1.

1
What is Economics?

1 Introduction to
Microeconomics

Chapter objectives:
1. Define economics.

2. Understand the concepts of opportunity cost, scarcity and rational

choice

3. Make clear the difference between microeconomic and macroeconomic

concerns.

4. Distinguish between positive economics and normative economics.

5. Explain the value of the ceteris paribus assumption within the context of
economic modeling.

1
 1
Introduction to Microeconomics

1.1 What is Economics?

Many definitions In economic history, there were many definitions and disputes on econom-
on economics ics. Many economists have tried to give the right definition about their sci-
ence.

 Thomas Carlyle (1795 - 1881): Economics - “the dismal science”

 W. Stanley Jevons (1835 - 1882): Economics - "the mechanics of utility and
self interest."

 Alfred Marshall (1842 - 1924): Economics - "a study of mankind in the ordi-
nary business of life"

 Lionel Robbins (1898 - 1984): Economics – “a science which studies human

behavior as a relationship between ends and scarce means which have alternative
uses."

Though the exact wording differs from author to author, the standard defini-
tion is now something like this:

Economics is the social science that examines how people choose to use limited or
scarce resources in attempting to satisfy their unlimited wants.

study of rational The key word in this definition is “choose”. Economics studies how we
choice choose to use the existing resources that nature and previous generations
have provided to best satisfy society’s unlimited wants. In a sense, econom-
ics is the “scientific study of rational choice.”

Scarcity Resources are said to be scare because their supply is limited. Scarcity re-
quires choice. People must choose which of their desires they will satisfy and
which they will leave unsatisfied. When we choose more of something, scar-
city forces us to take less of something else. Economics is sometimes called
the study of scarcity because economic activity would not exist if scarcity
did not force people to make choices.

Opportunity The resources that we are talking about here could be labor, capital, and raw
Cost materials. That they are scarce means there are not enough resources to
produce everything we want. To get more of one thing, one has to give up
something else. If you, e.g., want to sleep an extra hour, it is impossible to do
so without giving up something else, such as an hour of studying. There is,
consequently, a sort of a hidden cost to sleeping longer. This type of cost is
called opportunity cost (or alternative cost).

No free lunch A classical saying in economics is that “there is no such thing as a free
lunch”. This means that, even if you do not actually pay for the lunch, you
always have give up at least the time when you could have done something
else. That is, you always have to pay the opportunity cost. Opportunity cost

2
 1.1
What is Economics?

is the best alternative that we forgo, or give up, when we make a choice or a
decision. Nearly all decisions involve trade-offs. The opportunity cost of the
trip to the zoo is the value you attach to that one activity you would other-
wise have chosen.

The subject matters of economics have been usually studied under two
broad branches: Microeconomics and Macroeconomics.

 Microeconomics studies the economic behavior of individual economic Microeconomics

decision makers, such as a consumer, a worker, a firm, or a manager in
the markets for different goods and services and tries to figure out how
prices and quantities of goods and services are determined through the
interaction of individuals in these markets. Another question addressed
by microeconomics is the allocation of goods and services, i.e. who gets
the products that are produced and why does poverty exist?

 Macroeconomics on the other hand analyzes how an entire national Macroeconomics

economy performs, focuses on aggregate measures such as total output,
employment and aggregate price level instead of trying to understand
what determines the output of a single firm or the consumption patterns
of a single household. Macroeconomics deals with broad economic vari-
ables such as national production, total consumer spending, and overall
price movements.

While these two studies of economics appear to be different, they are actual-
ly interdependent and complement one another since there are many over-
lapping issues between the two fields. For example, increased inflation
(macro effect) would cause the price of raw materials to increase for compa-
nies and in turn affect the end product's price (micro effect) charged to the
public.

Economics also contains a number of subfields, such as international eco- Subfields

nomics, labor economics, and industrial organization. of economics

 International economics studies trade flows among countries and inter-

national financial institutions. What ware the advantages and disad-
vantages of trade? Why is the domestic currency strong or weak?

 Labor economics deals with the factors that determine wages, employ-
ment and unemployment. Which effects do taxes on the decision wheth-
er to work, how much to work and at what kind of job? Which roles do
the unions play?

 Industrial organization analyzes the structure and performance of indus-

tries and firms within a economy. How do firms compete? Who are the
winners and the losers of such competition?

3
 1
Introduction to Microeconomics

Figure 1-1 Microeconomic and macroeconomic questions

Production Prices Income Employment

Microeconomics Production/ Out- Price of Indi- Distribution of Employment

put in single vidual Goods Income and by single firms
firms and Services Wealth Jobs in the car
How much raw Price of medical Wages in the industry
material? care auto industry Number of
How many offic- Price of gaso- Minimum employees in a
es? line wages firm
How many cars Food prices Poverty
Apartment
rents
Macroeconomics National Produc- Aggregate National In- Total employ-
tion/Output Price Level come ment of coun-
Total Industrial Consumer Total wages try
Output Price Level and salaries Total number
Gross Domestic Producer Price Total corporate of jobs
Product Level profits Unemploy-
Growth of Out- Rate of Infla- ment rate
put tion

Economists classify issues as either positive or normative.

Positive  Positive questions explore the behavior of the economy and its partici-
Economics pants without judging whether the behavior is good or bad. For example,
an economist engaged in positive analysis might investigate why and
how the American health care industry uses the quantities of capital, la-
bor, and land that are currently devoted to providing medical services.
Positive economics collects data that describe economic phenomena (de-
scriptive economics) and constructs testable (cause-and-effect) theories
to explain the phenomena (economic theory). Positive economic state-
ments can be verified by empirical research.

Normative  Normative economic questions evaluate the results of behavior and ex-
Economics plore whether the outcomes might be improved. When economists advo-
cate that more resources should be allocated to health, they have implic-
itly moved into normative analysis. In the normative economics, we need
a judgment about what is good or bad for the society. Why is it better for
the society when people get more children? Which tax rates should be
raised to avoid the extent of pollution?

4
 1.2
Economic Models

1.2 Economic Models

A model is a formal statement of a theory. Models are descriptions of the Formal state-
relationship between two or more variables (for example the dependence ment of a theory
between income and consumption or tax and unemployment). The number
of economic models in current use is, of course, very large.

Specific assumptions used and the degree of detail provided vary greatly
depending on the problem being addressed. The types of models employed
to explain the overall level of economic activity in the United States, for
example, must be considerably more aggregated and complex than those
that seek to interpret the pricing of Arizona strawberries.

Despite this variety, however, practically all economic models incorporate

three common elements: (1) the ceteris paribus (other things the same) as-
sumption; (2) the supposition that economic decision-makers seek to opti-
mize something; and (3) the comparative statics method.

(1) The Ceteris Paribus Assumption

As is the case in most sciences, models used in economics attempt to portray Simple
relatively simple relationships. Economists (and other scientists) construct relationships
models – formal statements of relationships between variables of interest –
that simplify and abstract from reality. Graphs, words, or equations can be
used to express a model. In testing the relationships between variables with-
in a model it is convenient to assume ceteris paribus, that all other variables
have been held constant.

A model of the market for wheat, for example, might seek to explain wheat
prices with a small number of quantifiable variables, such as wages of farm
workers, rainfall, and consumer incomes. This parsimony in model specifica-
tion permits the study of wheat pricing in a simplified setting in which it is
possible to understand how the specific forces operate. Although any re-
searcher will recognize that many "outside" forces (presence of wheat dis-
eases, changes in the prices of fertilizers or of tractors, or shifts in consumer
attitudes about eating bread) affect the price of wheat, these other forces are
held constant in the construction of the model.

It is important to recognize that economists are not assuming that other

factors do not affect wheat prices, but rather, such other variables are as-
sumed to be unchanged during the period of study. In this way the effect of
only a few forces can be studied in a simplified setting. Such ceteris paribus
(other things equal) assumptions are used in all economic modeling.

5
 1
Introduction to Microeconomics

Difficulties for Use of the ceteris paribus assumption does pose some difficulties for the
empirical verifi- empirical verification of economic models from real-world data. With a few
cation notable exceptions, economists have not been able to conduct controlled
experiments to test their models. Instead, economists have been forced to
rely on various statistical methods to control for other forces when testing
their theories. Although these statistical methods are in principle as valid as
the controlled experiment methods used by other scientists, in practice they
raise a number of thorny issues. For that reason, the limitations and precise
meaning of the ceteris paribus assumption in economics are subject to
somewhat greater controversy than in the laboratory sciences.

(2) Optimization Assumptions

Optimizing Many economic models start from the assumption that the economic actors
behavior being studied are rationally pursuing some goal. For example, consumers
maximize their own wellbeing (utility), firms minimize costs, and govern-
ment regulators attempt to maximize public welfare. Although, as we will
show, all of these assumptions are somewhat controversial, all have won
widespread acceptance as good starting places for developing economic
models. There seem to be two reasons for this acceptance.

 First, the optimization assumptions are very useful for generating pre-
cise, solvable models. A primary reason for this is that such models can
draw on a variety of mathematical techniques suitable for optimization
problems.

 A second reason for the popularity of optimization models concerns their

apparent empirical validity. As some empirical research shows, such op-
timization models seem to be fairly good at explaining reality.

In all, optimization models have come to occupy a prominent position in

modern economic theory.

(3) The comparative statics method

How do things in the model change as things in the environment around it

change? The comparative statics method provides a way of dealing with this
issue. It is built on the concept of equilibrium and focuses on the relationship
between the equilibrium itself and one or more key parameters.

It is not a description of a process but is more like a set of snapshots of dif-

ferent instances of equilibrium that leave a trace of a process. The compara-
tive statics method is sometimes incorporated into specific relationships that
are used as a short hand to characterize the behavior of an economic agent.

6
 1.3
Learning by doing

The example of this is demand and supply functions - collectively referred to

as response functions.

1.3 Learning by doing

1. Which one of the following best describes the study of economics? Eco-
nomics analyzes

(a) how businesses can make profits.

(b) how society uses its scarce resources to satisfy its unlimited desires.

(c) how the government controls the economy and how people earn a liv-
ing.

(d) the allocation of income among different sectors of the economy.

2. Your opportunity cost of attending college does not include

(a) the money you spend on travelling between home and college.

(b) the money you spend on meals while at college.

(c) your tuition fees.

(d) the income you could have earned if you’d been employed full-time.

3. Macroeconomics approaches the study of economics from the perspec-

tive of

(a) individual consumers.

(b) the government.

(c) the entire economy.

(d) the operation of specific markets.

4. Microeconomics approaches the study of economics from the viewpoint

of

(a) the entire economy.

7
1
Introduction to Microeconomics

(b) the government.

(c) the operation of specific markets.

(d) the stock market.

5. Which of the following is most appropriately a microeconomic issue?

(a) The study of the relationship between the unemployment rate and the
inflation rate

(b) The determination of total output in the economy

(c) The forces determining the price level in an individual market

(d) The aggregate behavior of all decision-making units in the economy

6. Which of the following statements is true?

(a) Microeconomics studies consumer behavior, whereas macroeconom-

ics studies producer behavior.

(b) Microeconomics studies producer behavior, whereas macroeconomics

studies consumer behavior.

(c) Microeconomics studies behavior of individual households and firms,

whereas macroeconomics studies national aggregates.

(d) Microeconomics studies inflation and opportunity costs, whereas

macro-economics studies unemployment and sunk costs.

7. A difference between positive statements and normative statements is

that

(a) only positive statements are subject to empirical verification.

(b) positive statements are true by definition.

(c) economists use positive statements and politicians use normative

statements when discussing economic matters.

(d) positive statements require value judgments.

8. “An increase in the price of shampoo will cause less shampoo to be de-
manded, ceteris paribus.” Ceteris paribus means that

8
 1.3
Learning by doing

(a) there is a negative relationship between the price and quantity de-
manded of shampoo.

(b) the price of shampoo is the only factor that can affect the amount of
shampoo demanded.

(c) the price of shampoo is equal for all buyers.

(d) other factors may affect the amount of shampoo demanded but that
these are assumed not to change in this analysis.

9. The ceteris paribus assumption is used to

(a) make economic theory more realistic.

(b) make economic analysis more realistic.

(c) avoid the fallacy of composition.

(d) focus the analysis on the effect of a single factor.

10. Local farmers reduce the price of their tomatoes at the farmers’ market.
The price of corn is $30 per ear. A passing economist theorizes that, ceter-
is paribus, buyers will purchase more tomatoes than before. Which of the
following is TRUE? The economist is

(a) implying that the price of tomatoes will fall even further.

(b) assuming that the price of corn will remain at $30 per ear.

(c) assuming that tomatoes are of a better quality than before.

(d) implying that corn is of a poorer quality than before.

9
2
Demand, Supply and Market Equilibrium

2 Demand, Supply and Market

Equilibrium

Chapter objectives:
1. Define and apply quantity demanded and quantity supplied, and state
the law of demand and the law of supply.

2. Identify the determinants of demand and supply and indicate how each
must change for demand and supply to increase or decrease.

3. Determine equilibrium price and quantity and detail the process by

which the market moves from one equilibrium situation to another when
demand or supply shifts.

4. Explain concepts of elasticities.

5. Define excess demand (shortage) and excess supply (surplus) and pre-
dict their effects on the existing price level.

10
 2.1
Demand Curve

2.1 Demand Curve

Quantity demanded is the amount of a product that a household would buy, Quantity
in a given period, if it could buy all it wanted at the current price. The quan- demanded
tity demanded of a good usually is a strong function of its price. Suppose an
experiment is run to determine the quantity demanded of a particular prod-
uct at different price levels, holding everything else constant, ceteris paribus.

The relation between price and quantity demanded is described by the de- Demand curve
mand curve. By convention, the demand curve displays quantity demanded
as the independent variable (the x axis) and price as the dependent variable
(the y axis), as shown in Figure 2-1.

Demand Curve Figure 2-1

Price

Quantity
0 demanded

From Figure 2-1, we can see that the demand curve has a downward slope. It
means if price of good increases, quantity demanded will decrease, which is
called the law of demand.

The law of demand states that there is a negative relationship between the The law of de-
price and the quantity demanded of a product. In other words, quantity mand
demanded moves in the opposite direction of price (all other things held
constant), and this effect is observed in the downward slope of the demand
curve.

For basic analysis, the demand curve often is approximated as a straight line.
A demand function can be written to describe the demand curve. Let’s call
quantity demanded is QD, and price is P. Demand functions for a straight-
line demand curve take the following form:

11
 2
Demand, Supply and Market Equilibrium

QD = f(P) or

QD = a + bP

where a and b are constants that must be determined for each particular
demand curve. Value of b is negative because when price increases, the re-
sult is a decrease in quantity demanded as one moves along the demand
curve.

Shifts in the Demand Curve

When there is a change in an influencing factor other than price such as

above factors, there may be a shift in the demand curve to the left or to the
right, as the quantity demanded increases or decreases at a given price. For
example, if there is a positive news report about the product, the quantity
demanded at each price may increase, as demonstrated by the demand
curve shifting to the right, from QD to QD’:

Figure 2-2 Demand Curve Shift

Price

QD QD ’
Quantity
0 demanded

Some demand- A number of factors may influence the demand for a product, and changes
shifting factors in one or more of those factors may cause a shift in the demand curve. Some
of these demand-shifting factors are:

 Customer preference
 Prices of related goods

12
 2.2
Supply Curve

 Complements - an increase in the price of a complement reduces

demand, shifting the demand curve to the left.

 Substitutes - an increase in the price of a substitute product increas-

es demand, shifting the demand curve to the right.

 Income - an increase in income shifts the demand curve of normal goods

to the right.

 Number of potential buyers - an increase in population or market size

shifts the demand curve to the right.

 Expectations of a price change - a news report predicting higher prices in

the future can increase the current demand as customers increase the
quantity they purchase in anticipation of the price change.

2.2 Supply Curve

Quantity supplied is the amount of a product that a firm would be willing Quantity
and able to offer for sale at a particular price during a given time period. supplied
Price usually is a major determinant in the quantity supplied. Graphically,
Supply curve
the supply curve is shown in Figure 2-3, below.

Supply Curve Figure 2-3

Price

Quantity
0 supplied

As with the demand curve, the convention of the supply curve is to display
quantity supplied on the x-axis as the independent variable and price on the
y-axis as the dependent variable.

13
2
Demand, Supply and Market Equilibrium

The law of The law of supply states that there is a positive relationship between the price
supply and the quantity supplied of a product. In other words, the higher the price,
the larger the quantity supplied, all other things constant. The law of supply
is demonstrated by the upward slope of the supply curve.

The supply curve often is approximated as a straight line to simplify analy-

sis. A straight-line supply function would have the following structure:

QS = f (P) or

QS = a + bP

where QS is quantity supplied, a and b are constant for each supply curve.
Value of b is positive because when price increases, the result is a increase in
quantity supplied as one moves along the supply curve.

Shifts in the Supply Curve

While changes in price result in movement along the supply curve, changes
in other relevant factors cause a shift in supply, that is, a shift of the supply
curve to the left or right. Such a shift results in a change in quantity supplied
for a given price level. If the change causes an increase in the quantity sup-
plied at each price, the supply curve would shift to the right:

Figure 2-4 Supply Curve Shift

Price
QS QS’

Quantity
0 supplied

There are several factors that may cause a shift in a good's supply curve.
Some supply-shifting factors include:

14
 2.3
Market Equilibrium

 Prices of other goods - the supply of one good may decrease if the price Some supply
of another good increases, causing producers to reallocate resources to shifting factors
produce larger quantities of the more profitable good.

 Number of sellers - more sellers result in more supply, shifting the sup-
ply curve to the right.

 Prices of relevant inputs - if the cost of resources used to produce a good

increase, sellers will be less inclined to supply the same quantity at a giv-
en price, and the supply curve will shift to the left.

 Technology - technological advances that increase production efficiency

shift the supply curve to the right.

 Expectations - if sellers expect prices to increase, they may decrease the

quantity currently supplied at a given price in order to be able to supply
more when the price increases, resulting in a supply curve shift to the
left.

2.3 Market Equilibrium

Consumer-buyers and producer-sellers make decisions independent of each
other, but markets coordinate their choices and influence their actions. Let’s
combine the market demand and supply curves on one diagram to see this
co-ordination:

Supply and Demand Figure 2-5

Price QS
Excess supply

E
P*

Excess demand QD
Quantity
0 Q*

15
 2
Demand, Supply and Market Equilibrium

Equilibrium At a price of P*, the quantity demanded is the same as the quantity supplied
point at Q*. At this price, the amount that consumers wish to buy is the same as
the amount that producers wish to sell. This price is called the equilibrium
price and the quantity being bought and sold is called the equilibrium
quantity. The point of equilibrium, E, can be seen on Figure 2-5 at the point
where the demand and supply curves cross.

Excess supply At price levels above P*, the quantity that producers wish to supply is great-
er than the quantity consumers wish to buy. There is excess supply (surplus)
Excess demand
and the market is a buyer’s market. At prices less than P*, consumers wish to
buy more than producers wish to supply. There is excess demand (shortage)
and the market is a seller’s market.

As we have learned, a higher price will reduce the quantity of a good de-
manded by consumers. On the other hand, a higher price will increase the
quantity of a good supplied by producers. The market price of a good will
tend to change in a direction that will bring the quantity of a good consum-
ers want to buy into balance with the quantity producers want to sell. If the
price is too high, the quantity supplied by producers will exceed the quanti-
ty demanded. They will be unable to sell as much as they would like unless
they reduce their price. Alternatively, if the price is too low, the quantity
demanded by consumers will exceed the quantity supplied. Some consum-
ers will be unable to get as much as they would like, unless they are willing
to pay a higher price to bid some of the good away from other potential
customers. Thus, there will be a tendency for the price in a market to move
toward the price that brings the two into balance.

Case Study 2-1 How to Find the Equilibrium Point Mathematically

Problem

Supply and demand can be written as mathematical functions, and in simple

examples, they are often straight lines.

QD = 185 – 20P

QS = 85 + 30P

What are equilibrium price and equilibrium quantity?

Solution

We now want to find the price, P*, that makes QD = QS. If the left-hand sides
above are equal, the right-hand sides must also be so. Therefore, set the
right-hand sides equal to each other:

185 – 20P = 85 + 30P or

16
 2.4
Price Elasticity of Demand

50P = 100

The result is that:

P = 2 = P*

We can substitute the result of P into either supply or demand function

above to know the equilibrium quantity, Q*:

QD = 185 – 20P = 185 – 20x2 = 145 = Q*

QS = 85 + 30P = 85 + 30x2 = 145 = Q*

Consequently, we have the equilibrium price, P* = 2, and the equilibrium

quantity, Q* = 145.

2.4 Price Elasticity of Demand

The price elasticity of demand measures the responsiveness of quantity
demanded to a change in price, with all other factors held constant.

The price elasticity of demand,  Q ,P , is defined as the magnitude of:

Q
x100%
Q
 Q ,P  or
P
x100%
P
Q P
 Q ,P 
P Q

Since the quantity demanded decreases when the price increases, this ratio is
negative; however, the absolute value usually is taken and  Q ,P is reported as
a positive number.

Elastic versus Inelastic

 If the  Q ,P is greater than one, the good is elastic. Demand is responsive to Characteristics of
a change in price. If for example a 15% fall in price leads to a 30% in- the price elastici-
crease in quantity demanded, the price elasticity = 2.0, as shown in Fig- ty of demand
ure 2-6 (a).

17
2
Demand, Supply and Market Equilibrium

 If the  Q ,P is less than one, the good is inelastic. Demand is not very re-
sponsive to changes in price. If for example a 20% increase in price leads
to a 5% fall in quantity demanded, the price elasticity = 0.25, as shown in
Figure 2-6 (b).

 If the  Q ,P is equal to one, the good has unit elasticity. The percentage
change in quantity demanded is equal to the percentage change in price.
Demand changes proportionately to a price change, as shown in Figure
2-6 (c).

 If the  Q ,P is equal to zero, the good is perfectly inelastic. A change in price

will have no influence on quantity demanded. The demand curve for
such a product will be vertical, as shown in Figure 2-6 (d).

 If the  Q ,P is infinity, the good is perfectly elastic. Any change in price will
see quantity demanded fall to zero. This demand curve is associated with
firms operating in perfectly competitive markets, as shown in Figure 2-6
(e).

Figure 2-6 Shapes of The Price Elasticity of Demand

Price

P1
(a) The good
is elastic P2
QD
∆Q > ∆P

Quantity
0 Q1 Q2

Price

(b) The good P1

is inelastic
P2
∆Q < ∆P
QD
Quantity
0 Q1 Q2

18
 2.4
Price Elasticity of Demand

Price

P1

(c) The good has

P2
unit elasticity
QD
∆Q = ∆P
Quantity
0 Q1 Q2

Price QD

P1
(d) The good is
perfectly inelastic
P2
∆P>0, ∆Q=0

Quantity
0 Q

Price

(e) The good is P QD

perfectly elastic

∆P=0, ∆Q>0
Quantity
0 Q1 Q2

Price Elasticity of Demand Case Study 2-2

Problem

Suppose price is initially $5, and the corresponding quantity demanded is

1000 units. Suppose, too, that if the price rises to $5.75, the quantity de-

19
 2
Demand, Supply and Market Equilibrium

manded will fall to 800 units. What is the price elasticity of demand over this
region of the demand curve? Is demand elastic or inelastic?

Solution

In this case, ∆P = 5.75 – 5 = $0.75, and ∆Q = 800 – 1000 = -200, so

Q P 200 $5
 Q ,P    1.33
P Q $0.75 1000

Thus, over the range of prices between $5 and $5.75, quantity demanded
falls at a rate of 1.33 percent for every 1 percent increase in price. Because the
price elasticity of demand is between -1 and -∞, demand is elastic.

2.5 Other Elasticities

There are two common elasticities in addition to the price elasticity of de-
mand that are the income elasticity of demand and the cross-price elasticity
of demand.

Income elasticity Income elasticity of demand measures the relationship between a change in
of demand quantity demanded and a change in income. The basic formula for calculat-
ing the coefficient of income elasticity is:
Q
x100%
Q
 Q ,I  or
I
x100%
I
Q I
 Q ,I 
I Q

The income elasticity of demand for goods depends on the nature of the
goods. These are normal goods or inferior goods.

Normal goods  Normal goods have a positive income elasticity of demand so as income
rise more is demand at each price level. We make a distinction between
normal necessities and normal luxuries (both have a positive coefficient
of income elasticity). Necessities have an income elasticity of demand of
between 0 and +1. Demand rises with income, but less than proportion-
ately, such as food, toothpaste or newspapers. Luxuries on the other
hand are said to have an income elasticity of demand > +1. Demand rises

20
 2.5
Other Elasticities

more than proportionate to a change in income, such as international air

travel, antique furniture or designer clothes.

 Inferior goods have a negative income elasticity of demand. Demand Inferior goods
falls as income rises, such as bus travel, bicycle or canned goods.

Cross-price elasticity, often simply called just cross-elasticity, measures Cross-price elas-
whether goods are substitutes or complements. It looks at the response of ticity
people in buying one product when the price of another product changes.
The formula for cross-price elasticity is:
Q i
x100%
Qi
 Qi ,Pj  or
Pj
x100%
Pj

Q i Pj
 Qi ,Pj 
Pj Q i

Where Pj denotes the initial price of good j, and Qi denotes the initial quanti-
ty of good i demanded.

If goods are complements, cross-price elasticity will be negative. For example, Complementary
if the price of gasoline rises, the sales of large cars will decline. The positive goods
change in the denominator (bottom) is matched with a negative change in
the numerator (top) of the equation. The result is therefore negative. If goods
are substitutes, cross-price elasticity will be positive. For example, sales of Substitutive
Coke will fall if the price of Pepsi falls because some Coke drinkers will goods
switch from Coke to Pepsi.

Price Elasticity of Supply

The price elasticity of supply measures the sensitivity of quantity supplied Price elasticity of
QS to price. The price elasticity of supply, denoted by  Qs ,P , tells us the per- supply
centage change in quantity supplied for each percent change in price:
Q S
x100%
QS
 Qs ,P  or
P
x100%
P
Q S P
 Qs ,P 
P Q S

This formula applies to both the firm level and the market level. The firm
level price elasticity of supply tells us the sensitivity of an individual firm’s

21
2
Demand, Supply and Market Equilibrium

supply to price, while the market-level price elasticity of supply tells us the
sensitivity of market supply to price.

2.6 Learning by doing

1. Consider Pepsi and Coca-Cola. If the price of Coca-Cola increases then
we would expect that:

(a) The demand for Pepsi will increase.

(b) The demand for Coca-Cola will increase.

(c) The supply for Coca-Cola will decrease.

(d) None of the above.

2. The price of coffee beans fell by 10% last year and quantity purchased
fell by 10%. Which of the following statements best explains this?

(a) There was in increase in supply as a result of new production tech-

niques.

(b) Consumer incomes increased and coffee is a normal good.

(c) There was a major frost in Brazil that reduced its coffee production,
and Brazil has a big share of world production.

(d) Results of large-scale health survey suggested that coffee drinkers

were more likely to get cancer than non-coffee drinkers and this led to a
fall in demand.

3. There has been an increase in input prices for all producers of a good
sold in a competitive industry. Assuming this is a normal good, the ef-
fect of this change on market price and quantity will be:

(a) A higher price and higher quantity.

(b) A higher price and lower quantity.

(c) A lower price and higher quantity.

(d) A lower price and lower quantity.

22
 2.6
Learning by doing

4. The price of orange juice fell by 20% last year and quantity purchased
increased by 10%. Which of the following statements best explains this?

(a) There was an increase in supply as a result of a bumper orange har-

vest.

(b) Consumer incomes fell and orange juice is a normal good.

(c) The price of apple juice fell and this is a substitute for orange juice.

(d) There was bad weather in Florida that reduced the production in a
major growing area.

5. Theatre attendances fell last year even though prices of admission also
fell. Which one of the following on its own could explain this:

(a) There was an increase in the supply of theatre seats but no change af-
fecting demand conditions.

(b) There was an increased demand from consumers for visits to the
theatre as the quality of TV deteriorated.

(c) There was a reduction in prices of cinema seats as many new multi-
plex cinemas opened up.

(d) There was an increase in incomes of the theatre-going public.

6. When a price is below the market equilibrium price, there is a:

(a) Shortage and quantity demanded exceeds quantity supplied

(b) Shortage and quantity supplied exceeds quantity demanded

(c) Surplus and quantity demanded exceeds quantity supplied

(d) Surplus and quantity supplied exceeds quantity demanded

7. To protect animal, people are persuaded to stop using clothes from

animal, more and more people prefer to wear synthetic materials in-
stead of actual fur. What happens in the market for fur coats?

(a) The supply falls, causing the price of fur coats to rise.

(b) The supply of fur coats will now exceed the demand.

(c) The price of fur coats falls in response to lower demand

23
2
Demand, Supply and Market Equilibrium

(d) The equilibrium quantity falls, but there will be no change in the
equilibrium price

8. If the supply of turkeys in a particular November turned out to be unu-

sually small, do you think a turkey shortage would result? Why or why
not?

9. In 1999, after nearly 20 years of rent control in Berkeley, the elimination

of the law led to an estimated rise in rents of 40%. Using supply-and-
demand diagrams, illustrate how the law and then its elimination af-
fected the rental housing market. Discuss the effects on the equilibrium
rental price and the quantity of housing rented.

10. Suppose the market for good A can be described by the following equa-
tions:

Demand: P = 250 - Q

Supply: P = 10 + 3Q

Where P is the price in dollars per unit, and Q is the quantity in units.

(a) What is the free market price and quantity? Illustrate with a graph.

(b) Suppose the government imposes the maximum price of this good is
Pmax = $160. Is there exceed supply or exceed demand?

24
 2.6
Learning by doing

3 Preferences and Utility

Chapter objectives:
1. Understand the concept of preferences and state the assumptions about
consumer preferences

2. Explain the meaning of indifference curves and their properties. Show

how to graph indifference curves.

3. Explain the concept of utility function and state some special utility func-
tions.

4. Distinguish between total utility and marginal utility. State the law of
diminishing marginal utility.

5. Understand the concept of marginal rate of substitution and explain the

meaning of diminishing marginal rate of substitution.

25
 3
Preferences and Utility

3.1 Representation of Preferences

Basket of goods In modern society, consumers can purchase at vast array of goods and ser-
vices. We begin by considering a basket of goods (let’s say food and cloth-
ing) that an individual can consume. If we have to describe how this indi-
vidual allocates his budget between the two products we need to know
about his preferences toward these two goods.

Figure 3-1 illustrates the idea of this basket. There are seven possible con-
sumption baskets: A, B, D, E, G, H, J. For example, if the consumer buys
basket E, he consumes 20 units of food and 30 units of clothing per week. If
he chooses basket B instead, he weekly consumption would include 60 units
of food and 10 units of clothing. A basket might contain only one good, such
as basket J (only food) or basket H (only clothing).

Figure 3-1 Baskets of Food and Clothing

Units of clothing

H E A
30 ● ● ●

G
20● ●

D
10● ● ●B

J
●
0 20 40 60
Units of food

From millions of products and services that are available, each of us manag-
es somehow to sort out a set of goods and services to buy. When we make
our choices about what to consume, we make implicitly specific judgments
about the relative worth of things that are very different. In other words, we
express our preferences for certain baskets of goods and services by buying
them.

26
 3.2
Indifference Curve

Assumptions about Consumer Preferences

Our study of consumer preferences begins with three basic assumptions that Basic
underlie the theory of consumer choice. assumptions

1. Preferences are complete. That is, the consumer is able to rank any two
baskets and can state his preferences about the baskets according to one
of the following possibilities:

 “A is preferred to B” (written A  B)

 “B is preferred to A” (written B  A)

 “A and B are equally attractive” (written A  B)

2. Preferences are transitive. If an individual reports that "A is preferred to B"

and that "B is preferred to D," then he must also report that "A is pre-
ferred to D".

3. More is better. In other words, having more of a good is better for the
consumer. For example, in Figure 3-1, he would prefer basket A to E or H
because he receives the same amount of clothing with these three bas-
kets, but more food at A. He would prefer basket A to B or J because he
receives the same amount of food in these three baskets, but more cloth-
ing at A. Therefore, among the seven baskets, his most preferred basket
is A. However, without further information about the consumer’s prefer-
ences, we do not know how he would rank every pair of baskets, such as,
E and G because he would receive more clothing but less food at E and
more food but less clothing at G.

3.2 Indifference Curve

Now, suppose that basket A contains some quantity of food and clothing for Indifference curve
the consumer. Let us take all other baskets (i.e., all other combinations of
food and clothing) for which the consumer is indifferent between these bas-
kets and baskets A. This set of baskets forms the indifference set or indif-
ference curve of the consumer. If the consumer is not indifferent between
basket A and basket B, then A and B will be on different indifference curves.
A set of indifference curves representing the consumer’s preference with
respect to all possible baskets of goods is called an indifference map.

Figure 3-2

27
 3
Preferences and Utility

Indifference Curve

Clothing

●B Indifference curves
●A

●C

0 Food

Properties of Indifference curves on an indifference map share the following four proper-
indifference ties:
curves
1. When the consumer likes both goods, the indifference curves will have a negative
slope.
Indifference curves are usually drawn as negatively-sloping curves that
are convex to the origin and do not cross or touch. If more really is better,
then the indifference curves must have a negative slope. This is because
any basket, such as B, that lies to the northeast of basket A has more of
both goods and must therefore be preferable to A.

2. Indifference curves can not intersect.

Suppose that basket B is preferable to basket A, but the indifference
curves going through A and B respectively cross at D, as shown on Fig-
ure 3-3. This leads us to a contradiction. On the one hand, B is preferable
to B. On the other hand, the consumer must be indifference between B
and D since they are on the same indifference curve. Similarly, the con-
sumer must be indifference between A and D since D is also on the indif-
ference curve going through A. By transitivity then, the consumer should
be indifferent between A and B, which contradicts our initial assumption.

Figure 3-3

28
 3.2
Indifference Curve

Indifference Curves can not intersect

Clothing

●B
●A

●
D

0 Food

3. Every consumption basket lies on one and only one indifference curve.
Every point in the positive quadrant represents a possible consumption
basket. Our assumption that the consumer is capable of ranking all bas-
kets means that each basket has a unique level of utility. Other baskets
with that level of utility will be on the same indifference curve.

4. Indifference curves are not “thick”.

If they were, we could find two points A and B so that B lies to the north-
east of A and the consumer is indifferent between A and B since they are
on the same indifference curve. Since “more is (strictly) better”, B must
be preferable to A and can not therefore lie on the same indifference
curve.

Indifference Curves are not “thick” Figure 3-4

Clothing

●●B
A

0 Food

29
 3
Preferences and Utility

3.3 Utility Functions

The three assumptions – preferences are complete, they are transitive, and
more is better – allow us to represent preferences with a utility function. A
utility function measures the level of satisfaction that a consumer receives
from any basket of goods.

3.3.1 Total Utility

Level of To illustrate the concept of a utility function, let’s begin with a simple utility
satisfaction function for only one good, hamburgers. Let x denote the number of ham-
burgers he purchases each week, and let U(x) measure the level of satisfac-
tion (or total utility - U) that he derives from purchasing x hamburgers.

U = U(x)

If there are many goods, such as hamburgers, apple, shirt,…

U = U(x1, x2, x3,…,xn)

With x1, x2, x3,…,xn are quantities of each of goods – hamburgers, apple,
shirt,…and n – that might be consumed in a period.

Figure 3-5 Total and Marginal Utilities of Hamburgers

Quantity consumed Total Utility Marginal Utility

(x)
U( x)  10 x MU x  5 / x

0 0 0

1 10.00 5.00

2 14.14 3.54

3 17.32 2.89

4 20.00 2.50

5 22.36 2.24

6 24.49 2.04

30
 3.3
Utility Functions

3.3.2 Marginal Utility

While studying consumer behavior, we will often want to know how the Additional
level of satisfaction will change (∆U) in response to a change in the level of satisfaction
consumption (∆x where ∆ is read as “change in”). Economists refer to the
ratio ∆U/∆x as the marginal utility (MU), the rate at which total utility
changes as the level of consumption rises. We will denote the marginal utili-
ty of good x as MUx. Marginal utility is the additional satisfaction a consum-
er gains from consuming one more unit of a good

Suppose x measures the number of hamburgers a consumer has eaten this

week. If he is like most people, the additional satisfaction he receives from
eating another hamburger (the marginal utility) will depend on how many
hamburgers he has already consumed. The following equations express this
relationship, where ∆x is understood to be a small change in consumption.
U dU
MU x  
x dx

U(x  x)  U(x)

MU x  (3.1)
x

When the total utility from consuming x hamburgers is U( x)  10 x , the

marginal utility will be MU x  5 / x .We could use equation (3.1) to derive
the expression for MUx given U(x).1

In the example in Figure 3-5 and 3-6, marginal utility declines as the con- Law of
sumer eats more hamburgers. This trend illustrates the principle of dimin- diminishing
ishing marginal utility. This law states that for any good or service, the marginal utility
marginal utility of that good or service decrease as the quantity of the good
increase, ceteris paribus. In other words: The more of one good consumed in
a given period, the less satisfaction (utility) generated by consuming each
additional (marginal) unit of the same good. Although the first few potato
chips may provide quite significant increases in utility, subsequent chips are
likely to provide progressively less additional utility (as your hunger de-
creases and your guilt increases).

