Professional Documents
Culture Documents
Do Not Confess
B
A Confess
Do Not 1. 4.
Row Player’s
Confess (10,9) ( 7,11)
Preference:
2>1>3>4 Confess 2. 3.
(12,6) (8,8)
D Contribute Do Not
G Contribute
Row Player’s
Preference: Contribute 1. 4.
2>1>4>3 (3, 3) (2, 3.5)
Do Not 2. 3.
Contribute (3.5, 2) (1, 1)
If D knows that G would much rather pay full cost and have the
fence than not have it at all – he is likely to ‘pre-commit’ and say
he doesn’t need the fence and will not pay.
Since D knows that G will anyways build the fence irrespective
of whether he contributes or not then D has no incentive to reveal
his true preference – D can act as FREE RIDER!
G too has a similar incentive to not reveal his true preference.
Cost sharing may not occur & danger is that fence may NOT
get built –
MARKET FAILURE DUE TO FREE RIDER PROBLEM!
EXAMPLE 2
A Congested Highway
The cost of an additional vehicle is high
Greater wear and tear of road – greater waiting time for the
others already on the road – greater pollution – greater chance
of accident
But cost of exclusion is high – we have to keep someone to
collect the toll – his salary will have to be paid
Here it is desirable to exclude but not feasible !
MC of use
Impure Public Goods - Pure Private Good -
Congested Highway Health service
Ease of Exclusion
The first theorem guarantees that a competitive market will exhaust all of the gains
from trade – an equilibrium allocation achieved by a set of competitive markets
will necessarily be Pareto efficient.
Such an allocation may not have any other desirable properties but will necessarily
be EFFICIENT
Let us consider the First Welfare Theorem. This says that any competitive equilibrium is Pareto
efficient. There are hardly any explicit assumptions in this theorem—it follows almost entirely from
the definitions. But there are some implicit assumptions.
One major assumption is that agents only care about their own consumption of goods, and not
about what other agents consume. If one agent does care about another agent’s consumption, we
say that there is a consumption externality. We shall see that when consumption externalities are
present, a competitive equilibrium need not be Pareto efficient.
To take a simple example, suppose that agent A cares about agent B’s consumption of cigars. Then
there is no particular reason why each agent choosing his or her own consumption bundle at the
market prices will result in a Pareto efficient allocation. After each person has purchased the best
bundle he or she can afford, there may still be ways to make both of them better off—such as A
paying B to smoke fewer cigars.
Another important implicit assumption in the First Welfare Theorem is that agents
actually behave competitively. If there really were only two agents, as in the Edgeworth box
example, then it is unlikely that they would each take price as given. Instead, the agents would
probably recognize their market power and would attempt to use their market power to
improve their own positions. The concept of competitive equilibrium only makes sense when
there are enough agents to ensure that each behaves competitively.
Finally, the First Welfare Theorem is only of interest if a competitive equilibrium actually
exists. As we have argued above, this will be the case if the consumers are sufficiently small
relative to the size of the market.
Given these provisos, the First Welfare Theorem is a pretty strong result: a private market,
with each agent seeking to maximize his or her own utility, will result in an allocation
that achieves Pareto efficiency. The importance of the First Welfare Theorem is that it
gives a general mechanism—the competitive market—that we can use to ensure Pareto
efficient outcomes.
The First Welfare Theorem shows that the particular structure of competitive markets has the
desirable property of achieving a Pareto efficient allocation.
Implications of the Second Welfare Theorem
The Second Theorem of Welfare Economics asserts that under certain conditions,
every Pareto efficient allocation can be achieved as a competitive equilibrium.
The Second Theorem is concerned with the converse of the First and its focus can
be summarized as follows:
Given a Pareto optimum, can a competitive economy be constructed for which the
Pareto optimum is a competitive equilibrium? In terms of an Edgeworth box, the
same question can be formulated as asking whether it is possible to decentralize all
points on the contract curve. Using the diagram, it can be seen that this is possible
in an exchange economy if the households’ indifference curves are convex. With
convexity, the common tangent at a Pareto optimum provides the equilibrium
prices. To decentralize the economy, a point on this budget line is chosen as the
initial endowment point.
The Second Theorem has important policy implication:
In designing policy, it is almost certain that a policy maker would wish to achieve a
Pareto optimum, otherwise welfare could be increased at no cost.
The theorem demonstrates that the objective of the policy maker can be achieved by
making the economy competitive selecting the equilibrium that is to be decentralized
and providing each household with sufficient income to afford their allocation.
The lump-sum nature of the transfer is due to the fact that neither
household can alter the size of the transfer by changes in their behavior;
there is simply no scope for such changes in the economy described.
If at the equilibrium e the prices of the two goods are p1 and p2, then the value of
the transfer in Figure is p 1𝑥11 . The notional income of household 1 prior to
taxation is
𝑀1 = 𝑝1 𝜔11 + 𝑝2 𝜔12
Now, rather than actually redistributing quantities of the goods, the government
could tax household 1 an amount which reduces the income to:
The important point of this reinterpretation is that the tax scheme consisting of the taxes {T1,−T1} is
equivalent in its effect to the original transfer of endowment {−𝑥11 ,𝑥11 }.
This equivalence demonstrates that it is possible to view the transfers needed to achieve the
decentralizations as taking the form of either real transfers of goods between households or as
transfers of income in the form of lump-sum taxes. This description of lump-sum taxes can be easily
generalized to an H household economy in which a chosen equilibrium would be decentralized by a
vector of lump-sum taxes.
Since these taxes sum to zero, they represent
a simple redistribution of resources.
Lump-sum taxes have a central role in public economics due to their efficiency in achieving
distributional objectives.
It should be clear from the discussion above that the economy’s total endowment is not reduced by
the application of the lump-sum taxes. This point applies to lump-sum taxes in general.
As households cannot affect the level of the tax by changing their behavior, lump- sum taxes do not
lead to any inefficiency.
There are no resources lost due to the imposition of lump-sum taxes and redistribution is achieved
with no efficiency cost.
Having identified the nature and value of optimal lump-sum taxes, the question of their applicability
is now considered.
In practice, the endowment of most households is simply their future labour supply.
In order for a transfer, or tax, to be lump-sum the household involved must not be able to
affect the size of the transfer by changing their behavior.
It is clear that lump-sum taxes can be used, for example by taxing each household some fixed
amount a lump-sum tax is imposed.
Setting aside minor details, this was effectively the case of the U.K. poll tax.
This example motivates the following important observation. The efficiency of lump-sum
taxation rests partly on the fact that their imposition is costless but this was far from the case
with the U.K. poll tax.
In fact, the difficulties of actually collecting and maintaining information on the residential
address of all households made the imposition of a uniform lump-sum tax prohibitively
expensive.
Therefore, although the structure of lump-sum taxes makes them appear deceptively simple to
collect, this may not be the case in practice since the tax base, people, is highly mobile and
evasive.
However, the costs of collection are only part of the issue. What is the primary concern here is
the use of optimal lump-sum taxes - Optimality requires the tax to be based on all relevant
economic characteristics and households must not be able to alter these characteristics in
response to the taxes.
It may be possible to differentiate lump-sum taxes according to sex, age or eye-colour for instance,
but these are unlikely to be the relevant characteristics on which to base the tax.
For the exchange economy examples, the characteristics were the endowments and preferences of
the households.
More generally they may be the expected future labour incomes of the households or the
determinants of each household’s human capital.
Such characteristics are unlikely to be directly observable by the government and either it must
either rely on households honestly reporting their characteristics or the characteristics must be
inferred from the actions of households.
In the latter case, there is invariably scope for changes in market behavior, which implies the taxes
are no longer lump-sum.
When reports are the sole source of information, unobserved characteristics cannot form a basis for
taxation unless the tax scheme is such that there is an individual gain to truthful revelation.
As an example of the interaction between taxes and reporting, consider the following.
