Public Goods Theory: Advanced

Treatment
A. Damodaran
Lecture PPTs
 Pareto Concepts
• Pareto improvement (PI) is related to being Pareto
Efficient (PE) and being Pareto optimum (PO)
• PI=PO + PE
• A change in allocation is Pareto optimal if it leads to one
fellow being better off without leaving another worse off
• A change in allocation is Pareto efficient if it leads to one
fellow being better off without leaving another worse off
(Pareto optimal) and leaves no scope for waste
• A given allocation is Pareto Optimal and efficient if there
is no scope for Pareto improvement
• All the three do not however convey equality as a norm of
optimality
 An Example Question on Pareto
Concept
• Jay and Madhavi get to split Rs100. Suppose
that Jay gets x Rs, that Madhavi gets y Rs,
and that 100-x-y Rs is thrown away. For which
values of x and y is the allocation Pareto-
efficient?"
 Solution
• In the case of this question, giving Jay Re 1 and Madhavi Re 1
would not be a Pareto Efficient outcome, because we would have
Rs 98 left over. We could just as easily give each of them Rs2,
making them both better off without making anyone worse off
(PO).
• However Pareto efficient (PE) outcomes are when there are no
wastes
• If Jay and Madhavi's combined allocation is equal to Rs 100.
• But this could include situations where Jay gets Rs 100 and Madhavi
gets nothing. We cannot make Madhavi better off without taking
money away from Jay
• PI is when Jay and Madhavi's combined allocation is Rs 100 (or Rs
50 each or Rs 49 and Rs 1 each) as compared to the first situation
of 1 :1 (PO) and wastage is zero (PE)
 Criticism of Pareto
• The example above is at the heart of the
biggest criticism of the concept of Pareto
Efficiency - it says nothing about equality.
• The outcome where they each get Rs 50 and
the outcome where one party gets all the
money are Pareto Efficient.
Production Possibility Frontier Conveys
• Scarcity of resources
• Opportunity Costs (or marginal rate of
transformation or MRT),
• Productive Efficiency
• Allocative Efficiency
• Economies of Scale
PPF and Pareto Optimality/Efficient
• Means that a given PPF function should be set
equal to zero
 Samuelson Approach:
mrtx,G – mrs2x,G = mrs1x,G or mrs2x,G + mrs1x,G = mrtx,G
 Basic Assumptions In Samuelson
Theory
• Concave PPF between Private good and Public
Good
• Sacrifice by 1 consumer of private good is
required for Public Good to increase in supply
• Consumer 2 sacrifices his pvt good to enable
Consumer 1 and himself to enjoy
• However Consumer 2 sets the limit for consumer
1’s consumption possibility frontier by deciding
how much of private consumption he will make
 Assumptions
• Let there be two consumers 1 and 2 for a
Public good ‘x’ and a global public good ‘g’
 Pareto Optimality and Pure Public
Goods
• mrtx,G – mrs2x,G = mrs1x,G or mrs2x,G + mrs1x,G = mrtx,G
• In other words ‘mrs’ of both consumers should
be equal to ‘mrt’
• This is logical since mrs1xG is the residue of mrs2 and
Consumer 2 sacrifices to enable himself and
Consumer 1 to enjoy public goods which he also
enjoys at the same time