Figure 3-6

1 With U( x)  10 x  MU  ( U( x  x)  U( x)) / x  10( x  x  10 x ) / x

x

or 10( x  x  10 x )( x  x  x ) / x( x  x  x )  10( x  x  x) / x( x  x  x )

 10 /( x  x  x ) For small values of x => MUx  10 / 2 x  5 x

31
 3
Preferences and Utility

Total and Marginal Utility of Hamburgers

U, total utility
22.36 U  10 x
C
20.00
B
R

10.00
A

x, weekly
consumption
0 1 2 3 4 5 6 of hamburgers
MU, marginal utility

A’
5.00

B’
2.50 C’
2.24

x, weekly
consumption
0 1 2 3 4 5 6 of hamburgers

Marginal utility The concept of marginal utility is easily extended to the case of multiple
of multiple goods goods. The marginal utility of any one good is the rate at which total utility
changes as the level of consumption of that good rises, holding constant the
levels of consumption of all other goods.

For example, in the case in which only two goods are consumed and utility
function is U(x,y), the marginal utility of food (MUx) measures how the level
of satisfaction will change (∆U) in response to a change in the consumption of
food (∆x), holding the level of y constant. Similarly, the marginal utility of
clothing (MUy) measures how the level of satisfaction will change (∆U) in
response to a change in the consumption of clothing (∆y), holding constant
the level of food (x).
U U
MU x   y is held constant (3.2)
x x

32
 3.3
Utility Functions

U U
MU y   x is held constant (3.3)
y x

For example, when the total utility from consuming a bundle (x,y) is
U  xy , the marginal utilities are:

Equation (3.2) tells us that:

U U 1 y
MU x   
x x 2 x

U U 1 x
MU y   
y y 2 y

One possible way of measuring cardinal utility is by asking people “How

happy are you?” Such survey tends to confirm that marginal utility is posi-
tive and that people’ preferences are consistent with the principle of dimin-
ishing marginal utility.

Income and Happiness? Case Study 3-1

In a study by Richard Easterlin in 1994, American respondents were asked to

rate themselves on a happiness scale. As shown in the following Table 1,
“life satisfaction” was reported as rising sharply with income.

Table 1: Relative income and life satisfaction in the United State, 1994

Total household income “Very happy” “Pretty “Not too

(thousands) happy” happy”

Less than $10 16 62 23

$10-20 21 64 15

$20-30 27 61 12

$30-40 31 61 8

$40-50 31 59 10

$50-75 36 58 7

Greater than $75 44 49 6

Source: Adapted from Table 1 in Easterlin (2001)

33
 3
Preferences and Utility

Table 2 is derived from a 1984 survey by the same author comparing satisfac-
tion scores for 24 different nations, as related to per capita GNP (b). The
second study shows that, across different nations, greater average per capita
incomes once again were associated with higher reported levels of satisfac-
tion.

Table 2: Absolute income and life satisfaction (across nations), 1984

GNP per capita Number of nations Median “satisfaction” score

<$2,000 1 5.5

$2,000-4,000 3 6.6

$4,000-8,000 6 7.0

$8,000-16,000 14 7.4

Source: Estimated visually from Easterlin (1995), Figure 3.4

Both tables of the case study suggest that utility – as measured by reported
“satisfaction” – consistently rises with income. It means that the marginal
utility of income is positive. There is also an indication of diminishing mar-
ginal utility. In Table 2, for example, a comparison of the second and third
rows shows that a doubling of per capita GNP is associated with only a
small rise in national satisfaction level, from 6.6 to 7.0 – an improvement of
only about 6%.

3.3.3 Marginal Rate of Substitution

Rate of willing The marginal rate of substitution (between food and clothing) is the rate at
to substitute which the consumer would be willing to substitute a little more food for a
little less clothing. If the consumer is currently consuming at basket A, the
marginal rate of substitution is the increase in clothing that he would require
in exchange for a small decrease in food to leave him just indifferent be-
tween consuming basket A and the new basket that lies very close by.

Formally, the marginal rate of substitution is defined as the negative of the

slope of the indifference curve:
clothing MU f
MRS f ,c   
food MU c

(For a constant level of preference)

34
 3.3
Utility Functions

Depending on which basket the consumer is now consuming, the marginal

rate of substitution can change. For example, in order to compensate for a
small loss of clothing at basket B, below, the consumer needs a much larger
increase in food than at basket A.

To illustrate the concept of marginal rate of substitution, let us start with a

basket on the indifference curve:

U0 = U(x,y)

Now suppose the consumer changed the level of consumption of x and y by

∆x and ∆y, respectively. The corresponding impact on utility ∆U would be:

∆U = MUx(∆x) + MUy(∆y)

Changes in x and y that moves us along the indifference curve U0 must keep
utility unchanged, so that ∆U = 0:

0 = MUx(∆x) + MUy(∆y) or

MUy(∆y) = – MUx(∆x)

This can be rewritten:

y MU x
   MRS x ,y
x MU y

with holding utility constant.

Diminishing Marginal Rate of Substitution Figure 3-7

Clothing

∆CA
∆CA > ∆CB
●A
∆F

∆CB B
∆F
●

0 Food

35
 3
Preferences and Utility

Diminishing For preference such as Figure 3-7 we have a diminishing marginal rate of
marginal rate of substitution. The marginal rate of substitution of x for y declines as the
substitution consumer increases his consumption of x along an indifference curve. In
other words, the more food you have the more food you are willing to give
up to get a little clothing. This, again, reflects the consumer’s preference for
variety. Graphically, this means that the indifference curves get flatter as we
go toward the extremes on the horizontal axis and steeper as we go toward
the extremes on the vertical axis.

The rate of substitution that leaves the consumer on the same indifference
curves can be quite different for large changes than for very small changes.
This is shown Figure 3-8, where the rate of substitution for small changes
would be x/a and the rate of substitution for large changes would be (x+y)/b.

Figure 3-8 The Rate of Substitution for small Changes and for large Changes

Clothing

x/a > (x+y)/b

●A
x
a
y

0
Food

To see how you can use information about the total and marginal utilities to
understand the shape of a consumer’s indifference curves, suppose a con-
sumer has preferences between two goods that can be represented by the
utility function U = xy. For this utility function, MUx = y and MUy = x.

Case Study 3-2

36
 3.3
Utility Functions

Graphing Indifference Curves

Problem

(a) On a graph illustrate the shape of the indifference curve, U1 = 128. Then
answer the following questions:

1. Does the indifference curve intersect either axis?

2. Does the shape of the indifference curve indicate that the MRSx,y is
diminishing?

On the same graph draw a second indifference curve, U2 = 200

(b) Show how MRSx,y depends on x and y, and use this information to de-
termine if there is diminishing MRSx,y

Solution

(a) First note that both MUx and MUy are positive whenever the consumer
has positive amounts of x and y. Therefore, indifference curves will be nega-
tively sloped. This means that as the consumer increases x along an indiffer-
ence curve, y must decrease.

Can the indifference curve U1 intersect either axis? If so, it must be possible
to achieve a positive level of utility when one of the goods is not consumed.
Since U = xy, to achieve a positive level of utility, such as U1, the consumer
must buy positive amounts of both goods. If either x = 0 or y = 0, then U = 0.
So indifference curves for this utility function do not intersect either axis.

To draw the indifference curve U1 = 128, we look for combinations of x and y

that give xy = 128. One such basket is (x,y) = (8,16), labeled as basket G on
the graph. Two other baskets on the indifference curve are basket H, with
(x,y) = (16,8), and basket I, with (x,y) = (32,4).

The below figure illustrates that this indifference curve is bowed in toward
the origin, and therefore there is a diminishing MRSx,y. The indifference
curve corresponding to U2 = 200 lies up and to the right of U1 = 128, confirm-
ing that the consumer likes more of both goods.

(b) We know that MRSx,y = MUx/MUy = y/x. As we move along the indiffer-
ence curve by increasing x and decreasing y, then MRSx,y = y/x will decrease.
So we have diminishing marginal rate of substitution of x for y. We can nu-
merically verify that there is a diminishing MRSx,y. At basket G, MRSx,y =
MUx/MUy = 16/8 = 2. So the slope of the indifference curve through basket G
is -2. IF we move along the indifference curve to basket I, the marginal rate
of substitution diminishes to MRSx,y = MUx/MUy = 4/32 = 1/8. The slope of the
indifference curve through basket H is -1/8.

37
3
Preferences and Utility

y
16 ●G Preference
directions

H
8 ●
U2 = 200
4 ●I
U1 = 128

0 8 16 32 x

3.3.4 Special Utility Functions

Some particularly useful functional forms of utility are the following:

1. Cobb-Douglas, generally represented as U  Ax  y  , where A,  ,  are

positive constants.

The Cobb-Douglas utility function has three properties that make it of in-
terest in the study of consumer choice:

o The marginal utilities are positive for both goods. The marginal utili-
ties are MU x  Ax  1 y  and MU y  Ax  y 1 , and both MUx and
MUy are positive when A,  ,  are positive constant. This means that
“the more is better” assumption is satisfied.

o Since the marginal utilities are both positive, the indifference curves
will be downward sloping.

o The Cobb-Douglas utility function also exhibits a diminishing mar-

ginal rate of substitution. The indifference curves will therefore be
bowed in toward the origin.

2. Perfect Substitution, represented as U = x + Ay. In this case, the consumer

cares only about the weighted total amount of x and y he consumes: one
unit more of x would be exchanged against A units of y in order to leave
the consumer at the same preference level. In this sense, x and y are per-
fect substitutes. The indifference curves for this type of function are neg-
atively-sloped (as long as A is positive) and straight.

38
 3.3
Utility Functions

3. Perfect Complements, represented as U = min(x,y/A). In this case, the

consumer cares about one good only to the extent that he has the other
good as well. In other words, the only units of y that are valuable to the
consumer are those that can be paired with units of x in a proportion of A
units of y to one unit of x. In this sense, x and y are perfect complements.
The indifference curves for this type of function are right angles lie on a
ray of slope A going through the origin.

4. Quasi-Linear Preference, represented as U = v(x) + Ay, where v(x) is an

increasing function and A is a positive constant. The indifference curves
for this type of function have the same marginal rate of substitution as
one moves due North on the indifference map with y on the vertical axis.
To see this, notice that MUx = v’(x) and MUy = A so that the marginal rate
of substitution is equal to v’(x)/A. Since this expression is independent of
y, the MRS is constant for any given value of x.

Taste Tests and Consumer Preferences Case Study 3-3

If you listen to advertisements on television, you might believe that beer is a

highly differentiated product and that most consumers have strong prefer-
ences for one beer over another. To be sure, there are differences among
beers, and not all brands are sold at the same price. However, are brands so
different that one brewer could raise the price of its product without losing a
significant portion of its sales?

In looking at the U.S. beer industry, Kenneth Elzinga observed, “Several

studies indicate that, at least under blindfold test conditions, most beer
drinkers cannot distinguish between brands of beer”. He also noted that
brewers have devoted “considerable talent and resources… to publicizing
real or imagined differences in beers, with the hope of producing product
differentiation”.

In the end, Elzinga suggested that, despite brewers’ efforts to differentiate

their products from those of their competitors, most consumers are quite
willing to substitute one brand of beer for another, especially if one brand
were to raise its price significantly.

Elzinga’s analysis does not suggest that all consumers regard brands of beer
as perfect substitutes. However, when a consumer does not have a strong
preference for on beer over another, then the marginal rate of substitution of
brand A for brand B might be nearly constant, and probably near 1, since a
consumer would give up one glass of brand A for one glass of brand B.

39
3
Preferences and Utility

3.4 Learning by doing

1. Eva likes to take medicine when she is sick since it makes her fell better;
however, she tells her doctor that she does not care whether she gets a
prescription for generic or brand name drugs since they both contain the
same amount of the important active ingredients. For Eva, prescription
and brand name drugs are:

(a) perfect substitutes

(b) perfect complements

2. Which of the following statements about indifference curves is NOT

correct?

(a) Indifference curves plot combinations of baskets of goods between

which a consumer has the same level of total utility.

(b) Indifference curves are generally negatively sloped if the goods in-
volved are genuinely "goods" but they could be positively sloped if one
of the products involves is a "bad".

(c) Without utility being quantifiable we can say that one indifference
curve is higher than (or preferred to) another but we cannot say by how
much.

(d) Two different indifference curves can intersect but only once.

3. The law of diminishing marginal utility indicates that, after some point,
marginal utility

(a) decreases at a constant rate.

(b) decreases at an increasing rate.

(c) decreases at a decreasing rate.

(d) decreases.

4. Mario plays video games. Although he is experiencing diminishing mar-

ginal utility, his marginal utility remains positive. We can say that Mar-
io’s total utility is

(a) increasing at an increasing rate.

(b) increasing at a decreasing rate.

40
 3.4
Learning by doing

(c) decreasing at an increasing rate.

(d) decreasing at a decreasing rate.

5. Candy has eaten 10 Hershey’s Kisses and admits that each additional
Kiss has been less enjoyable than the previous one. We can deduce that,
for Candy,

(a) the marginal utility of Kisses is positive but decreasing.

(b) the marginal utility of Kisses is negative.

(c) the total utility of Kisses is diminishing.

(d) the total utility of Kisses has peaked.

6. Angela will buy additional units of a good (apples) if the value of the
good’s

(a) total utility exceeds the price.

(b) total utility is less than the price.

(c) marginal utility exceeds the price.

(d) marginal utility is less than the price.

7. An indifference curve plots all the combinations of Good X and Good Y

(a) that may be bought with a given income level and given prices for
Good X and Good Y.

(b) that give maximum satisfaction as income level changes.

(c) that give the same marginal utility.

(d) that give the same total utility.

8. An indifference curve is downward sloping because

(a) more is preferred to less.

(b) preferences remain constant as income level increases.

(c) the marginal rate of substitution increases as one moves along the in-
difference curve.

41
3
Preferences and Utility

(d) the marginal rate of substitution decreases as one moves along the
indifference curve.

9. Consider the utility function U(x,y) = y x with the marginal utilities

y
MU x  and MU y  x .
2 x

(a) Does the consumer believe that more it better for each good?

(b) Do the consumer’s preferences exhibit a diminishing marginal utility

of x? Is the marginal utility of y diminishing?

10. Consider the utility function U(x,y) = 3x + y, with MU x = 3 and MUy = 1.

(a) Is the assumption that more is better satisfied for both goods?

(b) Does the marginal utility of x diminish, remain constant, or increase

as the consumer buys more x? Explain.

(c) What is the MRSx,y? Is the MRSx,y diminishing, constant, or increasing

as the consumer substitutes x for y along an indifference curve?

(d) On a graph with x on the horizontal axis and y on the vertical axis,
draw a typical indifference curve (it need not be exactly to scale, but it
needs to reflect accurately whether there is a diminishing MRSx,y). Also
indicate on your graph whether the indifference curve will intersect ei-
ther or both axes. Label the curve U1. On the same graph draw a second
indifference curve U2, with U2 > U1.

42
 3.4
Learning by doing

4 Consumer Choice

Chapter objectives:
1. Draw and explain the meaning of a budget constraint diagram, given
price and income data.

2. Indicate the constraints on a household’s consumption choices and relate

these to the concept of the opportunity set.

3. Distinguish between the income effect and the price effect on the budget
lines.

4. Understand the concept of optimal choice and calculate the optimal

basket of goods.

5. Explain the concept of revealed preference.

43
 4
Consumer Choice

4.1 The Budget Constraints

Each household operates within constraints (income, wealth, and the prices
of goods) and, therefore, must make three choices:

 Which combination of goods to buy?

 How many hours of work to provide?
 How much money to save rather than to spend?
Maximum The budget constraint defines the maximum possible amount of goods that
possible amount can be bought and is determined by the household’s income, wealth, and
of goods prices. Normally, households don’t have the power to control the market
prices. Therefore, they have to buy goods and services at market-determined
prices. The amount that a household can spend depends on its income and
wealth.

Let us return to the example of Chapter 3 where food and clothing are the
only goods available for consumption. Further, let x be the number of units
of food he purchases each month and y the number of units of clothing. The
price of a unit of food is Px, and the price of a unit of clothing is Py. He has a
fixed income of I dollars per month for two above goods.

His total monthly expenditure on food will be Px x (the price of a unit of food
times the amount of food purchased). Similarly, his total monthly expendi-
ture on clothing will be Py y (the price of a unit of clothing times the number
of units of clothing purchased).

In order for the basket (food and clothing) to be affordable, this expenditure
cannot exceed his income. Then this basket is affordable if:

Px x + Py y ≤ I

Budget set The above inequality is the consumer’s budget constraint, which defines the
Opportunity set set of baskets that a consumer can purchase with a limited amount of income
This constraint determines the consumer’s budget set or opportunity set,
that is the set of baskets that are affordable. A budget constraint separates
those combinations of goods and services that are available, given limited
income, from those that are not.

Budget line The budget line is set of baskets that are just affordable, that is the set of
baskets that could purchased if the consumer expended all of his income.
Mathematically, the budget line is the set of baskets for which the inequality
above becomes equality.

Px x + Py y = I (4.1)

44
 4.1
The Budget Constraints

Let’s look at the graph of the budget line in Figure 4-1. Eric has an income of
I = $800 per month for food and clothing. The price of food is Px = $20 per
unit and the price of clothing is Py=$40 per unit. If he spends all $800 on
food, he will be able to buy, at most, I/Px = 800/20 = 40 units of food. So the
horizontal intercept of the budget line will be at x = 40. Similarly, if Eric buys
only clothing, he will be able to buy at most I/Py = 800/40 = 20 units of cloth-
ing. So the vertical intercept of the budget line will be at y = 20.

Using equation (4.1), we can plot the whole budget line. If I = $800, P x = $20,
and Py = $40:

$20x + $40y = $800

Budgetline Figure 4-1

I
 20
PY
Budget set
Budget line:
A
20

●
y, units of clothing

Income = $800 per month

B
● ●G Px 1
15

Slope =  
PY 2
F
● ●C
10

●D
5

E
●
0 10 20 30 40
I
x, units of food  40
PX

Eric’s income permits him to buy any basket on or inside the budget line
(Budget set), such as A, B, C, D, E, F. However, Eric cannot buy a basket
outside the budget line, such as G. To buy G he would need to spend $1000,
which is more than his monthly income.

In Figure 4-1, the slope of the budget line tells us how many units of clothing
(the good on the vertical axis), he must give up to obtain an additional unit of
food (the good on the horizontal axis). In another word, the slope of the budget

45
 4
Consumer Choice

line measures the magnitude of opportunity cost of consuming food. If Eric

moves from basket B to basket C, he gives up 5 units of clothing (∆y = -5) to
purchase 10 more units of food (∆x = +10). The slope of the budget line is
therefore ∆y/∆x = -1/2.

4.2 Income and Price Effects

The position of the budget line depends on the level of income and on the
price of both goods. If income changes, the vertical and horizontal intercepts
change but the slope of the line remains the same since the prices are still the
same. Hence a change in level of income shifts the budget line up or down
in a parallel fashion.

Figure 4-2 Effect of Income on the Budget Line

Clothing

I/Py

I/Py’

Food
0 I/Px’ I/Px

Purchasing pow- For example, as income rises, the vertical intercept increases and the set of
er affordable basket grows. In other words, the consumer’s purchasing power
(= the amount of goods and services that can be bought by consumers) in-
creases. If income falls, the vertical intercept falls and the budget line shifts
in. The set of affordable baskets then shrinks.

A change in the relative price of food, Px/Py, changes the slope of the budget
line. An increase in the relative price of food makes the budget line steeper

46
 4.2
Income and Price Effects

while a decrease makes the budget line flatter. The effect of a price change
on the vertical and horizontal intercepts of the budget line depends on the
specific change considered.

An increase in the price of clothing reduces the vertical intercept since, for a
given income I, a consumer spending all of his income on clothing could
now buy fewer units. Similarly, an increase in the price of food would reduce
the value of the horizontal intercept. Hence, changes in the relative price of
the two goods affect both the slope and the intercept(s) of the budget line.

Suppose, for example, that the price of clothing increases while the price of
food stays the same. The budget line becomes flatter as Px/Py has decreased
and the vertical intercept decreases since fewer units of clothing can be pur-
chased with the whole income I. On the other hand, the horizontal intercept
is unchanged: since the price of food and the level of income have not
changed, the number of units of food that can be bought with the whole
income has not changed either. This means that an increase in P y with no
change in either Px or I, rotates the budget line inward around its horizonal
intercept (see Figure 4-3, a) so that the budget set of the consumer shrinks.

As a second example, (see Figure 4-3, b), suppose that the price of food de-
creases while both Py and I remain unchanged. The slope of the budget line
becomes flatter as Px/Py decreases. The horizontal intercept increases since
more units of food can now be obtained by using the whole income. On the
other hand, the vertical intercept is not affected since both I and P y have not
changed. This means that the budget constraint rotates outward around its
vertical intercept so that the budget set of the consumer expands.

Effect of a Price on the Budget Line Figure 4-3

Clothing

I/Py
(a)

I/Py’

0 I/Px Food

47
4
Consumer Choice

Clothing

I/Py
(b)

0 I/Px’ I/Px Food

Problem

What would happen to the number of baskets available to the consumer if

the prices of both goods and income double? The consumer views the dou-
bling of income as good news because it increases his purchasing power.
However, the doubling of prices is bad news because it decreases his pur-
chasing power. What is the net effect of the good and bad news?

Solution

You can answer this question by observing how the budget line changes as
prices and income double. Initially, the intercept on the y axis is I/P y. If both I
and Py double, then the vertical intercept will be unchanged. Similarly, if
both I and Px double, then the intercept along the x axis (I/Px) will be un-
changed. And the slope of the budget line –(Px/Py) will also be unaffected
because both prices double. In short, the location of the budget line will be
unaffected if all (in this case, both) prices and income double. The consum-
er’s purchasing power is unaffected because the set of baskets available to
him does not change.

4.3 Optimal Choice

If we know a consumer’s preferences and budget constraint, we can deter-
mine the optimal amount of each good to purchase. We will assume that a
consumer makes these choices rationally. More precisely, optimal choice
means that the consumer chooses a basket of goods that (1) maximizes his

48
 4.3
Optimal Choice

Optimal choice
problem
satisfaction (utility) while (2) allowing him to live within his budget con-
straint.

To state the problem of optimal consumer choice, let U(x,y) represent the
consumer’s utility from purchasing x units of food and y units of clothing.
The consumer chooses x and y, but must do so while satisfying the budget
constraint Pxx + Pyy ≤ I. The optimal choice problem for the consumer is
expressed like this:

max U( x, y) (4.2)
( x ,y )

Subject to Pxx + Pyy ≤ I

Where the notation “ max U( x, y) ” means “choose x and y to maximize

( x ,y )

utility”, and the notation “Subject to Pxx + Pyy ≤ I” means “the expenditures
on x and y must not exceed the consumer’s income”. If the consumer likes
more of both goods, the marginal utilities of food and clothing are positive.
At an optimal basket all income will be spent (that is, the consumer will
choose a basket on the budget line Pxx + Pyy = I).

If the consumer maximizes utility while satisfying that budget constraint, he

will choose the basket that allows him to reach the highest indifference curve
while being on or inside the budget line. In Figure 4-4 that optimal basket is
A, where he achieves a level of utility U2. Any other point on (B, D) or inside
(C, F) the budget line will leave him with a lower level of utility.

Optimal Choice: Maximizing Utility with a Given Budget Figure 4-4

Preference
y, units of clothing

directions
20

B
● ●E
15

U3
F
● ●A
10

●
C U2
●D U1
5

x, units
0 10 20 30 40 of food

49
 4
Consumer Choice

To understand why basket A is the optimal choice, let’s explore why other
baskets are not optimal.

 First, baskets outside (i.e., to the northeast of) the budget line, such as E,
cannot be optimal because he cannot afford them. We can therefore re-
strict our attention to baskets on or inside the budget line.

 Any basket inside the budget line, such as C, F, will also not be optimal.
As indicated by the preference directions on the graph, the consumer
likes more of both goods (the marginal utilities are both positive). Since
there are affordable baskets to the northeast of baskets such as C and F, C
and F cannot be optimal. Thus, an optimal basket must lie exactly on the
budget line.

If he was to move along the budget line away from A, even by a small
amount, his utility would fall because the indifference curves are bowed in
toward the origin or, in economic terms, because there is diminishing mar-
ginal rate of substitution of x for y (MRSx,y). At the optimal basket A, the
budget line is just tangent to the indifference curve U2. This means that the
slope of the budget line (-Px/Py) and the slope of the indifference curve are
equal. We had equation of the slope of the indifference curve is -MUx/MUy
(that is, -MRSx,y). Thus, at the optimal basket A, the tangency condition re-
quires that:
MU x Px

MU y Py

or, equivalently, that:

Px
MRS x ,y 
Py

Interior optimum In Figure 4-4, the optimal basket A is said to be an interior optimum, that is,
an optimum at which the consumer will be purchasing both commodities
(x>0 and y>0). The optimum occurs at a point of tangency between the
budget line and the indifference curve. In other words, at an interior optimal
basket, the consumer will choose the commodities so that the ratio of the
marginal utilities (that is, the marginal rate of substitution) will equal the
ratio of the prices of the goods.

We can also express the tangency condition by rewriting above equation as

follows:

MU x MU y

Px Py

50
 4.3
Optimal Choice

This form of the tangency condition states that, at an interior optimal bas-
ket, he will choose commodities so that the marginal utility per dollar spent
on each commodity will be the same. Put another way, at an interior opti-
mum, the extra utility per dollar spent on good x is equal to the extra utility
per dollar spent on good y.

Although we have focused on the case in which the consumer purchases

only two goods, such as food and clothing, the consumer’s optimal choice
problem can also be analyzed when the consumer buys more than two
goods. For example, suppose the consumer chooses among baskets of three
commodities (x, y, z) to maximize utility U(x, y, z) while living within the
budget constraint Pxx + Pyy + Pzz ≤ I. If all of the goods have positive margin-
al utilities, then an optimal basket will be on the budget line. At an interior
optimal basket, the consumer will choose the goods so that the marginal
utility per dollar spent on all three goods will be the same, that is, so that
MUx/Px = MUy/Py = MUz/Pz. The same principles apply to the case in which
the consumer buys any larger number of goods.

Finding an Interior Optimum Case Study 4-1

Problem

Eric purchases food (measured by x) and clothing (measured by y) and has

the utility function U(x,y) = xy. His marginal utilities are MUx = y and MUy =
x. He has a monthly income of $800. The price of food is Px = $20, and the
price of clothing is Py = $40. Find his optimal consumption bundle.

Solution

We knew two conditions that must be satisfied at an optimum:

- An optimal basket will be on the budget line. This means that Pxx + Pyy = I,
or, with the given information:

20x + 40y = 800

- Since the optimum is interior, the indifference curve must be tangent to the
budget line. The tangency condition requires that MUx/MUy = Px/Py, or, with
the given information,
y 20

x 40

or, more simply,

x = 2y

51
 4
Consumer Choice

So we have two equations with two unknowns: 20x + 40y = 800 (coming from
the budget line) and x = 2y (coming from the tangency condition). If we
substitute x = 2y into the equation for the budget line, we get 20(2y) + 40y =
800. So y = 10 and x = 20. Eric’s optimal basket involves the purchase of 20
units of food and 10 units of clothing each month.

Maximize Utility The consumer’s optimal choice can also be seen as the result of minimizing
or Minimize the expenditure required to attain a given level of satisfaction. In this ap-
Expenditure proach, rather than moving to the highest indifference curve that still allows
tangency with the fixed budget line, one move to the lowest budget line that
still allows tangency with the fixed indifference curve. This rephrasing of the
problem is called the expenditure minimization problem and results in the
same optimal basket as the utility maximization problem we have analyzed
so far if utility is constrained to be at the optimal level of the utility maximi-
zation problem.

For example, in Figure 4-5, if the given level of utility is constrained to be the
same level as was attained in Figure 4-4, then the tangency point in Figure 4-
5 will be the same as tangency point A in Figure 4-4. Since expenditure min-
imization looks at the same problem as utility maximization, only from “the
other side”, we say that it is the dual of the utility maximization problem.

Figure 4-5 Minimize Expenditure

y, units of clothing

Increasing levels
of Expenditure

● Given Level
A of Utility

0
x, units of food

52
 4.4
Consumer choice with composite goods

4.4 Consumer choice with composite goods

Although consumers typically purchase many goods and services, econo- Composite good
mists often want to focus on the consumer’s selection of a particular good or
service, such as the consumer’s choice of clothing or food. In that case, it is
useful to present the consumer’s choice problem using a two-dimensional
graph with the amount of the commodity of interest (clothing) on the hori-
zontal axis, and the amount of expenditures on all other goods on the verti-
cal axis. The good represented on the vertical axis is called a composite good
because it represents the collective expenditures on all other good (measured
by y), with the price of a unit of the composite good being P y = $1.

Optimal Choice of Clothing Figure 4-6

y, units of the composite good and amount

Preference
of expenditure on all other goods

I directions

yA A

Budget line slope

U1
P
  c  Pc
Py

C, units of
CA I/Pc clothing

Let’s begin by considering Figure 4-6. Here we are interested in the consum-
er’s choice of clothing. On the horizontal axis are the units of clothing C. The
price of clothing is Pc. In the vertical axis is the composite goods, measured
in units by y, and with a price Py = $1. If the consumer spends all his income
on clothing, he could purchase at most I/Pc units of clothing, the intercept of
the budget line on the horizontal axis. If he spends all of his income on other
goods, he could purchase at most I units of the composite goods, the inter-
cept of the budget line on the vertical axis. The optimal basket will be A in
the graph.

53
 4
Consumer Choice

Case Study 4-2 Germans wary of Car Scrapping Scheme

By Ray Furlong, BBC correspondent, Hamburg

Kurt Kroeger is a happy man - his business has gone from bust to boom in the space
of a few months, thanks, he says, to Germany's car scrapping scheme.

"Our sales are up 100% on last

year," he says, showing me
around his gleaming showroom
on the edge of Hamburg.

"When scrap page was first intro-

duced, we had hordes of people
here. We had to work round the
clock. The customers were served
coffee and cakes to calm them
down, while they waited for a free
salesman." Germany is paying motorists to trade
in old cars for new ones
Downstairs, Mr. Kroeger shows
me a garage with about 20 cars in it - all being picked up by satisfied cus-
tomers today. His experience is not unique. In the first three months of this
year, Volkswagen reported a recession-busting 36% increase in new car reg-
istrations. Opel claimed a 50% increase in orders in the first quarter of 2009
compared with the previous year.

'Old banger bonus'

And Germany's scrapping scheme is clearly responsible. Called the "Envi-

ronment Bonus" by the government and the "Old Banger Bonus" by every-
one else, it's seductively simple - anyone with a car that is at least nine years
old can apply for the scheme. The car is scrapped, and they receive 2,500
euros off the price of a brand-new one.

Earlier this month, the German government extended the scheme until the
end of the year 2009, raising the budget for it from 1.5bn to 5bn euros. And
yet, ever since, there's been a rising tide of criticism.

"Since it's been introduced, people are buying cars like mad," says Joerg
Maltzen, a journalist at Bild-Auto, Europe's biggest-selling car magazine.

"But it's just a temporary effect. The people who are buying the cars now
will not buy next year - so we're all expecting a sharp slump in 2010.

"Also, it's mainly small cars in the lower price segment that are selling.
BMW, Mercedes and Audi are profiting very little."

54
 4.4
Consumer choice with composite goods

Economists and even government backbench MPs have complained about

this, arguing that the scheme is actually benefitting foreign carmakers. Oth-
ers say it's just diverting money from other sectors of the economy - leaving
them short-changed. One of the highest profile critics is Justus Haucap, head
of Germany's monopoly commission. "The economic benefit is minuscule
compared to the cost that it brings. The money could have been spent much
more wisely on other things, such as infrastructure and education," he says.

Let’s use the theory of consumer choice to examine how the German gov-
ernment program might increase the amount of the new car chosen by a
consumer. Suppose the consumer has preferences for the new car and other
goods (without the government’s program) as shown by the indifference
curves in Figure 4-8. The consumer has an income I and must pay a price P c
for a new car he buys and Py = 1 for each unit of the composite “other
goods”. The budget line is the segment KJ. When the consumer spends all
his income on other goods, he could buy I/Py = I units of the composite good.
With his preferences and the budget line KJ, he chooses basket A with C A
“quality” of a new car.

Optimal Choice of a new Car: Subsidy and Voucher Figure 4-8

I+S M
y, units of the composite goods and
amount of expenditure on all other goods

E
I+V

T
F
U4
K
I U3
R
A U2

U1

J G N
CA CF CB I IV IS C, “quality“
Pc Pc Pc of a new car

55
4
Consumer Choice

With this program, the consumer will receive 2.500 euro from the govern-
ment if he replaces his old car with a new car. Let’s suppose 2.500 euro is a
new car voucher worth V euro, because the consumer can use this voucher
only for purchasing for a new car (not for any other good).

Would it matter to the consumer or to the government whether the consum-

er receives a cash income subsidy of S euro or a new car voucher worth V
euro (2.500 euro)? (S>V)

For the government:

They could induce the consumer to choose CB “quality” of a new car with
either of the following two programs:

1. Give him an income subsidy of S euro, moving the budget line to MN.
The consumer chooses basket T with indifference curve U4 and CB.

2. Give him a new car voucher worth V euro that can be spent only on a
new car, moving the budget line to KRG. The consumer choose basket R
with U2 and CB.

If the government’s primary goal is to increase the consumption of new cars

to CB with high quality and high price (such as BMW, Mercedes, Audi,...), the
voucher (V) will cost the government less than an income subsidy (S) would.
The government can save (S - V) euro if it uses the voucher program instead
of an income subsidy.

For the consumer:

It is easy to know that the consumer will prefer S to V, because S is much

than V and with S, he can purchase for any good he buys. However, one can
also ask how the consumer would act if given a cash subsidy of V euro that
redeem for a voucher. Then the budget line would be EG. The consumer
would choose basket F and reach the indifference curve U3. The consumer
would prefer the cash subsidy of V euro (allowing him to choose basket F
and reach indifference curve U3) to the voucher worth V euro (allowing him
to choose basket R and reach indifference curve U2). However, with a vouch-
er for V euro, the consumer’s choice of a new car (CF) is below the govern-
ment’s target level (CB).

56
 4.5
Advertising and indifference curve analysis

4.5 Advertising and indifference curve

analysis
Advertising can be generally thought of as being either informative (telling Endorsement
potential customers of a product’s availability, characteristics or price) or as
persuasive. A persuasive advert might portray a product as being in vogue,
appealing to those of us wishing to be considered fashionable. Similarly,
hiring a well-known sports or media personality to endorse a product might
persuade consumers of the acceptability or quality of a product. Endorse-
ment is recommendation of the purchase of a good to others. For example, a
well-known footballer might endorse a brand of football boots

In reality it is difficult to classify adverts as being either solely informative or

persuasive. Most adverts contain elements of both, and what might be seen
as informative could also be considered persuasive. For example, informing
consumers that a new breakfast cereal has added fibre could be seen as per-
suasive, as there is an implicit message that the consumer should pay addi-
tional attention to their daily fibre intake.

How may we illustrate the impact of advertising through indifference curve analy-
sis?

The Impact of Advertising Figure 4-7

Consumption of Y

A
per unit of time

Y1 ●a

●c IC1
●b
IC2* IC3* Consumption
IC1* of X per unit of
0 X1 C X2 B time

57
4
Consumer Choice

Figure 4-7 shows a budget line AB and the consumer reaching equilibrium at
point a on indifference curve IC1 consuming X1 of good X and Y1 of good Y.
In equilibrium:

MU X PX
  MRS X ,Y
MU Y PY

Imagine that producer of good X now embarks on an advertising campaign

that successfully improves the product’s image, shifting the demand curve to
the right. This increases the relative attractiveness of X in terms of Y, the
consumer now being willing to sacrifice additional units of Y to obtain X. In
short, MRSx,Y increases for any value of X and we now have a new indiffer-
ence map IC1*, IC2*, IC3*.

The advertising campaign therefore increases the gradient of the indifference

curves. The more effective the campaign, the greater the increase in MRSX,Y
and the steeper the gradient. Point a is no longer a point of equilibrium on
IC2* as MRSX,Y is greater than Px/PY. The consumer regains equilibrium by
moving to point b on IC3*. The advertising campaign therefore allows the
firm to sell an additional X2 – X1 units at the original price.

Alternatively, the firm might wish to use the advertising to enable it to sell
the original quantity X1 at a higher price. On Figure 4-7 the price of X could
be raised, pivoting the budget line to AC, allowing the consumer to obtain
equilibrium at point c on the lower indifference curve IC1*. Finally, the firm
might raise price less than the above and sell between X1 and X2. Note that,
in reality, the firm might have difficulty in accurately predicting both the
influence of advertising upon sales and the price increase necessary to main-
tain sales following the advertising. Nevertheless, our basic analysis remains
valid.