If the level of tax was inversely related to IQ and if all households had to complete IQ tests, then the tax
system would not be manipulated since the incentive would always be to maximize the score on the test.
In contrast, if taxes were positively related to IQ, a testing procedure could easily be manipulated and the
mean level of tested IQ would be expected to fall considerably. This indicates the potential for mis-
revelation of characteristics.
These ideas have been developed formally by Mirrlees (1986) who presents theorems on the (im)
possibility of designing non-manipulable lump-sum tax schemes. The central theorem considers a
population of households who each have the utility function
Since the cs are the only differentiating characteristic between households, any optimal set of lump
sum taxes must be based on these characteristics.
Now assume that the government cannot observe the c vectors, that households only truly report
their characteristics when they do not lose by doing so and that misrepresentation can only take
place by a household claiming that the values of the characteristics are above their true values.
With this formulation, a tax policy T = T (c) conditional on the characteristics can only be
administered, in the sense that it generates truthful revelation from the households, if the final
utility allocation generated by the taxes is non-increasing in c.
U h = U h (T (c)) = 𝑈 h (c)
The Mirrlees theorem shows that
𝑈h (c) will be increasing in ci with the optimal tax policy if good i is normal.
Since some goods must be normal, the policy optimal policy cannot be administered.
Externalities: Problems and Solutions
1
OUTLINE
1. Externality Theory
5 Conclusion
2
EXTERNALITIES: PROBLEMS AND SOLUTIONS
3
EXTERNALITY THEORY: ECONOMICS OF
NEGATIVE PRODUCTION EXTERNALITIES
Example: steel plant pollutes a river but plant does not face
any pollution regulation (and hence ignores pollution when
deciding how much to produce)
4
5.1
Externality Theory
Economics of Negative Production Externalities
Chapter 5 Externalities: Problems and Solutions
© 2007 Worth Publishers Public Finance and Public Policy, 2/e, Jonathan Gruber 6 of 33
EXTERNALITY THEORY: ECONOMICS OF
NEGATIVE CONSUMPTION EXTERNALITIES
© 2007 Worth Publishers Public Finance and Public Policy, 2/e, Jonathan Gruber 10 of 33
A P P L I C A T I O N
Chapter 5 Externalities: Problems and Solutions
© 2007 Worth Publishers Public Finance and Public Policy, 2/e, Jonathan Gruber 11 of 33
EXTERNALITY THEORY: POSITIVE
EXTERNALITIES
8
5.1
Externality Theory
Positive Externalities
Chapter 5 Externalities: Problems and Solutions
© 2007 Worth Publishers Public Finance and Public Policy, 2/e, Jonathan Gruber 13 of 33
EXTERNALITY THEORY: MARKET OUTCOME IS
INEFFICIENT
12
PRIVATE-SECTOR SOLUTIONS TO NEGATIVE
EXTERNALITIES: COASE THEOREM
13
COASE THEOREM EXAMPLE
© 2007 Worth Publishers Public Finance and Public Policy, 2/e, Jonathan Gruber 16 of 33
THE PROBLEMS WITH COASIAN SOLUTIONS
17
THE PROBLEMS WITH COASIAN SOLUTIONS:
BOTTOM LINE
18
PUBLIC SECTOR REMEDIES FOR
EXTERNALITIES
19
5.3
Public-Sector Remedies for Externalities
Corrective Taxation
Chapter 5 Externalities: Problems and Solutions
© 2007 Worth Publishers Public Finance and Public Policy, 2/e, Jonathan Gruber 22 of 33
5.3
Public-Sector Remedies for Externalities
Subsidies
Chapter 5 Externalities: Problems and Solutions
© 2007 Worth Publishers Public Finance and Public Policy, 2/e, Jonathan Gruber 24 of 33
PUBLIC SECTOR REMEDIES FOR
EXTERNALITIES: REGULATION
22
5.4
Distinctions Between Price and Quantity
Approaches to Addressing Externalities
Chapter 5 Externalities: Problems and Solutions
Basic Model
© 2007 Worth Publishers Public Finance and Public Policy, 2/e, Jonathan Gruber 26 of 33
MODEL WITH HETEROGENEOUS COSTS
V = max q H + q L − cH (q H ) − cL(q L) ⇒
qH ,qL
M CH = 1, M CL = 1 ⇒ q H = 1/3, q L = 2/3
Optimum outcome is to have the low cost firm do more pol-
lution reduction than the high cost firm
24
TAX VERSUS REGULATION SOLUTION
25
Quantity Regulation with Trading Permits
© 2007 Worth Publishers Public Finance and Public Policy, 2/e, Jonathan Gruber 27 of 33
MULTIPLE PLANTS WITH DIFFERENT
REDUCTION COSTS
28
CORRECTIVE TAXES VS. TRADABLE PERMITS
© 2007 Worth Publishers Public Finance and Public Policy, 2/e, Jonathan Gruber 29 of 33
5.4
Distinctions Between Price and Quantity
Approaches to Addressing Externalities
Chapter 5 Externalities: Problems and Solutions
© 2007 Worth Publishers Public Finance and Public Policy, 2/e, Jonathan Gruber 30 of 33
CONCLUSION
The State can be envisaged as coming into existence to satisfy the needs of ALL members of the
community or to help gratify the wants of only a PART of it. The fist corresponds to allocative efficiency
and the second to redistribution.
Market process enables a society to move from a point inside the Pareto Possibility Frontier (PPF) to a
point on it. This is allocative efficiency even though the point on the PPF is chosen arbitrarily
To obtain Pareto Efficiency in the allocation of Public Good, a collective choice process that is less
anarchic than the market is required.
A conscious choice of the quantities of Public Good to be produced must be made and along with it the
choice of means of paying for it.
The issue of the distribution of gains from collective action is more clearly visible in the allocation
of Public Good by a political process than it is in the allocation of Public Good by the market
exchange process
Redistribution as Insurance
Assume that there will be two income classes in the post constitutional society, with every
member of a given class having the same income, Yi and Y2 > Y1. Let r be the number of rich
in class 2 and p the number of poor in class 1.
An individual uncertain of her future position chooses a tax of T on the rich and a benefit
subsidy B to the poor so as to maximize the following objective function:
where π2 and π1 are the probabilities that she will be in classes 2 and 1, respectively
(π2 = r/(r + p), π1 = p/(r + p)). Assuming zero transaction costs in transferring income
rT = pB. (3.2)
Substituting for π1, π2, and T into (3.1) and maximizing with respect to B, we obtain
From this it follows
An individual who maximizes her expected utility given that she is uncertain over whether
she will be rich or poor will support redistributive taxes that equate the marginal utilities of
representative members of each group. If all individuals have the same utility functions, she
chooses taxes and subsidies to equate incomes across all individuals
In creating institutions to redistribute from the rich to the poor, the uncertain
individual insures herself against the possibility that she will be one of the poor.
Uncertainty over future position could lead to unanimous agreement to include institutions
for redistribution in the constitution. In this case the constitution becomes a kind of
insurance contract.
Question which arises is why should there be state provision of such an insurance instead of
market. Two main reasons have been given.
(1) The amount of risk borne by any single member of an insurance pool declines as the
membership of the pool grows. When the risks associated with new members are the same as
those attached to existing members, the optimal size of the membership in the pool is infinity.
Insurance becomes a sort of “natural monopoly” with the optimal size of the “insurance club”
being all members of society
(2) The risks of being poor are not the same across all individuals, however. Those who are of
below average intelligence or ambition have higher probabilities of being poor than the
average person; higher intelligence, more ambitious people have lower probabilities. If it is
possible for an individual to determine his own probability of being poor, but it is not possible
for a private insurance company to make this determination, the sale of insurance by a private
company could lead to an adverse selection problem.
Redistribution as a Public good
Under the second hypothesis, the rich are seen as transferring income to the poor, not because they are
uncertain about whether they might become poor, but out of empathy or similar altruistic motivation.
