost variable cost 67–8
(TSLRIC) 362 Verizon 268
trade secrecy 170 vertical boundaries 10, 346
trademarks 170 vertical integration 267, 330–1, 345
tradeoff 318 very large crude carriers (VLCCs) 96–7
transfer pricing 125, 144–6; definition 144; Victoria, Australia 185
full capacity 145–6; perfectly competi- Virginia Smith Trust 287
tive market 145; Sinopec 127 VivoCity 279–80, 289; advertising and
truck drivers: monitoring with onboard promotion 296; resolving externalities
computers 338 289–90
trucking service market 182–3 Vodafone, 4G pricing 257
trunk telecommunications, price regula- Volkswagen 365
tions 362 voluntary retirement 314
Tuas Sewers 320 Vought Aircraft Industries 330, 351
tunnels, demand for 371–2
Tversky, Amos 7 wage rate 79, 322
two-part pricing, buyer surplus and 35 Wall Street Journal Asia 208
Walt Disney Company 290
uncertain consequences 231 Watson Pharmaceuticals 168
unconditional strategic move 235 weak incentives 338, 340
uniform pricing 194–8; common miscon- winner’s curse 319–20
ceptions 197–8; definition 195; demand Wintershall 346
and cost changes 196–7; price elasticity wireless telecommunications providers 155
195; profit-maximizing price 195–6; Woods, Tiger 318
shortcomings 198–9
unions 219–20, 237 Xerox Palo Alto Research Centre (PARC)
Unites States, furniture supply 89 286
University of California 287 Xiang Yu 234
unobservable action 335
US mobile telecommunications 268 Zennstrom, Niklas 11
user expectations 291–2 zero-sum game 227

Managerial Economics, 5th Edition PDF

Uploaded by

Copyright:

Available Formats

Managerial Economics, 5th Edition PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Managerial Economics, 5th Edition PDF

Uploaded by

Copyright:

Available Formats

What are some of the features of the 5th edition of the book?

What are some of the features of the 5th edition of the book?

What do professors say about using this book in their classes?

What do professors say about using this book in their classes?

Managerial Economics

• an introductory chapter emphasizing decision-making and behavioral biases,

As always, the text presents the key concepts of microeconomics intuitively,

Ivan Png is a Distinguished Professor in the School of Business and Department

“Ivan Png’s Managerial Economics has made my life easy as a lecturer! It is a

ISBN: 978-1-138-81025-9 (hbk)

Part I Competitive Markets 19

Part II Market Power 137

Part III Imperfect Markets 277

14 Incentives and Organization 329

1.1 Value added 4

7.3 Economies of scale 152

15.3 Economic efficiency in emissions 370

2.1 Individual demand 23

Managerial economics is the science of directing scarce resources in the manage-

• emphasizes simple, practical ideas,

This book is aimed at business students as well as practitioners. Accordingly, it

• Simple, practical ideas for business decision-making

Online support for this book can be found at https://sites.google.com/site/pngecon/.

Ivan Png is a Distinguished Professor in the School of Business and Department

1. What Is Managerial Economics?

Value added = Buyer benefit − Seller cost

FIGURE 1.1 Value added.

The two fundamental decisions in business can be stated simply as participa-

Which and How Much?

• in decisions on participation – which market, which product, which job – the

• Sunk-cost fallacy. The psychologists Hal Arkes and Catherine Blumer

to one group of students and the problem

Managerial economics analysis includes two types of models. Static models

Net Present Value

$50,000 $95,000 $95,000 $95,000

So, Max would get a higher NPV from the MBA.

With regard to vertical boundaries, a cable TV provider that produces movies is

NEW BUSINESS ORGANIZATION: PEER-TO-PEER

Some of the fastest-growing businesses challenge conventional thinking

One concept of managerial economics – the market – is so fundamental that it

THE EXTENT OF E-COMMERCE MARKETS

A conventional bricks-and-mortar retail store serves a geographical area

• Managerial economics is the science of cost-effective management of scarce

1. Explain the difference between value added and economic profit.

1. Mercy Hospital provides healthcare at subsidized prices and is so popular

4. In each of the following instances, discuss whether horizontal or vertical

You are the consultant!

Table 2.1 Individual demand

Price ($ per movie) Quantity (movies per month)

FIGURE 2.1 Individual demand curve.

GASOLINE PRICES AND DRIVING

Brittany and Danny Schulz live in Murfreesboro, Tennessee. Early in November

3. Demand and Income

Table 2.2 Individual demand with lower income

20 Demand curve with

Demand curve with

FIGURE 2.2 Individual demand curve with lower income.

Normal and Inferior Products

CHINA MOBILE: GROWING SUBSCRIBERS, SHRINKING ARPU