We saw above how the firm uses advertising to increase consumer prefer-
ence and raise MRS relative to other goods. In many instances, however,
advertising might seek to maintain market position in the face of competitive
advertising or other marketing initiatives from rival firms. For example,
firms X’s advertising might have been in retaliation to a simultaneous cam-
paign by firm Y which could have tilted consumer preferences towards Y
and away from X, make the indifference map shallower, decrease MRSX,Y
and increase the consumption of Y at the expense of X. Firm X’s retaliatory
advertising campaign might aim as a minimum to maintain the status and
the original indifference map, or hopefully on balance shift preference to-
wards X and increase MRSX,Y.

58
 4.6
Revealed Preference

4.6 Revealed Preference

Suppose you do not know the consumer’s indifference map. Can you infer
how he ranks baskets by observing his behavior as his budget line changes?
The main idea behind revealed preference is simple. We have observations
about consumer choice with two different budget lines:

1. When the budget line is BL1, the consumer chooses basket A.

2. When the budget line is BL2, the consumer chooses basket B.

What does the consumer’s behavior reveal about his preference?

Revealed Preference Figure 4-8

G
F Preference
directions

●C
C, units of clothing

●B
●
A J
●
E

BL2
BL1
H
●

h, units of housing

First, the consumer chooses A when he could afford any other basket on or
inside BL1. Therefore, A is at least as preferred as B (A ≥ B). But we can make
an even stronger statement about how the consumer ranks A and B. Consid-
er basket C, on BL1 to the “northeast” of B. The consumer chooses A when he
can afford C; therefore A ≥ C. Moreover, since C lies to the “northeast” of B,
C must be strongly preferred to B (C ≥ B). Then by transitivity A must be
strongly preferred to B (A ≥ B).

59
4
Consumer Choice

All baskets to the north, east, or northeast of A are strongly preferred to A

(including baskets in the dark shaded area). A is strongly preferred to all
baskets in the light shaded region, strongly preferred to any basket on the
segment EH, and at least as preferred as any other basket between F and E.
Therefore, although we do not know exactly where the indifference curve
through A lies, it must pass somewhere through the white area between GAJ
and FEH, perhaps including baskets on EF.

4.7 Learning by doing

1. The budget constraint is the limit imposed on household choices by the
household’s

(a) expectations about future income, wealth, and prices.

(b) wealth.

(c) income, wealth, and prices.

(d) preferences and monthly spending plan.

2. Gretchen’s opportunity set can be increased by

(a) a decrease in prices.

(b) a decrease in income.

(c) an increase in quantity demanded.

(d) a decrease in quantity demanded.

3. Jenny the Junior buys only two goods: meals and movies. This semester
she has $500 in a savings account that she plans to spend, but no income.
Meals cost $5 each and movies cost $10 each. Jenny’s maximum spending
on meals is and her maximum spending on movies is .

(a) $500; $0

(b) $0; $500

(c) $250; $250

(d) $500; $500

60
 4.7
Learning by doing

4. Jenny the Junior buys only two goods: meals and movies. This semester
she has $500 in a savings account that she plans to spend, but no income.
Meals cost $5 each and movies cost $10 each. Jenny can buy a maximum
of meals and a maximum of movies.

(a) 100; 0

(b) 100; 100

(c) 0; 50

(d) 100; 50

5. Good A and Good B cost $3 and $4, respectively. Jo, spending all of her
income, buys 4 units of Good A and 3 units of Good B. The final unit of
each good gives her 12 units of utility.

(a) Jo is maximizing her utility.

(b) Jo should buy more of Good A and less of Good B to maximize utility.

(c) Jo should buy less of Good A and more of Good B to maximize utility.

(d) Jo should buy less of both goods to maximize utility.

6. A given budget constraint will swivel

(a) out if the price of one of the goods increases.

(b) in if income falls.

(c) out if income falls.

(d) in if the price of one of the goods increases.

7. If prices double and income doubles, the budget constraint will

(a) double.

(b) move inward.

(c) move outward by 50%.

(d) not shift position.

8. Illustrate the budget constraint for each case below. For each case, write
down relative prices (slope); the maximum amount of X that can be pur-
chased, and the maximum amount of Y that can be purchased.

61
4
Consumer Choice

(a) Px = 1, Py = 2, Income = 50

(b) Px = 1, Py = 2, Income = 80

(c) Px = 2, Py = 1, Income = 50

(d) Px = 1 for the first two units, then Px = 0.5 for each extra unit over 2,
Py = 2, Income = 50

9. Optimal Choice

(a) At current consumption levels, Joe’s marginal utility from beer is 5

and his marginal utility from Coke is 4. The price of a can of beer is $2
and the price of a can of Coke $1. Is Tony’s consumption choice at an “in-
terior” optimum? Explain. If Joe could, would he want to trade beer for
Coke, Coke for beer, or stay at the current consumption point?

(b) Suppose that beer and Coke cost the same as they did in part (a).
Could Joe be consuming optimally if, for Joe, beer and Coke are perfect
substitutes for Joe, what mix of beer and Coke would you expect to be
consumed?

10. Utility Maximizing

(a) When prices are (p1,p2) = (3,2), a consumer demands (x1,x2) = (0,6) and
when prices are (q1,q2) = (2,3), a consumer demands (y1,y2) = (6,0). Can
you tell from this information which bundle (X or Y) is preferred by the
consumer?

(b) Is the behavior of part (a) consistent with the model of utility maxim-
izing behavior?

(c) How would your answers to (a) and (b) change if the bundle con-
sumed at prices (p1,p2) were (x1,x2) = (4,0) and the bundle consumed at
prices (q1,q2) were (y1,y2) = (1,4)?

62
 4.7
Learning by doing

5 Theory of Demand

Chapter objectives:
1. Draw and explain the meaning of demand curves and Engel curves.

2. Distinguish and calculate the income effect and the substitution effect.

3. Analyze the shape of the labor-supply curve using the income and sub-
stitution effects.

4. Identify consumer surplus.

5. Derive the market demand curve from individual demand curve. Ex-
plain the effects of network externalities on the demand curve.

63
 5
Theory of Demand

5.1 Individual Demand Curves and Engel

Curves
Price consump- Similarly the market demand curve, the individual demand curve traces out
tion curve an individual’s desired consumption of a good as the price of that good
changes holding income and all other prices constant. The price consump-
tion curve depicts the bundles of goods demanded when the price of one of
the goods changes ceteris paribus.

Now let’s see Figure 5-1(a) and Figure 5-1(b) (Note that value of vertical axis is
units of clothing in Figure 5-1a and price of food in Figure 5-1b; value of horizontal
axis is units of Food in both Figure 5-1a and Figure 5-1b).

Figure 5-1 The Effects of Changes in the Price of a Good Consumption

Units of Clothing

(a)

A Price consumption curve

CA C
CC
B
CB

PF=4 PF=2 PF=1

▪ ▪ ▪
FA FB FC Units of Food

4 A’ (b)
Price of Food

2 B’

1 C’

0 FA FB FC Units of Food

64
 5.1
Individual Demand Curves and Engel Curves

Suppose that, given a set of prices and an income level, the consumer opti-
mally chooses basket B of food (with quantity: F B and price: PF = 2) and cloth-
ing in Figure 5-1(a). Now, let the price of food increase (from PF = 2 to PF = 4),
keeping the price of clothing and income fixed. As we saw in Chapter 4, the
budget constraint of the consumer rotates in. Given this new budget con-
straint, basket A is now optimal. If the price of food were to fall (from P F = 2
to PF = 1), while the price of clothing and income remained fixed, the budget con-
straint of the consumer would rotate out so that basket C would be optimal.
The curve joins baskets A, B, C called a price consumption curve.

Because we know how much food is contained in each basket that is pur-
chased, we can also read off this graph the quantity of food desired for every
price of food that the consumer can face. We can then graph the demand
curve of this consumer, as in Figure 5-1(b). This demand curve joins A’, B’,
C’.

The income consumption curve

When the price of food changes, the consumer’s desired quantity moves
along the demand curve for food, but what if the income of the consumer
changes? The income consumption curve depicts the bundles of goods de-
manded at different income levels ceteris paribus.

Now, let’s see Figure 5-2(a) and Figure 5-2(b). Starting from basket B, that is Income consump-
the optimal basket with the income I2. When the income decreases, becomes tion curve
I1, the consumer would move to a new optimal basket such as A. Similarly, if
we were to increase income for the consumer, becoming I3, the consumer
would move to a new optimal basket such C. The curve that connects points
A, B, C is called the income consumption curve. The amount of food con-
sumed in each of these baskets is plotted in Figure 5-2(b) against the (fixed)
price of food so that points A’, B’, C’ correspond to points A, B and C. For
each of these levels of income, we can observe the changes in the optimal
basket purchased as the price of food changes and trace out a demand curve
for each of these levels of income through points A’, B’, and C’.

Figure 5-2

65
 5
Theory of Demand

The Effects of Changes in Income on Consumption

(a)

Units of Clothing
Income consump-
tion curve
C
CC
B
CB
A
CA
I1 I2 I3
10 15 20
Units of Food
Price of Food

(b)
A’ B’ C’
$2

D3
D2
D1

0 10 15 20
Units of Food

Case Study 5-1 Elasticity of Demand for Cable Television (TV) Subscriptions

The cable TV industry is one of the most important sources of programming

for households in the United States. Competitors include traditional broad-
cast stations, direct broadcast satellites, wireless cable, and video cassettes.
However, about two-thirds of all households subscribe to cable TV.

66
 5.1
Individual Demand Curves and Engel Curves

Public policy toward the cable TV industry has changed repeatedly during
the last two decades. In 1984 the industry was deregulated, and cable sys-
tems rapidly expanded the services they offered. However, by the early
1990s, Congress was concerned that local cable operators were charging
unacceptably high prices, and that many homeowners lacked adequate
access to alternative programming. In 1992, over President Bush’s veto, Con-
gress passed a sweeping set of regulations for the industry. However, in 1996
Congress again removed regulation from much of the cable TV industry,
recognizing that competition to provide programming had increased.
Public policy debates on this subject often focus on the nature of the demand
for cable TV. How much will consumers pay for basic cable TV services?
How sensitive are consumers to changes in the prices charged? In a study of
the demand for cable TV with data from 1992, Robert Crandall and Harold
Furchtgott-Roth found the price elasticity of demand to be about -0.8 for the
basic service offered by a typical cable TV system. Thus, a 10 percent in-
crease in the price of a basic subscription would lead to a loss of 8 percent of
the subscribers. Some of those who drop their subscriptions might opt for
other forms of programming, while other might choose no programming at
all. As competition from other sources of programming (including the Inter-
net) intensifies over time, the demand for cable TV will become more elastic.

The Engel curve

Sometimes, it is convenient to ralate income to the consumer’s desired quan- Engel curve
tity of food directly. This relation is called an Engel curve. An Engel curve
relates the quantities of a commodity demanded to different income levels
ceteris paribus. This time, the prices of all goods remain fixed, but the in-
come of the consumer changes.

Let’s see Figure 5-3(a) and Figure 5-3(b). As income changes, the optimal
consumption basket will change. As in the case of individual demand, utility
is maximized along the Engel curve and increases as we move to higher
income levels.

Figure 5-3

67
 5
Theory of Demand

Engel Curve

Monthly consumption
(a)

of other goods C

I1 I2 I3
10 15 18
Hot dogs per month
Monthly income

(b)

$400 C’

$300 B’

A’
$200

0 10 15 18 Hot dogs per month

Normal good/ When the income consumption curve has a positive slope, the optimal con-
Inferior good sumption of the good increases with income. Such a good is called a normal
good. If an increase in income causes the optimal consumption level of a
good to decrease, that good is called an inferior good. A single good may
have ranges of income for which it is normal and other ranges for which it is
inferior. In Figure 5-3b, the good is normal until income $300, after which it
is inferior.

68
 5.2
Income and Substitution Effect

5.2 Income and Substitution Effect

A decrease in the price of one good has two effects on the consumer’s choice Two effects of
of an optimal basket. price changes

 First, the real purchasing power of the consumer rises. The effect of this Income effect
increase in real income on the quantities of each good chosen by the con-
sumer is called an income effect. The income effect may be positive (in
the case of a normal good) so that the consumer desires to purchase more
of the good as purchasing power rises or negative (in the case of an infe-
rior good) so that the consumer desires to purchase less of the good as
purchasing power rises.

 Second, a decrease in the price of one good makes that good relatively Substitution
cheaper. This leads the consumer to adjust his optimal consumption bas- effect
ket to contain more of the relatively cheaper good. In this sense the
change in price also has a substitution effect. The sign of the substitution
is not ambiguous: it will always cause substitution toward the good that
has become relatively cheaper. Since these two effects usually occur to-
gether in response to a price change, a move along the demand curve will
be the net result of an income and a substitution effect.

Note that when we refer to an income effect, we are not referring to the ef-
fect on demand of a change in the consumer’s exogenous income I (which
would cause the demand to shift). Rather, we are referring to a secondary
effect of a change in the price of the good, which affects the purchasing
power of the consumer and helps explain moves along the demand curve.

More precisely the (own) substitution effects is defined as the change in the
amount of a good consumed as the price of that good changes, holding con-
stant the level of utility, other prices and the consumer’s nominal income. This
is illustrated in Figure 5-4, where a decrease in the price of food changes the
equilibrium choice of the consumer from A to C.

The substitution effect measures the move from A to B along the indifference Decomposition
curve that was attainable at the original basket. B must be on the same indif- basket
ference curve as A in order to hold the level of utility constant. B must also be at
a point where the slope of that indifference curve is equal to the new relative
price of food. In other words, B is the basket that minimizes the expenditure
required to remain at the original level of utility under the new price regime.
Point B is often referred to as the decomposition basket: it is a fictitious
construction used to analyze the change in demand and not a basket that is
on the demand curve of the consumer in reality.

69
 5
Theory of Demand

Figure 5-4 Income and Substitution Effects

Clothing

●A

B ●C
●
U2
U1

BL1 BL2
0 FA FB FC Food

The income effect accounts for move from point B to the new equilibrium
position C. It is measured by the shift in purchasing power that would be
required to move the budget constraint out from the level that goes though
point B to the level that goes through point C. In other words, once the effect
of the change in relative prices has been taken into account, we need to take
into account the additional effect of the change in prices on purchasing pow-
er. As we saw above, an increase in purchasing power will cause an increase
in consumption of a good if the good is normal, but will cause a decrease in
consumption of the good if the good is inferior.

Giffen good If the good is so inferior that the net effect of a price decrease is to decrease
the demand for the good, then the good is referred to as a Giffen good. This
good is so strongly inferior that the income effect outweighs the substitution
effects, resulting in an upward sloping demand curve over some region of
prices

70
 5.2
Income and Substitution Effect

How to calculate income and substitution effects numerically?

The method we use to calculate income and substitution effects on X due to

a change in the price of X from an initial level, PX, to a final level, PX’, for a
given level of income, I, a fixed price of good Y, PY, and a utility function,
U(X,Y), is the following:

1. Calculate the original consumption basket by solving the original budget

constraint, PXX + PYY = I and the tangency condition at the original pric-
es, MUX/MUY = PX/PY for unknowns X and Y. Call the answer to this cal-
culation the original optimal consumption basket (XA,YA).

2. Substitute (XA,YA) into the utility function to get the utility level at the
original optimal basket, (XA,YA). Call this utility level UA.

3. At the new price for good X, PX’, calculate the tangency condition when
the utility is at the original level, UA by solving equations MUX/MUY =
PX’/PY and UA = U(X,Y). Call the answer to this calculation the decompo-
sition basket (XB,YB).

4. At the new price of good X, PX’, calculate the optimal consumption bas-
ket by solving the new budget constraint PX’X + PYY = I and the tangency
condition, MUX/MUY = PX’/PY. Call the answer to this calculation the final
consumption basket (XC,YC).

5. The substitution effect is the change XB - XA. The income effect is the
change XC - XB.

Calculating Income and Substitution Effects Case Study 5-2

Problem

A consumer’s tastes are represented by U = cb, with MUb = c and MUc = b.

Calculate the income and substitution effects of a move from prices pb = 1
and pc = 2 to prices pb = 2 and pc = 2. Income, I = 20.

Solution

We use above steps to answer the question. The original consumption basket
is obtained by simultaneously solving.

Step 1:

b + 2c = 20
MU b p b c 1
 or 
MU c pc b 2

71
5
Theory of Demand

Calculate two equations, we have the original consumption basket contains:

b = 10 and c = 5.

Step 2: Substitute (b = 10; c = 5) into the utility function: U = cb and get U =

5x10 = 50.

Step 3: At the new price pb = 2, calculate the tangency condition when utility
is at the original level, U = 50:

The decomposition basket can be obtained by simultaneously solving the

c 2
equations:  and bc = 50, we have: b = 501/2 and c = 501/2.
b 2

Step 4: Calculate the optimal consumption basket by solving the new budget
c 2
constraint 2b + 2c = 20 and the tangency condition,  . The final con-
b 2
sumption basket is (b = 5; c = 5); U = 25.

Step 5: The substitution effect on b is measured by the difference between

the amount of b consumed in the decomposition basket and the amount
consumed in the original consumption basket: ∆b = 501/2 – 10 = -3 (approxi-
mately). In other words, as the price of b rises, the substitution effect is that
the consumer chooses to include less of b in the optimal consumption bas-
ket. The income effect on b is measured by the difference between the
amount of b consumed in the final consumption basket and the amount
consumed in the decomposition basket: ∆b = 5 - 501/2 = -2 (approximately). In
other words, as the price of b rises, the income effect is that the consumer
chooses to include less of b in the optimal consumption basket. The two
effects together add up to -5: the total change in the consumption of b as we
move from the original basket to the final consumption basket. As an aside,
we can see from the sign of the income effect that b is a normal good.

5.3 Applications of the Decomposition into

Income and Substitution Effect

5.3.1 The Labor-Leisure Tradeoff

A household chooses how much labor to supply (and how much leisure and
unpaid work time to relinquish). The opportunity cost of paid work is lei-

72
 5.3
Applications of the Decomposition into Income and Substitution Effect

sure and unpaid work. Think of the wage rate as the “price” of unpaid work
or leisure.

 If the wage rate rises, the substitution effect encourages supplying addi-
tional labor (the opportunity cost of leisure is now higher), but

 The income effect discourages additional work (higher income makes the
worker want to increase consumption of normal goods, including lei-
sure).

When the income effect overwhelms the substitution effect, the labor-supply
curve bends backwards.

Let’s divide the day into two parts, the hours when an individual works and
the hours when he pursues leisure. Why does the consumer work at all?
Because he works, he earns an income, and he uses the income to pay for the
activities he enjoys in his leisure time. The term leisure includes all non-work
activities, such as eating, sleeping, recreation, and entertainment. We assume
that the consumer likes leisure activities.

Let’s suppose the consumer chooses to work L hours per day. Since a day has
24 hours, the time available for leisure will be the time that remains after
work, that is, 24 – L hours. The consumer is paid an hourly wage rate w.
Thus, his total daily income will be wL. He uses the income to purchase
units of a composite good measured by y. The price of each unit of the com-
posite good is $1.

The consumer’s utility U depends on the amount of leisure time and the
number of units of the composite good he can buy. We can represent the
consumer’s decision on the optimal choice diagram in Figure 5-5. On the
horizontal axis, we plot the number of hours of leisure each day, which must
be no greater than 24 hours. On the vertical axis we represent the number of
units of the composite good that he may purchase from his income. Since the
price of the composite good is $1, the vertical axis also measures the con-
sumer’s income.

Figure 5-5

73
5
Theory of Demand

Optimal Choice of Labor and Leisure

600 -

Daily Income and Units of the Composite Good w = 25

480 -

w = 20

360 -
w = 15 U5
U4
U3
240 - I
w = 10 ●H ●
U2
●G

120 - w=5 U1 F
●

E
●
A
●
0 13 14 15 16 24
Hours of Leisure

To find an optimal choice of leisure and other good, we need a set of indif-
ference curves and a budget constraint. The figure shows a set indifference
curves for which the marginal utility of leisure and the composite good are
both positive. Thus: U5 > U4 > U3 > U2 > U1. The indifference curves are
bowed in toward the origin, so there is also a diminishing marginal rate of
substitution.

The consumer’s budget line for this problem will tell us all the combinations
of the composite good y and hours of leisure (24 - L) that the consumer can
choose. If the consumer does no work, he will have 24 hours of leisure, but
no income to spend on the composite good. This corresponds to point A on
the budget line in the graph.

The location of the rest of the budget line depends on the wage rate w. Sup-
pose the wage is $5 per hour. This means that for every unit of leisure the
consumer gives up to work, he can buy 5 units of the composite good. The
budget line will have a slope of -5. If the consumer works 24 hours per day,

74
 5.3
Applications of the Decomposition into Income and Substitution Effect

his income would be $120 and could buy 120 units of the composite good.
This is B on the budget line. The consumer’s optimal choice will then be at
basket E. The diagram tells us that when the wage rate is $5, the consumer
will work 8 hours.

For any wage rate, the slope of the budget line is –w. In the figure, budget
lines are drawn for five different values of the wage rate ($5, $10, $15, $20
and $25). The graph shows the optimal choice for each wage rate. As the
wage rate rises from $5 to $15, the number of hours of leisure falls. However,
as the wage rate continues to rise, the consumer begins in increase his choice
of leisure time.

5.3.2 Consumer Surplus

Income and substitution effects also help us evaluate measures of consumer Consumer
welfare which refers to the individual benefits derived from the consump- welfare
tion of goods and services. Consider a proposal for a new government pro-
gram that changes the relative prices of goods. The government wishes to
know whether the benefits to consumers of the program justify its costs. In
order to answer this question, we need to have some measure of how much
better off consumers are when they make a purchase at the new prices com-
pared to the old prices. We can use the theory of demand that we have de-
veloped in order to answer this question.

The net economic benefit to the consumer from a purchase is the difference
between the maximum amounts the consumer would be willing to pay for a
purchase minus the actual amounts the consumer does have to pay (the
price). For example, if you are willing to pay $10 for a good, but the actual
price is only $5 you realize a net economic benefit of $5 on the purchase. To
compute such a difference we need to know how much the consumer would
be willing to pay for a purchase. This information can be obtained from the
individual demand curve.

Although we usually interpret the demand curve as telling us how many

units of the good the consumer would wish to buy at a given price, we can
also read it as telling us how much a consumer would be willing to pay to
purchase an additional unit of the good. For example, in Figure 5-6, the con-
sumer is willing to pay $10 for the first unit of the good, but only $9 for the
second, $8 for the third, and so on. Notice that the demand curve shows the
willingness to pay for additional or marginal units. This is because the con-
sumer chooses this optimal consumption basket to equate his marginal will-
ingness to exchange one good against the other to the market’s rate of ex-
change of one good for the other.

75
5
Theory of Demand

In other words, since demand is derived from the consumer’s optimization

problem, the marginal rate of substitution equals the price ratio all along the
demand curve. It follows that the consumer’s total willingness to pay for a
given quantity Q of a good is equal to the sum of his marginal willingness to
pay for each individual unit up to Q. In Figure 5-6, for example, the total
willingness to pay for five units of the good would be equal to $10 + $9 + $8 +
$7 + $6 = $40. This corresponds to the area underneath the demand curve
and to the left of the total quantity consumed Q.

Figure 5-6 Consumer Surplus

10
9
8
7
6
5

0 Q

Consumer Since the consumer pays the same market price for all of the units consumed,
surplus the actual sum paid is equal to the market price P times Q. In Figures 5-6, the
consumer would pay $6 for 5 units. This amounts to an expenditure of 5 x $6
= $30 and corresponds to the area underneath the price line and to the left of
Q. We can, then, think of the area under an ordinary demand curve and
above the price as a measure of the net economic benefit (or consumer sur-
plus) of a purchase. The consumer surplus is the difference between the
maximum amount a consumer is willing to pay for a good and the amount
he must actually pay when he purchases it. In our example, this net benefit
would be equal to $40 - $30 = $10.

This benefit can then be used to calculate whether a program that stimulates
purchase of a good is worth undertaking by measuring the consumer sur-

76
 5.4
Market Demand

plus before and after the change in prices and comparing the increase to the
cost of the program.

5.4 Market Demand

Suppose that a business wishes to sell a product for which there is infor-
mation on demand from two different market segments. If the business is
interested in the total market demand for the product, it must know how to
aggregate the information on the individual demands for the market seg-
ments into a total market demand curve. The market demand tells us the
total quantities demanded for any given price. This total quantity is equal to
the quantity demanded by the first segment at this price plus the quantity
demanded by the second segment at this price. In other words, the market
demand curve is the horizontal sum of the individual (or the segment) de-
mands.

Market and Segment Demand Curves Figure 5-7

P, (dollars per liter)

$5 Demand
(Casual Consumer)
4

Demand
A
3 ● (Health-Conscious
Consumer)
2
Market Demand
1

0 4 6 12 15 16 21

Q, (liters of orange juice per month)

77
 5
Theory of Demand

In the example of Figure 5-7, only two consumers of orange juice are in the
market, a casual consumer and a health-conscious consumer. The market
demand curve (the bold curve) is found by adding the demand curves for
the individual consumers horizontally. For example, if the price is $1 per
liter, the casual consumer will buy 4 liters per month and the health con-
scious buyer will purchase 12 liters; the total monthly demand for orange
juice will be 16 liters.

5.5 Network Externalities

We say that there are network externalities when a consumer’s demand for a
good depends on the number of other consumers also purchasing the good.
There are positive network externalities if a consumer’s demand increases
with the number of other buyers and negative network externalities if a
consumer’s demand decreases as the number of other buyers increases.

There are three main sources of positive network externalities:

 Physical network externalities: these arise mostly in communication and

transportation networks such as telephony, the Internet or railroads,
where consumers want to be able to reach as many other consumers or
destinations as possible.

 Virtual network externalities arise because the possibilities of exchange

or the quality and variety of complementary products that is available
increase with the number of consumers purchasing the good. A typical
example is video-game platforms: the more Nintendo consumers there
are the better the possibilities of borrowing games and the better the of-
fering of Nintendo-compatible games.

 Fads, that is, instances when consumers derive psychological benefits

from buying what everyone else buys.

Negative network externalities come mostly from snob appeal, that is, situa-
tions where consumers derive psychological benefits from purchasing goods
that are only bought by very few others.

Bandwagon Positive network externalities make the market demand curve more elastic.
effect As for any non-Giffen good, a decrease in price leads to an increase in the
quantity demanded. With positive network externalities, however, this in-
crease itself makes the product more attractive so that sales expand further.
This extra positive effect on the quantity demanded is called the bandwagon
effect. With negative network externalities, the bandwagon effect is nega-
tive, making the market demand curve less elastic.

78
 5.6
Learning by doing

5.6 Learning by doing

1. The substitution effect occurs when

(a) a decrease in the price of Good A makes the good relatively cheaper
and encourages consumers to buy more.

(b) a decrease in the price of Good A encourages consumers to buy more

of substitute Good B.

(c) a decrease in the price of Good A makes consumers better off so that
they can buy more of the good.

(d) an increase in the price of Good A encourages consumers to buy less

of substitute Good B.

2. The income effect helps to explain why

(a) the demand curve for a normal good shifts to the right when income
increases.

(b) the quantity demanded of a good increases when the price of that
good decreases.

(c) the demand curve of an inferior good shifts to the left when income
decreases.

(d) normal goods have higher prices than inferior goods do.

3. The substitution effect helps to explain why when the price of Good A
rises.

(a) sellers switch production and increase quantity supplied of Good A

(b) the demand curve for Good A is sloped as it is

(c) demand for another Good B, rises

(d) price elasticity increases along the curve

4. The price of Froot Loops cereal falls. Consumers switch over from other
cereals to Froot Loops because its price is relatively lower. This is the in
operation.

(a) income effect

(b) substitution effect

79
5
Theory of Demand

(c) ceteris paribus effect

(d) quantity demanded effect

5. The price of Froot Loops cereal falls. Wanda finds that she has some
money left over after buying her usual quantity of Froot Loops. She
spends some of the extra money on more Froot Loops. This is the

(a) income effect in operation.

(b) substitution effect in operation.

(c) normal effect in operation.

(d) inferior effect in operation.

6. The wage rate rises. If the quantity of labor supplied falls, the most likely
explanation is that the substitution effect

(a) and the income effect are both positive.

(b) is negative and the income effect is positive.

(c) and the income effect are both negative.

(d) is positive and the income effect is negative.

7. In Crantown, an increase in the wage of deckhands results in an increase

in the number of deckhands making themselves available for work. This
indicates that the

(a) income effect is stronger than the substitution effect.

(b) substitution effect is stronger than the income effect.

(c) income effect is positive.

(d) substitution effect is negative.

8. Jill is maximizing her utility. The price of Good A falls. Jill will

(a) buy more of Good A because it is relatively cheaper—the substitution

effect.

(b) buy less of Good A because her marginal utility is diminishing.

(c) buy more of Good A because her marginal utility is increasing.

80
 5.6
Learning by doing

(d) buy more of Good B—the substitution effect.

9. Individual demand for an inferior good that is not a Giffen good is:

(a) Downward-sloping

(b) Upward-sloping

(c) Flat

Whereas individual demand for a Giffen good is:

(a) Downward-sloping

(b) Upward-sloping

(c) Flat

10. A consumer who purchases two goods, food and clothing, has the utility
function U(x,y) = xy, where x denotes the amount of food consumed and
y the amount of clothing. His marginal utilities are MUx = y and MUy =
x. Suppose that he has an income of $72 per week and that the price of
clothing is Py = $1 per unit. Suppose that the price of food is initially Px1
= $9 per unit, and that the price subsequently falls to Px2 = $4 per unit.

Find the values of income and substitution effects on food consumption.

81
6
Inputs and Production Function

6 Inputs and Production

Function

Chapter objectives:
1. Explain how the production function relates inputs to outputs. Define
the production set of a firm.

2. Understand and graph the isoquants of a given production function.

3. Explain the relationships between marginal, total, and average product

and, given a production function graph, interpret the behavior of mar-
ginal and average product.

4. Define the marginal rate of substitution and explain its relationship to

the marginal utility.

5. Explain the concepts of returns to scale.

82
 6.1
Inputs and Production Function

6.1 Inputs and Production Function

Production of goods and services involves transforming resources – such as Transforming
labor power, raw materials, and the services provided by facilities and ma- resources
chines – into finished products. The productive resources, such as labor and
capital equipment, that firms use to manufacture goods and services are
called inputs or factors of production, and the amount of goods and ser-
vices produced is the firm’s output.

The maximum possible output that can be attained by the firm for any given Production
quantity of inputs is given by the production function of the firm. A produc- function
tion function is a mathematical or numerical representation of the relation-
ship between inputs and outputs. Mathematically, the production function
can be written as follows:

Q = f(K,L)

Where Q is the output, K is capital input, L is labor input. This expression

tells us that the maximum quantity of output the firm can get depends on
the quantities of capital and labor it employs for a given time period. There-
fore, labor is measured in terms of man-hours used over this period, and
capital is measured as the machine-hours used to produce the output over
the same period of time. We could have listed more categories of inputs, but
many of the important tradeoffs that real firms face involve choices between
capital and labor. Moreover, we can develop the main ideas of production
theory using just these two categories of inputs.

The production function in above equation tells us the maximum quantity of

output a firm could get from a given combination of capital and labor. Of
course, inefficient management could reduce output from what is technolog-
ically possible.

It’s essential to differentiate between the long run and the short run produc- Long run/
tion function. Short run

 The long run is a time period long enough for the firm to alter any and
all of its factors of production (capital and labor).

 The short run is the time period less than that – the period in which each
firm has a fixed scale of production with at least one resource fixed in
quantity.

In the short run perspective, the amount of capital K (for example: number
of machines) is fixed so that a higher output can only achieved by a higher
amount of labor (overtime). The short run production therefore depends
only on L (labor), so that we have the short run production function

83
 6
Inputs and Production Function

Q = f(L)

Production set Figure 6-1 depicts this possibility by showing the production function for a
single input, labor: Q = f(L). Points on or below the production function
make the firm’s production set that is the technically feasible combinations
of inputs and outputs. Points, such as A and B, in the production set are
technically inefficient (i.e., at these points the firm gets less output from its
labor than it should). Points, such as C and D, on the boundary of the pro-
duction set are technically efficient. At these points, the firm produces as
much output as it possibly can be given the amount of labor it employs.

Figure 6-1 Technical Efficiency and Inefficiency

Technically
Q, quantity of output (units per year)

efficient

D Q = f(L)
●
C
●
●B
Technically
inefficient
●A

Production set

L (units of labor per year)

If we invert the production function, we get a function L = g(Q), which tells

us the minimum amount of labor that is required to produce a given amount
of output Q. This function is the labor requirements function.

6.2 Isoquant
The production function is, in many ways, analogous to the utility function
of the consumer: just as the utility function told us the preference level asso-

84
 6.2
Isoquant

ciated with various combinations of purchases, the production function tells

us the level of output associated with various combinations of inputs. The
utility function was derived from preferences, and the properties of these
preferences determined the properties of the utility functions. Similarly, the
production function is derived from technologies and the basic properties of
the technologies determine the properties of the production function.

In consumer theory, an indifference curve was the set of combinations of the Isoquant
goods that resulted in the same level of satisfaction for the consumer. In
production theory an isoquant tells us the set of all possible combinations of
the inputs that are just sufficient to produce a given amount of output. Each
point on an isoquant corresponds to a technically efficient combination of
inputs, that is, to a point on the production function.

To illustrate, let’s once again consider the production function described in

Table 6.1. From this table we see that two different combinations of labor and
capital, A (L = 6, K = 18), B (L = 18, K = 6), result in an output of Q = 25 units
(where each “unit” of output represents a thousand semiconductors). Thus,
each of these input combinations is on the Q = 25 isoquant.

Production Function for Semiconductors Figure 6-2

0 6 12 18 24 30

0 0 0 0 0 0 0

6 0 5 15 25 30 23
L 12 0 15 48 81 96 75

18 0 25 81 137 162 127

24 0 30 96 162 192 150

30 0 23 75 127 150 117

85
 6
Inputs and Production Function

K
Isoquants

18 A

●C Q = 30

B
6 Q = 25

0 L
6 18

Isoquants located farther to the northeast correspond to higher levels of

output. Like indifference curves, isoquants are convex to the origin, reflect-
ing the fact that average levels of inputs (point C) result in higher levels of
output than extremes (points A and B).

Like indifference curves, isoquants cannot cross, since crossing would be in

conflict with the assumption that output is produced efficiently. For exam-
ple, if an isoquant that produced 10 units of output crossed with an isoquant
producing only 5 units, then at the point of intersection the same input com-
bination would at the same time produce 5 units of outputs and 10 units!
The isoquant representing only 5 units of output could not, then, represent
an efficient use of inputs.

6.3 Marginal and Average Product

Maginal product The marginal product of an input is the change in output that results from a
small change in an input, holding the levels of all other inputs constant. For
example, we have:

 Marginal product of labor: MPL = ∆Q/∆L

(holding constant all other inputs)

 Marginal product of capital: MPK = ∆Q/∆K

(holding constant all other inputs)

Average product The marginal product of an input must be distinguished carefully from its
average product. The average product of an input is simply equal to the total

86
 6.3
Marginal and Average Product

output produced divided by the quantity of the input that is used in its pro-
duction:

 Average product of labor: APL = Q/L

 Average product of capital: APK = Q/K
Graphically, the total product, average product, and marginal product can be
related as follows for a production function Q = f(L), where Q is output and
L is labor:

Relationship among Total, Average, and Marginal Product Function Figure 6-3

Total
product
function

0 L

APL

MPL

APL

0 L
MPL

87
 6
Inputs and Production Function

Notes:

1. The slope of a ray from the origin attains its maximal value and be tan-
gent with Total product function (TPL), as APL reaches its peak.

2. TPL increases, decreases, or remains unchanged as MPL is positive, nega-

tive, or equal to zero.

3. APL increases, decreases, or stays the same as MPL is greater, smaller, or

equal to APL.

Law of The marginal product of an input typically declines as we use more and
diminishing more of this input. The law of diminishing returns states that, when extra
marginal returns units of a variable input are combined with fixed inputs, marginal product
will eventually decline. This is a short-run concept—in the long run there are
no fixed inputs. For example, the amount of wheat that can be grown on 20
acres of land increases drastically if one employs two workers than one, but
it would increase very little if the number of workers rises from 50 to 51. This
relation is so pervasive that it is referred to as the law of diminishing mar-
ginal returns, even though it is simply an empirical regularity and not a law
based on any theoretical principle.

Generally, we will only be concerned with the case of marginal products that
remain positive (even if declining). The region where the marginal product
becomes negative (so that the total returns to the input are decreasing) is
called the uneconomic region of production. No rational producer would
choose to operate in this region since one could at the same time increase
output and decrease cost by using less of the input with a negative marginal
product.