Each member of the highest income group is envisaged as gaining some satisfaction from the utility gains of
members of the lower classes. The highest income group acts as a sort of club that unanimously agrees to
transfer income from itself to members of the lower group(s)
Assuming three groups, with Y3 > Y2 > Y1, then each member of group 3, when voting, can be seen as
maximizing an objective function consisting of a weighted sum of the utilities of its own members and
those of members of lower-income groups
Assuming three groups, with Y3 > Y2 > Y1, then each member of group 3, when voting, can be seen as
maximizing an objective function consisting of a weighted sum of the utilities of its own members and
those of members of lower-income groups:
where n3, n2, and n1 are the numbers of individuals in groups 3, 2, and 1, respectively; T is the tax imposed
on the richest group, and B1 and B2 are the per capita subsidies to the other two groups.
Each member of the richest group places full weight on the utility of each member of its own group,
and partial weights (α1 ≤ 1, α2 ≤ 1) on the utilities of members of other groups. Substituting from the
budget constraint
If a member of the richest class places the same weight on the utilities of members of classes 1 and 2
(α1 = α2) and assumes that each derives the same utility from income, then (3.13) implies subsidies to
members of classes 1 and 2 so as to equate their marginal utilities of income.
Since Y1 < Y2, if the marginal utility of income falls with increasing income, then the incomes of the lowest
class must be raised to equality with those of class 2 before any transfers are made to class 2
A saintly altruist who placed equal weight on her own utility as on that of others (α1 = α2 = 1) would vote to
equate everyone’s income.
Everyday altruists who place more weight on their own utility than on the utilities of others (0 < α < 1)
will not favor transfers so large as to bring their own incomes into equality with those to whom they make
transfers.
Since charity is a purely voluntary act, whereas government redistribution programs are not, one wonders
why, if all the members of group 3 do favor redistribution, reliance is not made on private charities (clubs) for
redistribution.
An argument for government intervention relies again on the free-rider problem. If a member of group 3
wishes to see the welfare of all individuals in group 1 raised, and not just a few whom she knows personally,
she cannot achieve her goal alone.
If all members of group 3 feel likewise, they can achieve their goal by joint-collective action. But if a voluntary
association is employed, free-riding may ensue, and less than the Pareto-optimal amount of redistribution
may occur.
Redistribution as Taking/Grabbing
Almost no democratic system makes its collective decisions using the unanimity rule.
Once government action can be taken despite the opposition of some citizens, redistribution can take on
the form of pure involuntary transfers from the losers to the winners under the political process.
Let us assume again two groups, whose members obtain utility from income, and that own political
resources that they can spend to obtain additional income in the form of government subsidies.
Of course, only one group can obtain positive subsidies, so that the other group must use its political
resources to reduce its taxes.
U1 = U1(Y1 + B, R1),
Where,
(∂U1/∂Y1) > 0, (∂2U1/∂Y12 ) < 0, Having to use political resources (Ri) lowers utility
(∂U1/∂R1) < 0, and (∂2U1/∂R12) < 0
For Group 2:
For group 2 we have U2 = U2(Y2 − T, R2)
B = B(R1, R2),
Political resources can take many forms. In a democracy there might be effort exerted by a group for
one party (handing out leaflets, stuffing envelops, telephoning) to bring about its victory. Here one
might expect groups with low opportunity costs of time (unemployed, retired) to do well at winning
subsidies.
Involuntary redistribution must make someone worse off, and can make everyone worse off.
We usually think of involuntary redistribution as money flowing from one group to the government and
out again to a second group, with the first group being made worse off and the second better off. Such
a situation would definitely occur through a pure tax/subsidy scheme if only one group expended
resources to
win the subsidy.
The fact that it was willing to spend its resources would imply that its gross benefits exceed the
resources spent.
If both groups expend resources to win subsidies, the end result may be that they are both worse off
than they would have been had they each not attempted to obtain a subsidy.
To see this assume that both groups spend money lobbying for a subsidy and that their efforts perfectly
offset one another. Neither group obtains any benefits from their lobbying, and both are worse off by
the amount of resources spent on lobbying.
Voting Rules
Social Welfare Function
and
Arrow Impossibility Theorem
Since all can benefit from the provision of a public good – obvious voting rule would be
unanimous consent.
Knut Wicksell (1896) was the first to link the potential for all to benefit from collective
action to unanimity rule.
The Unanimity Rule is the ONLY voting rule certain to lead to a Pareto preferred public
good quantity and tax share.
However, there are two main criticisms levelled against it
Criticisms of the Unanimity Rule
The unanimity rule is the only voting rule certain to lead to Pareto preferred public good
quantities and tax shares.
(1) A groping search for a point on the contract curve might take considerable time particularly
in a large community of heterogenous tastes.
The loss in time might easily outweigh the gains to those who are saved from paying a tax
exceeding their benefits from the public good.
An individual who is uncertain about whether he would be so ‘exploited’ under a less than
unanimity rule might easily prefer such a rule rather than spend the time required to attain full
unanimity.
(2) The second objection under unanimity rule is that it encourages strategic behaviour.
If A knows the maximum share of taxes that B will assume rather than go without the public
good, A can force B to point C on the contract curve by voting against all shares greater than
tc - All gains then accrue to A
If B behaves the same – the final outcome is dependent on the bargaining strengths of the
two individuals.
Bargaining can further delay the attainment of the agreement as each player has to “test”
the others willingness to make concessions.
Up
T W
Majority rule can lead to redistribution other than via the obvious route of direct cash transfers.
Tullock described a community of 100 farmers in which access to the main highway is via small trunk roads,
each of which serves only 4 or 5 farmers. The issue comes up as to whether the entire community of 100 should
finance the repair of all of the trunk roads out of a tax on the entire community.
Obviously one can envisage a level of repairs and set of taxes on the individual farmers under which such a
proposal would be unanimously adopted.
But under majority rule it is to the greater advantage of some to propose that only one half of the roads are
repaired out of a tax falling on the entire population. Thus, one can envisage a coalition of 51 of the farmers
forming and proposing that only the roads serving them are repaired out of the community’s general tax
revenue (Tullock discusses other possible outcomes, which we take up shortly).
Such a proposal would pass under majority rule, and obviously involves a redistribution from the 49 farmers
who pay taxes and receive no road repairs to the 51 farmers whose taxes cover only slightly more than one half
of the cost of the road repair.
In the Tullock example, redistribution to the 51 farmers in the majority coalition takes place through the
inclusion in the entire community’s budget of a good that benefits only a subset of the community.
CYCLING
V2
Voters X Y Z X
V3
1 > > <
V1
2 > < >
Plurality Rule: Choose the candidate who is ranked first by the largest no. of votes.
Condorcet criterion: choose the candidate who defeats all others in pairwise elections using
majority rule.
The Coombs System: Each voter indicates the candidate he ranks lowest of the m candidates.
Remove from the list of candidates the one ranked lowest by the most voters. Repeat the
procedure for the remaining (m-1) candidates. Continue until only one candidate remains.
Declare this candidate the winner.
Borda Count:
Give each of the m candidates a score of 1 to m based on the candidate’s ranking in a voter’s
preference ordering; that is, the candidate ranked first receives m points, the second one m-1,…,
the lowest ranked candidate one point. The candidate with the highest points is declared the
winner
Approval Voting:
Each voter votes for the k candidates (1 ≤ k ≤ m) he ranks the highest of the m candidates,
where k can vary from voter to voter. The candidate with the most votes is the winner.
Social Welfare Function
The traditional means for representing the values of the community in economics is to use a social
welfare function (SWF).
The seminal paper on SWFs is by Bergson (1938), with the most significant further explication by
Samuelson (1947, ch. 8). The SWF can be written as follows:
The objective is to define a W and set of zi s, and the constraints thereon, to yield meaningful first- and
second-order conditions for a maximum W.