• This does not happen in a market economy

since each consumer would equate his own
‘mrs’ to ‘mrt’
Where public good is an intermediate good and
is used by producers in a non rival way
• Samuelson condition says that the aggregate
Marginal Product of the public good input
should be equal to the amount of private good
foregone by producing the marginal unit of
the public good
• Thus in place of adding the mrs of various
consumers we add the marginal valuations of
different firms using the public goods input.
 Overcoming free- rider problem
• Incentive compatibility – when consumption is not
competitive one cannot use prices to prevent
consumers from consuming the public good
• Taxes may not be proportionate to Marginal Value of
the good to an individual
• Individuals understate their preferences for public
goods for fear of attracting taxes
• They would like to free ride
• If they do not reveal their preferences then how does
financing of public goods happen?
• Solution to free riding has come from three sources
 The Lindahl-Wicksell Solution
• Aims to provide a market solution to the problem
of public goods
• Assumed a N person exchange economy in which
there is one commodity which can be used either
as a private good or (costlessly transformed) into
a public good
• Hence the ‘mrt’ between the private and the
public goods is 1 and the conditions of Pareto
optimum supply of the public good is that the
sum of the ‘mrs’ between the private and the
public be equal to 1
 Lindahl-Wicksell
• The Lindahl equilibrium is a set of personalized
tax rates and an amount of public good say’g’
such that (1) each individual wants the same
amount (G) of the public good (ii) G is a Pareto
optimal supply of the public good
• The LW logic is powerful but does not solve the
free rider problem well since it assumes that an
individual will truly reveal her true preferences
 Groves-Loeb Mechanism
• Public Board set up to supply the public good
• Asks people for their Vi schedule
• This may not be truly given
• But the Board also lets it known that a person’s
tax liability will be decided by G supplied – sum
of utilities (from the good)
• So no incentive to understate utilities
• But will the Board’s budget be good enough to
provide for desired supplies without deficits
 Groves Ledyard Mechanism
• Similar to G-LO but consumers are asked not to state V
schedules but only desired increments – whether 5
more fighter planes or 3
• Does provide a Pareto optimal supply of the public
good and is incentive compatible. For this however we
must initially have mutually consistent desired
increments

• But is based on the assumption that one person’s

increment is not affected by the other’s increment that
need not hold
• A Cournot - Nash assumption may also be a reality
 Clark-Groves- Vickrey mechanism
• Asks people for information on their utility
schedules just as in G-Loeb mechanism.
• The announced utility schedule is different
from the true utility
• Will create revenues to the Board to cover
production cost and also imposes an
externality tax
• This will create a surplus for the Board which
is not an optimal situation
Voting Models: Single Peaked Vs Non
 Origins of Median Voting Theory
• Condorcet (1785) an eminent French mathematician
and philosopher discovered the idea of a pivotal voter
and also noted how the accuracy of decisions can be
improved by majority decisions in juries, but includes
no clear statement of the median voter theorem.
• Political pundits have noted (and lamented) tendencies
for candidate positions to converge in democratic
elections, as did the occasional economic theorist,
Hotelling (1929),but the median voter theorem
awaited Duncan Black's work on majority voting
(1948), and Anthony Downs' extension to
representative democracy (1957).
 Median Voter
• If the median voter gets more or less what he or she wants, then anything
that affects the median voter's assessment of the relative merits of
alternative policies or candidates will also affect political outcomes.
• For example, the median voter's age, sex, income, information, ideology
and expectations should all systematically affect public policy.
• To the extent that these predictions are largely borne out by empirical
research, the median voter model can be regarded not only as a
convenient method of discussing majoritarian politics and a fruitful engine
of analysis, but also a fundamental property of democracy.
• Note that a median voter always votes in favor of the outcome that wins
the election.
• If voters cast their vote for the party or candidate closest to their most
preferred feasible policy, it turns out that the candidate who is closest to
the median voter always wins the election. This follows because the
candidate closest to the median voter is also closest to the ideal points of
more than half of the electorate
 Median Voting Theorem (7 types of voters: assume
each voter votes for 1 outcome at a time and votes are
counted by majority vote, then the best outcome is O4
as this is what is preferred by the median voter
Absence of single peak leads to Condorcet Paradox of
cyclical majorities
 Log Rolling: Splitting Ranks
• Under majority voting it is possible to adopt
policies that are actually opposed by a
majority of voters
• Logrolling is useful in a situation where a
minority group support some options
supported by sections of the majority groups
in order to get the support of these sections
for minority causes
 Rational Voter ignorance
• Usually on public decision making issue like
National Health Policies, voters are not
informed about public decision making as
transaction costs of supplying, collecting info
and analyzing and understanding the issues is
high
• How does RTI Act solve this problem?
 Bureaucracy and Lobbyists
• Apart from politicians , bureaucrats also affect
supply of public goods
• Another set of entities are lobbyists – some of
them lobby for resource allocation for non
public goods