Figure 6-4 Economic and Uneconomic Regions of Production

Uneconomic
region

Economic
region
Q2
Q1

0 L

88
 6.4
Marginal Rate of Technical Substitution

In dark region, the backward-bending and upward-sloping regions are the

uneconomic region of production. In this region, the cost for K and/or L
increases while the total product is unchanged (isoquant Q1 or Q2).

6.4 Marginal Rate of Technical Substitution

The marginal rate of substitution measured the amount of good x the con- Marginal rate of
sumer would require in exchange for giving up a small amount of good y to technical substi-
remain just indifference between the new and the old basket. Similarly, the tution
marginal rate of technical substitution of labor for capital (MRTS) measures
the extra amount of input K the firm would require in exchange for using a
little less of input L in order to just be able to produce the same output as
before. The MRTS is equal to minus the slope of the isoquant in the same
way as the marginal rate of substitution was minus the slope of the indiffer-
ence curve:

MRTSL,K = -∆K/∆L (for a constant level of output)

= - slope of isoquant

Marginal Rate of Technical Substitution of Labor for Capital along an Isoquant Figure 6-5

MRTS changes
along isoquant
●

● Q = 100

0 L

89
 6
Inputs and Production Function

In the same way as the marginal rate of substitution could be related to the
marginal utility, the marginal rate of technical substitution can be related to
the marginal product of the inputs. A move along the isoquant is defined by
a change in L and K that leaves total output unchanged. Mathematically, a
small change in L will result in a total change in output of MP L(∆L) and a
small change in K will result in a total change in output of MP K(∆K). In order
to stay on the same isoquant, we want the sum of these changes in output to
be zero or,

MPL(∆L) + MPK(∆K) = 0

Rearranging terms, this gives us a relation between the marginal products

and the marginal rate of technical substitution:

MPL/MPK = MRTSL,K

If the marginal products decline as we use more of a single input, then the
marginal rate of technical substitution of labor for capital, MRTSL,K falls as
one moves down along the isoquant: as some capital is replaced by labor,
MPL decreases and MPK increases so that MRTS = MPL/MPK becomes small-
er. In other words, isoquants with the convex shape that we often see have a
diminishing marginal rate of technical substitution. Thus, as we increase
the use of labor, we must offset any increase in labor by an ever-smaller
decrease in capital in order to keep output constant along the isoquant. As
we decrease the use of labor, we must offset any increase in labor by an ever-
larger decrease in capital in order to keep output constant along the
isoquant.

6.5 Returns to Scale

When inputs have positive marginal products, a firm’s total output must
increase when the quantities of all inputs are increased simultaneously – that
is, when a firm’s scale of operations increases. The concept of returns to scale
tells us the percentage increase in output when a firm increases all of its
input quantities by a given percentage amount:

Returns to scale = %∆ (quantity of output)/ %∆ (quantity of all input)

Returns to scale: the proportionate increase in a country's or firm's output as

a result of increases in all its inputs.

Types of returns 1. If a 1% increase in all inputs results in a greater than 1% increase in out-
to scale put, then the production function exhibits increasing returns to scale
(IRS).

90
 6.6
Learning by doing

2. If a 1% increase in all inputs results in exactly a 1% increase in output,

then the production function exhibits constant returns to scale (CRS).

3. If a 1% increase in all inputs results in a less than 1% increase in output,

then the production function exhibits decreasing returns to scale (DRS).

Constant, Increasing, Decreasing Returns to scale Figure 6-6

8 Q4 = 60

4 Q3 = 50

2
Q2 = 20
1
Q1 = 10

0 1 2 4 8 L

In Figure 6-6, the returns to scale are constant from Q1 to Q2 (since doubling
the inputs doubles the output), increasing from Q2 to Q3 (since doubling the
inputs more than doubles the output) and decreasing from Q 3 to Q4 (since
doubling the inputs less than doubles the output).

6.6 Learning by doing

1. In the long run

(a) the majority of resources are not fixed.

(b) all firms will make positive economic profits.

91
6
Inputs and Production Function

(c) a firm can vary all of its inputs, but can’t change its mix of inputs.

(d) the firm can leave the industry if it so chooses.

2. In the short run, a firm

(a) can shut down and leave the industry.

(b) can shut down but cannot leave the industry.

(c) cannot shut down.

(d) cannot vary its output level.


3. A graph showing all the combinations of labor and capital that can be
used to produce a given level of output is called

(a) an isocost line.

(b) an isoquant.

(c) an isotech line.

(d) a production function.

4. Four workers produce 160 units of output and 5 workers produce 180.
The marginal product of the fifth worker is

(a) 32 units of output.

(b) 4 units of output.

(c) 20 units of output.

(d) 36 units of output.

5. The law of diminishing returns

(a) applies in the short run but not in the long run.

(b) requires that all factors of production must diminish in equal propor-
tions.

(c) requires that all factors of production must diminish in unequal pro-
portions.

(d) states that marginal product must always be less than average prod-
uct.

92
 6.6
Learning by doing

6. The short run is a period of time during which

(a) all resources are fixed.

(b) all resources are variable.

(c) the scale of production is fixed.

(d) the scale of production is variable.

7. If diminishing returns have set in, a firm that doubles the number of
workers will see total production

(a) decrease.

(b) less than double.

(c) more than double.

(d) decrease by 50%.

8. Consider the production function Q = f(L,K) = 4L + 3K, where L and K

are the amounts of labor and capital used respectively.

(a) Draw a typical isoquant.

(b) Calculate the marginal products and MRTS.

9. Let Q = 12(KL)2 – K4

(a) What is the average product of capital?

(b) Let L = 2. By plugging in values of K between 1 and 5, can you tell at

what level of K the average product of capital is the highest?

(c) Let L = 2. Using only the information you have been given so far, at
what level K does the marginal product of capital equal the average
product of capital?

10. Suppose that a production function can be written Q = 9KL 2 – L3.

(a) Graph the function for K = 1 with Q on the vertical axis and L on the
horizontal axis. On your graph, show the regions for which the margin-
al product of labor is rising, constant, falling, and negative. At what val-
ue of L does Q reach a maximum?

93
6
Inputs and Production Function

(b) For K = 1, show the regions for which the average product of labor is
rising, constant, falling, and negative.

(c) Calculate the formula for the average product of labor. At what level
of L does the average product of labor equal the marginal product of la-
bor?

(d) At what level of labor does the marginal product of labor become
negative? How does this relate to the point at which the total product is
at a maximum?

94
 6.6
Learning by doing

7 The Theory of Cost

Chapter objectives:
1. Define, distinguish, and explain cost concepts.

2. State how firms determine the cost-minimization solution.

3. State, distinguish long-run total cost curve and short-run total cost curve.

4. Explain the relationships among average and marginal cost in long-run

and short-run.

95
 7
The Theory of Cost

7.1 Cost Concepts for Decision Making

Before we can discuss the theory of costs, some difficulties about the proper
definition of "costs" must be cleaned up.

Opportunity cost

Consider a decision maker who has to choose from a set of mutually exclu-
sive alternatives, each of which entails a particular monetary payoff. The
opportunity cost of a particular alternative is the payoff associated with the
best of the alternatives that are not chosen.

Many examples of opportunity cost exist at the level of the individual, the
household, the firm, the government and the economy:

 The opportunity cost of deciding not to work is the lost wages foregone
 The opportunity cost of spending money on a foreign holiday is the lost
opportunity to buy a new dishwasher or the chance to enjoy two short
breaks inside the United Kingdom

 The opportunity cost of the government spending £20 billion on interest

payments on the national debt is the extra money it might have allocated
to the National Health Service

 The opportunity cost of an economy investing its resources in new capi-

tal goods is the current production of consumer goods that is given up

Explicit versus Implicit Costs

Explicit cost/ Opportunity costs do not necessarily involve direct outlays of cash by the
firm. For example, the opportunity cost of the time that an owner devotes to
Implicit cost
his business does not entail a cash outlay. For this reason, economists distin-
guish between explicit cost and implicit costs. Explicit costs involve a direct
monetary outlay. Purchases of labor or materials entail explicit costs. Implicit
costs do not involve outlays of cash. The opportunity costs associated with
the use of a firm’s capital assets are often implicit costs.

Economic versus Accounting Costs

Economic cost/ Closely related to the distinction between explicit and implicit costs is the
distinction between economic costs and accounting costs. Accounting
Accounting cost
statements typically show historical expenses: How much did the firm
spend on labor or on materials in a particular year? Accounting statements

96
 7.1
Cost Concepts for Decision Making

are designed to serve an audience outside the firm, such as lenders and
equity investors. Because accounting numbers must be objective and verifia-
ble, historical costs serve the purpose best. By contrast, economists empha-
size the use of costs for decision making. To economists, decision-relevant
costs are opportunity costs. Economic costs are therefore the sum of explicit
and implicit costs. By contrast, accounting costs include explicit costs but do
not include implicit costs.

Economic Costs versus Accounting Costs Case Study 7-1

Imagine that two years after receiving your college degree your annual sala-
ry as an assistant store manager is $28,000, you own a building that rents for
$10,000 yearly, and your financial assets generate $3,000 per year in interest.
On New Year’s Day, after deciding to be your own boss, you quit your job,
evict your tenants, and use your financial assets to establish a pogo-stick
shop. At the end of the year, your books tell the following story:
Total Sales Revenue $130,000
Cost of pogo sticks 85,000
Employees’ wages 20,000
Utilities 5,000
Taxes 5,000
Advertising expenses 10,000
Total (Explicit) Costs –125,000
Net (Accounting) Profit $5,000
Being in this business caused you to lose as income:
Salary –28,000
Rent –10,000
Interest –3,000
Total Implicit Costs –41,000
Net (Economic) Profit –$36.000
“Therefore, I’ve had an economic profit that’s negative, a loss of – 36,000.
This harebrained business is a loser!”

Sunk versus Non-sunk Costs

To analyze costs we also need to distinguish between sunk and non-sunk Sunk cost/
costs. When assessing the costs of a decision, the decision maker should
Non-sunk cost
consider only those costs that the decision actually affects. Some costs must
be incurred no matter what decision is made. They are costs that have al-
ready been incurred and cannot be avoided. These are called sunk costs. By

97
 7
The Theory of Cost

contrast, non-sunk costs are costs that are incurred only if a particular deci-
sion is made and are thus avoided if the decision is not made (For these
reason non-sunk costs are also called avoidable costs). When evaluating alter-
native decisions, the decision maker should ignore sunk costs and consider
only non-sunk costs.

For example, a baseball team decides whether to sign a player to a contract,

the value of this contract is a non-sunk cost. At the time the decision is being
considered, the team can avoid spending money for him. However, if this
team has signed a player to a guaranteed contract, then the player’s salary
should not affect the decision to make him a starter. Since the salary will
have to be paid anyway, the decision should only be based on the player’s
expected contribution to the team. In other words, the total payment to the
player under the guaranteed contract is a sunk cost and should be ignored.

Case Study 7-2 Managing a Major League Baseball Team

Sparky Anderson was one of the greatest managers in the history of major
league baseball. He managed the Cincinnati Reds from 1970 to 1978 and the
Detroit Tigers from 1979 to 1995. His teams won the World Series three dif-
ferent times (the Reds in 1975 and 1976 and the Tigers in 1984), and he is still
the only manager in baseball history who has won the World Series in both
the National and American Leagues.

Despite his greatness, though, late in his career, Sparky Anderson made a
critical “sunk cost” mistake in deciding who should pitch for the Detroit
Tigers. In late 1992, the Tigers signed a pitcher named Mike Moore to a
guaranteed contract of $5 million per year, an enormous amount for a pitch-
er whose career until then had been less than spectacular. No matter how
poorly Moore performed, no matter how little he pitched, the Tigers would
have to pay his $5 million annual salary. For the next three seasons, Moore
pitched very ineffectively. He lost more games than he won, and each sea-
sons his earned run average (ERA) was over 5.00, a poor performance by
major league baseball standards of that period. Still, between 1993 and 1995,
Sparky used Moore regularly. He explained to reporters that the Tigers were
paying this guy a huge salary, so they had to pitch him regularly.

This is a sunk cost fallacy. Sparky should have realized that the Tigers had to
pay Moore whether he pitched or not. In deciding whether to let Moore
pitch a game, Moore’s salary was a sunk cost. It should not have affected
Sparky’s decision.

98
 7.2
The Cost-Minimization Problem

Why did Sparky do this? It’s hard to say. He was a shrewd manager
throughout his career who rarely made decisions that hurt his teams. Maybe
he was doing what he felt top Tiger management wanted him to do, reason-
ing that it would make them look bad to pay a pitcher who doesn’t play on a
regular basis such a large salary. Or maybe he felt that Moore would eventu-
ally break out of his slump if he continued to pitch regularly. (He didn’t). But
if he really did continue to pitch Moore because of Moore’s salary, then there
is an important lesson here, even for those of you who don’t aspire to be
baseball managers. Don’t be like Sparky and let sunk costs influence your
decision.

7.2 The Cost-Minimization Problem

We will assume that the firm’s owners have decided to produce some output
level, Q0. He wants to minimize the cost of achieving this level of output
subject to the constraint that they may only produce in technologically feasi-
ble ways. This is an example of a constrained optimization problem. Alge-
braically, let the technological possibilities of the firm be represented by the
production function Q = f(L,K). The total cost (TC) of using L units of labor
at wage rate w per unit, and K units of capital at rental rate r per unit is: TC =
wL + rK. Now, we can state the decision problem as follows:
MinTC  rK  wL
K ,L

Subject to Q0 = f(K,L)

This can be read as, “Choose the levels capital and labor to minimize the
total cost of production subject to the constraint that output be Q 0 units”.

Let’s us remember the expenditure minimization problem, the consumer Isocost line
chose to minimize the total expenditures, subject to attaining a given level of
utility.

Graphically, the consumer shifted in his budget line as much as possible,

subject to the constraint of remaining on the target indifference curve. In the
cost minimization problem, the firm minimizes expenditures necessary to
attain a given level of output. This means that the firm will shift in his isocost
line, all the input combinations that lead to the same production costs, as
much as possible subject to the constraint of remaining on the isoquant that
corresponds to the target output, such as Q0 at the cost level TC0.

99
7
The Theory of Cost

TC0 = rK + wL

Rewriting this equation, we have:

K = (TC0/r) – (w/r)L

This is the equation of a line with vertical intercept TC0/r and slope –w/r
when K is measured on the vertical axis and L is measured on the horizontal
axis. Such a line is called an isocost line. The higher the cost level associated
with input levels (K,L), the farther out from the origin the isocost line is
(Figure 7-1(a)). A change in the wage rate, w, or in the rental rate of capital, r,
affects the slope of isocost line.

For example, if the wage rate rises, the isocost line rotates in around its verti-
cal intercept (Figure 7-1(b)). In other words, the maximum number of units
of labor that can be purchased at a given cost decreases while the maximum
number of units of capital that can be purchased at that cost remains un-
changed.

Figure 7-1 Isocost Lines

K
(a)
TC1/r

TC0/r
Isocost lines,

Slope = -w/r

0 TC0/w TC1/w L

TC0 < TC1

100
 7.2
The Cost-Minimization Problem

K
Isocost lines

TC0/r Slope = -w0/r

(b)

Slope = -w1/r

0 TC0/w1 TC0/w0 L

w1 > w0

Our constraint is that we must continue to produce at least output level Q 0.

The cheapest possible way of producing Q0 must, then, be the point on the
isoquant that lies on the lowest isocost curve. If the isoquant is a smooth
curve, then the solution to the constrained optimization problem will be at
the tangency point where the slope of the isoquant (the marginal rate of
technical substitution) exactly equals the slope of isocost line (the ratio of the
wage to the rental rate of capital):
MPL w
MRTS L ,K   
MPK r

MPL MPK
Or, rewriting this equation, 
w r

If we wanted to solve for the internal optimum algebraically, we would

simply note that our solution must both lie on the target isoquant and be at a
tangency point. These conditions give us two equations that we can use to
solve for the two unknowns, K and L:

Q0 = f(K,l)
MPL MPK

w r

Figure 7-2

101
 7
The Theory of Cost

Cost Minimization

K
Isocost line with slope –w/r

Solution point where

slope of isocost = slope of isoquant

K*

Isoquant Q0 = f(K,l) with

slope –MPL/MPK

0 L* L

Case Study 7-3 Cost minimizing combination

Problem

Let the production function be Q = K1/2L1/2 with MPL = (½)L-1/2K1/2 and MPK =
(½)K-1/2L1/2. Let w = 1 and r = 1. What is the cost minimizing combination of
labor and capital, (L*,K*), that yields an output level of 100?

Solution

We must solve two equations:

(1) 100 = K1/2L1/2

1 1 / 2 1 / 2 1 1 / 2 1 / 2
L K K L
(2) 2  2
1 1

Equation (2) simplifies to K = L. Substituting this into equation (1), we obtain

L = K = 100. You can check that both equations (1) and (2) hold at this point
by substituting K = L = 100 into both.

102
 7.2
The Cost-Minimization Problem

The discussion up to this point has only been concerned with long-run costs, Long-run/
that is, with the minimum cost of producing a given level of output when all
Short-run
factors of production can be varied as much as the firm would like. However in the
short-run, at least one of the factors of production is fixed.

For example, suppose that capital is fixed at level K . Then the short-run total
cost function (STC) can be written as: In production, short-run at the firm
level is the period when at least one input is not variable long-run at the
firm level is the period when all inputs are variable

STC  wL  rK

Where the total fixed cost is simply r K and the total variable cost is wL. Our Fixed cost/
problem now is much simpler than before: We must choose the level of labor
Variable cost
that minimizes the total cost of obtaining the target level of output, Q0, with
the constraint that capital must be set at K :

Min STC  wL  rK
L

Subject to Q0 = f(L, K )

The solution to this problem gives us a short-run input demand function for
labor. For isoquants that look like a smooth curve, there is only a single level
of labor that can be combined with this fixed amount of capital, K , to effi-
ciently produce the target level of output for this problem.

Shor- Run Input Demand Function Figure 7-3

Q = Q0

0 L = L(Q0, K ) L

103
7
The Theory of Cost

Short-run conditional factor demand function with more than two inputs

Suppose that there are N factors of production K1, K2,...KN that are fixed in
the short-run and two that are variable, L1 and L2. Define w1 and w2 as the
unit price of L1 and L2, respectively.

The equation of the isoquant corresponding to a level of output Q 0 is Q0 =

F(L1,L2;K1,K2,…KN). From this equation we can also obtain the marginal
product for each of the variable factors:

MPL1(L1,L2,Q0;K1,K2,…KN) and

MPL2((L1,L2,Q0;K1,K2,…KN)

We can then express the required tangency condition as w 1/w2 = MPL1/MPL2.

Solving this equation together with the equation of the isoquant gives the
short-run conditional factor demands

L1(Q0;K1,K2…KN) and

L2(Q0;K1,K2…KN).

7.3 Long-Run Total Cost Curve

The long-run total cost curve shows how the minimized total cost varies as
we vary output, while keeping factor prices fixed. Algebraically, we can
write total cost as:

TC = wL + rK

Where L and K are the conditional input demands. These input demands are
functions of the input price w, r and of the level of output, Q, that the firm
wishes to produce, that is:

L = L(w,r,Q)

K = K(w,r,Q)

Substituting these two functions into the equation for total cost give us:

TC = wL(w,r,Q) + rK(w,r,Q)

So that total cost is itself a function of w, r and Q. We can write this as:

TC = C(w,r,Q)

Where C(w,r,Q) is the long-run cost function of the firm.

104
 7.3
Long-Run Total Cost Curve

Minimizing Total Cost Case Study 7-4

Suppose that the production function is Q = 50L1/2K1/2.

Problem

(a) How does minimized total cost depend on the output Q and the input
prices w and r for this production function?

(b) What is the graph of the long-run total curve when w = 25 and r = 100?

Solution

(a) Let’s begin with the tangency condition MPL/MPK = w/r.

Using the expressions for the marginal products of labor and capital, we
have:
MPL K

MPK L

Thus, our tangency condition is:

K w
 or
L r
r
L K
w

Let’s now substitute this into the production function and solve for K in
terms of Q, w, and r:
1
 r 2
1
Q  50 K  K 2
w 

which yields the demand curve for capital:

1
Q  w 2
K  
50  r 

Since L = (r/w)K, it follows that the demand curve for labor is:
1
Q  r 2
L  
50  w 

To find the minimized total cost, we calculate the total cost the firm incurs
when it uses this cost-minimizing input combination:

TC = wL + rK

105
 7
The Theory of Cost

1 1 1 1
Q  r 2 Q  w  2 Q 21 21 Q 21 21 w 2 r 2
TC  w    r    w r  w r  Q (7.1)
50  w  50  r  50 50 25

(b) The below figure shows that the graph of the long-run total cost curve is
a straight line. We derive it by substituting w = 25 and r = 100 into expression
(7.1) to get:

TC(Q) = 2Q

TC TC(Q) = 2Q

0 1 2 Q

Since all factors of production are variable in the long-run, total cost must be
zero when output is zero, that is, the long-run total cost curve must go
through the origin of the graph. Producing a higher level of output involves
a higher cost: In order to reach a higher isoquant, one must move to a higher
(parallel) isocost line (Figure 7-4). This implies that the total cost function
must be increasing in the level of output.

Figure 7-4

106
 7.3
Long-Run Total Cost Curve

The Total Cost increases when the level of output increases

K
TC2/r

TC1/r

Q2

Q1

0 L

In demand theory, the demand curve showed how output demanded was
related to the price of output, all else held constant. As we changed the price,
we moved along the demand curve. Otherwise, when one of the variables
that had been held constant changed, the demand curve shifted. Similarly, the
total cost curve shows how total cost relates to output, all else held constant.
As we change the firm’s desired level of output from Q0 to Q1, we move
along total cost curve Q(w+r) from TC0 to TC1 in Figure 7-5.

On the other hand, when one of the variables that had been held constant
changes, the total cost curve shifts. For example, if the wage changes, the
total cost curve shifts. This need not be a parallel shift, however. Let us ex-
amine how total cost curve (w+r)Q changes as w changes. As w rises to w’
(>w), the total cost curve rotates upward: it still passes through zero, but as
Q rises and we use more and more labor, the total cost at the higher wage
level diverges more and more from the total cost at the lower wage level.

Figure 7-5

107
 7
The Theory of Cost

The total cost changes when output of production or price of input changes

TC
Total cost curve = (w’+r)Q

Slope = w’ + r
Total cost curve = (w+r)Q
TC1

TC0
Slope = w + r

0 Q0 Q1 Q

Long-run average and marginal cost

Long-run average cost is the firm’s cost per unit of output. It equals long-run
total cost divided by Q:
TC(Q)
AC(Q) 
Q

Long-run marginal cost is the rate of change at which long-run total cost
changes with respect to output:
TC(Q  Q)  TC(Q) TC
MC (Q)  
Q Q

Although long-run average and marginal cost are both derived from the
firm’s long-run total cost curve, the two costs are generally different. Aver-
age cost is the cost per unit that the firm incurs in producing an additional
unit of output.

Figure 7-6

108
 7.3
Long-Run Total Cost Curve

Average and Marginal Cost

AC, MC

MC AC

ACmin, AC(Q) = MC(Q)

AC is decreasing, AC is increasing,
so MC(Q) < AC(Q) so MC(Q) > AC(Q)

0 Q

Relationship between total, average, and marginal cost is the same as the
relationship between any “total”, “average”, ”marginal” magnitude. For
example, suppose that two students have handed in their tests and have
received marks of 85 and 90. The average is (85 + 90)/2 = 87.5. Now, you
hand in your test and receive a mark of 95. This last, or marginal, test is high-
er than the current average of 87.5 so that it raises the average class mark to
90 = (85 + 90 + 95)/3. On the other hand, if you had received a mark of 80,
your marginal test would have been below the current average and would
have lowered the average class mark to 80.

Hence we can conclude:

 When average cost is decreasing in quantity, marginal cost is less than av- Relationships
erage cost. That is, if AC(Q) decreases in Q, MC(Q) < AC(Q) between average
cost and margin-
 When average cost is increasing in quantity, marginal cost is greater than al cost
average cost. That is, if AC(Q) increases in Q, MC(Q) > AC(Q)

 When average cost neither increases nor decreases in quantity, either because
its graph is flat or we are at a point at which AC(Q) is at a minimum,
then marginal cost is equal to average cost.

109
 7
The Theory of Cost

Case Study 7-5 Calculating Average Cost

Problem

Let TC(Q) = 100Q – 5Q2 + 3Q3, with MC(Q) = 100 – 10Q + 9Q2. At what level
of output does the average cost curve reach a minimum? What is the average
cost at its minimum?

Solution

Average cost, AC(Q) = TC(Q)/Q = 100 – 5Q + 3Q2

This equals marginal cost when AC = 100 – 5Q + 3Q2 = 100 – 10Q + 9Q2 or,
solving for Q, when Q* = 5/6. Substituting into AC, we have AC min = 100 –
11
5Q* + 3Q*2 = 100 – 25/12 or AC  97 .
12

Economies and Diseconomies of Scale

Economies/ Dise- When average cost falls as output rises, we say that there are economies of
conomies of scale scale. If average cost rises as output rises, we say that there are disecono-
mies of scale. In other words, the left-hand side of the “U” on a typical aver-
age cost curve is the region of economies of scale. The right-hand side of the
“U” is the region of diseconomies of scale. The point at which the long-run
average cost curve attains its minimum is often called the minimum effi-
cient scale of production.

Figure 7-7 Real-World Average Cost Curve

AC
AC(Q)

0 Q’ = minimum Q” Q
efficient scale Q

110
 7.3
Long-Run Total Cost Curve

This average cost curve typifies many real-world production processes.

There are economies of scale for output less than Q’. Average costs are flat
between Q’ and Q”, and there are diseconomies of scale thereafter. The out-
put level Q’ at which the economies of scale are exhausted is called the min-
imum efficient scale.

Where do economies or diseconomies of scale come from? Economies of Indivisible input

scale may result from indivisible inputs. In other words, an input may be
available only in a certain minimum size. Even if the output is small, the
quantity of this input cannot be scaled down. On the other hand, disecono-
mies of scale may result from managerial diseconomies. In other words, they
result when an increase in output requires a more than proportional increase
in spending on managerial services.

Economies of Scale in Alumina Refining Case Study 7-6

Manufacturing aluminum involves several steps, one of which is alumina

refining. There are substantial economies of scale in the refining of alumina.
The below table - drawn from John Stuckey’s study of the aluminum indus-
try – shows estimated long-run average costs as a function of the capacity of
an alumina refinery. As plant capacity doubles from 150,000 tons per year to
300,000 tons per year, long-run average cost declines by about 12 percent.
Stuckey reports that average costs in alumina refining may continue to fall
up to capacities of 500,000. If so, then the minimum efficient scale of an alu-
mina refinery would occur at an output of 500,000 tons per year.

If firms understand this, we would expect most alumina plants to have ca-
pacities of at least 500,000 tons per year. In fact, this is true. In 1979, the aver-
age capacity of the 10 alumina refineries in North America was 800,000 tons
per year, and only two were under 500,000 tons per year. No alumina refin-
ery’s capacity exceeded 1.3 million tons per year. This suggests that disecon-
omies of scale set in at about this level of output.

Plant Capacity (tons) Index of Average Cost

(equals 100 at 300,000 tons)
55,000 139
90,000 124
150,000 114
300,000 100

Source: Table 1-1 in Stuckey, Vertical Integration and Joint Ventures in the Alumi-
num Industry (Cambridge, MA: Harvard University Press, 1983)

111
7
The Theory of Cost

7.4 Short-Run Total Cost Curve

The long-run total cost curve shows how the firm’s minimized total cost
varies with output when the firm is free to adjust all its inputs. The short-
run total cost curve, STC(Q), tells us the minimized total cost of producing
Q units of output when at least one input is fixed at a particular level.

In the following discussion we assume that the amount of capital used by

the firm is fixed at K . The short-run total cost curve is the sum of two com-
ponents: the total variable cost curve, TVC(Q), and the total fixed cost
curve, TFC.

STC(Q) = TVC(Q) + TFC

The total variable cost curve, TVC(Q), is the sm of expenditures on variable

inputs, such as labor and materials, at the short-run cost-minimizing input
combination. Total fixed cost is equal to the cost of the fixed capital services
(i.e., TFC = r K ) and thus does not vary with output.

Figure 7-8 Short-Run Total Cost Curve

STC(Q)
TC
TVC(Q)

TFC =r K

TFC

rK

0
Q

Total fixed cost is equal to the cost, r K , of the fixed capital services. Since
that cost is independent of output, the total fixed cost curve is a horizontal
line. At every quantity Q, the vertical distance between the total variable cost
curve and the short-run total cost curve is equal to total fixed cost.

112
 7.4
Short-Run Total Cost Curve

Short-run marginal and average costs

Just as we can define long-run average and long-run marginal costs, we can
also define short-run average cost (SAC) and short-run marginal cost
(SMC):
STC(Q)
SAC(Q) 
Q

STC(Q  Q)  STC(Q) STC

SMC(Q)  
Q Q

Because we can break short-run total cost into two pieces (total variable cost
and total fixed cost), we can also break short-run average cost into two piec-
es: average variable cost (AVC) and average fixed cost (AFC):

STC = TVC + TFC

SAC = AVC + AFC

Put another way, average fixed cost is total fixed cost per unit of output, i.e.,
AFC = TFC/Q. Average variable cost is total variable cost per unit of output,
that is, AVC = TVC/Q.

Short-Run Marginal and Average Cost Curves Figure 7-9

Cost
SMC(Q) SAC(Q)
AVC(Q)

SACmin

AVCmin
AFC(Q)
0 Q

113
7
The Theory of Cost

Figure 7-9 illustrates typical graphs of the short-run marginal cost, SMC(Q),
short-run average cost, SAC(Q), average variable cost, AVC(Q), and average
fixed cost, AFC(Q).

We obtain the SAC curve by “vertically summing”2 the AVC curve and the
AFC curve. The AFC curve decreases everywhere and approaches the hori-
zontal axis as Q becomes very large. This reflects the fact that as output
increases, fixed capital cost are “spread out” over an increasingly large vol-
ume of output, driving fixed costs per unit downward toward zero. Because
AFC becomes smaller and smaller as Q increases, the AVC and SAC curves
get closer and closer together. The SMC curve intersects the SAC curve and
the AVC curve at minimum point of each curve.

7.5 Learning by doing

1. Which of the following statements about fixed costs is true?

(a) Fixed costs increase as time goes by.

(b) Average fixed cost graphs as a U-shaped curve.

(c) Fixed costs are zero in the long run.

(d) Fixed costs are zero when the firm decides to produce no output.

2. As output increases, total fixed costs

(a) increase.

(b) remain constant.

(c) decrease.

(d) decrease and then increase.

3. The curve decreases continuously as output increases.

(a) average fixed cost

(b) average variable cost

(c) total fixed cost

2 Vertically summing means that, for any Q, we find the height of the SAC curve by
adding together the heights of the AVC and AFC curves at that quantity.

114
 7.5
Learning by doing

(d) total variable cost

4. Marginal cost can be defined as

(a) the value of total cost divided by the value of quantity produced.

(b) the change in total variable cost divided by the change in quantity
produced.

(c) the change in average total cost divided by the change in quantity
produced.

(d) the change in average variable cost divided by the change in quantity
produced.

5. When average cost is greater than marginal cost,

(a) average cost is rising.

(b) average cost is falling.

(c) marginal cost is rising.

(d) marginal cost is falling.

6. If a firm experiences constant returns to scale, its long-run average cost

curve will be

(a) horizontal.

(b) U-shaped.

(c) upward sloping.

(d) downward sloping.

7. Jill and John Pantera produce earthenware mugs. They can sell their
mugs at $2 each. They find that the marginal cost of production for the
first, second, third, fourth, and fifth mug is 50¢, $1.00, $1.50, $2.00, and
$2.50 respectively. Assuming that Jill and John do produce some mugs,
which of the following statements is true?

(a) The profit-maximizing output level is three mugs.

(b) Jill and John can make a positive profit from selling their mugs.

(c) Jill and John should produce the fourth mug.

115
7
The Theory of Cost

(d) Because marginal cost is increasing by 50 cents per mug, there is a

constant rate of increase in total cost.

8. Suppose that average earnings of an adult in the United Kingdom are

£16,000 per year. A recent policy debate has centered on waiting times for
patients in the National Health System. The difficulty is that many pa-
tients must wait for operations. Suppose that these waits average three
months. During this waiting time 50% of the patients are unable to work
and, in addition, impose considerable burden on their families since they
often are unable to care for themselves. Assume that the number of peo-
ple waiting for operations is approximately 50,000. A new budget is pro-
posed that allocates £1 billion per year to the National Health Service,
with the aim of reducing the waiting time to a minimal amount, which
you can assume to be zero for the purposes of this question.

(a) What is the total opportunity cost of waiting time for people who
need operations?

(b) Would you vote for this budget?

9. Let Q = L1/3K2/3, where Q is output, L is labor, and K is capital.

(a) Let w = 1 and r = 2, where w is the wage rate and r is the rental rate of
capital. Solve for the optimal input combination to produce 10 units of
output.

(b) For arbitrary values of w, r, and Q, solve for the (long-run) input de-
mand curves.

(c) Suppose that the firm’s capital is fixed in the short-run at K . Find the
firm’s short-run demand for labor.

10. The total cost of a firm can be written TC = a + bQ + cQ 2, with a, b, and c

all non-zero.

(a) Write down the formula for average cost for this firm and draw aver-
age cost on a diagram. Where are the regions of decreasing and increas-
ing economies of scale?

(b) Compare this to a case where the total cost of a firm can be written
TC = a + bQ with MC = b. What can you say about economies of scale for
this total cost curve?

116
 7.5
Learning by doing

(c) If you wished to estimate the cost curve for a firm, but did not know
whether the function used in part (a) or in part (b) was correct, why
would it be a better choice to estimate a translog cost curve than a con-
stant elasticity cost curve? (The translog production function is a general-
ization of the Cobb-Douglas production function. The name stands for
“transcendental logarithmic”.)

117
8
Perfectly Competitive Markets

8 Perfectly Competitive Markets

Chapter objectives:
1. List each of the conditions underpinning the perfectly competitive model
and indicate how perfect competition produces efficiently.

2. Identify four causes of market failure and describe how they arise.

3. Explain and calculate the profit maximization of a price-taking firm.

4. Outline how the market price is determined in a short-run and long-run

equilibrium.

5. Explain economic rent and producer surplus.

118
 8.1
What is a Perfectly Competitive Market?

8.1 What is a Perfectly Competitive

Market?
Let us now proceed with the analysis of market equilibrium, which is the
subject of this chapter. The type of market we will analyze is a perfectly
competitive market. A perfectly competitive market has four characteristics:

1. Each buyer’s purchases are so small that he/she has an imperceptible Characteristics of
effect on market price. Further, each seller’s sales are so small that he/she perfectly compet-
has an imperceptible effect on market price. When both buyers and itive market
sellers are small in this way, we call the industry fragmented.

2. Firms produce undifferentiated products in the sense that consumers

perceive the products to be identical. In other words, the products may
be physically the same or, if they are not the same, the differences do not
in any way affect consumers’ demand (or valuation) of the product.

3. Consumers have perfect information about the prices all sellers in the
market charge.

4. Finally, all firms (both current industry participants and potential new
entrants to the industry) have equal access to resources (such as tech-
nology and raw materials).

These four characteristics have three important consequences:

1. Characteristic (1) implies that buyers and sellers take the price of the Price taker
product as given when making their purchase and output decisions,
since each decision has an imperceptible effect on market price. Because
of this, sellers and buyers in such markets are termed price takers.

2. Characteristics (2) and (3) imply that there is a single price at which trans-
actions occur: any output in this market that is priced lower than any
other would immediately attract the entire market demand, since con-
sumers view all output as exactly the same except for price. Similarly, any
output priced higher than any other would attract no demand at all. This
is implication is often called the Law of one price.

3. Characteristic (4) implies that all firms have identical long-run cost func-
tions. This consequence is called free entry.

In the real world, the stringent conditions of the perfectly competitive model Market failure
might not be met. In such a situation a market failure occurs and inefficien-
cies creep into the system of unregulated markets.

119
8
Perfectly Competitive Markets

 Imperfect competition, such as monopoly, arises when firms have some

measure of control over price and/or competition. Economic profits may
be maintained indefinitely if competitors can be barred from entry. The
consequences are higher prices and more restricted production than
would occur in a competitive environment—consumers lose.

 Public goods provide benefits to more than one consumer at a time.

Because one can consume without having to pay, there is a powerful in-
centive not to pay, i.e., to be a free rider. Although private firms might
produce most public goods, the problem is that they will not produce
enough of them, because they cannot compel consumers to pay their fair
share. By using tax dollars, the government can produce a sufficient
quantity of public goods and improve efficiency.

 Externalities (costs or benefits encountered by a party outside a transac-

tion) do not enter into the calculations of profit- and utility-maximizing
firms and consumers. If externalities are a factor, the allocation of re-
sources is likely to be suboptimal.

 Misinformation or poor information affects the market participant’s abil-

ity to make an informed choice. The less informed he or she is, the more
likely a suboptimal outcome is.