Although in principle any variables that are related to a society’s well-being (e.g., crime statistics,
weather data, years of schooling) might be included in the SWF, economists have focused on economic
variables.
The normative issue to be resolved with the help of the SWF is which point along the generalized
Pareto-possibility frontier should be chosen; what set of lump-sum taxes and subsidies is optimal.
Both Bergson and Samuelson speak of solving this question with the help of a variant of the SWF in
which the utility indexes of each individual are direct arguments in the welfare function
W = W(U1,U2, . . . ,Us ).
In particular how would the function W (or more generally the social ordering R) depend upon
individual preference orderings? Or in other words what should be the collective choice rule?
A Collective Choice Rule is a functional relation f such that for any set of n individual orderings
(R1………Rn ) (One ordering for each individual)
Arrow’s use of the expression SWF is different from that of Bergson and Samuelson. A Collective
Choice Rule that specifies ordering for society is called a SWF by Arrow. Any ordering for the society
is a Bergson Samuelson SWF.
An Arrow SWF determines a Bergson Samuelson SWF on the basis of individual ordering.
Since it is possible to impose to set up a SWF or Collective Choice Rule in any manner consistent or
inconsistent, reasonable or wild, it is useful to impose some mild conditions on the form of the SWF.
In his “General Possibility Theorem” or since the theorem is negative, “General Impossibility
Theorem”, Arrow proved that a set of very mild looking conditions are altogether so restrictive that
they rule out not some but every possible SWF
AXIOM A: Completeness For all x and y either xRy or yRx where R stands for preferred/indifferent to
AXIOM B: Transitivity For all x, y and z xRy and yRz implies xRz
The set of 4 conditions are:
2. NON-DICTATORSHIP:
No individual enjoys a position such that whenever he or she expresses a preference between any
two alternatives and all other individuals express the opposite preference, his or her preference is
always preferred in the social ordering
3. UNRESTRICTED DOMAIN:
There are at least 3 among all the alternatives under consideration for which all logically possible
individual ordering of these three alternatives is admissible (elaborated upon in next slide)
The requirement that the rule must work for every logically possible
configuration of individual preference orderings is called the
condition of unrestricted domain
V1 = xPyPz
V2 = yPzPx
V3 = zPxPy
1. The MAJORITY RULE is an appealing collective choice rule (CCR). However, it fails to satisfy the condition of
unrestricted domain (i.e. problem of cycling exists).
2. The method of BORDA COUNT also fails because it does not pass the condition of Independence of Irrelevant
Alternatives.
Suppose there are 3 alternatives x, y, z and three individuals. According to the method 3 marks are assigned to
the alternative ranked first; 2 marks are assigned to the alternative ranked second and 1 mark to the third
alternative.
The individual orderings are
V1: xPyPz
V2: zPxPy x y z
V3: zPxPy V1 3 2 1
V2 2 1 3
Here x and z receive 7
points and there is a TIE
V3 2 1 3
TOTAL 7 4 7
Now suppose that everyone’s ranking of x vis-à-vis z remains the same, but
individual 1 changes his/her mind about an irrelevant alternative viz., y and has
the following ordering
While everybody’s ordering of x and z are still the same, the social choice between x and z is not the same
Definition of a Decisive Set: A set of individuals D is decisive, for alternatives x and y in a given social
welfare function, if the function yields a social preference for x over y, whenever all individuals in D prefer x
to y and all others prefer y to x
• The unrestricted domain assumption allows any possible constellation of ordinal preferences.
• When unanimously preferred alternative does not emerge, some method of choosing among the
Pareto Preferred alternatives must be found.
• The Independence of Irrelevant Alternatives assumption restricts attention to the ordinal
preferences of the individuals for any two issues, when deciding those issues
• But as we seen in our discussions of majority rule it is all too easy to construct rules that yield
choices between two alternatives but produce a cycle when three successive pair-wise choices are
made
• The transitivity postulate forces a choice among the three.
• The Social Choice process is not to be left indecisive. But with the information at hand i.e. individual
ordinal rankings of issue pairs – there is no method for making such a choice that is not imposed or
DICTATORIAL
Significance of Arrow’s Results
It has been known for a long time that some methods of combining individual preferences into social
preferences lead to inconsistencies.
Condorcet (1785) had noted intransitivities of majority decision almost two centuries ago
The most discussed case of such inconsistency is the so called “Paradox of Voting”. This example gives
a good introduction to the nature of the problem
The importance of the General Possibility Theorem lies in the fact that we can predict the result in each
case i.e. NO SOCIAL WELFARE FUNCTION will satisfy the conditions.
The theorem is completely general and saves a long (and perhaps endless0 search
Public Goods and Publicly
Provided Private Goods
We need to make an important distinction, between public production and public provision. The two
are often confused, though both logically and in practice they are distinct.
The government provides for the National Defence, yet much of the production of the goods
purchased for national defence is within the private sector.
The government has, in many countries, a monopoly of the mail service, yet it charges for the use of
mail in a manner little different from that of private enterprise.
It may be impossible, or extremely costly, to charge for the use of a specified commodity. In other words, it
may not be possible to exclude non-contributors. This is essentially a technical question, and depends on
the available technology.
In the case of television, calculation of the extent of use depends on it being possible to determine from
outside whether the receiver is in operation or on the employment of scrambling devices.
It has been suggested that automatic metering devices could be installed to record the passage of
vehicles through the highways system and that with large-scale computer networks it would be feasible to
charge for actual usage.
For some goods, such as national defence, it is hard to imagine that even future developments in information
processing will allow individual benefit to be determined; so that for these exclusion is indeed impossible.
Where exclusion is not technically impossible, it may still be decided to supply the
good publicly for certain reasons…
The first is that it may not be desirable on efficiency grounds to use prices to govern the usage of a
commodity. The effects of charging depend on (1) the conditions of demand and (2) the conditions on which
the good can be supplied to an additional individual.
If the demand is highly inelastic, then pricing has little effect on usage. In the extreme case, if demand is
completely inelastic, there is no efficiency loss from not charging for the commodity (although there may be
other arguments, such as raising revenue, as we have seen).
Many places do not charge for the quantity of water used, because it is judged that the benefits for metering
would be relatively small, demands not being very elastic, and insufficient to warrant the installation of
metering devices. (There may also be external economies in consumption—at least, that was an important
historical reason for public provision)
Standard discussions tend in effect to focus on the second aspect—that usage by one person does not
reduce the amount that others can consume. In other words, the cost of supplying a fixed quantity to
another individual is zero.
Examples typically given include television programmes (my listening to a TV programme transmitted over
the airwaves does not detract from others listening); information (my knowing something does not detract
from others knowing the same thing); and national defence. These are extreme cases, and are referred to
as pure public goods, where “each individual’s consumption of such a good leads to no subtraction
from any other individual’s consumption” (Samuelson, 1954)
More generally, there is a range of commodities that have the property that an increase in one person’s
consumption (keeping aggregate expenditure on the commodity constant) may not decrease the
consumption of other people by the same amount. If one person travels on a little-used highway, the
benefits of the road to others are reduced only slightly.
On this view, private goods are at one extreme of a spectrum, where an increase of one unit in the
consumption by Mr X reduces the consumption available to others by one unit; and pure public goods at
the other extreme, where an increase in Mr X’s consumption leads to no reduction for others. These polar
cases are sometimes characterized in the following way.
Let be the consumption by household h of the ith commodity. Then for private goods
It may be noted that this assumes no free disposal. For many public goods, such as defence, this may not be an
unreasonable assumption; on the other hand, for goods such as television, free disposal is possible, and should be
replaced by
The intermediate cases are somewhat harder to characterize, and various approaches have been
suggested in the literature. One is to write the consumption possibility frontier for the economy as being
for good i:
One is to write the consumption possibility frontier for the economy as being for good i:
𝜕𝜒/𝑋𝑖1
=0
𝜕𝜒/𝑋𝑖2
An alternative approach is in terms of consumption externalities, and this has been developed by
Samuelson (1969). In this case the purchase of good i by household h may enter the utility function of
other individuals.
In this case the purchase of good i by household h may enter the utility function of other individuals.
In both cases, we have a problem of defining what it is that is being consumed, and how it is to be
measured. For instance, for television and radio broadcasts, the obvious unit to measure consumption is
“programmes listened to”. In this case, the first approach seems more natural.
On the other hand, if individuals privately purchase protective services (e.g., police guards), utility may be
a function of the level of “safety” in the community, which may be a function of the aggregate expenditure
on protective services, as well as on the private level of protection.
Individuals, in providing protection for themselves (and thus lowering the return to crime), are providing a
public good (safety), and the consumption externalities representation seems natural.
For instance, if P represents the total number of policemen available, and Ph represents the number of
policemen assigned to (“consumed by”) the household h, then ΣPh = P, and police appear to be a private
good, yielding consumption externalities.
If however what is consumed (negatively) is the expected number of crimes suffered by household h, denoted
by Ch, then we have a consumption possibilities curve
The third set of reasons for public provision relates to distributional objectives. This may stem either
from a general distributional goal, for example embodied in a social welfare function, or from principles
of specific egalitarianism.
Thus, distributional reasons are probably the primary rationale for the public provision of education —
either because it reduces inequality of endowments, or because access to at least a minimum level of
education is an objective in itself.
Optimum Provision of Pure Public Goods—Efficiency
In this section we consider the optimum level of provision of a single, pure public good, consumed in
quantity G by everyone. There is an aggregate production relationship:
F(X,G) = 0
First-Best Allocation
The government of a fully controlled economy is assumed to choose the level of G, and the
allocation of private goods Xh to household h (where h = 1, . . . , H) to maximize an individualistic
social welfare function. If the individual utility function is Uh (Xh, G), then the social welfare
function may be written as
This is the basic condition for the optimum supply of public goods: the sum of the marginal rates of
substitution between the public good (and some private good) must equal the marginal rate of
transformation (ΣMRS = MRT).
There is a clear intuitive interpretation of these conditions for a full optimum. The marginal benefit of an
extra unit of a public good is the benefit that person 1 gets, plus the benefit that person 2 gets, etc. In
contrast, an extra unit of a private good is either given to person 1 or given to person 2.
The solution may be illustrated diagrammatically for the case where there are two individuals and two goods (X =
private good, G = pure public good).
There are two individuals and Figure shows in the upper part the indifference curves
two goods (X = private good, for citizen I and the production constraint AB.
G = pure public good).
Suppose we fix citizen I on the indifference curve UI.
The possibilities for citizen II are shown in the lower
part of Fig. by CD (the difference between AB and UI).
i.e.
INDIVIDUAL BEHAVIOUR
The individual's utility function, U(y, Q),
Exogenous money income, I, can be used either to purchase units of the private good or to
acquire additional units of the public good. The number of units of the public good acquired by the
individual is denoted by q. If there are n individuals in the community
From the point of view of the utility function, the individual's own contribution and that of the rest of the
community are perfect substitutes.
However, the individual will be keenly aware of the distinction, since the former involves an opportunity
cost in terms of the private good forgone in acquiring q, whereas the latter involves no such cost.
Like the individual, we find the distinction important and find it useful to refer to the rest of the
community's contribution separately. This we denote by
Maximization of utility is subject to constraints. The simplest formulation assumes a linear trade-off
between y and q:
y + PQ q = /
This is a budget constraint, with PQ being the given money price of a unit of acquisition of the public good
It may equally well be thought of as reflecting any constant cost technology that converts the given quantity / of
primary resource (e.g., labor services) into either of two final goods, in which case pQ is the marginal rate of
transformation between the public good and the private good.
The total contribution, Q, may be thought of as producing the final public good, Z, with decreasing returns to scale:
The indifference map is truncated by the vertical line BB, which corresponds to the value of q
that would, by itself, completely exhaust the consumer's budget.
The individual's choice of q cannot be determined
without reference to the value of 𝑄෨
We know that movement along any IC curve implies by definition unchanged utility
Their preferences and income levels may differ, but they face the same price p^ for the public good.
We shall also suppose that there is one private good, whose price is unity throughout. Each
individual has the problem
is Nash Equilibrium if
That is, each individual's chosen contribution is a "best response" to the other's.
A possible justification of such an equilibrium concept runs as follows. Suppose each is to choose a
contribution level. At the time when this choice is made, each knows the other's budget constraint and
preferences. Moreover, this is common knowledge – not only does each know, but each knows that the
other knows, and so on.
Individual A may now think: "If I expect B to contribute 𝑞ത 𝐵 , my most preferred contribution will be 𝑞ത 𝐴 .
Now, if B expects me to contribute 𝑞ത 𝐴 , Bs most preferred contribution — given that I know her
preferences and that she is a utility-maximizer, will be 𝑞ധ 𝐵 . If 𝑞ത 𝐵 ≠ 𝑞ധ 𝐵 then there is an inconsistency
between belief and action for at least one of the individuals.
In this two-person example, the rest of the
community's contribution is simply the other
individual's provision.
First, it is typically not Pareto-efficient. The shaded area of Figure 6.4 constitutes a
region of mutual advantage with respect to E, consisting as it does of allocations that
Pareto-dominate the Nash-Cournot equilibrium.
1 1 1
𝑝𝑥 𝑚 − 1 = 𝑝𝑥 = +
𝑝𝑥 𝑛 − 1 𝑛 𝑚
(𝑝𝑥 𝑚 − 1)(𝑝𝑥 𝑛 − 1) = 1
1 1
𝑝𝑥 = 𝐵 + 𝐴
𝑈𝑦 𝑈𝑦
𝑝𝑥2 𝑚𝑛 − 𝑝𝑥 𝑚 − 𝑝𝑥 𝑛 + 1 = 1 𝑈𝑄𝐵 𝑈𝑄𝐴
𝑝𝑥 𝑚𝑛 − 𝑚 − 𝑛 = 0 𝑝𝑥 =MRSB + MRSA
𝑝𝑥 𝑚𝑛 = 𝑚 + 𝑛
This is the familiar Samuelson condition, which says that provision of a public good should be
taken up to the point at which the marginal rate of transformation equals the sum over all
individuals of the marginal rates of substitution between the public and the private goods
𝑀𝑅𝑆 = 𝑀𝑅𝑇
Comparing this with the expression below clarifies the sub-optimality in the Nash case
In Nash case, individuals adjust their own contributions independently of their neighbours and
therefore will add to the public good only to the point at which their private marginal rates of
transformation 𝒑𝑸 𝐞𝐪𝐮𝐚𝐥𝐬 𝐭𝐡𝐞𝐢𝐫 𝐩𝐫𝐢𝐯𝐚𝐭𝐞 𝐌𝐑𝐒
Voluntary provision of public goods with Mueller p.18
constant returns to scale
Letting Gi be the contribution to the public good of individual i , then the total quantity of public good supplied is
G = G1 + G2 + G3 +· · · Gn. (1)
Let each individual’s utility function be given as Ui (Xi , G), where Xi is the quantity of private good i consumes
Now consider the decision of i as to how much of the public good to supply, that is, the optimal Gi , given her
budget constraint
Yi = Px Xi + PgGi
where Yi is her income and Px and Pg are prices of the private and public goods, respectively.
In the absence of an institution for coordinating the quantities of public good supplied, each individual must
decide independently of the other individuals how much of the public good to supply.
In making this decision, it is reasonable to assume that the individual takes the supply of the public good by
the rest of the community as fixed.
Each i chooses the Gi that maximizes Ui , given the values of Gj chosen by all other individuals j .