8.2 Profit Maximization by a Price-Taking

Firm
We distinguished between accounting cost and economic cost. The key dif-
ference between the two cost concepts is that economic cost measures the
opportunity cost of the resources that the firm uses to produce and sell its
products, whereas accounting cost measures the historical expenses the firm
incurred to produce and sell its output.

We will now make a similar distinction between accounting profit and eco-
nomic profit:

Accounting profit = sales revenue – accounting costs

Economic profit = sales revenue – economic costs

That is, economic profit is the difference between a firm’s sales revenue and
the totality of its economic costs, including all relevant opportunity costs.

Having defined economic profit, we can now study the problem of a price-
taking firm that seeks to maximize its economic profit. We denote profit by

120
 8.2
Profit Maximization by a Price-Taking Firm

letter π, and it equals the difference between total revenue, TR, and total
cost, TC:

max π = TR(Q) – TC(Q)

Total revenue, in turn, equals the market price P multiplied by the quantity
of output Q produced by the firm: TR(Q) = P x Q. Total cost TC(Q) is the
total cost curve that we discussed in Chapter 7 and tells us the total cost of
producing Q units of output, assuming that the firm chooses the input com-
bination to produce that output at minimum cost.

The solution to this problem is that the firm wishes to set q at a level that
equates marginal revenue, MR, defined as the additional revenue earned
when output increases slightly, to marginal cost.
TR(Q  Q)  TR(Q) TR
MR  
Q Q

However, a price-taking firm anticipates that each additional unit sold will
earn the market price. Therefore, the marginal revenue of a price-taking firm
exactly equals price.

MR = P

Example: Total Revenue, Cost, and Profit for a Price-taking Rose Producer Case Study 8-1

Nakao is a rose-growing company. They anticipates that the market price

for fresh-cut roses will be P = $1 per rose. This below table shows total reve-
nue, total cost, and total profit for various output levels, and the top diagram
in Figure 8.1 graphs these numbers.

Q TR TC π
(thousands of roses (thousands of $ (thousands of $ (thousands of $
per month) per month) per month) per month)
0 0 0 0
60 60 95 -35
120 120 140 -20
180 180 155 25
240 240 170 70
300 300 210 90
360 360 300 60
420 420 460 -40

121
8
Perfectly Competitive Markets

TR, TC, π
TR TC
$300

210

90 π

0 60 300 Q

P MC

$1 MR = P

0 60 300 Q

This Figure shows that:

 For quantities between Q = 60 and the profit-maximizing quantity Q =

300, producing more roses increases profit. Increasing the quantity in this
range increases total revenue faster than total cost:
TR TC
 or P > MC
Q Q

 For quantities greater than Q = 300, producing fewer roses increases profit.
Decreasing quantity in this range decreases total cost faster than it de-
creases total revenue.
TR TC
 or P < MC
Q Q

122
 8.3
How the Market Price is determined: Short-Run Equilibrium

 If this company can increase its profit when either P > MC or P < MC,
quantities at which these inequalities hold cannot maximize their profit.
It must be the case, then, that at the profit-maximizing output:

P = MC

What is the difference between Q = 60 and Q = 300? At Q = 300, the marginal

cost curve is rising, while at Q = 60 the marginal cost curve is falling.

From this example, we can conclude: At a profit-maximizing quantity, two

conditions must hold:

 P = MC
 MC must be increasing
These are the profit-maximization conditions for a price-taking firm.

8.3 How the Market Price is determined:

Short-Run Equilibrium
The firm’s short-run supply curve tells us how the profit maximizing output,
q*, changes as the market price changes in the short-run. Offhand, it appears
that the firm’s short-run supply curve is defined by the equation P = MC.
After all, this equation does tell us how the profit maximizing output chang-
es as the market price changes. There are two reasons why this is not the
case:

 The short-run is the period of time in which the firm’s plant size is fixed Shut-down price
and the number of firms in the industry is fixed. Hence, the marginal
cost that we would want to use in this equation is the short-run marginal
cost, SMC.

 We have implicitly assumed in our profit maximization conditions that

the firm does better producing a positive output than not producing at
all. We now define PS as the price below which the firm would prefer to
shut its doors. PS is called the shut-down price.

Taking these two considerations into account, the firm’s short-run supply
curve is defined by the following equations:

1. P = SMC, where SMC slopes upward as long as P ≥ PS

2. Q = 0, where P < PS

123
8
Perfectly Competitive Markets

Figure 8-1 The Short-run Supply Curve

SMC
Supply curve
above PS SAC

PS
Supply
curve
below PS
0 Q

Short-run supply curve for a firm: Some fixed costs sunk, some non-sunk

In order to determine PS, we must compare the profits that the firm could
earn if it produced such that P = SMC to the profits the firm would earn by
producing nothing. What the firm could earn by producing nothing depends,
in turn, on the portion of the firm’s short-run total costs that are fixed and
sunk (SFC), and the portion that are fixed but non-sunk (NSFC).

Sunk fixed costs are the costs that are output insensitive when output is
positive and unavoidable even if output is zero. Non-sunk fixed costs are the
costs that are output-insensitive when output is positive, but avoidable when
output is zero.

 If all the firm’s fixed costs are sunk, then it must pay the same fixed costs
whether its output is positive or zero. The firm will choose to produce a
positive output if the total revenue from production covers its total varia-
ble costs, TR(Q) > TVC(Q) or P > AVC(Q). The shut-down price is defined,
then, as the lowest point on the average variable cost curve: If price were
to exceed AVC at any positive output, the firm could earn profit from the
output and would therefore do better than producing nothing.

 If none of the firm’s fixed costs are sunk, the firm can avoid all costs by
producing nothing. On the other hand, it must cover all costs, fixed and
variable, if it produces a positive amount. Therefore, the firm will pro-

124
 8.3
How the Market Price is determined: Short-Run Equilibrium

duce a positive output only if the total revenue from production covers
its total variable and fixed costs, TR(Q) > TC(Q) or P > SAC(Q) as the
production condition. The shut-down price is the lowest point on the av-
erage cost curve.

 If only some of the fixed costs are sunk, the firm will choose to produce a
positive output if the total revenue from production covers its total vari-
able costs plus the portion of fixed costs that are non-sunk, TR(Q) >
TVC(Q) + NSFC. Defining the average non-sunk cost curve as ANSC as
ANSC(Q) = [TVC(Q) + NSFC]/Q, we have P > ANSC(Q) as the production
condition. The shut-down price is the minimum of the average non-sunk
cost curve.

These conditions can be summarized as follows:

All fixed costs Some fixed cost

No sunk costs
sunk sunk

Produce if… P ≥ SAC(Q) P ≥ AVC(Q) P ≥ ANSC(Q)

PS Min(SAC) Min(AVC) Min(ANSC)

It is helpful to remember, when computing PS, that the minimum of the SAC
(or ANSC or the AVC) curve occurs where SMC equals SAC (or ANSC or
AVC).

Short-run market supply curve

We have just derived the short-run supply curve for an individual price-
taking firm. Let’s now see how to go from the firm’s supply curve to the
supply curve for the entire industry.

The short-run market supply curve at any price is the sum of the quantities
supplies by each firm at that price. Graphically, this is the horizontal sum of
the firm’s short-run supply curves since the quantities each firm supplies are
measured on the horizontal axis. Let this sum be denoted Q S(P).

Figure 8-2

125
 8
Perfectly Competitive Markets

The Short-run Market Supply Curve

Supply Curve Supply Curve

of Firm 1 of Firm 2 Market Supply,
P P QS(P)

PS2

PS1

0 Q1 Q1* Q2 Q2* Q 0 Q1 Q1*+Q2 Q2+Q2* Q

Short-run perfectly competitive equilibrium

A short-run perfectly competitive equilibrium occurs when quantity de-

manded by the market at given price equals the quantity supplied by the mar-
ket at that same price. If QD(P) is market demand at price P, then, the short-
run perfectly competitive equilibrium is defined by the equation:

QS(P) = QD(P)

Graphically, the short-run perfectly competitive equilibrium occurs where

market supply equals market demand. It is very important that you remem-
ber that market and not firm supply equals market demand in the market
equilibrium and that the number of firms in the market is fixed in the short-
run.

Figure 8-3

126
 8.3
How the Market Price is determined: Short-Run Equilibrium

Short-run Perfectly Competitive Equilibrium

Firm Supply
Market Supply
SMC
P P

P*
SAC

Market Demand

PS

0 q* Firm 0 Q* Firm
Output, q Output, Q

To summarize, then, the perfectly competitive equilibrium for a given num-

ber firms, n, in the industry and a market demand, QD(P), is defined as the
firm output level, q*, and the price level, P* such that:

 Every firm is producing on its short-run supply curve, P* =

SMC(q*)

 Industry supply equals industry demand, nq* = QD(P*)

Short-run market equilibrium Case Study 8-2

The market consists of 300 identical firms, and the market demand curve is
given by QD(P) = 60 – P. Each firm has a short-run total cost curve STC = 0.1
+ 150Q2, and all fixed costs are sunk. The corresponding short-run marginal
cost curve is SMC = 300Q. The corresponding average variable cost curve is
AVC = 150Q. You should verify (e.g., by sketching the SMC and AVC curves)
that the minimum level of AVC is 0. Thus, a firm will continue to produce as
long as price is positive.

Problem

(a) Find the short-run equilibrium in this market

127
8
Perfectly Competitive Markets

(b) At the market equilibrium, do firms make positive economic profit?

Solution

(a) We first derive the supply curve qS(P) of an individual firm. Each firm’s
profit-maximizing quantity is given by equating marginal cost and price:

300Q = P

which implies

qS(P) = P/300

The short-run equilibrium thus occurs where market supply (300 S(P)) equals
market demand or:

300(P/300) = 60 – P or

P = 30

At a market price of $30 per unit, each firm produce 30/300 = 0.1 unit per
year. Total market demand and supply is 300 x 0.1 = 30 units per year.

(b) A firm’s short-run average cost function is given by:

STC 0.1
SAC    150Q
Q Q

When each firm produces Q = 0.1, short-run average cost is:

0.1
SAC   (150x0.1)  16
0.1

Each firm’s cost per unit is $16, while the market price is $30, so P > SAC.
Each firm thus makes a positive economic profit.

8.4 How the Market Price is determined:

Long-Run Equilibrium
The long-run is the period of time in which all the firm’s inputs, including
plant size, can be adjusted. The number of firms in the industry can change
as well. Therefore, while the day to day output decisions of the firm might
be determined by the solution to the firm’s short-run supply problem, its
decisions on target production levels in the future would be based on the
firm’s long-run equilibrium output.

The firm’s long-run supply curve

128
 8.4
How the Market Price is determined: Long-Run Equilibrium

The firm’s long-run supply decision will be based on its long-run total costs.
All costs are avoidable in the long-run, therefore there are no sunk costs.
The appropriate long-run level of the shut-down price is the price that just
equals the minimum of the long-run average cost curve. At any price lower
than this, the firm would make negative profits, since total revenue would
be less than total cost. Summarizing, the firm’s long-run supply curve is
described by the two following conditions:

 If P ≥ PS, the firm sets output so that P = MC, where MC is long-run mar-
ginal cost.

 If P < PS, the firm sets output to zero.

 PS is the minimum price that touches the long-run AC curve.

Long-run perfectly competitive equilibrium

In our short-run analysis of the perfectly competitive equilibrium, we as-

sumed that the number of firms in the industry was fixed. But in the long-
run, new firms can enter the industry. A firm will enter the industry if, given
the market price, it can earn positive economic profits. Positive economic
profit indicates that there is an opportunity for an entrant to create wealth
for its owners by participating in the industry.

Long-run Equilibrium in a Perfectly Competitive Market Figure 8-4

P SMC MC P
SAC
AC
Market
demand, QD(P)

P*

0 q* 0 nq*=QD(P*)
Firm Market
output output

129
 8
Perfectly Competitive Markets

A long-run perfectly competitive equilibrium occurs at a price at which

supply equals demand and firms have no incentive to enter or exit the indus-
try. More specifically, a long-run perfectly competitive equilibrium is charac-
terized by a market price P*, a number of firms, n*, and quantity of output
Q* per firm that satisfies three conditions:

 P* = MC(Q*). Long-run profit is maximized with respect to output and

plant size.

 P* = AC(Q*). Economic profit is zero. The existing firms are earning zero
economic profits so that there is no incentive for more firms to enter.
Moreover, an active firm cannot earn negative economic profit by partic-
ipating in this industry.

 QD(P*) = n*q*. This condition ensures that aggregate demand equals

aggregate supply so that there is no pressure for market price to change.

Case Study 8-3 Long-run market equilibrium

Problem

In this market, each firm and potential entrant has a long-run average cost
curve:

AC(Q) = 40 – Q + 0.01Q2

and a corresponding long-run marginal cost curve:

MC(Q) = 40 – 2Q + 0.03Q2

where Q is thousands of units per year. The market demand curve is:

QD(P) = 25,000 – 1,000P

where QD(P) is also measured in thousands of units. Find the long-run equi-
librium price, quantity per firm, and number of firms.

Solution

Let asterisks denote equilibrium values. The long-run competitive equilibri-

um satisfies the following three equations.

P* = 40 – 2Q* + 0.03Q*2 (profit maximization)

P* = 40 – Q* + 0.01Q*2 (zero profit)

 25,000  1,000P * 
n* =   (supply equals demand)
 Q* 

130
 8.4
How the Market Price is determined: Long-Run Equilibrium

By combining the first two equations, we can solve for the quantity per firm:

40 – 2Q* + 0.03Q*2 = 40 – Q* + 0.01Q*2

0.03Q*2 – 0.01Q*2 = 2Q* – Q*

0.02Q*2 = Q*

Thus, Q* = 50, so each firm in equilibrium produces 50,000 units per year.
Note that this quantity is a firm’s minimum efficient scale. Substituting Q* =
50 bask into the average or marginal cost function gives us the minimum
level of average cost and thus the equilibrium price:

P* = 40 – 50 + 0.01(50)2 = 15

The equilibrium price of $15 per unit corresponds to a firm’s minimum level
of average cost. By substituting P* into the demand function we can find the
market demand at $15. It is given by:

25,000 – 1,000(15) = 10,000

or 10 million units per year. The equilibrium number of firms is equilibrium

market demand divided by minimum efficient scale:

n* = 10,000/50 = 200

To summarize, the long-run equilibrium price is $15 per unit. The quantity
supplied by an individual firm is 50 thousand units per year. The equilibri-
um number of firms is 200. Market demand and market supply in equilibri-
um equal 10 million units per year.

The long-run market supply curve

The long-run market supply curve tells us the total quantity of output that
will be supplied at various market prices, assuming that all long-run ad-
justments take place. It cannot be derived as the sum of the firm’s long-run
supply curves because the number of firms in the industry can vary in the
long-run. This means that we do not know how many firms’ supply curves
to add up. Instead, the long-run market supply curve is derived from the
following reasoning.

If price exceeds the minimum of the average cost curve, then entry would
occur. This entry would shift out industry supply until the market equilibri-
um price fell back to the minimum of average cost.

If price were to fall below the minimum of the average cost curve, then firms
would be earning negative profits and would exit the industry. Exit would

131
 8
Perfectly Competitive Markets

shift back the industry supply curve until market equilibrium price rose
back to the minimum of the average cost curve.

Therefore, output can only be supplied so as to set price equal to the mini-
mum of average cost in the long-run. In other words, the long-run market
supply curve is flat along the line P = ACmin.

Increasing-cost, and decreasing-cost and constant-cost industries

When constructing the long-run supply curve in the previous section, we

assumed that entry of new firms into the industry does not change the cost
of inputs to production. This assumption need not always hold, however.

For example, the long-run market supply curve is upward-sloping when the
average cost curve shifts up as input prices rise due to an increase in indus-
try output. Such an industry is called an increasing-cost industry. This is
more likely to be the case when firms use industry specific inputs, which
are scarce inputs that only firms in this industry use.

In some situations an increase in industry output can lead to a decrease in

the price of an input. We then have a decreasing-cost industry. For example,
a larger industry may be able to better organize transportation activities to
take advantage of lower transport prices when shipping larger volumes.

In contrast, when changes in industry output have no effect on input prices,

we call the industry a constant-cost industry. Here, the long-run market
supply curve is flat.

8.5 Economic Rent and Producer Surplus

Reservation Economic rent measures the economic surplus that is attributable to an ex-
value of an input traordinarily productive input whose supply is limited. Specifically, econom-
ic rent is equal to the difference between the maximum amount a firm is
willing to pay for the services of the input and the input’s reservation value.
The input’s reservation value, in turn, is the return that the input owner
could get by deploying the input in its best alternative use outside the indus-
try. Putting the pieces of this definition together we thus have:

economic rent = A – B

where:

A = maximum amount firm is willing to pay for services of input

132
 8.5
Economic Rent and Producer Surplus

B = return that input owner gets by deploying the input in its best
alternative use outside the industry.

To illustrate this definition, suppose that the maximum amount that a rose
firm would be to pay to hire an extraordinary master grower – the A term in
our definition of economic rent – is equal to $100,000. Suppose further that
the grower’s best available employment opportunity outside the rose indus-
try is to work as a grower in the tulip industry for an annual salary of
$70,000. This is the B term in our definition. The economic rent attributable
to the extraordinary master grower is thus $100,000 - $70,000 = $30,000 per
year.

Producer surplus

So far, economic profit has been our main measure of producer benefit from Producer surplus
production. However, producer surplus is an alternative measure of produc-
er benefit. In other words, if we think of the supply curve as the price, P(Q),
that would be just high enough to induce producers to supply Q units of
output, producer surplus measures the amount by which market price ex-
ceeds this minimum. It can be thought of as the area between the supply
curve and the market price. Since the supply curve is the MC curve above
the minimum of average non-sunk costs, the producer surplus corresponds
to the shaded area in Figure 8-5, below. We can see the relationship between
producer surplus and economic profit as follow:

In short-run:

Producer surplus = total revenues – total non-sunk costs.

Economic profit = total revenues – total costs.

or, Producer surplus = Economic profit + total sunk fixed costs.

In long-run: (all costs are non-sunk costs)

Producer surplus = Economic profit = total revenues – total costs.

Figure 8-5

133
8
Perfectly Competitive Markets

Producer Surplus

P P
Market
Firm MC supply
supply
AC

P*

PS

0 0
Firm Market
output output

Another very useful way to think of producer surplus is as a gain from

trade. In the short-run, the firm earns its total revenue minus all (sunk and
non-sunk costs) if it produces so that trade occurs. If the firm shuts down, so
that trade does not occur, it loses its sunk costs. The producer’s surplus is
equal to the gain from trading versus not trading: the profit from production
minus the profit from producing zero. This difference simply equals the total
revenue minus all non-sunk costs.

8.6 Learning by doing

1. Characteristics of perfectly competitive markets do not include:

(a) homogeneous products.

(b) large numbers of potential buyers.

(c) large numbers of potential sellers.

(d) the ability of sellers to set prices.

2. In the long run for a perfectly competitive market:

(a) some firms exit so that others may earn more than normal profits.

134
 8.6
Learning by doing

(b) established firms reap higher profits than newer firms.

(c) all resources are fixed for the industry as a whole.

(d) firms can enter or leave the market because all resources are variable

3. In the market for apples there is a consumer surplus, and a short-run

economic profit is being made. This is conclusive evidence of

(a) externalities.

(b) market failure.

(c) apples being a public good.

(d) None of the above

4. In short-run, if all the firm’s fixed cost are sunk, the shutdown price is:

(a) PS = Min(SAC)

(b) PS = Min(AVC)

(c) PS = Min(ANSC)

(d) there is not the shutdown price

5. The firms in a perfectly competitive industry merge into one big firm and
impose barriers to entry into the industry. We can say that

(a) we can expect that externalities will begin to appear.

(b) the product of this industry is a public good.

(c) the price of the product will be higher, and the output lower, than the
values under perfect competition.

(d) this firm will be unable to maximize profits.

Use the following information for the next three questions: Arboc has only
two products—ground nuts and goat cheese. These goods are substitutes
and are produced by perfectly competitive firms. Initially, both markets are
in long-run equilibrium. Now consumer preferences shift away from goat
cheese and toward ground nuts.

6. Given the preceding information, which of the following will not occur?

135
8
Perfectly Competitive Markets

(a) In the short run, goat cheese producers will incur losses.

(b) In the short run, there will be an increase in the demand for ground
nut workers.

(c) In the long run, more firms will enter the ground nut industry.

(d) More capital will flow into the production of goat cheese.

7. Given the preceding information, we would expect

(a) short-run profits in the ground nut industry.

(b) long-run losses in the goat cheese industry.

(c) long-run profits in the ground nut industry.

(d) short-run losses in the ground nut industry.

8. As a producer of goat cheese, your best short-run strategy is to

(a) leave the industry and enter the ground nut industry.

(b) switch over to ground nut production.

(c) set your output level to equalize marginal cost and the market price.

(d) cut your price to become more competitive and to increase your mar-
ket share.

9. Three regions of the world are able to produce cocoa beans: Africa, South
America, and Central America. In each of these regions the production of
cocoa is ensured by a large number of small plantations. None of these
regions has pecuniary effects in cocoa production. The cost of production
varies across these three regions, however. Let the following be true:

Africa: MCA = 4QA; minimum of average cost = min(ACA) = 10

South America: MCS = 5QS; min(ACS) = 15

Central America: MCC = 6QC; min(ACC) = 20

Where: QA, QS, QC are the outputs of cocoa beans in Africa, South Amer-
ica and Central America, respectively. MCi and ACi (i = A, S, C) are the
long-run marginal and average costs for each region. Further, assume
that no region can produce more than 30 units of cocoa, even in the
long-run. The worldwide market for cocoa beans is perfectly competi-
tive.

136
 8.6
Learning by doing

(a) Write the equation for a typical firm’s long-run supply curve in each
region.

(b) Write the equation for the long-run worldwide market supply curve.
Draw this supply curve on a graph.

10. Let total cost be TC(Q) = 100 + Q2 for positive q and TC(Q) = 0 for Q = 0,
marginal cost be MC(Q) = 2Q and demand be QD = 10,000 – 100P. Answer
the following question.

(a) What will be the industry output in long-run equilibrium?

(b) Suppose that demand is QD = 5,000 – 50P. What is the industry out-
put in long-run equilibrium? How many firms will there be in the in-
dustry?

(c) Suppose that demand now shifts to QD = 6,000 – 50P but that it is not
possible to manufacture more (industry) output than the long-run equi-
librium output in (b). What is the new equilibrium price of output?
How much profit does each manufacturer earn if each sets its output
optimally given the new price and constraints in this setup?

(d) In the long-run, how many new firms will enter this industry when
demand shifts as in (c)?

137
9
Competitive Markets: Applications

9 Competitive Markets:
Applications

Chapter objectives:
Analysis the consequences of government intervention in competitive mar-
ket:

1. Imposing excise taxes

2. Granting subsidies to producers

3. Regulating the maximum and minimum price producers may charge

4. Setting quotas limiting the amount that may be produced in a market

5. Imposing tariffs or quotas on imports

138
 9.1
The invisible Hand

9.1 The invisible Hand

We learned how equilibrium was determined in perfectly competitive mar- Invisible hand
kets. In this chapter, we learn how to analyze the consequences of govern-
ment intervention in such market. Let us first show why a perfectly competi-
tive market without intervention maximizes total surplus. A perfectly com-
petitive equilibrium market price and quantity occur at the intersection of
demand and supply, (P*, Q*), in Figure 9-1(a,b). Consumer surplus is the
area of triangle ABC, while producer surplus is the area of triangle DBC.
Total surplus is the sum of these, area ADC.

Consider now an output level in Figure 9-1(a), Q1, which is lower than Q*. At
Q1, sellers would be willing to sell an extra unit at price PS, while consumers
would be willing to buy at price PD. PD exceeds PS so that any price between
these two at which trade occurs, call it P*, results in a gain in consumers’
surplus of PD – P* and a gain in producers’ surplus of P* - PS. In other words,
there are unrealized gains from trade. The gain in total surplus is P D – PS
(which is independent of P*). Therefore, increasing output above Q1 increas-
es total surplus. This will be true as long as PD exceeds PS or, equivalently, as
long as demand lies above supply.

Consider now a level of output that exceeds Q* (e.g., Q 2 in Figure 9-2(b)). At

Q2 the price at which sellers would be willing to supply an extra unit, PS,
exceeds the price at which consumers would be willing to buy, PD. Any price
between these two at which trade occurs, say P*, results in a loss in consum-
ers’ surplus of P* - PD and a loss in producers’ surplus of PS – P*. The loss in
total surplus is PS – PD. Therefore, reducing output below Q2 increases sur-
plus. This will be true as long as PS exceeds PD or, equivalently, as long as
supply lies above demand.

Figure 9-1

139
 9
Competitive Markets: Applications

Total Surplus

A
Supply

PD

B C
(a) P*

PS

D Demand

0 Q1 Q* Q

A
Supply

PS
B C
(b) P*

PD

D Demand

0 Q* Q2 Q

Deadweight loss It follows that total surplus can only be maximized at the competitive equi-
librium point (P*, Q*). Any other point involves a loss in total surplus called
a deadweight loss. For example, the deadweight losses in Figure 9-1(a,b) are
the shaded areas. In Figure 9-1(a), there is a deadweight loss because too
little is produced so that there are unrealized gains to trade. In Figure 9-1(b),
there is a deadweight loss because too much is produced so that there are
losses to trade that occur.

140
 9.2
Impact of an Excise Tax

Economic efficiency means that the total surplus is maximized or the output
in the competitive market is the one that maximizes net economic benefits.
As Adam Smith described in his classic treatise in 1776 (An Inquiry into the
Nature and Causes of the Wealth of Nations), it is as though there is an “Invisi-
ble Hand” guiding a competitive market to the efficient level of production
and consumption.

9.2 Impact of an Excise Tax

An excise tax is an amount paid by either the consumer or the producer per
unit of the good at the point of sale. This is sometimes called a specific tax. In
another words, the total amount paid by the demander exceeds the total
amount received by the seller by the tax, T. Consider the effect of a tax that
must be paid by suppliers, illustrated in Figure 9-2. In order to induce pro-
ducers to supply the same amount as before, they must receive $T more than
before in order to compensate for the tax. Geometrically, this means that the
supply curve shifts up by $T to S’. The market equilibrium occurs at output
Q1, where S’ intersects demand. At this new equilibrium, buyers pay PD and
sellers receive, for each unit of the good, an after tax revenue of PD – T = PS.
The equilibrium conditions for this market are now that quantity demanded
must equal quantity supplied, QD = QS, and PD – T = PS.

Excise Tax Figure 9-2

S’
P

PD
T
P*
PS

Demand

0 Q1 Q* Q

141
9
Competitive Markets: Applications

The equilibrium before the tax was at output level Q* so that the shaded area
in the figure was part of consumers’ and producers’ surplus before the tax
was imposed. Now, the shaded area is a deadweight loss to society since
these units are no longer produced in equilibrium.

The equilibrium price paid by buyers, PD exceeds the original equilibrium

price, P*, by less than the amount of the tax. In other words, P D is less than
P* + T. This is because consumers reduce their purchases as price rises. The
equilibrium price received by sellers, PS, falls short of the original equilibri-
um price, P*, by less than the entire amount of the tax. In other words, P S
exceeds P* - T. This is because sellers reduce their supply as the price falls.
The amount by which the price paid by buyers, PD, rises over the non-tax
equilibrium price, P*, is the incidence of the tax on consumers; the amount
by which the price received by sellers, PS, falls below P* is called the inci-
dence of the tax on producers. In the special case where supply is flat, the
equilibrium price will rise by the entire amount of the tax, so that the entire
tax burden falls on consumers. In the special case where supply is vertical,
supply cannot shift when the tax is imposed so that the entire tax burden
falls on producers.

In general, as long as the demand curve slopes downward and the supply
curve slopes upward, the tax raises the equilibrium price to buyers and low-
ers the price to sellers. Figure 9-3 suggest that the incidence of the tax on
consumers and producers depends on the relative elasticties of demand and
supply. In fact, it can be shown that the relative change in the buyers’ price,
PD, and the sellers’ price, PS, due to the tax equals the relative elasticities of
supply,  QS ,P , and demand  QD ,P .

∆PD/∆PS =  QS ,P /  QD ,P

142
 9.3
Subsidies

Incidence of the Tax Figure 9-3

PD = P* + T S’
T
PS = P* S

D
0 Q

P
S

PS = P*
T
PD = P* - T

D
0 Q

9.3 Subsidies
A subsidy is a negative tax. In other words, a subsidy, T, paid to producers
means that the price producers receive, PS, is the price paid by consumers,
PD, plus the subsidies. Not surprisingly, then, the effects are opposite from
those of a tax. The subsidy shifts supply down by a vertical distance T and
the equilibrium quantity traded rises. The price paid by buyers falls and the
price received by sellers rises as long as supply slopes upward and demand

143
 9
Competitive Markets: Applications

slopes downward. Here, the deadweight loss occurs because output rises
above the no-subsidy level. The subsidy is depicted in Figure 9-4, below,
where the deadweight loss is the shaded area. The equilibrium conditions in
this market are now that QD = QS and PD + T = PS.

Figure 9-4 Subsidies

P
S

PS T S’

PD

0 Q* Q2 Q

9.4 Maximum Price Ceilings for Producers

Price ceiling A price ceiling is a legal maximum on the price per unit that a producer can
receive. If the price ceiling is below the pre-control competitive equilibrium
price, then the ceiling is called binding. Let the price ceiling be Pmax in Figure
9-5. The quantity traded in this market will be limited to the amount that
producers are willing to supply at this price, QS, even though demanders
would like to purchase a much higher quantity, QD. Because QD exceeds QS,
there is excess demand. Because the quantity traded does not equal the pre-
control competitive quantity, Q*, there is a deadweight loss: some consumers
who value the good more than its marginal cost in production are unable to
purchase. If we assume that the consumers who value the good most highly
are the ones who actually receive the supply that producers make available,

144
 9.5
Minimum Price Floors

the deadweight loss is the shaded area in Figure 9-5, below. The equilibrium
conditions for this market are now that P = Pmax; QS = QS(Pmax) and QD =
QD(Pmax).

Price Ceiling Figure 9-5

Pmax

Excess De-
mand
D

0 QS Q* QD Q

9.5 Minimum Price Floors

A price floor is a minimum price that consumers can legally pay for a good. Price floor
Price floors sometimes are referred to as price supports. If the price floor is
above the pre-control competitive equilibrium price, it is said to be binding.
The quantity traded in this market will be limited to the amount that con-
sumers are willing to purchase at this price, QD, even though producers
would like to supply a much higher quantity, QS. Because QS exceeds QD,
there is excess supply. Because the quantity traded does not equal the pre-
control competitive quantity, Q*, there is a deadweight loss: consumers who
value the good more than its marginal cost in production are unwilling to
purchase at the inflated price. If we make the further assumption that the
most efficiently produced units are the ones that actually are consumed,
then the effect of the binding price ceiling is the shaded area in Figure 9-6,

145
9
Competitive Markets: Applications

below. The equilibrium conditions of this market are now that P = P min; QS =
QS(Pmin); QD = QD(Pmin).

Figure 9-6 Price Floor

P
Excess S
Supply

Pmin

0 QD Q* QS Q

9.6 Production Quotas

A production quota is a limit on the total quantity that producers can supply
to the market. Imposing a quota creates a vertical segment in the supply
curve at the level of the quota: even if the price increases dramatically, firms
cannot increase the quantity supplied. The new equilibrium occurs where
the new supply curve intersects demand. The production quota is said to be
binding if the limit on production is below the competitive equilibrium out-
put level in the absence of the quota. A binding quota will, then, result in a
deadweight loss since some consumers who are willing to pay more than the
marginal cost of production cannot purchase the good. Since the original
supply curve represents the opportunity cost (marginal cost) of resources to
produce the output, the appropriate measure of welfare loss due to unreal-
ized gains from trade is the difference between the demand curve and the
original supply curve over the range of output curtailed by the quota. This is

146
 9.7
Import Tariffs and Quotas

illustrated in Figure 9-7, below, where the deadweight loss is the shaded
area. The equilibrium conditions are now that Q = Qmax; PD = PD(Qmax); PS =
PS(Qmax).

Production Quota Figure 9-7

P Supply with
quota
Original
Supply

P*

0 Qmax Q* Q

9.7 Import Tariffs and Quotas

Tariffs are taxes levied by the government on goods imported into the coun-
try. Tariffs sometimes are called duties. An import quota is a limit on the
total number of units of a good that can be imported into the country. In
order to examine the effects of these instruments, we must first discuss the
underlying supply and demand conditions that generate imports. Imports
will be observed any time the world price of a good is below the equilibrium
price for the domestic market in the absence of imports. For example, sup-
pose that the country can import as much of a good as it wants at price P W in
Figure 9-8(a) (i.e., foreign supply is flat at PW). This price is below P*, the
price that would prevail in the domestic market if imports were not possible.
The quantity imported will equal Q4 – Q1 at price PW and Q1 will be pro-
duced locally. Consumers’ surplus will be the area of triangle APWB and

147
9
Competitive Markets: Applications

domestic producers’ surplus will be the area of triangle DPWC. Hence the
total surplus of the country is ABCD.

Figure 9-8 Import Tariffs and Quotas

P
A Domestic
Supply

(a) P*
C B Foreign Supply
PW
D Domestic
Demand
0 Q1 Q4 Q

P
Domestic
Supply

(b) PW + T
PW D C T
A B
Domestic
Demand
0 Q1 Q2 Q3 Q4 Q

If a tariff of T per unit is imposed on imports, the effective world price in-
creases form PW to PW + T, as shown in Figure 9-8(b), below. At this new
price, quantity Q3 is demanded and quantity Q2 is supplied domestically.
Imports shrink to the quantity Q3 – Q2. Since total consumption decreases
from Q4 to Q3, there is a deadweight loss from this distortion in consumption
equal to area B. There is also a net opportunity cost to expanding domestic
output from Q1 to Q2: units of output that could be obtained at a cost of P W
under free trade must now be produced at a higher cost. This “production”
deadweight loss is equal to area A. Area C represents the domestic govern-

148
 9.8
Learning by doing

ment’s tariff revenue: T times the number of units imported. This is a net
gain for the domestic government. But this tariff revenue is entirely paid by
local consumers who must now pay T more for each of the (Q3 – Q2) units
imported. In the end, then, C is just a transfer from consumers to their gov-
ernment and does not affect the country’s welfare. Area B was part of con-
sumers’ surplus before the tariff was imposed, but is part of the domestic
producers’ surplus after the tariff is imposed. Again, this is a simple transfer
that does not affect the welfare of the country as a whole. Therefore, the net
effect of the tariff on the country’s welfare is equal to (-A - B), which is un-
ambiguously negative. Equilibrium conditions for this market are now that P
= PW + T; QD = QD(Pw + T); QS = QS(Pw + T).

9.8 Learning by doing

1. A price ceiling on a good or service_____.

(a) increases the price of the good or service

(b) creates a surplus of the good and service

(c) creates a shortage of the good or service

(d) benefits everyone who plans to buy the good or service

2. What is the incidence of an excise tax in the long-run in a constant cost

perfectly competitive industry?

(a) The entire burden of the tax falls on producers

(b) The entire burden of the tax falls on consumers

(c) The burden is shared between consumers and producers

3. A subsidy is:

(a) a payment that acts as a negative tax to reduce the buyer's price in
order to increase the quantity demanded.

(b) a payment that reduces the market price and quantity supplied of a
good by the seller.

(c) a payment that reduces the buyer's price to below the seller's price.

(d) Both (a) and (c).

149
9
Competitive Markets: Applications

4. If the government imposes a tariff on motor vehicles imported into an

economy:

(a) Government collects revenue equal to the tariff times the quantity
consumed.

(b) The change in producer surplus is equal to the tariff times the quan-
tity supplied by domestic producers.

(c) Consumers lose consumer surplus equal to the tariff times the quan-
tity of motor vehicles consumed.

(d) None of the above.

5. When the government imposes a specific tax on the sale of a good or

service:

(a) the price the buyer pays must exceed the price the seller received by
the amount of the tax.

(b) the price the buyer pays will be higher than the untaxed market
equilibrium price by the amount of the tax.

(c) the price the seller received will be lower than the untaxed market
equilibrium price by the amount of the tax.

(d) Both 1 and 2

6. In a competitive market with no government intervention, the equilibri-

um price is $10 and the equilibrium quantity is 10,000 units. Explain
whether the market will clear under each of the following forms of gov-
ernment intervention:

(a) The government imposes an excise tax of $1 per unit

(b) The government pays a subsidy of $5 per unit produced

(c) The government sets a price floor of $12

(d) The government sets a price ceiling of $8

(e) The government sets a production quota, allowing only 5,000 units
to be produced.

150
 9.8
Learning by doing

7. Assume that a competitive market has an upward-sloping supply curve

and a downward-sloping demand curve, both of which are linear. A tax
of size $T is currently imposed in the market. Suppose the tax is dou-
bled. By what multiple will the deadweight loss increase? (You may as-
sume that at the new tax, the equilibrium quantity is positive).