Individual i ’s objective function is thus
(2)
(3)
(4)
(5)
This equilibrium is often referred to as a Cournot or Nash equilibrium, as it resembles the behavioral
assumption Cournot made concerning the supply of a homogeneous private good in an oligopolistic
market.
Now let us contrast (2.5) with the condition for Pareto optimality. To obtain this, we maximize the
following welfare function
(6)
where all γi > 0.
Given the positive weights on all individual utilities, any allocation that is not Pareto optimal – that is,
from which one person’s utility can be increased without lowering anyone else’s – cannot be at a
maximum for W
(7)
(8)
𝜆𝑃𝑥
From (9) 𝛾𝑖 = (B)
𝜕𝑈𝑖
For i= 1……n ൗ𝜕𝑋
𝑖
𝜆𝑃𝑥 (C)
Hence σ 𝛾𝑖 = Σ 𝜕𝑈𝑖
ൗ𝜕𝑋
𝑖
𝜕𝑈𝑖ൗ 𝜕𝑈𝑖ൗ
𝜕𝐺 = 𝜆𝑃𝑔 𝜕𝐺 𝑃𝑔 (11)
or ` =
𝜕𝑈𝑖 𝜆𝑃𝑥 𝜕𝑈𝑖 𝑃𝑥
ൗ𝜕𝑋 ൗ𝜕𝑋
𝑖 𝑖
It can be shown that public goods provision in (11) i.e. under Pareto provision, is greater than in (5) i.e. under Nash
𝜕𝑈𝑖ൗ 𝜕𝑈𝑗
ൗ
𝜕𝐺 = 𝑃𝑔 − 𝜕𝐺
𝜕𝑈𝑖 𝑃𝑥 𝜕𝑈𝑗 (12)
ൗ𝜕𝑋 𝑗≠𝑖 ൘𝜕𝑋
𝑖 𝑗
𝜕𝑈𝑗
If X and G are Normal goods then ൗ
𝜕𝐺 > 0
𝜕𝑈𝑗
𝑗≠𝑖 ൘𝜕𝑋
𝑗
MRS of Public for Private good under MRS of Public for Private good under Pareto = 6/1
Nash = 8/1 i.e. 8 units of public good i.e. 6 units of public good given up for 1 unit of
given up for 1 unit of private good X private good X
Since 8 has been given up under NASH and 6 is given up under PARETO.
This implies that greater G is being provided under PARETO than under NASH
Thus for a community greater than one individual, Gn/Gp will fall i.e.
less than Pareto optimal quantity of Public Good is supplied voluntarily.
These three decisions are mutually interdependent and must be rendered jointly
Let us consider a community of 2 taxpayers A and B and one type of social good only
Its supply furnishes benefits to both, A & B whose benefit shares may be considered joint products
Jointly A & B must contribute enough to cover the total cost of whatever volume of social goods is
supplied – individually each will have to pay less as the other contributes more
B’s offer to contribute certain percentage of the total cost of various amounts of social goods may be
interpreted from A’s point of view as a supply schedule of social goods and A’s offer may be interpreted
similarly from the viewpoint of B.
In the figure the volume of social goods is
measured on the horizontal axis
Here, A contributes ED, and B contributes DH per cent. For any amount
in excess of OE, the combined cost shares that A and B are willing to
accept fall short of 100 %
For any supply below OE, both A and B are willing to offer better terms
than the other demands. At ON, for instance, total offers exceed costs
by RZ per cent.
If B contributes TZ, A will purchase this amount for NZ, even though he
would be willing to con tribute NR.
10. For output < OE along UDY and to the left of it,
combined contributions exceed total costs. Exact shares
of A and B will depend on their bargaining skills.
The problem will not vanish when the number of voters increases. Individuals can free ride on the
contributions of others
For large numbers each individual’s contribution is so small that it cannot affect supply.
The preceding critique of the voluntary exchange model focused on the assumption that true
preferences will be revealed.
The second flaw arises from the partial equilibrium setting of the model in which the satisfaction of
social wants is considered independently of private wants – Samuelson said that the must be re-stated
in General equilibrium terms.
Restatement in Terms of General Equilibrium - Optimal Solution with Known Preferences
In order to examine this aspect of the problem, let us assume that true preferences are revealed and
known. How, then, can the government arrange for an optimal allocation of resources between
private and social wants?
In Figure 4-3 the total output of social goods is measured on the vertical axis,
and that of private goods, on the horizontal axis. The curve FE is a
transformation schedule, showing what combinations of social and private goods
may be produced.
The combination o he chosen will depend upon the preferences of our two
consumers, A and B, and upon the distribution of income between them.
In order to define this distribution let us assume that only private goods are
produced.
We then specify that A’s income in terms of private goods equals OC and B’s
income equals OD, where OC+OD=OE (total output)
Now, let OC be A's income terms of private goods and i1C his
indifference curve through C.
On balance, A is as well off as in the absence of We thus obtain the greatest gain that A can derive from the
social goods since he has remained on i1C, while supply of social goods provided that B’s initial position is not
B's position is improved since he has moved harmed.
from i1D to the higher indifference curve i2W.
We also obtain the greatest gain that B can derive without
We now reverse the argument and obtain curve harming A.
CN in figure 2 as the path travelled by A while B
moves up along i1D in figure 3
This need does not arise in the allocation of The General Equilibrium view thus points to a second flaw
resources between various private wants where in the voluntary-exchange model. Even if all preferences
various individuals may consume different amounts of are revealed, there is no single best solution analogous to
any one product. the Pareto optimum in the satisfaction of purely private
wants.
The simple welfare condition leads to a single solution
and the optimal allocation of resources is determined Instead we arc confronted with large number of solutions,
uniquely on the basis of a given state of distribution. all of which are optimal in the Pareto sense.
A major flaw in the Voluntary Exchange Models is that it crucially hinges on voluntary
revelation of preferences – which is an unrealistic assumption
One line of advance from this is towards CLUB GOODS where by volunteering to be a
member of a club preferences are revealed
or LOCAL PUBLIC GOODS in which the act of residing in a location reveals preferences
for local public goods being provided
X’ X0
Suppose j mis-reports MV to MVj’ in the
adjacent figure. The supply of public
goods rises to X’.
CLUB GOODS
A club is a voluntary group of individuals who derive mutual benefit from sharing one or more of the
following: production costs, the members' characteristics, or a good characterized by excludable
benefits.
Privately owned and operated clubs must be voluntary; members choose to belong because they
anticipate a net benefit from membership. Thus, the utility jointly derived from membership and from the
consumption of other goods must exceed the utility associated with Non-membership status. This
voluntarism serves as the first characteristic by which to distinguish between pure public goods and club
goods.
Clubs involve sharing, whether it be the use of an impure public good or the enjoyment of the desirable
attributes of the members. Sharing often leads to a partial rivalry of benefits as larger memberships
crowd one another, detracting from the quality of the services received.
Crowding or congestion depends on some measure of utilization, which could include the number of
members, the total number of visits to the club's facilities
As membership size expands, both costs and benefits arise: Costs involve increased congestion, while
benefits result from cost reductions owing to the sharing of provision expense associated with the club
good.
By adding a cost offset to the benefits derived from expanding the membership size, crowding leads to
finite memberships, a second characteristic serving to distinguish club goods from pure public goods.
For the latter, crowding costs are zero, and therefore new users can be included always at a net benefit
owing to reduced per-person assessments.
Because clubs are often exclusive, with finite memberships that are subsets of the population, the
disposition of the nonmembers of a given club is a third distinguishing characteristic of club goods For
pure public goods, all individuals can be members without crowding taking place, so that nonmembers
do not exist.
The entire population is in a single provision association for pure public goods. For club goods,
nonmembers to a given club have two options: They can join another club providing the same club good,
or they may not join any club offering the club good.