8. Let demand be QD = 100 – P and supply be QS = P.

a. Calculate the competitive market equilibrium, consumers’ surplus,

producers’ surplus, and total surplus.

b. Calculate consumers’ surplus, producers’ surplus, and total surplus

if a production quota of Q = 45 is set. Is this more or less than your
answer to (a)? How does you answer change if the quota is set at Q
= 55?

c. Calculate consumers’ surplus, producers’ surplus, and total surplus

if an excise tax of T = 10 is imposed on suppliers. How does your
answer compare to (b)? How does government revenue compare
under the quota and the tax? What if this is a subsidy for suppliers
of T = 10?

d. Calculate consumers’ surplus, producers’ surplus, and total surplus

if a price ceiling of P = 45 is imposed in this market. Calculate con-
sumers’ surplus, producers’ surplus, and total surplus if a price
floor of P = 55 is imposed on this market. How do these surpluses
compare?

9. Explain why foreign producers might prefer quotas to tariffs.

10. Suppose the market demand for cigarettes is: QD = 10 - P, and the supply
of cigarettes is: QS = -2 + P, where P is the price per pack of cigarettes. If
the government imposes a cigarette tax of $1 per pack:

(a) What is the price paid by consumers? What is the price faced by
suppliers?

(b) What is the government revenue from the tax?

(c) How much is consumers' tax and producers' tax burden?

(d) What is the deadweight loss of the tax?

151
10
Monopoly and Monopsony

10 Monopoly and Monopsony

Chapter objectives:
1. List the “perfectly competitive” assumptions that are not met by a mo-
nopolist and relate these to the features found in such a market. Name
three types of barriers to entry.

2. Define marginal revenue and explain why the monopolist’s marginal

revenue decreases as output increases.

3. Draw and interpret a diagram representing the price and output choices
of a profit-maximizing monopolist.

4. Explain why monopoly market exist.

5. Distinguish a natural monopoly from other monopolies.

152
 10.1
Profit Maximization by a Monopolist

10.1 Profit Maximization by a Monopolist

Until now, we have analyzed markets characterized by many small buyers
and sellers. For example: Travel to Europe and experience “Bed and Break-
fast” in a private household offering cheap overnight accommodation to
tourists. This feature of tourist travel is a reasonable approximation of a
perfectly competitive market. The key elements include

 many sellers and buyers;

 fairly homogeneous products; and
 ease of entry (all you need is a spare bed!).
Private motels at the beach are a similar example in the United States. When Pure monopoly
these competitive elements are absent, imperfect competition (monopoly,
oligopoly, or monopolistic competition) is present. A pure monopoly occurs
when there is a single firm in an industry producing a product with no close
substitutes. Barriers to entry are the reason that this situation can arise and
endure. Different forms of barriers to entry are: legal (franchises, patents),
economies of scale, and exclusive ownership of a necessary input. Govern-
ment franchises and economies of scale are typical reasons for the emer-
gence of natural monopolies.

We will now examine production and pricing decisions in monopoly mar- Monopolist
kets, which are markets with many small buyers but only a single seller. Since
the monopolist is the only seller in his market, he faces the entire market
demand, which we can write as P(Q). We will assume that the monopolist
chooses output so as to maximize profit. We knew that a firm’s profit maxi-
mization problem can be written as:
Max (Q)  TR(Q)  TC(Q)
Q

where Q is the monopolist’s output, TR(Q) is the total revenue of producing

Q units, and TC(Q) is the total cost of producing Q units.

The profit maximizing output is the output such that marginal revenue
equals marginal cost. In other words, as long as an additional unit earns
more than it costs, MR(Q) > MC(Q), the firm can increase profits by increas-
ing production. As soon as the additional unit costs more than it earns,
MR(Q) < MC(Q), the firm can increase profits by decreasing production.
Only at the output level where MR(Q) = MC(Q) can the monopolist’s profits
be maximized. In other words, the profit maximization condition for a mo-
nopolist is:

MR(Q) = MC(Q)

153
 10
Monopoly and Monopsony

Case Study 10-1 Total Revenue, Cost and Profit for the Beryllium Monopolist

Q P TR TC Profit
(million ($/oz.) ($ million) ($ million) ($ million)
ounces)
0 12 0 0 0
1 11 11.00 0.50 10.50
2 10 20.00 2.00 18.00
3 9 27.00 4.50 22.50
4 8 32.00 8.00 24.00
5 7 35.00 12.50 22.50
6 6 36.00 18.00 18.00
7 5 35.00 24.50 10.50
8 4 32.00 32.00 0
9 3 27.00 40.50 13.50
10 2 20.00 50.00 -30.00
TR, TC, and π (millions

TC
of dollars per year)

(a)

Profit Quantity (mil-

TR
(π) lions of ounces
0 4 per year)
Price (dollars per ounce)

MC

(b)

MR D Quantity (mil-
lions of ounces
0 4 per year)

154
 10.1
Profit Maximization by a Monopolist

Panel (a) shows the monopolist’s total revenue curve, TR, total cost curve,
TC, and total profit curve, π. Total cost increases as Q increases. Total reve-
nue first increases and then decreases. Similarly, total profit first rises and
then falls. The monopolist’s profit is maximized at Q = 4 million ounces.
Panel (b) shows the monopolist’s profit-maximization condition, MR = MC,
using the marginal curve and marginal cost curve that correspond to the
total revenue and total cost curve in panel (a). The marginal revenue and
marginal cost curve intersect at the point where total profit is at its maxi-
mum.

Marginal Revenue and Average Revenue

For a firm in a perfectly competitive market, marginal revenue was equal to Market power
price: since each perfectly competitive firm takes the price as given, the extra
revenue provided by the sale of one more unit of output is the price itself.
For a monopolist, this is not true: marginal revenue does not equal price
because the monopolist realizes that its output is so large that its supply
decisions do affect price. The fact that the monopolist can influence market
price by its output decisions is called market power. Perfectly competitive
firms do not have market power. Monopolists do.

For a firm with market power, marginal revenue is less than price. When the
monopolist increases production by one unit, it earns the market price for
that unit. On the other hand, in order to sell the additional unit, the monopo-
list realizes that it must reduce price slightly. This reduction in price affects
all units the monopolist wishes to sell and not just the last unit. All these
units except the last unit are called the inframarginal units. Hence, the mo-
nopolist suffers a loss due to the price reduction on all the inframarginal
units. Therefore, for a small increase in output, the marginal revenue from
the additional unit sold has two components:

MR(Q) = P + Q(∆P/∆Q)

Since the price that demanders are willing to pay falls as consumption rises,
(∆P/∆Q) is negative. Therefore, MR(Q) must be less than price. When few
items are affected by the price cut, the increase in sales increases total reve-
nue, and marginal revenue is positive. However, as progressively more
items suffer a lower price, the increase in revenue from the new sales is off-
set by the decrease in revenue from existing sales, and total revenue decreas-
es – marginal revenue becomes negative.

If total revenue can be written as the output sold times the price demanders Marginal revenue
are willing to pay to consumer that output, TR(Q) = QP(Q), then it follows
Average revenue
that average revenue is:

155
10
Monopoly and Monopsony

TR PxQ
AR =  P
Q Q

Therefore, the inverse demand curve is precisely the average revenue curve.
Total revenue, average revenue, marginal revenue, and market demand are
related as shown in Figure 10-1.

Figure 10-1 Total, Average, and Marginal Revenue

TR

TR
0 Q

AR,
MR

AR = Market Demand

MR

0 Q

The Monopolist’s Profit-Maximization Condition shown graphically

Figure 10-2 illustrates the profit-maximization condition MR = MC for our

beryllium monopolist. The marginal revenue curve MR is a decreasing line
that lies below the demand curve (which is also the average revenue curve)
for all positive output levels. The firm’s marginal cost curve MC is a straight
line from the origin. This Figure also shows the monopolist’s average cost

156
 10.1
Profit Maximization by a Monopolist

curve, AC. For all positive output level, the firm’s marginal cost curve lies
above this average cost curve.

The Monopolist’s Profit-Maximization Condition Figure 10-2

12
MC

A
8

B
AC
4
E
2
F MR D
0 4 Q

The profit-maximizing quantity is where the marginal revenue curve inter-

sects the marginal cost curve. This occurs at a quantity of 4 million ounces
per year. We find the monopolist’s profit-maximizing price by referring back
to the market demand curve. In Figure 10-2, this price is $8 per ounce, the
price at which quantity demanded is equal to 4 million ounces.

The monopolist’s profit is the difference between total revenue and total cost.
Total revenue is price (or average revenue) times quantity, which corre-
sponds to area B + E + F in Figure 10-2. Total cost is average cost times quan-
tity, which corresponds to area F. Total profit is thus B + E + F – F = B + E.
This area equals $24 million.

There are three important points about the equilibrium in a monopoly mar-
ket:

 First, the monopolist’s profit-maximizing price ($8) exceeds the marginal

cost of the last unit supplied ($4). This differs from the outcome in a per-
fectly competitive market, in which price equals the marginal cost of the
last unit supplied.

157
 10
Monopoly and Monopsony

 Second, in contrast to a perfectly competitive firm in a long-run equilib-

rium, the monopolist’s economic profits can be positive. This is because
the monopolist does not face the threat of free entry that drives economic
profits to zero in competitive markets.

 Third, even though the monopolist raises price above marginal cost and
earns positive economic profits, consumers still enjoy some benefits at
the monopoly equilibrium. The consumer surplus at the equilibrium in
Figure 10-2 is area A, the area between price and demand curve. In this
example, consumer surplus equals $8 million. The total economic benefit
at the monopoly equilibrium is thus the sum of consumer surplus and
the monopolist’s profit, which is equal to A + B + E, or $32 million per
year.

Case Study 10-2 Profit-maximizing quantity and price for the monopolist

Problem

The equation of the monopolist’s demand curve is P = 12 – Q, while the equa-

tion of marginal cost is given by MC = Q, where Q is expressed in millions of
ounces. What is the profit-maximizing quantity and price for the monopo-
list?

Solution

Our approach to solve this problem proceeds in three steps: First, we find
the marginal revenue curve. Second, we equate marginal revenue to margin-
al cost to find the profit-maximizing quantity. Third, we substitute this quan-
tity back into the demand curve to find the profit-maximizing monopoly
price.

We have total revenue:

TR = PQ = 12Q – Q2.

Thus, MR = 12 – 2Q

If we now equate this to marginal cost, we can solve for the profit-
maximizing quantity of output:

MR = MC

12 – 2Q = Q, which implies

Q=4

We can now substitute this back into the market demand curve to determine
the monopolist’s optimal price:

158
 10.2
The Inverse Elasticity Pricing Rule

P = 12 – 4 = 8

The monopolist’s profit-maximizing price is thus $8 per ounce, and its profit-
maximizing quantity is 4 million ounces, just as we found when we solved
the monopolist’s problem graphically in Figure 10-2.

10.2 The Inverse Elasticity Pricing Rule

It is sometimes useful to rephrase the profit maximization condition of the
monopolist in terms of the price elasticity of demand, εQ,P = (∆Q/∆P)(P/Q).
Manipulating our expression for marginal revenue, we have:

MR(Q) = P + Q(∆P/∆Q) = P[1 + (Q/P)(∆P/∆Q)] or

MR(Q) = P(1 + 1/εQ,P ) (10.1)

From Equation 10.1, we can conclude an important set of relationship:

 When demand is elastic, that is, when εQ,P is between -1 and - ∞, margin- Relationships
al revenue is positive. This is because when εQ,P is between – 1 and - ∞, between elastici-
the expression (1 + 1/εQ,P ) is greater than zero. Thus, when the monopo- ty of demand and
list operates on the elastic region of the demand curve, it can increase to- revenue
tal revenue by increasing quantity through a lower price.

 When demand is inelastic, that is, when εQ,P is between 0 and - 1, mar-
ginal revenue is negative. This is because when εQ,P is between 0 and - 1,
the expression (1 + 1/εQ,P ) is less than zero. Thus, when the monopolist
operates on the inelastic region of the demand curve, it can increase total
revenue by decreasing quantity through a higher price.

 When demand is unitary elastic, that is, when εQ,P = -1, marginal revenue
is zero. This is because when εQ,P = –1 implies (1 + 1/εQ,P) = 0.

Hence, we can also rewrite the profit maximization condition as:

MR(Q) = P(1 + 1/εQ,P ) = MC(Q) or

(P - MC)/P = - 1/εQ,P (10.2)

The entire final expression is called the inverse elasticity pricing rule (or IEPR).
The IEPR states that the monopolist’s optimal markup of price above mar-
ginal cost is equal to minus the inverse of the price elasticity of demand.
Why it is important? The IEPR tells us that the price elasticity of demand

159
10
Monopoly and Monopsony

plays a vital role in determining what price a monopolist should charge to

maximize profits.

10.3 Monopoly Comparative Statics

We turn now to the comparative statics of the monopolist’s profit maximiza-
tion problem. If the market demand curve shifts up, then the monopolist’s
equilibrium output increases. From Figure 10-3, below, we can see that an
upward shift in demand increases the price that customers are willing to pay
for each level of output, Q. This increases marginal revenue. Since marginal
revenue rises but marginal cost does not, units that previously were unprof-
itable to produce now are profitable. Therefore, equilibrium output increas-
es. The point of intersection of marginal revenue and marginal cost must
move to the right. If MC is constant or increases with quantity, then equilib-
rium price rises as well. If MC is downward sloping, however, then equilib-
rium price may fall as shown in the second panel of Figure 10-3.

Figure 10-3 The Comparative Statics of the Monopolist’s Profit Maximization Problem

P1
P0
MC
D1

D0

0 Q0 Q1 MR0 MR1 Q

160
 10.3
Monopoly Comparative Statics

P0
P1

MR1
D1

MR0 D0
MC
0 Q0 Q1 Q

An upward shift in marginal cost will decrease the profit maximizing quan-
tity and increase the profit maximizing price for the monopolist. The profit-
maximizing quantity decreases because units that were previously just prof-
itable are unprofitable at the new higher marginal cost. If equilibrium quan-
tity decreases, however, then market price will rise as long as market de-
mand slopes downward.

Marginal Cost and the Profit-Maximization Problem Figure 10-4

P1
P0
MC1

MC0

MR D

0 Q1 Q0 Q

The effect of the price increase or decrease on revenues depends on the elas-
tic of demand. If demand is elastic, revenues increase when price decreases.
But if demand is inelastic, revenues decrease when price decreases.

161
10
Monopoly and Monopsony

10.4 The Welfare Economics of Monopoly

We showed that the perfectly competitive equilibrium maximized total sur-
plus. Because the monopoly equilibrium does not correspond to the perfect-
ly competitive equilibrium, the monopoly equilibrium generally entails a
deadweight loss. For example, consider a market that is perfectly competi-
tive. What would be the deadweight los from converting production to mo-
nopoly?

The deadweight loss due to monopoly is shown as the shaded area in Fig-
ure 10-5, below, where the perfectly competitive equilibrium is point eC and
the monopoly equilibrium is point eM. The deadweight loss results because
demanders who value the product at more than its opportunity cost (i.e.,
marginal cost) are unable to buy when the monopolist restricts supply in
order to force up prices (and profits). In other words, the deadweight loss is
due to unrealized gains from trade.

Figure 10-5 Deadweight Loss due to Monopoly

MC
eM
PM
eC
PC

MR D
0 Q M Q C Q

10.5 Why do Monopoly Markets exist?

We have studied how a profit-maximizing monopolist determines its quanti-
ty and price. And because quantity and price differ from the perfectly com-

162
 10.5
Why do Monopoly Markets exist?

petitive equilibrium, we have seen that the monopoly equilibrium creates a

deadweight loss. How do monopolies arise in the first place?

Natural Monopoly

Suppose that the total cost incurred by a single firm to produce output is less
than the combined total cost of two or more firms if they divided that output
among them. More formally:

TC(Q) < TC(q1) + TC(q2) + … TC(qN) for all N ≥ 2

where Q = q1 + q2 + … qN is the sum of the outputs of N ≥ 2 firms and C(qi) is

the cost to firm i of producing qi. A market with this type of underlying cost
condition is called a natural monopoly because it is more efficient for single
firm to produce than for many firms to produce the same total quantity of
output. As a result, a monopoly producer could always “drive out” a set of
many small producers by pricing based on its lower cost. Therefore, it would
be “natural” for monopoly to arise. Such a market is illustrated in Figure 10-
6.

Natural Monopoly Market Figure 10-6

PM

AC
MR MC
D
0 QM Q

A natural monopoly occurs where average costs decrease as output levels

rise because of long-lasting economies of scale. In such circumstances, a
single-firm industry can be the most efficient way to organize production.

163
 10
Monopoly and Monopsony

Here it would be undesirable to break up the monopoly, because having one

producer is the “best” way to organize the market.

Natural monopoly can only arise if average costs are sufficiently decreasing,
that is, if they are decreasing up to the point where they meet the demand
curve or if they do not start increasing much before that point. We know that
when average costs are decreasing marginal cost must be less than average
cost. This is why this cost structure is not compatible with perfect competi-
tion: any small firm would be producing in the range where MC < AC so
that the competitive rule of P = MC would leave the firm with negative prof-
its. There are two important points about natural monopoly markets:

 A necessary condition for natural monopoly is that the average cost

curve must decrease with output over some range. That is, natural mo-
nopoly markets must involve economies of scale.

 Whether a market is a natural monopoly depends not only on technolog-

ical conditions (the shape of the AC curve) but also on demand condi-
tions. A market might be a natural monopoly when demand is low but
not when demand is high.

Barriers to entry

Barriers to entry A natural monopoly is an example of a more general phenomenon known as

barriers to entry which are obstacles in the path of a firm which wants to
enter a given market. Barriers to entry are factors that allow an incumbent
firm to earn positive economic profits, while at the same time making it
unprofitable for newcomers to enter the industry.

Perfectly competitive markets have no barriers to entry: When incumbent

firms earn positive profits, new firms enter the industry, driving profits to
zero. But barriers to entry are essential for a firm to remain a monopolist.
Without the protection of barriers to entry, a monopoly or cartel (a group of
producers that collusively determines the price and output in a market) that
earned positive economic profits would attract new market entry, and com-
petition would then dissipate industry profit. Barriers to entry can be struc-
tural, legal, or strategic.

Structural  Structural barriers to entry exist when incumbent firms have cost or
barriers marketing advantages that would make it unattractive for a new firm to
enter the industry and compete against it. The interaction of economies
of scale and market demand that gives rise to a natural monopoly market
is an example of a structural barrier to entry.

Legal barriers  Legal barriers to entry exist when an incumbent firm is legally protected
against competition. For example, patents are an important legal barrier
to entry.

164
 10.5
Why do Monopoly Markets exist?

 Strategic barriers to entry result when the incumbent firm takes explicit Strategic
steps to deter entry. An example of a strategic barrier to entry would be barriers
the development of a reputation over time as a firm that will aggressive-
ly defend its market against encroachment by new entrants (e.g., by start-
ing a price war if a new firm chooses to come into the market).

United States of America versus Microsoft: The Applications Barrier to Entry Case Study 10-3

Between October 1998 and June 1999, one of America’s best known and most
successful companies, Microsoft, went on trial for violating the U.S. antitrust
statutes. The U.S. government accused Microsoft of employing tactics aimed
at monopolizing the market for operating systems for personal computers
(PCs). In the opinion of the U.S. District Court, “Microsoft … engaged in a
concerted series of actions designed to protect the applications barrier to
entry, and hence its monopoly power, from a variety of … threats, including
Netscape’s Web browser and Sun’s implementation of Java. Many of these
actions have harmed consumers in ways that are immediate and easily dis-
cernible”.3

What does the Court mean by the term applications barrier to entry? This
phrase appears repeatedly in the Court’s opinion in this case. The Court uses
the term applications barrier to entry to describe a barrier to entry in the mar-
ket for PC operating systems based on positive network externalities. This
barrier, in the Court’s opinion, allowed Microsoft Windows to monopolize
the market for operating systems for Intel-compatible PCs. The Court de-
scribed the applications barrier to entry this way:

The fact that there is a multitude of people using Windows makes the products
more attractive to consumers. The large installed base attracts corporate custom-
ers who want to use an operating system that new employees are already likely
to know how to use, and it attracts academic consumers who want to use soft-
ware that will allow them to share files easily with colleagues at other institu-
tion. The main reason that demand for Windows experiences positive network ef-
fects, however, is that the size of Windows’ installed base impels ISVs [compa-
nies that write software applications] to write applications first and foremost to
Windows … The large body of applications thus reinforces the demand for Win-
dows, augmenting Microsoft’s dominant position and thereby perpetuating ISV
incentives to write applications principally for Windows. This self-reinforcing
cycle is often referred to as a “positive feedback loop”.

3 This quote comes from p.204 of the United States of America v. Microsoft, United States
District Court for the District of Columbia, Findings of Fact.

165
 10
Monopoly and Monopsony

What for Microsoft is a positive feedback loop is for would be competitors a vi-
cious cycle? For just as Microsoft’s large market share creates incentives for
ISVs to develop applications first and foremost for Windows, the small or non-
existent market share of an aspiring competitor makes it prohibitively expensive
for the aspirant to develop its PC operating system into an acceptable substitute
for Windows (pp. 18 - 19).

In the Court’s opinion – an opinion that Microsoft strongly disputes – many

of Microsoft’s actions toward competitors, such as Netscape and Sun, were
attempts to preserve this applications barrier to entry. For example, in the
summer of 1995, Microsoft attempted to convince Netscape to drop efforts to
develop a Web browser that could have served as a platform for Internet-
based software applications. The Court believed that Microsoft did this in
order to remove a threat to the applications barrier to entry that sustained
Windows’ dominance.

10.6 Monopsony
Monopsonist A monopsony market is a market consisting of a single buyer that can pur-
chase from many sellers. We call this single buyer a monopsonist. We will
derive the profit maximizing equilibrium for the monopsonist, derive the
equivalent of IEPR for this type of market, and analyze the welfare conse-
quences of monopsony. We will see that it is analogous to our analysis of
monopoly.

Monopsony Equilibrium

The basic profit maximization condition for the monopsonist can be ex-
pressed:

MRPL = MEL

where L is the input of which the monopsonist is the sole buyer. If L repre-
sents labor, then MRPL is called the marginal revenue product of labor. It is
the additional revenue that the monopsonist gets from using one more unit
of labor input. In other words, it is the benefit of increasing the use of labor
slightly. More precisely, the MRPL will be the marginal product of labor
times the revenue earned from this marginal product. In the simple case

166
 10.6
Monopsony

where the monopsonist sells its product in a competitive market, the mar-
ginal revenue product is, then:

MRPL = P x MPL

where P is the price of the output and MPL is the increase in units of output
when input L is increased slightly.

MEL is the marginal expenditure on labor. It is the increase in the firm’s total
cost when it employs an additional unit of labor. If labor is supplied at an
increasing price (so that, for example, the labor supply curve, w(L), slopes
upward), then the marginal expenditure on labor is composed of two terms.
The first term is the market wage. The second is the increase in wages that
occurs due to the unit increase in employment times the total labor forces
(since the increase in wages applies to all employed persons). Hence, we
have:

MEL = w + L(∆w/∆L)

The first term reflects the expenditure on the marginal unit, and the second
term reflects the increase in expenditure on the inframarginal units. The mar-
ginal expenditure on labor lies above the supply curve for labor. This can be
seen algebraically by noting that ∆w/∆L is positive when the supply curve
for labor is upward sloping and can be seen graphically in Figure 10-7, be-
low.

The Monopsony Equilibrium Figure 10-7

W
MEL

Supply of
labor W(L)
MRPL*

w*

MRPL
0
L* L

167
 10
Monopoly and Monopsony

An Inverse Elasticity Pricing Rule for Monopsony

Buyer power The fact that the monopsonist can influence its input prices is called buyer
power. Firms in perfectly competitive input markets do not have buyer
power. Monopsonists do. In fact, we could measure this buyer power by
rewriting the monopsony equilibrium condition, MRPL = MEL in the form of
an inverse elasticity pricing rule for monopsony in the same way as we devel-
oped an IEPR for monopoly. Hence, defining the elasticity of labor supply,
εL,w, as εL,w = (w/L)(∆L/∆w) we have:

MEL = w + L(∆w/∆L) = w[1 + (L/w)(∆w/∆L)] or

MEL = w(1 + 1/εL,w )

So that the profit maximization condition for the monopsonist becomes:

MRPL = MEL = w(1 + 1/εL,w ) or

(MRPL - w)/w = 1/εL,w

This final expression is the IEPR for the monopsonist. In other words, the
percent deviation between the marginal revenue product and the wage de-
pends, in equilibrium, on the elasticity of supply of labor in the same way as
the percentage deviation between the marginal cost and the price of the
monopolist depended on the price elasticity of demand. We can use this
equation, in turn, to express the amount of buyer power in the input market.

Deadweight Loss under Monopsony

The monopsonist’s surplus is measured as the area under the MRPL curve
and above the equilibrium wage. In other words, it is the benefit earned on
each unit of input net of the input’s price. Since the monopsonist is the buyer
of the input, this surplus is equivalent to consumers’ surplus in our model of
monopoly. The surplus earned by labor is measured as the area under the
equilibrium wage and above the supply curve of labor. In other words, it is
the benefit earned by each “unit of labor” net of the wage that would be
required to induce each “unit of labor” to work. Since labor is the input that
is supplied in this model, this surplus is equivalent to producers’ surplus in
our model of monopoly.

By comparing the monopsony solution, (L*, w*), to the competitive solution,

(LC, wC), we can deduce that the monopsony results in a deadweight loss of
area B + C in Figure 10-8. This represents the loss due to the input restriction
of the monopsonist: there are unrealized gains from trade since units of
labor that would be willing to work at a wage that is less than their marginal
revenue product go unemployed.

168
 10.7
Learning by doing

Deadweight Loss from Monopsony Figure 10-8

W MEL

Supply of
labor W(L)

B
wc
C
w*

MRPL
0
L* LC L

10.7 Learning by doing

1. A pure monopoly is an industry with a single firm that produces a prod-
uct that has _____ close substitutes and in which there are _____ barriers
to entry.

(a) many; significant

(b) many; no

(c) no; significant

(d) no; no

2. Your local electric company maintains its monopoly because of

(a) a patent.

(b) product differentiation.

(c) advertising.

(d) a government franchise.

169
10
Monopoly and Monopsony

3. In a monopoly,

(a) The market demand curve is above, and parallel to, the marginal rev-
enue curve

(b) The marginal revenue curve is downward sloping

(c) Increasing price will not result in a decrease in quantity demanded

(d) We assume that the demand curve is unknown

4. If a natural monopoly is split into a number of smaller competing firms,

price will

(a) increase, because smaller firms will have higher average costs.

(b) increase, because each firm will experience diseconomies of scale.

(c) decrease, because additional competition is taking place.

(d) increase in the short run because of the disruption, but decrease in
the long run because of the additional competition.

5. Which of the following is not a barrier to entry?

(a) Ownership of patent rights

(b) Ownership of private property

(c) The possession of a government franchise

(d) Substantial economies of scale 3.

6. The Monopolist MONO is seeking to maximize profits. Currently, he is

producing where marginal revenue is less than marginal cost. He should

(a) increase production.

(b) reduce price.

(c) reduce production.

(d) produce where price is equal to marginal cost.

7. The profit-maximizing monopolist must decide all of the following ex-

cept

(a) output level.

170
 10.7
Learning by doing

(b) price level.

(c) the wage level.

(d) the combination of inputs.

8. A monopolist is currently maximizing profits. We can conclude that

(a) he is maximizing total revenue and minimizing total cost.

(b) he has reduced the difference between marginal revenue and margin-
al cost to zero.

(c) he is maximizing total revenue and marginal revenue.

(d) he is producing where marginal revenue equals average cost.

9. Let a monopolist face inverse demand P(Q) = 100 – Q and marginal cost
MC(Q) = 20. What is the profit maximizing price and quantity for the
monopolist? Use IEPR to derive the price elasticity of demand at the
profit maximizing price.

10. Let a monopolist have total cost TC(Q) = 2Q and face an inverse demand
of P(Q) = 20 – Q.

(a) What is the profit maximizing output and price? At this point, what is
the profit of the monopolist? How does it compare to producer’s sur-
plus?

(b) Now, suppose that TC(Q) = 2 + 2Q. What is the profit maximizing
output and price? How do profit and producer’s surplus compare at this
point?

171
11
Capturing Surplus

11 Capturing Surplus

Chapter objectives:
1. Explain how a firm with market power can capture even more surplus
by engaging in price discrimination.

2. Identify three basic degrees of price discrimination.

3. Explain how price discrimination affects profits and the sum of consum-
er and producer surplus.

4. Explain how firms can capture more surplus with using bundling

5. Explain the use of advertising to create and capture surplus.

172
 11.1
Capturing Surplus

11.1 Capturing Surplus

In chapter 10 we saw how a monopolist will choose its quantity (and its
uniform price) to maximize its profit. With this quantity, monopolist can
maximize profit where marginal revenue equals marginal cost, as in Figure
11.1, below.

The Monopsony Equilibrium Figure 11-1

P
A

MC
eM
PM
eC
PC

MR D
0 QM QC Q

This condition results in an amount of consumer’s surplus, triangle AP MeM,

an amount of producer’s surplus, area 0PMeMB, and a deadweight loss equals
to the shaded area in the figure.

Imperfectly competitive firms, such as monopolies, may be able to practice Price

price discrimination. By charging different customers different prices, profits discrimination
can be boosted. Essentially, the seller must identify those buyers who are
willing to pay more for a good and those who will pay less. The seller is
appropriating a portion of the consumer surplus.

How can the monopolist capture even more surplus by charging more than
one price for its product, which is the essential feature of price discrimina-
tion? There are three basic types (traditionally called degrees) of price dis-
crimination:

173
 11
Capturing Surplus

 First-degree price discrimination

 Second-degree price discrimination
 Third-degree price discrimination

11.1.1 First-Degree Price Discrimination

To understand first-degree discrimination, it helps to think of the demand
schedule for a product as a willingness-to-pay schedule. After all, a demand
curve represents the amounts consumers are willing to pay for the units they
purchase. Since the demand curve slopes downward, the person buying the
first unit is willing to pay a higher price than the consumer buying the sec-
ond unit. The maximum willingness-to-pay declines with each successive
unit purchased.

The reservation First-degree price discrimination is ideal from the seller’s viewpoint. If the
(or reserve) price firm knows this maximum willingness to pay for each unit4, then a first-
degree price discriminating monopolist’s profit maximizing strategy is to
charge the maximum willingness to pay for each unit. For example, the
consumer who values a unit at 10 would be charged 10, while the consumer
who values a unit of the good at 9 would be charged 9 and so on. Alterna-
tively, a consumer wishing to buy 3 units would be charged 10 for the first, 9
for the second, and 8 for the third (we call that is consumer’s reservation
price for each unit), as shown in Figure 11-2.

With first-degree price discrimination, the marginal revenue generated by

each unit sold is exactly the price of that unit. Because the monopolist is now
able to charge a different price for each unit of the good, an additional unit
of the good can be sold without having to decrease the price of the other
(inframarginal) units. The sale of the additional unit can be secured by set-
ting the price for this unit equal to the consumer’s willingness to pay. As a
result, the monopolist continues to sell units as long as the price of that unit
exceeds the marginal cost of producing the unit.

Figure 11-2

4 Often sellers can learn something about willingness to pay based on knowledge of
where a person lives and works, how he dresses or speaks, the kind of car he
drives, or how much money he makes. The information may not perfectly reveal a
consumer’s willingness to pay, but it can help the seller to capture more surplus
than it could without such information.

174
 11.1
Capturing Surplus

Consumer’s reservation price

P
10

0 Q

The maximum willingness to pay for the last unit placed on the market is
exactly equal to marginal cost. The monopolist will not sell any more units
because the marginal cost exceeds the price received for any additional
units. Hence, the price discriminating monopolist maximizes profits by
selling Q* units in Figure 11-3.

Monopoly with First-Degree Discrimination Figure 11-3

P A

MC
eM
PM
e*
P*

MR D
0 QM Q* Q

175
11
Capturing Surplus

As usual, we calculate producer’s surplus as the difference between the

market price and the marginal cost for every unit sold. When the monopolist
price discriminates, however, market price is traced out by the demand
curve for every unit sold. Hence, at output level Q*, producer’s surplus is the
entire area below the demand curve and above marginal cost for all units up
to Q*. This is the area of triangle Ae*0. Consumer’s surplus is zero, since
each consumer pays a price equal to his maximum willingness to pay. In
other words, consumers will receive no surplus because the producer has
captured all of it. Moreover, the deadweight loss is also zero because output
is sold up to the point where the willingness to pay is just equal to the op-
portunity cost of resources. In fact, Q* is exactly the same output level as the
perfectly competitive equilibrium output level.

11.1.2 Second-Degree Price Discrimination

If a seller does not know each consumer’s exact willingness to pay for each
unit of the good, then it is not possible to pursue a policy of first degree price
discrimination: all consumers would have an incentive to claim that their
willingness to pay is very low so as to obtain the product cheaply. On the
other hand, the monopolist may still be able to follow a policy of second-
degree price discrimination. Under second-degree price discrimination, the
firm offers consumers a quantity discount. The amount you pay per unit
depends on the number of units you purchase5. For example, the price you
pay for a piece of computer software often depends on the number of copies
you buy. A firm may offer you a computer game for $50 if you buy only one
copy. But it may also offer you 10 copies for $400, or only $40 per copy.

With second-degree price discrimination, the amount you pay for a good or
service actually depends on two or more prices. A common example of sec-
ond-degree discrimination is a block tariff, which is a particular type of
quantity discount. To illustrate, a block tariff with two “blocks” means that a
consumer pays a price for the first x units consumed and a different price for
any additional units consumed.

For example, many consumers buy their telephone service under a multipart
tariff. You may pay $20 per month to be hooked up to the telephone system,

5 While we observe quantity discounts in everyday life, not every form of quantity
discounting is the result of price discrimination. Often sellers offer quantity dis-
counts because it costs them less to sell a larger quantity. For example, a pizza that
serves four people usually sells for less than twice the price of a pizza for two
people. Labor, cooking, and packaging costs are not very sensitive to the size of
the pizza. The price reflects the fact that the cost per ounce is lower for a large piz-
za.

176
 11.1
Capturing Surplus

even if you never make a telephone call. The $20 fee is called a subscriber
charge. You may also pay a usage charge – that is, an additional amount (say,
five cents) for each local call you make. If you make only one call, your
monthly total telephone bill will be $20.05. If you make 100 calls, your total
bill will be $25.00. Note that your average expenditure per telephone call
falls when you make more calls. The average price you pay is $20.05 if you
make only one call per month, but the average price falls to $0.25 if you
make 100 calls.

Declining Block Tariffs for Electricity Case Study 11-1

When a power company sells electricity with a block tariff, it does not know
each individual’s demand schedule. However, it does know that some cus-
tomers have larger demands for electricity than others. It also knows that
each consumer’s demand curve is downward sloping, so that a lower price
will stimulate that consumer to purchase more electricity.

Suppose the market has two customers in the market, Mr. Large and Mr.
Small, the below Figure shows their demand curves for electricity D Large and
DSmall. If the company charges a uniform price P1 for all units of electricity
sold, Mr. Small will buy Q1S units of electricity, and Mr. Large will purchase
Q1L units. Now let the company introduce a block tariff. It announces that it
will sell up to Q1L units (the first block) at a price of P1. Any customer who
buys more than Q1L units can purchase the additional units at a lower block
price P2.

How will the block pricing affect Mr. Small, Mr. Large, and the electric pow-
er company? Mr. Small’s purchases are unchanged because he does not pur-
chase enough electricity to take advantage of the lower block price P 2. He
still buys Q1S units and a price P1, and his consumer surplus is therefore the
same as it was under the uniform price.

Given the lower price of the second block, Mr. Large will expand his con-
sumption of electricity from Q1L to Q2L units. His consumer surplus increases
by the lightly shaded area in the figure. Finally, the company is better off
with the block tariff because its producer surplus increases by the darkly
shaded box in the figure.

This example illustrates an important potential benefit of block tariffs. If we Pareto superior
start with a uniform tariff that is different from marginal cost, then introduc-
ing a block tariff leads to a Pareto superior allocation of resources. A Pareto

177
11
Capturing Surplus

superior allocation of resources makes at least one participant in the market

better off and no one else worse off 6.

P
Additional consumer
surplus
P1
Additional producer
P2 surplus

MC

DSmall DLarge

0 Q1S Q1L Q2L Q

11.1.3 Third-Degree Price Discrimination

A seller may know how the willingness to pay for each unit varies across
broad customer groups, even if the seller does not know the exact willing-
ness to pay of each consumer. For example, it may be possible to find out
that students and senior citizens are less willing to pay for certain goods
(like movies) than other people. In such a case, the monopolist sells output
to different consumers for different prices, but each unit sold to any one
consumer group carries the same price. This type of price discrimination
typically requires each consumer to be easily identifiable as belonging to a
certain group. For example, it may be possible to identify a student by re-
quiring a student identification card to be shown at the time of purchase.

6 R.D. Willig, “Pareto Superior Nonlinear Outlay Schedules”, Bell Journal of Economics,
9 (1987): 56 – 69. The argument for the Pareto superiority of nonlinear outlay
schedules is clearest when the consumers are end users of electricity (for example,
households). The argument is a bit more complex when the purchasers of the elec-
tricity are firms that compete with one another in some market. One of the com-
plications arises because quantity discounts from block pricing could conceivably
allow a large, less efficient firm to produce with lower costs than a smaller, more
efficient firm, because the larger firm can purchase electricity at a lower average
price. Pareto superiority is named for the Italian economist Vilfredo Pareto (1848 -
1923).

178
 11.1
Capturing Surplus

The profit maximizing price will be chosen to equate marginal revenue to

marginal cost for that group’s demand curve. For example, suppose that
output is sold in two markets, Germany and United States, and that con-
sumers cannot exchange goods (arbitrage) across the two countries. Demand
in the two markets is represented in Figure 11-4.

If the marginal cost is the same for each market, the optimal price will set the
marginal revenue for each market equal to the (same) marginal cost. This
yields price P1 in market 1 and price P2 in market 2 with a surplus of area
P1Q1 + area P2Q2 captured by the price discriminating monopolist. Note that
this raises the surplus captured by the producer compared to the case where
the same price must be charged to the two groups. For example, suppose
that the price that sets marginal revenue equal to marginal cost in market 1
is used in both markets. Hence, under uniform pricing, the monopolist cap-
tures surplus of only area P1Q1 + area P1Q’. It can be seen from the figure that
area P1Q’ is smaller than area P2Q2 so that the monopolist captures less sur-
plus under this scheme.

Third-Degree Price Discrimination Figure 11-4

P P

D2
P2

D1
P1 P1

MC
MR MR
0 Q 0
Q1 Q2 Q’ Q

If the marginal cost differs across the two markets, the profit maximizing
price will be chosen to equate each market’s marginal revenue equal to each
market’s marginal cost (MR1 = MC1 and MR2 = MC2).

Case Study 11-2

179
11
Capturing Surplus

Pricing Airline Ticket

Airlines typically sell tickets at a variety of fares. Third-degree price discrim-

ination is one of the strategies airlines use to fill the plane with travelers in
the most profitable way. Airlines often charge different prices for seats in the
same class of service, such as coach class, even though the marginal cost of
serving a passenger is about the same for all coach passengers. The reason
for this is that different customers are willing to pay different amounts for
tickets. For example, people traveling on vacation often can book their tick-
ets weeks or months in advance of the flight, and they are willing to shop
around for the best price. Vacation travelers often choose their destinations
based on the availability of relatively inexpensive airline tickets. They are
usually sensitive to price, especially if the vacation involves buying tickets
for a whole family.

On the other hand, passengers traveling on business are often less sensitive
to the price of the tickets. When business requires that a passenger be in
London for an important meeting on Monday morning at 8:00 A.M., the
traveler will make the trip even if the fare is expensive.

An airline knows that markets have different segments. It knows that busi-
ness customers typically have relatively inelastic demands, and that most
vacation travelers have relatively elastic demands. Using the inverse elastici-
ty pricing rule, an airline would like to charge a higher price for business
travelers.

How does the airline implement price discrimination? It often imposes re-
strictions on tickets sold at lower prices. For example, an airline knows that
business travelers often do not know about the meeting in London far in
advance, whereas the excursion traveler schedules a vacation months in
advance. Business travelers often cannot stay at the destination over Satur-
day night, whereas vacation travelers will be willing to stay over Saturday
night, especially if they can get a cheaper ticket by doing so. The airlines
therefore use these restrictions to screen passengers.

180
 11.2
Tying (Tie-in Sales)

11.2 Tying (Tie-in Sales)

Tying (tie-in sales) refers to a case when customers may buy one product Tying (tie-in
only if they agree to buy another product as well. There are two types of tie- sales)
in sales. The first, the requirements tie-in sale, requires that customers who
make a purchase of one product from a firm make all their purchases of
another product from the same firm. For example, a customer who buys a
fax machine from a firm might be required to buy paper for the fax machine
from the same firm. The second, bundling, is a sale in which many units of
the same good – or two (or more) goods – are sold in a package. Customers
cannot buy these products (or units) separately. For example, most new cars
are bundles of products that could theoretically be sold separately such as a
tires, a heating system, a radio, seats, and so on. If these different products
are available either as a package or as separate “options”, the firm practices
mixed bundling.

Requirements tie-in sale

If one product is essentially useless without the other (such as a fax machine
and paper), a requirement tie-in sale allows the firm to meter demand of the
products and thus practice second-degree price discrimination. For example,
if the firm has a fixed price, F, for the fax machine and a price, p, for each
sheet of paper used, then it can charge the equivalent of a subscription fee
(F) plus a usage charge (p) for the machine and its use. Low-level users
would choose contracts with low subscription fees and a relatively high
usage charges while high-level users would select plans offering high sub-
scription fees but low usage charges.

Bundling

Bundling refers to tie-in sales in which customers are required to purchase

goods in a package. The customer cannot buy the goods separately. For
example, when you subscribe to cable television, you typically have to buy a
“package” of channels together, rather than subscribing to each channel
individually.

Why do firms sometimes sell two or more items as package instead of sepa-
rately?

Let’s consider a company that sells two different products, a computer and a
computer monitor. The marginal cost of the computer is $1,000, and the
marginal cost of monitor is $300.

181
 11
Capturing Surplus

For simplicity, suppose only two customers are in the market, but the firm
cannot price discrimination. Table 11-3 shows how much each customer is
willing to pay for a computer and for a monitor. Both customers might like
to buy a new computer and a new monitor. However, either customer might
like to buy a new computer or a new monitor alone (perhaps already having
an old monitor or an old computer).

Case Study 11-3 Bundling Can Increase Profit When Customer Preferences are Negatively Correlated

Reservation Price
(Maximum Willingness to Pay)
Computer Monitor

Customer 1 $1,200 $600

Customer 2 $1,500 $400

Marginal cost $1,000 $300

 First, let’s see how much profit the firm can earn if it does not bundle the
computer and the monitor?

 If the firm sets price of a computer (Pc) is $1,500, it will sell only one
computer (to customer 2) and earn a profit of $500 ($1,500 - $1,000 = $500).
If Pc = $1,200, it will sell two computer (one to each customer) and earn a
profit of $400 ($200 from each computer). So it should set the price of the
computer at $1,500.

 If the firm sets price of a monitor (Pm) is $600, it will sell only one moni-
tor (to customer 1) and earn a profit of $300 ($600 - $300 = $300). If Pm =
$400, it will sell two monitor (one to each customer) and earn a profit of
$200 ($100 from each monitor). So it should set the price of the computer
at $600.

The best the firm can do without bundling is to set Pc = $1,500 and Pm = $600.
It will then earn a total profit of $800 ($500 from the computer sales and $300
from the monitor sales).

 Now consider the option to bundle the computer and the monitor, selling
the two components in a single package. Customer 1 would be willing to
pay up to $1,800 for the package, and customer 2 would pay up to $1,900.

182
 11.2
Tying (Tie-in Sales)

If the bundle is sold at Pb = $1,900, only customer 2 will buy the bundle and
the total profit would be $600 ($1,900 - $1,300 = $600). If Pb = $1,800, both
customers will buy the bundle, and the total profit will be $1,000 [2($1,800 -
$1,300) = $1,000].

Thus, the manufacturer will maximize profit by selling a bundle at Pb =

$1,800. Bundling has increased profit from $800 (without bundling) to $1,000
(with bundling).

Why does bundling work to increase profit? The key is that the customers’
demands are negatively correlated. The negative correlation means that cus-
tomer 1 is willing to pay more than customer 2 for the computer, while cus-
tomer 2 is willing to pay more than customer 1 for the monitor. By bundling
the goods, the manufacturer is inducing the consumers to take both prod-
ucts when they might not otherwise do so.

What happens if the customer demands are positively correlated? Suppose

the customer demands are as shown in Table 11-4. Here the customer prefer-
ences are positively correlated because customer 1 is willing to pay more for
a monitor, and more for a computer, than customer 2.

Bundling Does Not Increase Profit When Customer Preferences are Positively Cor- Case Study 11-4
related

Reservation Price
(Maximum Willingness to Pay)
Computer Monitor

Customer 1 $1,200 $400

Customer 2 $1,500 $600

Marginal cost $1,000 $300

If the manufacturer does not bundle, it maximizes profit by selling comput-

ers at $1,500, earning a profit of $500 from each computer sold (only custom-
er 2) and profit of $300 from each monitor sold (only customer 2 will buy at

183
 11
Capturing Surplus

$600). Total profit will be $800. (You should verify that it would be less than
profitable for the firm to sell either a computer or a monitor at a price low
enough to attract customer 1).

If the manufacturer offers the computer and monitor as a bundle, the best
the firm can do is to set the price at $2,100, earning a profit of $800. There-
fore, bundling does not increase the firm’s profits.

Mixed Bundling

In practice, firms often allow customers to purchase components individual-

ly, as well as offering a bundle.

Let’s consider the example illustrated in Table 11-5. Each of the four custom-
ers is willing to pay $1,700 for a bundle. Their demands are negatively corre-
lated because a customer who is willing to pay more for a computer is will-
ing to pay less fro a monitor.

Case Study 11-5 Mixed Bundling Can Increase Profit

Reservation Price
(Maximum Willingness to Pay)
Computer Monitor

Customer 1 $900 $800

Customer 2 $1,100 $600

Customer 3 $1,300 $400

Customer 4 $1,500 $200

Marginal cost $1,000 $300

To see what the optimal strategy will be, let’s consider three options:

Types of bundling  No bundling. If the manufacturer does not bundle, it maximizes profit by
selling computers at $1,300 and monitor at $600. When Pc = $1,300, cus-
tomers 3 and 4 will buy computers and the profit from computers will be
$600 [2($1,300 - $1,000) = $600]. When Pm = $600, customers 1 and 2 will

184
 11.3
Advertising

buy monitors and the profit from monitors will be $600 [2($600 - $300) =
$600]. The total profit will be $1,200.

 Pure bundling (selling only a bundle). If the manufacturer offers customers

the computer and monitor as a bundle, priced at $1,700, all four custom-
ers buy the bundle. On each bundle the profit will be $400 ($1,700 -
$1,300 = $400). The total profit will be $1,600.

 Mixed bundling.
o Customer 1: He is only willing to pay $900 for computer, which is less
than the marginal cost of the computer. It will therefore not be profit-
able for the firm to sell a computer to this customer. If he buys a bun-
dle at $1,700, the firm makes a profit of $400 ($1,700 - $1,300 = $400).
The customer’s surplus is zero. However, the firm can make more
profit from customer 1 by selling the monitor separately. If the man-
ufacturer prices the monitor separately at $799, customer 1 will buy
it. The profit is $499 ($799 - $300 = $499) for the firm. The customer’s
surplus is $1 ($800 - $799 = $1). So the firm should set Pm = $799.

o Similarly, customer 4 is only willing to pay $200 for monitor, which is

less than the marginal cost of the monitor. If the manufacture sets Pc =
$1,499, and gets a profit of $499 ($1,499 - $1,000 = $499). The custom-
er’s surplus is $1 ($1,500 - $1,499 = $1). So the firm should set Pc =
$1,499.

o Finally, customers 2 and 3 have negatively correlated demands. Fur-

ther, the amounts that they are willing to pay for each component ex-
ceed the marginal cost. The firm would therefore like to sell them a
bundle. It should offer a package with a computer and a monitor at
Pb = $1,700.

In sum, with mixed bundling, customer 4 buys the computer separately,

customer 1 takes the monitor alone, and customers 2 and 3 buy the bundle.
Total profit is $1,798. The profit is higher with mixed bundling than it would
be with no bundling ($1,200) or selling only a bundle ($1,600).

11.3 Advertising
One of the important functions of advertising is to shift out demand. Sup-
posing, then, that this is the only effect of advertising, we can depict demand
with and without advertising as in Figure 11.5. While we have drawn de-
mand as rotating out, we could also suppose that the effect of advertising is
to cause a parallel shift out in demand.

185
 11
Capturing Surplus

While the shift clearly increases the profit that can potentially be earned by
the firm, advertising is also costly. As usual, the profit maximizing amount
of advertising must balance this cost against the potential benefits that can
be earned. Suppose that advertising is a pure fixed cost. The profit maximiz-
ing conditions of a monopolist who can control both the output level and
advertising expenditures are as follows:

1. Output is chosen optimally, as before, by setting MR Q = MCQ, holding

constant the level of advertising. MRQ is defined as the change in revenue
due to a change in output, holding the level of advertising constant. MCQ,
similarly, is defined as the change in total cost due to a change in output
only.

2. Advertising is chosen optimally by setting the marginal revenue due to

an increase in advertising holding output constant equal to the marginal
cost of increasing advertising also holding output constant. We can express
this as MRA = MCA. The marginal revenue from an increase in advertis-
ing holding output constant is the change in revenues that occur as de-
mand is shifted out due to a small increase in advertising expenditures.
Similarly, the marginal cost from an increase in advertising holding output
constant is the change in total cost that occurs when advertising increases
slightly.

By manipulating these two profit maximization we can obtain the expression

that summarizes these two profit maximizing conditions for the advertising
monopolist:

A E
  Q ,A
PQ E Q ,P

Q A
EQ,A is the elasticity of output with respect to advertising, E Q ,A  , and
A Q
Q P
EQ,P is the price elasticity of demand, E Q ,P  .
P Q

This expression must hold at the profit maximizing advertising expenditure

and output. The expression says that the ratio of advertising expenditure to
sales revenues (the left-hand side of the equation) must vary according to the
relative responsiveness of demand to advertising expenditure (the right-
hand side of the equation).

Figure 11-5

186
 11.4
Learning by doing

Effects of Advertising

Demand without advertising

Demand with advertising

0 Q

11.4 Learning by doing

1. Which of the following is NOT necessary for a firm to be able to engage
in price discrimination?

(a) A firm must have some market power.

(b) A firm must have some information about its consumers’ willingness
to pay.

(c) A firm must be able to prevent arbitrage.

(d) A firm must be a price-taker.

Questions 2–3 are based on the following information.

Suppose demand for a monopolist’s product is given by p = 400 − 4Q and has

constant marginal cost MC = 80. If this monopolist engages in first-degree
price discrimination.

2. Total output will equal

(a) 20

187
11
Capturing Surplus

(b) 40

(c) 60

(d) 80

3. Consumer surplus will be

(a) 0

(b) 1,600

(c) 3,200

(d) 4,800

4. In order to practice any form of price discrimination, a monopoly must

be able to:

(a) identify the maximum price that each customer is willing to pay.

(b) separate its customers into distinct groups.

(c) prevent resale of its product.

(d) all of the above are necessary to successfully practice price discrimi-
nation.

5. With third-degree price discrimination

(a) The firm tries to price each unit at the consumer’s reservation price.

(b) The firm offers consumers a quantity discount.

(c) The firm charges different consumer groups or market segment a dif-
ferent price.

(d) A buyer can only purchase one product by agreeing to purchase some
other products as well.

6. What type of price discrimination typically involves plans that offer

quantity discounts to customers?

(a) First-degree price discrimination

(b) Second-degree price discrimination

(c) Third-degree price discrimination

188
 11.4
Learning by doing

(d) Fourth-degree price discrimination

7. Which of the following is a real-world example of third-degree price

discrimination?

(a) A railroad charges more to haul 100 tons of coal than it does to haul
100 tons of grain.

(b) A movie theater charges senior citizens a cheaper price for movie
tickets than it charges non-senior citizens for the same movie ticket.

(c) An airline charges a lower price for a coach ticket purchased four
weeks in advance than for the same type of ticket purchased three days
in advance.

(d) Sam's Club warehouses sell bulk quantities of macaroni and cheese
for a cheaper per unit price than a grocery store, but the boxes are pack-
aged together so that the customer must buy six boxes at a time.

8. Which of the following is an example of first-degree price discrimina-

tion?

(a) a movie theatre gives senior citizen discounts.

(b) an electric utility that gives quantity discounts.

(c) a grocery store that doubles the value of coupons.

(d) a prestigious regional liberal arts college charges each student

the maximum tuition that they are willing to pay.

9. Suppose that a monopolist has a constant marginal cost MC = 2. The firm

faces the demand curve P = 20 – Q. There are no fixed costs.

(a) Suppose price discrimination is not allowed (or is not possible). How
large will the producer’s surplus be?

(b) Suppose the firm can engage in perfect first-degree price discrimina-
tion. How large will the producer’s surplus be?

10. Suppose that a monopolist can sell in two markets. Demand in market 1
is Q1 = 120 – p1 while demand in market 2 is Q2 = 20 – p2. Marginal cost
is 10 in both markets.

189
11
Capturing Surplus

(a) What is the profit maximizing price and quantity to provide in the
two markets if the monopolist can conduct third-degree price discrimi-
nation?

(b) What is the profit maximizing price and quantity to produce if the
monopolist must charge a uniform price in the two markets? Will both
markets be served?

(c) If the price discrimination scheme for the two markets has an admin-
istrative fixed cost of 30, will the profit maximizing monopolist pursue
this scheme?

190
 11.4
Learning by doing

12 Market Structure and

Competition

Chapter objectives:
1. Identify characterizes of types of market structures.

2. Explain the models of Cournot and Bertrand in oligopoly market with

homogeneous products.

3. Identify characterizes of dominant firm markets.

4. Identify characterizes of oligopoly with differentiated products.

5. Identify the factors and analyze equilibrium in monopolistic competition.

191
 12
Market Structure and Competition

12.1 Oligopoly with Homogeneous Products

Markets differ in two important dimensions: the number of sellers and the
nature of product differentiation. With many firms producing a homogene-
ous product, we have a perfectly competitive market. With a single firm, the
market becomes a monopoly (with only one seller, monopoly, or one buyer
monopsony). Between these two extremes are

 markets with a small number of firms (oligopoly),

 markets with a dominant firm, and
 markets with many differentiated-products producers (monopolistic com-
petition).

Oligopoly Oligopoly is the market structure with many buyers bust just a “few” inter-
dependent firms, each having market power and exerting strong barriers to
entry. The behavior of one firm in an oligopolistic industry depends on the
reactions of the others. Because the actions of each firm depend on the ex-
pected reactions of its rivals, this market structure is notoriously complex.

Duopoly. There may be only two firms producing the product (a duopoly), or only
three or four firms, or as many as twenty or thirty. Oligopoly products can
be either homogeneous (as is the case with gasoline, aluminum or steel) or
differentiated (automobiles, computers, electronic sound equipment). With
only a few firms, each firm faces its own market demand and each firm is
able to set its own price. However, in so doing, each oligopoly firm must
take into consideration the behavior of the other firms in response to its own
actions. That is, in oligopolistic industries there exists a certain strategic
interdependence.

Cournot Oligopoly

Cournot oligopoly is a model of oligopoly by the mathematician and econ-

omist Antoine Cournot7. To take the simplest case of a Cournot duopoly,
imagine there are two firms selling a particular product. Each firm decides
independently how much to produce and put on the market. However, the
market price of the good is determined by the inverse demand function
applied to the sum of what both firms put on the market.

Case Study 12-1

7 Antoine Augustin Cournot (1801-1877)

192
 12.1
Oligopoly with Homogeneous Products

Cournot Oligopoly

Problem

Suppose that two firms, named A and B, are competing in choosing quanti-
ties. Suppose the market demand curve is Q = 100 - p, and each firm's mar-
ginal cost is 10. Each firm takes the other firm's output as given, and chooses
its own output to maximize its profit. What is the equilibrium in this mar-
ket?

Solution

Suppose the A's output is q, and the B's output is g, then market price would
be: p = 100 - (q+g). The A's total revenue, when it produces q, given the B is
producing g, is

TRA = q[100-(q+g)].

A's marginal revenue is

MRA = 100 - g - 2q

To be optimal for the A, given B's output, A chooses its output so that mar-
ginal revenue equals marginal cost:

100 - g - 2q = 10.

Thus, A's optimal output is a function of B's output, and is given by

q(g) = (90 - g)/2

This function is called A's reaction function. Its curve is called the reaction Reaction
curve. Function.

Now firm B should set its output in a similar way. Given A's output q, B's
marginal revenue is

MRB = 100 - q - 2g.

To be optimal, B also sets its output so that marginal revenue equals margin-
al cost, that is,

100 - q - 2g = 10.

Firm B's reaction function is thus

g(q) = (90 - q)/2.

In equilibrium, each firm should correctly anticipate what the other firm
would do, and therefore should set q = q(g), and g = g(q). That is

q = (90 -g)/2

193
12
Market Structure and Competition

g = (90 -q)/2

The solutions to the two equations above, or the interception point of the
two reaction curves, Q* = q = g = 30, are the equilibrium output of the two
firms. We can find the equilibrium market price P* by substituting these
quantities into the market demand curve: P* = 100 – (30 + 30) = 40.

B’s output, g(q)

90

A’s reaction
function

Cournot
45
equilibrium
30 B’s reaction
function

A’s output,
0 30 45 90 q(g)

Cournot From this example, we have four steps to finding a Cournot equilibrium in
equilibrium which each firm chooses a profit-maximizing output given the output cho-
sen by other firms

1. Find the equation of the residual demand curve for each firm

2. Find the equation of the residual marginal revenue curve for each firm

3. Set the residual marginal revenue curve equal to marginal cost for each
firm

4. Solve the set of equalities in step 3 simultaneously to obtain the equilib-

rium output for each firm. These equilibrium output levels are the
Cournot equilibrium outputs for the market.

194
 12.1
Oligopoly with Homogeneous Products

Bertrand competition8

In the Cournot model, each firm selects a quantity to produce and the result-
ing total output determines market price. In contrast, the Bertrand model
assumes that each firm selects a price to charge and simply produces the
quantity required to satisfy the demand that comes its way. In the simplest
case, we will derive equilibrium in a market with two firms (duopoly) and a
homogeneous product. Each firm chooses its price to maximize its profits
given the anticipated price of the other firm. The choice of price occurs sim-
ultaneously and non-cooperatively.

To illustrate Bertrand price competition, let’s return to Case study 12-1. A

Bertrand equilibrium occurs when each firm chooses a profit-maximizing
price, given the price set by the other firm. Taking B’s price as fixed at 40, A’s
demand curve is a broken line that corresponds to the market demand D M at
prices below 40 and the vertical axis at prices above 40. If A slightly undercut
B’s price by charging 39, it would steal all of B’s business and would also
stimulate one unit of additional demand as well. Thus, though A’s price is
lower than before, it more than compensates for the lower price by more
than doubling its volume. As a result, A enjoys an increase in profit given by
area B (the gain from the additional volume of output it sells) minus area A
(the reduction in profit due to the fact that it could have sold 30 units at the
higher price of 40).

Bertrand Price Competition Figure 12-1

P
100

A’s demand curve

40
39 A

B DM
10 MC
q(g)
0 30 60 61 90 100

8 Bertrand competition is named after Joseph Louis François Bertrand (1822-1900).

195
12
Market Structure and Competition

But note that P1 = 39 and P2 = 40 cannot be an equilibrium either. At these

price, B would gain by undercutting A’s price. Indeed, as long as both firms
set prices that exceed their common marginal cost of 10, one firm can always
increase its profits by slightly undercutting its competitor. This implies that
the only possible equilibrium in the Bertrand model is when each firm sets a
price equal to its marginal cost of 10. At this point, neither firm can do better
by changing its price. If either firm raises price, it would sell nothing. Thus,
in the Bertrand equilibrium, P = MC = 10, and the resulting market demand
is 90 units. Thus, unlike the Cournot equilibrium with two firms, the Ber-
trand equilibrium with two firms results in the same outcome as a perfectly
competitive market with a large number of firms.

12.2 Dominant Firm Markets

A dominant firm is one which accounts for a significant share of a given
market and has a significantly larger market share than its next largest rival.
Normally, the dominant firm faces a number of small competitors, referred
to as a competitive fringe (the “competitive fringe”). Fringe firms are as-
sumed they are so small that their own output decisions cannot affect mar-
ket price. Hence, they choose their output levels so as to maximize profits
taking market price as given. In other words, they maximize profits by set-
ting price equal to marginal cost.

For the dominant firm, we assume that it wishes to set a price that maximiz-
es its profits taking into account both market demand. It will calculate its
residual demand and then equates residual marginal to marginal cost in
order to determine its profit maximizing price. In other words, instead of
taking the (single) quantity or price of the other firm as given (similar to
oligopoly firm), the dominant firm computes the total output of the competi-
tive fringe for each possible price level.

Suppose that the market demand is DM, and the fringe supply cure is SF.
Then, the residual demand of the dominant is, DR, the market demand mi-
nus the fringe supply at each price.

For example, at a price of $35, market demand is 90 units, and the price-
taking fringe would supply 10 units. The dominant firm’s residual demand
at a price of $35 is thus 80 units, so the point A ($35, 80) is one point on the
residual demand curve. At prices less than $25, fringe firms will not supply
output, and the dominant firm’s residual demand curve coincides with the
market demand curve. At $75, de dominant firm’s residual demand shrinks
to zero, and fringe firms satisfy the entire market demand. The dominant

196
 12.3
Oligopoly with differentiated Products

firm maximizes profits by equating its marginal revenue MRR to its marginal
cost MC. Its profit-maximizing quantity is thus 50 units. At this price, the
fringe supplies 25 units.

Dominant Firm Market Figure 12-2

P
SF

75
Residual demand,
DR, is kinked line
50
A
35 ●
25 MC

MRR DM

0 10 25 50 75 80 90 Q

12.3 Oligopoly with differentiated Products

There are two types of product differentiation:

 Two products, 1 and 2, are vertically differentiated if, at the same price,
all consumers prefer to purchase 1 rather than 2. In other words, vertical
differentiation refers to a pure quality differences between goods 1 and 2.

 Two products, 1 and 2, are horizontal differentiated if, at the same price,
some customers prefer to purchase 1 rather than 2 at an equal price while
others prefer to purchase 2 rather than 1.

If a market contains products that are horizontally differentiated, then these

products can sell for different prices. Because these products are still (imper-

197
 12
Market Structure and Competition

fect) substitutes for each other, the demand for each product depends on the
prices of all the products in the market. For example, if products 1 and 2 are
horizontally differentiated, then the demand for product 1 decreases when
its price increases but increases as the price of product 2 increases. Hence, an
example of demands for products 1 and 2 might be the following:

Q1 = 100 – P1 + 0,5P2

Q2 = 100 – P2 + 0,5P1

In other words, as the price of 1 rises, less of 1 and more of 2 is consumed

while as the price of 2 rises, less of 2 and more of 1 is consumed. By treating
the two equations above as a system of equations in two unknowns, P 1 and
P2, and solving it, we can obtain the corresponding inverse demands:

P1 = 200 – (4/3)Q1 – (2/3)Q2

P1 = 200 – (2/3)Q1 – (4/3)Q2

Case Study 12-2 Oligopoly with Differentiated Products

Let the demands for two products, 1 and 2, be written:

q1 = 100 – P1 + 0,5P2

q2 = 100 – P2 + 0,5P1

Product 1 is produced by firm 1 at a marginal and average cost of 10. Prod-

uct 2 is produced by firm 2 at the same marginal and average cost. Each firm
wishes to set its price so as to maximize its own profits.

Problem

a. Compute the residual demand of firm 1. Illustrate the residual demand

on a graph with the price of good 1 on the vertical axis and the quantity
of good 1 on the horizontal axis. What is the vertical intercept of this re-
sidual demand? What happens to residual demand of good 1 as the
price of good 2 increases?

b. Compute the residual demand for firm 2. Compute the marginal reve-
nue curve of firm 1 and 2 based on residual demand.

c. Calculate and illustrate the reaction functions of the two firms. Using
your answer to part (a), explain the slopes of the reaction functions.

d. Calculate the Bertrand equilibrium prices, outputs, and profits for this
market.

198
 12.3
Oligopoly with differentiated Products

Solution

a. The residual demand of firm 1 is firm 1’s demand taking as given firm
2’s price: P1 = (100 + 0,5P2) – q1. The vertical intercept is the term in
brackets. When the price of good 2 rises, the demand for good 1 shifts
up. This is what one would expect if the goods are demand substitutes.

Price of good 1

100 + 0,5P2

Residual demand

0
Quantity of good 1

b. Similarly, the residual demand for good 2 is P2 = (100 + 0,5P1) – q2. The
marginal revenue curves for firms 1 and 2 are MRR1 = 100 + 0,5P2 – 2q1
and MRR2 = 100 + 0,5P2 – 2q2. These two marginal revenues are symmet-
ric (in the sense that they are the same functions with the subscripts re-
versed). This is not surprising since the demands for the two products
are symmetric as well. The marginal revenue curves also shift up when
the competitor’s price rises, reflecting the increase in demand that oc-
curs when a competitor raises its price.

c. Setting marginal revenue equal to marginal cost for firm 1, we have 100
+ 0,5P2 – 2q1 = 10. Hence, 45 + 0,25P2 = q1. Using our equation for de-
mand, then, we have P1 = (100 + 0,5P2) – (45 + 0,25P2) = 55 + 0,25P2. Hence
the reaction function of firm 1 is P1* = 55 + 0,25P2. Similarly, the reaction
function of firm 2 is P2* = 55 + 0,25P1. These are upward sloping in the
space of prices. For example, the best response of firm 1 to a price rise
by firm 2 is to raise its price as well. To explain this, recall from parts (a)
and (b) that the residual demand (and marginal revenue) of firm 1 shift-
ed out when the price of firm 2 rose. When P2 rises, firm 1 faces an in-
creased incentive to raise its own price because consumers are less in-
clined to buy firm 2’s product at the higher price. Hence, the gain on
firm 1’s inframarginal units when it increases price is balanced against a
smaller loss in revenue because fewer consumers defect to product 2.

199
12
Market Structure and Competition

This means that firm 1’s best response to a price increase by firm 2 is a
price increase as well.

Price of firm 2
Reaction function
of firm 1

Reaction function
of firm 2

●
Bertrand Equilibrium

55

0 55 Price of firm 1

d. Substituting form firm 2’s reaction function into firm 1’s reaction func-
tion, we obtain P1 = 55 + 0,25(55 + 0,25P1) or P1* = 73,333 = P2*. Using de-
mand, we obtain q1* = 100 – 73,333 + 0,5(73,333) = 36,667 = q2* and profits
of both firms are total revenue minus total cost or 73,333(36,667) –
10(36,667) = 2322,22.

12.4 Monopolistic Competition

An market that is monopolistically competitive shares two of the assump-
tions we make for perfectly competitive firms.

 First, as in perfect competition, there are many buyers and sellers in the
market. However, monopolistically competitive firms do not take the
price as given as do perfectly competitive firms.

 Second, there is ease of entry and exit by competing firms into and out of
this market.

 The third assumption, firms produce horizontally differentiated prod-

ucts. Each firm produces a product that is slightly different from all other
firms. Hence, each firm faces a separate demand curve for its product,
since its product is in a slightly different market than the other firms’

200
 12.4
Monopolistic Competition

products. This assumption is the basis for firms setting their own prices
for their products instead of taking the market price as given.

With these assumptions, monopolistic competition is characterized by a Product

large number of firms, none of which can control market price, but which differentiation
produce-differentiated products. There are no barriers to entry in monopo-
listic competition. Product differentiation (making one’s product appear
unique) is essential in monopolistic competition, and advertising can be an
important aid in making the firm’s or industry’s demand curve less elastic.
Supporters of advertising argue that the firm must make consumers aware
of their products and differentiate its product from those of its competitors;
opponents claim that advertising contains little information, distorts con-
sumer preferences, and is wasteful.

In the short-run, no new firms can establish themselves in the market (since
the quantity of capital, by the definition of the short-run, is fixed). In Figure
12-2, DS is the short-run demand curve an individual firm faces in a market
with monopolistic competition, and MRS is the corresponding marginal
revenue. Similar to a monopoly, the firm maximizes its profit by chossing the
quantity, q1*, that makes MC = MRS. Since the average cost, AC, is below th
price at that quantity, the firm makes a profit, q1*(p1* - AC), corresponding to
the grey rectangle in the figure.

Equilibrium in the Short-Run for Monopolistic Competition Figure 12-3

P
MC

p1*
AC

AC

MRS DS

0 q1* Q

In the long-run, there are no barriers to entry; new firms will establish them-
selves in the market. Thereby, the demand curve that the individual firm

201
12
Market Structure and Competition

faces changes so that at each price it is now possible to sell a smaller number
of goods. This means that in Figure 12-3, the demand curve, DL, and the
marginal revenue, MRL, have shifted inwards.

How far do they shift? They shift until there is no profit. Remember that, the
firms choose the quantity that maximizes profit at MR = MC. The demand
curve, DL, will consequently shift until the quantity where the firm maxim-
izes its profit, q2*, is such that the price the firm can take for the good, p 2*, is
exactly equal to the average cost, AC. At that point, the profit is q2*(p2 - AC)
= 0.

Figure 12-4 Equilibrium in the Long-Run for Monopolistic Competition

P
MC

AC
p2*

MRL DL
0 q2* Q

12.5 Learning by doing

1. A firm faces a small number of competitors. This firm is competing in

(a) a monopoly.

(b) monopolistic competition.

(c) an oligopoly.

(d) perfect competition

202
 12.5
Learning by doing

2. A monopolistically competitive firm is like a perfectly competitive firm

insofar as

(a) both face perfectly elastic demand.

(b) both earn an economic profit in the long run.

(c) both have MR curves that lie below their demand curves.

(d) neither is protected by high barriers to entry

3. In the long run, a monopolistically competitive firm’s economic profits

are zero because of

(a) product differentiation.

(b) the lack of barriers to entry.

(c) excess capacity.

(d) the downward-sloping demand curve of each firm.

4. The demand faced by the dominant firm in the price-leadership model is

(a) established first, then the demand for other firms is determined based
on total market demand.

(b) equal to the sum of the demand curves of the smaller firms.

(c) determined by subtracting the supply of the smaller firms from the
market demand curve.

(d) determined by subtracting the demand of the smaller firms from the
market supply curve.

5. Each of the following is a characteristic of monopolistic competition

except

(a) many firms.

(b) product differentiation.

(c) no barriers to entry.

(d) mutual interdependence.

6. In monopolistic competition, firms achieve some market power

203
12
Market Structure and Competition

(a) by growing larger.

(b) by merging with other firms into a cartel.

(c) by establishing barriers to exit from the industry.

(d) through product differentiation.

7. Monopolistic competition differs from perfect competition because, un-

like the perfect competitor, the monopolistically competitive firm

(a) faces a perfectly inelastic demand curve.

(b) can earn positive economic profit in the short run and in the long run.

(c) cannot earn positive economic profit even in the short run.

(d) does not have the same marginal revenue at every output level.

8. Unlike a monopolist, a monopolistically competitive firm

(a) can earn positive economic profit in the short run but not in the long
run.

(b) has a downward-sloping marginal revenue curve.

(c) can never cover its minimum average cost in the long run.

(d) may sell to many buyers.

9. In monopolistic competition, when profits are being maximized, the

price

(a) equals marginal revenue.

(b) exceeds marginal cost.

(c) is less than marginal revenue.

(d) equals marginal cost.

10. A homogeneous products duopoly faces a market demand function

given by P = 300 – 3Q, where Q = Q1 + Q2. Both firms have a constant
marginal cost MC = 100.

204
 12.5
Learning by doing

(a) What is Firm 1’s profit-maximizing quantity, given that Firm 2 pro-
duces an output of 50 units per year? What is Firm 1’s profit-maximizing
quantity when Firm 2 produces 20 units per year?

(b) Derive the equations of each firm’s reaction curve and then graph
these curves.

(c) What is the Cournot equilibrium quantity per firm and price in this
market?

(d) What would the equilibrium price in this market be if it were perfect-
ly competitive?

(e) What would the equilibrium price in this market be if the two firms
colluded to set the monopoly price?

(f) What is the Bertrand equilibrium price in this market?

(g) What are the Cournot equilibrium quantities and industry price when
one firm has a marginal cost of 100 but the other firm has a marginal cost
of 90?

205
13
Game Theory and Strategic Behavior

13 Game Theory and Strategic

Behavior

Chapter objectives:
1. Identify elements of a game

2. Identify the prisoners’ dilemma and explain their application in the busi-
ness world.

3. Understand the concept of Nash equilibrium.

4. Describe simple simultaneous-move games using payoff matrix, and to

explain concepts of dominant, dominated, pure and mixed strategies,
and best responses.

5. Find dominant strategy equilibrium, pure and mixed strategy Nash

equilibrium

206
 13.1
The Concept of Nash Equilibrium

13.1 The Concept of Nash Equilibrium

Let us start Game theory with a question: How do the players act in chess?
They only decide on their moves, but they do so depending on how they
believe that the opponent will respond. Game theory is concerned with the
actions of decision makers who are conscious that their actions affect each
other. The best way to understand which situations can be modeled as
games and which cannot is to think about examples like the following:
OPEC members choosing their annual output; Two electric companies de-
cide whether to build a new power plant; …

The Prisoners’ Dilemma

Two prisoners, A and B, have been arrested and are kept in isolation. A pros-
ecutor suggests A the following:

 If you confess and B does not, you will be set free as a sign of our grati-
tude. B will then get 10 years in prison.