If all population individuals are allocated among a set of clubs with no overlapping or Non-assigned
individuals, the population is partitioned by the set of clubs. The number of clubs then becomes an
important choice variable. When, however, some individuals do not belong to any club supplying the club
good, then the population is not partitioned.
A fourth distinguishing feature of club goods is the presence of an exclusion mechanism, whereby users'
rates of utilization can be monitored and nonmembers and/or nonpayers can be barred. Without such a
mechanism, there would be no incentives for members to join and to pay dues and other fees.
The operation and provision of an exclusion mechanism, such as a turnstile or a toll booth, must be at a
reasonable cost. By "reasonable" we mean that the associated cost of the exclusion mechanism must be
less than the benefits gained from allocating the shared good within a club arrangement.
A fifth distinguishing attribute of club goods concerns a dual decision. Since exclusion is practiced,
members with user privileges must be distinguished from nonmembers.
Moreover, the provision quantity of the shared good must be determined. Insofar as the membership
decision affects the provision choice, and vice versa, neither can be determined independently. For pure
public goods, however, only the provision decision needs be considered - the membership is the entire
population.
The kind of clubs that we deal with are homogeneous clubs whose members have identical tastes and
endowments. If either tastes or endowments doffer than the club is called heterogeneous or mixed.
BASIC MODEL: Homogenous Clubs with Fixed Utilisation Rates
In the basic model, we assume the existence of two goods: a private good (y) and a club
good (X)
The homogeneous members possess the same tastes and endowments. A representative member's taste is
represented by a utility function,
where
yi is the ith member's consumption of the private good
X is his or her consumption of the club good, and
s is the membership size.
Since the utilization rate of the club good is the same for all members, we have xi = X for all members,
where xi is the ith member's utilization rate of the club facility, and X is the size of the club facility. Hence,
each member is viewed as using what is available.
The Utility function satisfies standard requirements: In particular, an increase in either good will augment
utility - the indifference curves will be convex to the origin in goods space. The utility function will be twice
continuously differentiable. This latter assumption allows us to take up to two derivatives; hence, s is
implicitly assumed to be continuous.
The marginal utility derived from additional members may be positive for small memberships, owing to
camaraderie, but eventually crowding will occur, and marginal utility will become negative.
Congestion therefore results in a decrease in utility as membership size expands beyond some point s
The existence of both a costless exclusion mechanism and congestion implies that the club good is not a
pure public good in the Samuelson sense, even though club provision is consumed equally by all
members.
Since all members are identical, and since club costs are equally shared among members, a
membership increment will reduce resource expenditures for each member — that is
In other words, each member must pick up a smaller share of the club's total costs as membership
expands, if other things remain constant.
The representative member maximizes utility subject to the budget constraint in (2). The
following first-order conditions result from this maximization:
Equation (3) is the provision condition for the shared good, and it indicates that for each member the
marginal rate of substitution (MRS) between the club good and the private good must be equated to the
marginal rate of transformation (MRT) between these two goods.
Thus, for the club good, members equate their marginal benefit with their marginal cost.
The i superscript denotes the individual, and hence MRT^ refers to the individual's marginal cost ratio
between the two goods. If, at the margin, the club is breaking even in providing the public good, the
sum of the members' marginal costs (or payments) must equal the club's marginal cost of provision
Equation (3) then indicates that the usual Samuelson provision condition for public goods holds for the
club good
In the basic model, the provision condition for the club good does not differ significantly from that for a
pure public good, except in terms of the number of individuals aggregated by the summation index and
the interaction of the provision and membership conditions.
The novel aspect of club analysis shows up in the membership condition, expressed in (4)
For within-club optimality, a representative member equates the MRS between group size and the
private good [left-hand side of (4)] with the associated MRT [right-hand side of (4)], thereby achieving
equality between the marginal benefits and marginal costs from having another club member.
These marginal benefits are normally negative owing to crowding, and the corresponding marginal
costs are negative owing to cost reductions derived from cost sharing.
Since, by assumption, a whole member must be added (or removed) from the club, the membership
condition may not be satisfied as an equality. That is, going from s to s + 1 members may reverse an
inequality between the two sides of (4).
When this discreteness problem occurs, members should be added, provided that marginal benefits
exceed marginal costs. The membership size prior to the reversal of the inequality is optimal.
A pure public good club could accommodate the entire population, because marginal benefits from
new members are zero and therefore are always greater than the corresponding negative marginal
costs. Consequently, pure public goods do not require a membership restriction.
For the provision condition, the MRS between the
member's resource constraint two goods is equated with the individual's share of
the marginal costs of provision. Cross-multiplying
by s in (6) gives the standard Samuelson provision
condition, equating the sum of the MRS values
and the marginal cost.
L = U(y, X, s) +λ (I – y – C(X,s)/s)
Optimal membership requires an equality between
𝜕𝐿 𝐶 the relevant MRS and the marginal costs of
= 𝑈𝑥 − 𝜆 𝑥ൗ𝑠 = 0
𝜕𝑋 increasing the membership size. The latter
includes increased maintenance fees (i.e., Cs/s)
𝜕𝐿
= 𝑈𝑦 − 𝜆 = 0 and reduced membership fees, owing to sharing
𝜕𝑦 [i.e., - C(.)/s2 ]
𝜕𝐿 In this alternative representation, full financing
= 𝑈𝑥 − 𝜆[{𝑠𝐶𝑠 − 𝐶 . ]/𝑠 2 = 0
𝜕𝑋 always results, since the budget constraint divides
the club costs among the members.
Then
𝑈𝑠
= 𝐶𝑠Τ
𝑠 − 𝐶(𝑋,𝑠)ൗ𝑠2
𝑈𝑦
LOCAL PUBLIC GOODS
The theory of local public goods differs in that goods are assumed to be specific to a particular
geographical location, and consumers, in deciding on their location, can exercise choice with respect to
the quantity and types of public goods provided.
For some public goods there may be no spatial restriction (for example, the benefits from research and
development); but for others the benefits, although available at no additional cost to new residents, are
confined to one community (possibly with some spill-over to neighbouring communities).
The construction of sea defences benefits those protected by the sea wall; the transmission of a television
programme benefits those within a certain distance of the transmitter.
we examine some of the implications of the local nature of such public goods and their provision by local
communities.
The mobility of individuals between communities supplying local public goods has a number of major
implications. It is in particular relevant to the problem of the revelation of preferences.
Much of the interest in local public goods was stimulated by the intriguing suggestion of Tiebout (1956)
that, if there were enough communities, individuals would reveal their true preference for public goods
by the choice of community in which to live (in much the same way as individuals reveal their
preferences for private goods by their choices).
Where there is a wide range of choice, all those deciding to live in the same community would have
essentially the same tastes, and there would be no problem of reconciling conflicting preferences.
Moreover, it is often asserted that such a local public goods equilibrium would be Pareto-efficient.
Just as the consumer may be visualised as walking to a private market place to buy his goods, . . . we
place him in the position of walking to a community where the prices (taxes) of community services are
set. Both trips take the consumer to the market. There is no way in which the consumer can avoid
revealing his preferences in a spatial economy. [Tiebout, 1956, p. 422]
This parallel ignores however certain key characteristics of local public goods. One of the most
important of these is the essential non-convexity associated with the provision of such goods
In the case of local public goods, non-convexities are inherent in that the cost of supplying a given
quantity of a public good (e.g., a local radio programme) to an additional individual is zero (in the pure
case).
In a local public goods equilibrium, there may well be fewer communities to satisfy the needs of all
consumers.
There are the issues raised by redistribution. The rich may attempt to segregate themselves from the
poor, in part because there is a large element of redistribution involved in the provision of education and
other services of the local community. By moving to their own communities, the rich can avoid this
redistribution.
Optimum Provision of Local Public Goods
For a pure public good that is not spatially limited (such as the benefits from research and development),
the issue of the number and size of communities does not arise. Where however the benefits from a
public good are spatially restricted, this issue is important.