 If the both of you confess, you each get 2 years in prison.

 If B confesses and you do not, you get 10 years in prison while B is set
free.

 If none of you confesses, we will frame you for a petty crime and you
will each get 1 year in prison.

At the same time, B gets the same suggestion. The two players (as prisoners)
cannot communicate with each other and therefore must consider a solution
in solitude.

This game, like all games, has the following elements:

 The players. It could be individuals, firms, or countries. Often, there are

only two or three players. In our game, they are A and B.

 Actions. All possible actions the different players can choose to do, for
instance decide on quantity or price. For A: choose “Confess” or “Do not
confess”, and similarly for B: choose “Confess” or”Do not confess”.

 Information. What each player knows at different stages of the game. Both
A and B know that the other has received the same offer, but they do not
know how the other chooses.

 Strategies. A strategy is a complete description of what a player will

choose at each possible situation that could arise in a game. Both A and B

207
 13
Game Theory and Strategic Behavior

can only choose one of two different actions. Possible strategies for A are
then “choose confess” or “choose not to confess”, and similarly for B.

 Payoffs. The utility a player gets, given a certain outcome of the game. In
our example, we need to know the two players’ preferences. For simplici-
ty, we assume that they have the same preferences and that they are as
follows: 10 years in prison (-10), 2 years in prison (-2), 1 year in prison (-
1) and freedom (0).

We can summarize the decisions that the firms must take and the payoffs to
each decision in the following payoff matrix:

Case Study 13-1 Payoff Matrix for the Prisoners’ Dilemma

Player A

“Confess” “Do not confess”

“Confess” -2, -2 0, -10

Player B
“Do not confess” -10, 0 -1, -1

Nash equilibrium Now, we wish to know how players solve this game. Game theorists use the
concept of equilibrium to find out results of game. One commonly used type
of equilibrium is the Nash equilibrium9. At a Nash equilibria, each player
chooses a strategy that gives the highest payoff to that player, given the
strategy chosen by each other player in the game. In other words, each play-
er does the best he possibly can without taking into consideration the conse-
quences for the other player(s).

It is often easy to find the Nash equilibrium for a game in matrix form. Look
at the game in Table 13.1 again. Let us first look at the game from the per-
spective of player A. He does not know how player B will choose, but he
does know that player B will choose either “Confess” or “Do not confess”.
Say that player B would choose “Do not confess”. Then, obviously, the best
thing player A can do is to choose “Confess”, since he will then get a utility
of 0 (freedom) instead of -1 (1 year in prison). Now, say that player B chooses
“Confess“ instead. Then the best thing player A can do is still to choose

9 John Nash, “Equilibrium Points in w-Person Games”, Proceedings of the National

Academy of Sciences 36 (1950): 48-49.

208
 13.1
The Concept of Nash Equilibrium

“Confess”, since he will then get a utility of -2 (2 years in prison) instead of

-10 (10 years in prison).

Consequently, player A has a strategy that is the best one, independently of Dominant
what player B chooses. Such a strategy is called a dominant strategy. A dom- strategy.
inant strategy is a strategy that is better than any other strategy a player
might choose, no matter what strategy the other player follows When a
player has a dominant strategy, that strategy will be the player’s Nash equi-
librium strategy.

Player B’s problem is the same as player A’s, and hence it is a dominant strat-
egy for player B as well to choose “Confess”. As a result, they both choose
“Confess” and get 2 years in prison. This is so, even though it is possible for
them both to get away with 1 year in prison (if they both choose “Do not
confess”). This is the dilemma.

This game illustrates the important point that the Nash equilibrium does not
necessarily maximize the collective interest of the players. For both player A and B
it is individually rational to confess, but acting that way they achieve an
outcome worse that what is “collectively” possible. If they had been able to
cooperate, they would both have been able to reach a higher utility level.
Games where the players choose a set of payoffs that do not maximize the
aggregate payoffs of the players are called Prisoners’ Dilemma.

The Prisoner's Dilemma in Business

Business life is rife with prisoner's dilemmas including the employer-

employee relationship and that between vendor and customer.

If an employer offers a job and an employee takes it, they are both offering
an immediate sacrifice. The employer trusts the employee with the key to the
door, with money, with confidential information, with clients, but the em-
ployer should give up the opportunity to hire someone else. The employee
also trusts the employer not to leave him in a position where he will have a
hard time getting another job, but the employee should give up the oppor-
tunity to work somewhere else.

As a new employee, the employee may not be productive for a period of a

month or some months while you learn your trade. If the employee leaves
after the employer trains him but before he begins producing for the em-
ployer, he has a set of tools with which to go out and make money elsewhere
and the employer has received the sucker's payoff.

Similarly, a company engages in a prisoner's dilemma with its vendors and

its clients. We may see one large client repeatedly prune its vendor's list,
throwing out smaller vendors. This defection is possible because, given the

209
 13
Game Theory and Strategic Behavior

relative size and power of the parties, the future had no shadow for the
client; the defection of smaller vendors could not harm it, their cooperation
had no meaning for it, and the future had no shadow.

Dominant and Dominated Strategy

Dominated The opposite of a dominant strategy is a dominated strategy. This is a strat-

strategy. egy that gives a lower payoff than an alternative strategy, no matter what the
other player does. Let us see this example: Two firms, A and B, have two
advertising strategies. Their advertising choices might be in newspapers,
Economics or Investment. The Table 13-2 shows the payoffs from each of
these strategies. In this table, with just two strategies for each player, if one
strategy is dominant, then, the other must be dominated.

Case Study 13-2 Payoff Matrix for Firms A and B

Firm B

The Economics The Investment

The Economics 2 , 1 2 , 2
Firm A
The Investment 3 , 4 3 , 3

First, firm A supposes that firm B will choose to advertise in The Economics.
Then, the best solution (circles) firm A can choose is to advertise in The In-
vestment (3 > 2). Say that firm B will choose to advertise in The Investment.
Then, firm A will still choose to advertise in The Investment (3 > 2).

Now, similarly, firm B suppose that firm A would choose to advertise in The
Economics. Then, the best solution (squares) firm B can choose is to adver-
tise in The Investment (2 > 1). Say that firm A would to advertise in The
Investment. Then, firm B will choose to advertise in The Economics (4 > 3).

Hence, the Nash equilibrium for this game is that both firm A and firm B
will advertise in The Economics. The Nash equilibrium is (3, 4).

Here, we can see that firm A has a dominant strategy, because firm 1’s adver-
tising strategy in The Investment is better than any other strategy, no matter
which strategy firm B follows. However, firm B does not have a dominant

210
 13.1
The Concept of Nash Equilibrium

strategy: The best response of firm B changes depending on which strategy

firm A follows. In another hand, advertising strategy in The Economics is a
dominated strategy for firm A since advertising strategy in The Investment
gives a higher payoff to firm A, no matter what firm B does.

Games with More Than One Nash Equilibrium

Game theorists usually use game “Chicken” to illustrate for a game with
more than one Nash equilibrium. The name "Chicken" has its origins in a
game in which two drivers drive towards each other on a collision course:
one must swerve, or both may die in the crash, but if one driver swerves and
the other does not, the one who swerved will be called a "chicken," meaning
a coward. A formal version of the game of Chicken with payoffs is showed in
Table 13-3 which outcomes are represented in words. Where each player
would prefer to win over tying, prefer to tie over losing, and prefer to lose
over crashing.

The Game of Chicken Case Study 13-3

Player B

Swerve Straight

Swerve Tie , Tie Lose , Win

Player A
Straight Win , Lose Crash , Crash

Let us find Player A’s best response to two possible strategies of Player B:

 If Player B chooses “Swerve”, player A’s best response is “Straight”.

 If Player B chooses “Straight”, player A’s best response is “Swerve”.
Now, let us find Player B’s best response to two possible strategies of Player
A:

 If Player A chooses “Swerve”, player B’s best response is “Straight”.

 If Player A chooses “Straight”, player B’s best response is “Swerve”.

211
 13
Game Theory and Strategic Behavior

There are two Nash equilibriums in this game. The first is Lose – Win (Player
A chooses “Swerve” and Player B chooses “Straight”). The second is Win –
Lose (Player A chooses “Straight” and Player B chooses “Swerve”).

Because the loss of swerving is so trivial compared to the crash that occurs if
nobody swerves, the reasonable strategy would seem to be to swerve before
a crash is likely. If one believes one's opponent to be reasonable, one may
well decide not to swerve. In the belief that he is reasonable, he will decide
to swerve, leaving the other player the winner. Chicken game arises in eco-
nomics when two firms compete in a market that can profitably support
only one firm. The Nash equilibrium in the Chicken game tells us that one
firm will eventually exit the market and one firm will survive.

Mixed Strategies

A pure strategy provides a complete definition of how a player will play a

game. In particular, it determines the move a player will make for any situa-
tion they could face. A player's strategy set is the set of pure strategies avail-
able to that player. A mixed strategy is an assignment of a probability to
each pure strategy. This allows for a player to randomly select a pure strate-
gy.

For example, suppose that we play a game called Matching Pennies. It is a 2-

player simultaneous move game. Both players take a penny and, after a
three-count, place it on the table. If the 2 pennies match (both heads or both
tails) then player 1 wins $1 from player 2. If the 2 pennies do not match then
player 2 wins $1 from player 1. The normal form of this game is:

Case Study 13-4 The Game of Matching Pennies

Player 2

Heads Tails

Heads 1, -1 -1, 1
Player 1
Tails -1, 1 1, -1

212
 13.2
Repeated Games

There is no (pure strategy) Nash equilibria in this game. If we play this

game, we should be “unpredictable.” That is, we should randomize (or mix)
between strategies so that we do not get exploited.

Since this game is completely symmetric it is easy to see that at mixed strat-
egy Nash equilibria both players will choose Heads with 50% chance and
Tails with 50% chance. In this case the expected payoff to both players is
0.5×1 + 0.5×(-1) = 0 and neither can do better by deviating to another strategy
(regardless it is a mixed strategy or not). In general there is no guarantee that
mixing will be 50-50 at equilibrium.

A five-step approach to identifying the Nash equilibria of a simultaneous-

move game involving two players

1. If both players have a dominant strategy, these constitute their Nash

equilibrium strategies (and their dominant strategy equilibrium, as well).

2. If one player has a dominant strategy, then this constitutes the Nash
equilibrium strategy for this player. We then find the rival player’s best
response to the dominant strategy in order to find the rival’s Nash equi-
librium strategy.

3. If neither player has a dominant strategy, then we may search for and
eliminate any dominated strategies in order to simplify the game.

4. By using the “circles and squares” technique, we can indentify all the
Nash equilibria in pure strategies for the simplified game.

5. After finding the equilibria in pure strategies, we can look for the mixed
strategy equilibria by finding the probabilities to apply to each set of
payoffs so that the pure strategies yield the same expected payoffs for the
players.

13.2 Repeated Games

The Prisoners’ dilemma is a one-shot game, each agent must only decide on One-shot game/
an action once. With this game, we can find out a Nash equilibria for players
Repeated game
that is (“Confess”,”Confess”). However, many real life strategic situations
involve repeated interaction among economic agents. Such situation can be
analyzed in repeated game; a given one-shot game is played several times
by the same players.

213
 13
Game Theory and Strategic Behavior

In a repeated game, the strategy of always “Confess” is not a dominant

strategy, as it is in the one-shot game, because it is not the best response to
Grim strategy
various suboptimal strategies. We can find a simple perfect equilibrium for
the infinitely repeated Prisoner’s Dilemma in which both players cooperate -
a game in which both players adopt the Grim Strategy:

 Start by choosing “Do not confess”.

 Continue to choose “Do not confess” unless some player has chosen
“Confess”, in which case choose “Confess” forever.

Notice that the Grim Strategy says that even if a player is the first to deviate
and choose “Confess”, he continues to choose “Confess” thereafter.

If the player A uses the Grim Strategy, the Grim Strategy is weakly the player
B’s best response. In other words, if the player B cooperates, he will continue
to receive the high payoff (“Do not confess”, “Do not confess”) forever. If the
player B confesses, he will receive the higher payoff (“Confess”, “Do not
confess”) once, but the best he can hope for thereafter is the payoff (“Con-
fess”, “Confess”).

Our analysis of the repeated Prisoners’ dilemma game teaches an important

lesson: In competitive settings you must anticipate the reactions of your
competitors. For example, you are a business firm in a market, and you cut
price in order to increase your market share, you need to anticipate whether
your price cut will be detected, whether your competitor will respond by
matching the price, and if so, how long your competitor will take to match.
By ignoring the possibility of competitive responses, you run the risk of
overestimating the potential benefits that will accrue to you from various
forms of non-cooperative behavior. You also run the risk of plunging your
market into a costly price war that will erase any temporary gains you might
enjoy from having undercut the prices of your competitors.

13.3 Sequential-Move Games and Strategic

Moves
Game tree In a previous section, we examined the prisoner's dilemma and plotted it on
a matrix. That was an example of a simultaneous-move game. Games are
either simultaneous-move or sequential-move games. In simultaneous-move
games, both players make a move without knowledge of the other players'
moves. In sequential-move games, the players take turns moving, as in chess
or negotiations.

214
 13.3
Sequential-Move Games and Strategic Moves

For games in which both players move simultaneously, we must assume our
opponent is going to seek the best outcome possible. Therefore, we must
protect ourselves by also making the most advantageous move possible.
Would players choose different strategies in the prisoner's dilemma if it was
played sequentially instead of simultaneously?

While simultaneous-move games can be plotted on matrices, sequential-

move games can be plotted on game trees. The prisoner's dilemma mapped
out on a game tree would look like this:

Game Tree for the Prisoners’ Dilemma Figure 13-1

(-2,-2)

Player B

(0,-10)
Player A

(-10,0)

Player B

(-1,-1)

The tree is read from left to right. First, Player A chooses between the top
branch of the tree (“Confess”) and the bottom branch (“Do not confess”).
Next, Player B chooses between “Confess” and”Do not confess”, each repre-
sented by another branch on the tree. After two players have made their
decisions, each receives the payoff listed at the right-hand side of the tree.
The payoff to player A is listed first and the payoff to player B is listed sec-
ond in each set of brackets. Player A gets to choose first.

We're able to look at the game in this form and choose the best strategy for Backward
the player making the final move in the game, which is Player B. This allows induction
us to work backwards to determine what move Player A should make,
knowing how Player B will rationally behave in either scenario. This process
of looking at the game from right-to-left (or “end-to-beginning” in another
game tree graph), starting with the last set of branches on the tree, is called

215
13
Game Theory and Strategic Behavior

backward induction. With this method, you start from the terminal node
(the payoff) and work your way back to the beginning, eliminating subopti-
mal choices as you go, and thereby identifying an individual player's most
likely path

We see that Player B's best strategy is to choose “Confess”, regardless of

Player A's opening move. By moving our analysis further to left side of the
tree, we know that Player A must choose “Confess”.

13.4 Learning by doing

1. Game theory is about:

(a) Market structure in the entertainment industry

(b) Strategic interaction among players in the same market

(c) Abstractions with no application in the real world.

(d) An unnecessary exercise since all industries reach equilibrium solely

the supply and demand analysis

2. The study of game theory is not as applicable to firms that are perfect
competitors because:

(a) They cannot afford to hire strategists.

(b) The firms in perfect competition are too interdependent

(c) Perfect competitors can sell all that they produce at the market price

(d) They have to be concerned that the strategy that they would opt for
would generate a reaction by their competitors

3. Which of the following best describes a strategy?

(a) a plan to profit maximize and ignore competitors

(b) a plan to drop out of an industry

(c) a plan to dominate an industry

(d) A plan that describes the actions a firm will take given the actions of
other interdependent firms

216
 13.4
Learning by doing

4. A dominant strategy is one which:

(a) is best for a player no matter what strategy the other player chooses.

(b) is optimal given the other player's strategy, but may not be optimal
should the other player switch strategies.

(c) will lead to a Pareto-optimal outcome.

(d) in a sequential game, is optimal when the other player's best response
is taken into account

Questions 5–6 are based on the following information.

Given the payoff matrix for Buy Corporation and Able Corporation.

Buy Corporation

Increase Don’t Increase

Advertising Advertising

Increase 20,35 30,20

Advertising
Able
Corporation Don’t Increase 25,40 40,50
Advertising

5. Determine what is the highest profit available and for what corporation
it is available for, and determine the decisions by both companies that
would enable this highest profit.

(a) Profit is $40 million for Able Corporation, Able Corp.’s decision is "no
increase" and Buy Corporations decision is "increase"

(b) Profit is $40 million for Buy Corporation, Able Corp.’s decisions is "no
increase" and Buy Corp's decision is "increase"

(c) Profit is $50 million for Buy Corporation, Able Corp.’s decision is “no
increase” and Buy Corp.'s decision is "no increase"

(d) Profit is $50 million for Buy Corporation, Able Corp.'s decision is “no
increase” and Buy Corp.'s decision is "increase"

217
13
Game Theory and Strategic Behavior

6. Which of the following answers is correct regarding the payoff matrix for
Able Corp. and Buy Corp?

(a) Buy Corporation and Able Corporation both have dominant strate-
gies.

(b) Able Corporation has a dominant strategy; Buy Corporation does not.

(c) Buy Corporation has a dominant strategy; Able Corporation does not.

(d) Neither corporation has a dominant strategy

Questions 7–8 are based on the following information.

Given the payoff matrix for Deluxe Retail Company and Cordon Retail
Company.

Deluxe Retail Company

Lower prices Don’t change

prices

Lower prices 25,15 35,10

Cordon Retail
Company Don’t change 20,25 30,20
prices

7. Which of the following statements is accurate:

(a) Cordon has a dominant strategy but Deluxe does not.

(b) Deluxe has a dominant strategy but Cordon does not.

(c) Neither company has a dominant strategy

(d) Both companies have a dominant strategy

8. Given that we are going to use game strategy with the information given
to us, ceteris paribus, we can conclude that equilibrium will be reached
by:

(a) Cordon will lower prices and Deluxe will not change prices

(b) Cordon will lower prices and Deluxe will lower prices

(c) Cordon will not change prices and Deluxe will not change prices

218
 13.4
Learning by doing

(d) Cordon will not change prices and Deluxe will lower prices

9. Two individuals, A and B, who like each other, have arranged a date.
They will meet either at a pop concert or at a techno party. However, they
have not decided on which of the two. A prefers techno whereas B pre-
fers pop. However, they both prefer being at the same event as the other
to going alone to the pop concert or to the techno party.

Suppose they cannot communicate, and therefore must decide separate-

ly. Then the game can be represented as in the below table. The worst
outcome is that they end up alone at their least preferred event. The best
outcome for A is that they both go to the techno party, but that is only the
second best outcome for B. The best outcome for B (and the second best
for A) is that they both go to the pop concert.

Techno Pop

Techno 10, 9 2, 2
A
Pop 0, 0 9, 10

(a) What is a Nash equilibria? Give a definition in words

(b) Find all Nash equilibria in the game.

(c) To avoid this type of problems in the future, A and B decide on the
following rule: If a game such as the one in the above table arises, then
we go to the one that A prefers. Does that rule constitute an improve-
ment for B?

10. American and United airlines are rivals on the route from San Francisco
to Chicago. The market game described below in strategic form gives
the profits of each depending on the actions of both, with American's
profit listed first:

219
13
Game Theory and Strategic Behavior

UNITED

Full Announced Unannounced

Fare Discount Discount

AMERICAN Full Fare 12,12 12,12 5,15

Announced 15,5 11,9 10,10

Discount
Unannounced 14,8 14,8 9,11
Discount

(a) What is American’s strongly dominated strategy?

(b) What is United’s dominant strategy?

(c) Find Nash equilibria in the game.

220
 13.4
Learning by doing

14 Externalities and Public Goods

Chapter objectives:
1. Define externality and give examples of negative and positive externali-
ties.

2. Derive social marginal cost curve and social marginal benefit curve.

3. Explain the significance of property rights.

4. Identify and compare the social equilibrium with the private market
equilibrium.

5. Identify the two critical characteristics of public goods. Explain why a

free market would underproduce public goods and identify the two
“problems” associated with the production of public goods.

221
 14
Externalities and Public Goods

14.1 Externalities
Externalities occur when the actions or decisions of one agent impose a cost,
or bestow a benefit, on second or third parties. There is no incentive to figure
these costs (or benefits) into the decision to act. An externality is a situation
in which the consumption or the production of goods has positive or nega-
tive effects on other people’s utility where these effects are not reflected in
the price. One consumer’s actions may benefit or harm producers or other
consumers. Similarly, one producer’s actions may benefit or harm consumers
or other producers.

It is common to distinguish between positive and negative externalities:

Posi-  Externalities are positive, if they help other producers or consumers. For
tive/Negative example, a beekeeper keeps the bees for their honey. A side effect or ex-
externalities ternality associated with his activity is the pollination of surrounding
crops by the bees.

 Externalities are negative, if they impose costs on, or reduce benefits for,
other producers or consumers. For example, the harvesting by one fish-
ing company in the ocean depletes the stock of available fish for the other
companies and overfishing may be the result.

Negative externalities

With a negative externality a wedge is driven between marginal cost, as

faced by producers or consumers, and price. Producers look only at margin-
al private cost. A broader view would encompass all extra costs of produc-
tion—marginal social costs. The difference between the two is marginal
damage cost. When we fail to consider these additional costs (or benefits) to
society, inefficient outcomes emerge. When external costs are present, more
production occurs than society would prefer. When external benefits are
present, the market, on its own, will produce less than society would like.
Pollution is the “classic” example of an external cost.

Let us study the classical example of a negative externality: A firm produces

a good, but in doing so they also pollute the environment. First, we need to
define a few concepts:

 The marginal external cost, MEC. The change in the cost of the marginal
effect, when production is increased by one unit. The marginal external
cost curve rises because the incremental damage to the environment in-
creases as more pollution occurs.

 The marginal private cost, MPC. The marginal private cost measures the
firm’s marginal cost of producing one unit.

222
 14.1
Externalities

 Social cost. The sum of the cost of producing the good and the cost of the
external effect.

 Marginal social cost, MSC. The sum of the firm’s marginal cost and the
marginal external cost, MSC = MPC + MEC.

In the perfectly competitive market, MPC curve is also the supply curve of
the goods. The market equilibrium will be at the point where the market
supply curve intersects demand, at ec. The social optimum occurs at the
point where the demand curve lies above the marginal social cost, at e s.
Hence, the market equilibrium results in overproduction of amount qc – qs. It
results in a deadweight loss because, for all units between qs and qc, the
marginal social cost exceeds the willingness to pay (the demand curve).

The Effect of a Negative Externality Figure 14-1

p MSC = MPC + MEC

MPC

es
ps Deadweight loss
pc ec

D
MEC
q, (units of
good = units
0 qs qc of pollutant)

Regulations of Markets with Externalities

In order to eliminate this inefficiency, the government has several options.

 Emissions standards. Emissions standards are requirements that set specif-

ic limits to the amount of pollutants that can be released into the envi-
ronment. Many emissions standards focus on regulating pollutants re-
leased by automobiles (motor cars) and other powered vehicles but they
can also regulate emissions from industry, power plants, small equip-
ment such as lawn mowers and diesel generators.

223
 14
Externalities and Public Goods

Case Study 14-1 Emission Standard Level of U.S. Environmental Protection Agency

Emission Standard Level (listed from least clean to cleanest)

T1 Tier 1 The least stringent emission standards

TLEV Transitional Low More stringent than Tier 1 standards for

Emission Vehicle hydrocarbons (HC)

LEV Low Emission More stringent than TLEV standards for

Vehicle both hydrocarbons (HC) and nitrogen
oxides (NOx)

ULEV Ultra Low Emis- More stringent than LEV standards for
sion Vehicle HC

SULEV Super Ultra Low Even more stringent standards than

Emission Vehicle ULEV for both HC and NOx

ZEV Zero Emission The strictest emission standard, permit-

Vehicle ting no emissions

 Emissions fee. An emissions fee is a tax imposed on pollution that is re-

leased into the environment. This tax may be a fixed amount imposed on
firms or a tax, t(Q), varies with the quantity produced may be applied. If
the government imposes on firms, then it has the effect of shifting up the
MPC curve, the optimal quantity decreases and the pollution also de-
creases.

Common Property

A common property is a resource that anyone can access and no one can be
excluded from using common property. This is also a case of negative exter-
nality. For example, a lake in a public park can be used for fishing. Everyone
can fish at the lake. Then, if the number of fishermen grows, they deplete the
number of fish in the lake so that it is harder to catch a fish. In other words,
as the number of fish in the lake decreases, each fisherman must spend more
time to hook a fish. Hence, when fishing exceeds some minimum level, a
negative “congestion” externality sets in.

This situation is illustrated in Figure 14-2. The “congestion” externality is

represented as a MEC curve that is zero for low usage levels, but non-zero

224
 14.1
Externalities

after some minimum, q*, is exceeded. Hence, the MSC curve (“the marginal
cost of using the facility”) exceeds the MPC curve for usage exceeding q*.

Common Property Figure 14-2

p, price of lake
use (time)
MSC = MPC + MEC

MPC

ps es Deadweight loss
pc ec
MEC
●MPC

Marginal
●MEC benefit of use

0 q* qs qc q, (Lake usage)

Suppose that we measure the marginal benefit from lake usage as a decreas-
ing curve. In other words, an additional unit of “use” of the lake has less
benefit than the first unit of usage. This marginal benefit curve reflects the
demand, or the willingness to pay, for fishing. In the absence of any inter-
vention, the private market will result in an equilibrium level of usage where
the private marginal cost of using the lake equals the private marginal bene-
fit from using the facility. This occurs at ec. This results in a deadweight loss,
however, as the social marginal cost exceeds the marginal cost so that, for all
usage levels between qs and qc, the marginal social cost exceeds the marginal
benefit of using the lake.

A usage fee set at the vertical distance between MSC and MPC at the opti-
mum, qs, would result in the socially optimal use of the lake. Alternatively,
the government could restrict access to the common property (on, for exam-
ple, a first-come-first serve basis) in order to limit usage to a total of q s. This
is sometimes imposed by issuing licenses (for fishing, in our example).

225
14
Externalities and Public Goods

Positive Externality

When one person becomes vaccinated against a communicable disease, he

creates a positive externality because he reduces both his own chance of
catching the disease and the risk that any other person catches the disease as
well. When an individual consumer decides whether to obtain a new vac-
cine, for example, he takes into account his own benefit but not the benefit
he imparts to others. Hence, the marginal social benefit (MSB) exceeds the
marginal private benefit (MPB) of consuming the vaccine by the marginal
externality benefit of MEB.

We could depict this as shown in Figure 14-3 as the vertical distance between
the MSB and the MPB curve, if the marginal benefit of consuming additional
units falls for both MSB and MPB. The supply of vaccines can be represented
by the increasing curve, reflecting the marginal cost of provision of the vac-
cine. Hence, the social optimum occurs at the intersection of this marginal
cost curve with the MSB curve, es, and the market optimum occurs at the
intersection of the marginal cost curve with the MPB curve, ec. In other
words, the market equilibrium results in excessively low use of the vaccine
so that a deadweight loss occurs from underproduction.

Figure 14-3 Positive Externality

p, price of
vaccines

MC (supply curve)
Deadweight loss

ps es
pc ec

MEB
MPB MSB q, vaccines
0 qc qs given

226
 14.1
Externalities

A subsidy to the production (or the consumption) of vaccines could elimi-

nate this deadweight loss by shifting down the marginal cost curve so that it
intersects the MPB curve precisely at qs. This is shown in Figure 14-4, as
equilibrium es’ for a fixed subsidy of $S.

Optimal Subsidy with a Positive Externality Figure 14-4

p, price of
vaccines

MC (supply curve)
MC -S

ps es
pc ec S
es’

MEB
MPB MSB q, vaccines
0 qc qs given

Property Rights and the Coase Theorem

A property right is the exclusive control over the use of an asset or resource, Property rights
without interference by others.
The Coase Theorem 10states that when property rights are clear and en- The Coase Theo-
forceable, when all economic agents have full information, and when trans- rem
action costs are low, there is no need for government intervention to correct
externalities, because the economic agents can bargain to achieve a optimal
allocation of resources.

For example, there are two factories, one upstream and one downstream.
The upstream factory produces waste that is discharges into a river. The

10 A theory is named after Ronald Harry Coase (born 29 December 1910), who re-
ceived the Nobel Prize in Economics in 1991.

227
 14
Externalities and Public Goods

downstream factory uses clean water as an input, so the waste imposes ex-
ternal costs on the downstream factory.

The problem is the lack of property rights in the river and in clean water. If
either factory held these rights the two factories could bargain over the use
of water. If the upstream factory held the rights, the downstream factory
would offer to pay to reduce waste disposal up to the point where the mar-
ginal cost was equal to the marginal benefit. This would produce an efficient
outcome. If the downstream factory held the rights, the upstream factory
would offer to pay for the right to dispose of waste up to the point where the
marginal cost was equal to the marginal benefit. Again the outcome would
be efficient.

The problem is there are other potential difficulties with bargaining. If two
factories do not know the costs and benefits of reducing the externality, or if
they have different perceptions about these costs and benefits, then bargain-
ing may not lead to an efficient outcome. Finally, both factories must be
willing to enter into agreements that are mutually beneficial. If one of facto-
ries simply refuses to bargain, or refuses to give the other factory an ac-
ceptable compensation, it may not be possible to achieve an efficient re-
source allocation.

Case Study 14-2 The Coase Theorem

Problem

Farm A raises cattle, and the cattle occasionally stray onto the land of a
neighboring farm, Farm B, which raises crops. Farm A’s cattle impose a nega-
tive externality by damaging the crops on Farm B.

(a) Suppose it is costless for the parties to bargain. Verify the Coase Theorem
when the cost of the fence is $2,000 and the cost of the damage is $1,000.

(b) Verify the Coase Theorem if the fence costs $2,000 and the damage cost is
$4,000.

Solution

(a) Suppose the property rights are assigned to A. Owner B can either pay
for a fence costing $2,000, or live with the damage of $1,000. B therefore does
not find it worthwhile to pay for a fence, and the cattle will roam. Owner B
receives no compensation for the damage of $1,000.

Suppose the property rights are assigned to B. Owner A can either spend
$2,000 to build a fence to prevent damage or build no fence and pay $1,000
to owner B to compensate for damage. Owner A does not find it worthwhile

228
 14.2
Public Goods

to pay for a fence, and the cattle will roam. The damage to B is $1,000, but A
will compensate B.

With either property rights assignment, the outcome is the same: the cattle
will roam. It is economically efficient to build no fence because the fence
costs more than the damage from roaming cattle.

(b) Suppose the property rights are assigned to A. Owner B now finds it
worthwhile to pay for a fence, and the cattle will not roam.

Suppose the property rights are assigned to B. Owner A now finds it worth-
while to pay for a fence, and the cattle will not roam.

Once again, with either assignment of the property right, the outcome is the
same: The cattle will not roam. It is economically efficient to pay for the
fence because the fence costs less than the damage that would have occurred
from roaming cattle.

14.2 Public Goods

Public goods are goods whose benefits are received collectively by all mem-
bers of society (non-rival) and/or whose benefits can be denied to no one
(non-excludable), even if they refuse to pay. Because of these characteristics,
public goods will either be underproduced or not produced at all if the pri-
vate sector is made responsible for their provision.

A public good is a good that fulfills both of the following two criteria: Characteristics of
public goods
 Non-rival. One individual’s consumption of the good does not affect any
other individual’s consumption of the same unit of the good. Examples
include lighthouses, television signal, parks, ...

 Non-exclusive. It is not possible to exclude anyone from consuming the

good. The above examples are usually non-exclusive.

A private good is, instead, a good that does not fulfill any of the two criteria,
i.e. one that is both rival and exclusive. Most goods are private goods.

The Aggregate Willingness to Pay

To find the market’s demand curve for a public good we must know each
individual’s demand for it.

229
14
Externalities and Public Goods

Suppose we have two individuals, A and B, and that they each have a mar-
ginal private benefit regarding, say, a park, corresponding to MBA and MBB
in Figure 14-2. Suppose we produce one unit of the good and that A value
that unit to 10, whereas B values it to 15. Had it been private good, only one
of them could have consumed it. However, since a public good is non-rival,
both A and B can consume it at the same time. Consequently, the aggregate
willingness to pay for this unit is 10 + 15 = 25. Similarly for the second, third,
and all following units, and we sum the demand curves in Figure 14-2 verti-
cally. Then, the marginal social benefit in this case, corresponds to the thick
demand curve MSB.

Figure 14-5 Public Goods

p
MC
MSB

MBB

MBA

0 qA q* q

We see that, the optimal quantity, q*, is at the point where the marginal so-
cial benefit equal to the marginal cost. As a comparison, we have also indi-
cated what would happen if only one of the individuals had decided on the
quantity. If A had done so, the quantity qA would have been produced. As
compared to the optimum, q*, we would have seen a much smaller quantity.

Free Riding

We derived the marginal social benefit for the public good by summing the
individuals’ marginal private benefit. The problem is that we usually do not
know their marginal private benefit, or their marginal willingness to pay. For
private goods, this is not a problem, since it is optimal for the consumers to
pay a price up to their willingness to pay. For instance, if the price of milk is

230
 14.3
Learning by doing

$10 and a consumer buys three liters of milk, his marginal willingness to pay
for the first unit is $10 and his marginal willingness to pay for additional
units is less than 10. We do not need to know it beforehand, as he will reveal
it by his behavior.

However, for public goods it is not optimal for consumers to reveal their Free rider
willingness to pay. If he will later have to pay an amount equal to the one he
states, it is often individually better for him to understate her willingness to
pay. If the good is still produced, he will make a sort of profit: He receives
more utility than he has paid for. He is then said to be a free rider.

The free rider problem makes it difficult for a private market to provide
public goods efficiently. It is generally easier to organize effective efforts to
collect voluntary funding when the number of people involved in paying for
a project is small because each person recognizes that his contribution is
important. However, when the number of consumers of a public good be-
comes large, it is more likely that many consumers will act as free riders.
Public intervention may be necessary to ensure the provision of a socially
beneficial public good. The government therefore often produces a public
good itself, or subsidizes the enterprises that produce the good.

14.3 Learning by doing

1. If one person's consumption of a good diminishes other people's use of
the good, the good is said to be:

(a) rival.

(b) a good produced by a natural monopoly.

(c) a common resource.

(d) excludable.

2. ____ is the sum of the marginal costs of producing a good and the cor-
rectly measured damage costs involved in the process of production.

(a) Marginal damage cost

(b) Marginal social cost

(c) Marginal private cost

(d) Marginal external cost

231
14
Externalities and Public Goods

3. Negative externalities arise when:

(a) Social benefits are greater than social costs

(b) Social costs exceed private costs

(c) Private costs exceed social costs

(d) Social costs are at a minimum

4. When marginal social cost exceeds the firm’s marginal private cost, the
industry’s supply curve is too far to the ____ and ____ is being produced.

(a) right; too much

(b) right; too little

(c) left; too much

(d) left; too little

5. The Coase theorem implies that:

(a) Externalities lead to inefficient outcomes

(b) Bargaining at zero transaction cost between the two sides to an exter-
nality can lead to an efficient outcome if both agree voluntarily in their
own self interest

(c) Where externalities exist, efficient outcomes can only be achieved by

government intervention

(d) Externalities will always cause market failure

6. Public goods are ____ in consumption and non-purchasers ____ be ex-

cluded from their benefits.

(a) rival; can

(b) rival; cannot

(c) non-rival; can

(d) non-rival; cannot

232
 14.3
Learning by doing

7. Pure public goods are characterized as being: Non-rival and non-

excludable. Which of the two characteristics leads to the free rider’s prob-
lem? Explain

8. Because public goods are non-excludable, individuals are usually unwill-

ing to pay for them. This characteristic is known as the
(a) drop-in-the-bucket problem.

(b) impossibility theorem.

(c) Coase Theorem.

(d) free-rider problem.

9. As the number of recipients of a public good increases, the number of

free riders will tend to

(a) increase because, as the size of the group increases, it is more difficult
to detect free riders.

(b) increase because, as the size of the group increases, individuals be-
come rivals for the benefits of the good.

(c) decrease because, as the size of the group increases, the per person
payment will decrease.

(d) decrease because, as the size of the group increases, it is easier to ex-
clude non-contributors.

10. A firm that produces honey form bees is located next to a firm that pro-
duces apples from an apple orchard. The orchard requires pollination of
the trees by bees in order to produce apples. Currently, the bees are al-
lowed to collect pollen from any source, including the apple trees, at no
fee. This leaves some of the trees un-pollinated, however, imposing a loss
on the apple grower. This could be corrected by increasing the number of
hives in the beekeeper’s operation. On the other hand, the honey firm
recognizes that the apple blossoms make their honey very tasty, giving
them an edge on their competition.

(a) What are the externalities present in this problem?

(b) Is there an argument, based on externalities, for merging the opera-

tions of these two separate firms?

(c) After the merger, is there an externality?

(d) Relate this discussion to the Coase Theorem.

233