As far as the public good is concerned, it is indeed natural to ask why there should be more than one
community. If the addition of a person does not detract from the benefits enjoyed by others, then—from
this point of view—the optimum allocation would involve everyone living in the same community.
Against this, however, must be balanced the diminishing returns to labour with a fixed quantity of land, or
the declining utility arising from congestion (e.g., as residential density increases). Moreover, for some
public goods congestion may set in beyond a certain size of community.
We focus on a single pure public good, considering the balance between the increasing returns inherent
in its provision and the decreasing returns to labour as the population within a community is increased.
We assume at this stage that all individuals are identical and examine the optimum allocation over a
number of identical communities.
Basic Framework
The model is a highly simplified one, in which total output, Y, in a community can be used either for
private consumption (X per person) or for the public good, G, in that community. It is assumed
that output is an increasing, concave function of the number of workers in the community, N:
On the assumption that everyone in the community is identical and is treated the same, the aggregate production
constraint gives:
For fixed N, this defines the consumption opportunity set illustrated in Fig. 17–1.
We assume that individuals have identical preferences,
represented by the utility function U(X, G)
L = U(X, G) + 𝜆 (𝑌 − 𝑋𝑁 − 𝐺)
𝜕𝑈
= 𝑈𝑥 − 𝜆𝑁 = 0
𝜕𝑋
𝜕𝑈
= 𝑈𝐺 − 𝜆 = 0
𝜕𝐺
𝑈𝐺 𝜆
=
𝑈𝑥 𝜆𝑁
which is the conventional result that the sum of the marginal rates of
𝑈𝐺 1 𝑁𝑈𝐺
= substitution equal the marginal rate of transformation (ΣMRS = MRT ).
𝑈𝑥 𝑁 or =1
𝑈𝑥
#Note MRT = Px /Pg and assumption made was that same good can be
used either as public good or pvt good, hence price Px = Pg hence =1
As we increase N, output, and hence the maximum level of public goods, increases (since f' (N) > 0) but
the maximum level of consumption per capita ( f (N)/N ) decreases.
The variable N opportunity locus is the outer envelope of the fixed N opportunity loci (Fig. 17–2)
𝜕𝑌
= 𝑋 = 𝑓′(𝑁)
𝜕𝑁
Now substitute for X
𝑓′(𝑁)N+G=f(N)
G = f – f’N
G = f – f’N
This condition has an interesting interpretation
Since f' is the marginal product of labour, f –Nf‘ is output minus wage payments if workers are
paid their marginal product.
Thus if the level of public exp is fixed, but pop is variable, the pop that maximises consumption per
capita is such that rents equal public good expenditure
This has been dubbed the “Henry George” theorem (Stiglitz, 1977), since not only is the land tax
non-distortionary, but also it is the “single tax” required to finance the public good.
COST BENEFIT ANALYSIS
In principle Cost Benefit Analysis (CBA) is straight forward. An investment project can be viewed as
representing a perturbation of the economy from what it would have been had the project not been
undertaken.
To evaluate whether the project should be undertaken we look at the consumption levels of
individuals of all commodities at all dates, under the two different situations. If all individuals are
better (worse) off then the project should be accepted (rejected).
If some individuals are better off and some worse off, adoption of the project depends on how we
weight the gains and losses of different individuals.
Depending on whether the individual is a selfish rationalist or a pure altruist he would consider net
gains to himself or may totally disregard gains to himself.
Substituting benefits for advantages and costs for disadvantages we can set up a very useful
proposition. This is that CBA merely formalises a common-sense concept of rationality. This
essence of CBA, however, is that it is not confined to decisions that affect one individual. It relates to
social decisions that affect a group of individuals.
Does the characteristic of rationality remain if we extend it to the social context? The basic argument
underlying CBA is that this rationality does remain. That is, if we leave a group of individuals to work out
their own personal CBAs we can simply aggregate the results to secure a social evaluation.
What is being aggregated over here is not a set of political votes but rather a set of money valuations – all
positive valuations indicate net gains while negative valuations indicate net costs to the person expressing
them.
To analyse the idea of “collective rationality” we should be able to see that adding up “money votes” in this
way effectively allows for the two types of ‘balance of advantage’.
Second, once these individual are recorded some will show net benefits and some will show net costs.
The essence of collective rationality argument is that it is legitimate to aggregate the individual money
votes in such a way that the difference between the votes in favour and the votes against defines the
concept of Net Social Benefit.
Individual Benefits Costs Benefits - Costs
Consider the following:
1 10 11 -1
2 10 7 3
3 12 2 10
4 7 10 -3
5 6 14 -8
Total 45 44 +1
Individuals 1 to 5 have various (arbitrary) benefits and costs arising from an individual project.
It can be observed that three of personal net benefit are negative and two positive. If each individual is
treated as a single voting unit, a simple majority rule would lead to a rejection of the project.
But using money values has led to the result that society appears “better off” by accepting the project.
An important principle being made use of in the above example is the Kaldor-Hicks compensation
principle.
If individuals paid out 12 units of their 13 units net gains to individuals 1, 4 and 5 they would still have 1
unit left over between them while individuals 1, 4 and 5 would have zero net gains – they would be no
worse off than they were before.
Political Votes and Economic Votes
The example above illustrates the distinction between one person one vote principle of
political voting and the kind of vote recorded in CBA.
Money values permits some expression of the intensity of preferences in the vote. It is this
aspect that gives economic voting an advantage over political voting in some cases.
It must of course be remembered that Arrow’s Impossibility Theorem holds true here as
well.
While individuals may all behave rationally, collecting their votes together and securing a
net overall result may contradict some principles of rationality.
Value Judgements and Cost Benefit Analysis
CBA deals with economic votes. Since these are expressed in the market place they reflect
individuals Willingness to Pay (WTP). WTP for benefits is a familiar concept though WTP for costs is
non-sensical. What is meaningful, however, is the idea of WTP to avoid costs that would otherwise
be incurred or being prepared to accept compensation for costs actually suffered. Thus economic
votes relate to two things:
(i) An underlying political vote for or against the project in question.
(ii) The power of the individual to pay or be paid
The second aspect demonstrates that economic votes are dependent upon the individuals ability to
pay i.e. upon his income and wealth.
This is most obvious if we think of WTP, but not quite so obvious id we think of compensation.
One would expect compensation to be linked to not income but the loss involved.
However, if we measure that magnitude in terms of the sum of money necessary to restore the
individual to her original level of welfare, this will in general be correlated with her income.
On the basis of the above we can establish 2 value Judgements of CBA
Both of these statements are likely to face considerable disagreement on the following count:
(i) A decision which reflects individual preferences is a good decision but may not be agreeable to all. It can be
argued that some decision may need to be made ‘’on behalf of’’ the people since they are not always the best
judges of what is good for society.
(ii) A voting system in which those with higher income have a greater say poses even more problems. This is
clearly offensive to egalitarians. A change in (ii) will quite possibly change the outcome of our CBA. It says
that the prevailing income distribution should be used when evaluating projects. It is open to vary it and ask
what would happen if we evaluated the project as if the income distribution were different. It is clear in any
case that some judgement about a desirable income distribution is implicit in any CBA
Investment Criteria
The principle of CBA is to weigh up the advantages and disadvantages, the costs and benefits of any project.
Ideally these costs and benefits are expressed in money terms.
Each individual cost and benefit item i will have a quantity qic and qib and a price pic and pib . In any one year
costs are given by
𝑐 𝑐
𝐶𝑡 = 𝑞𝑖,𝑡 . 𝑝𝑖,𝑡
And benefits by
𝑏 𝑏
𝐵𝑡 = 𝑞𝑖,𝑡 . 𝑝𝑖,𝑡
We could simply add up the costs and benefits upto some time period T, which is called the time horizon.
T is set most frequently by the economic life of the investment which will in general be shorter than its physical life.
If we add up the costs and benefits we could obtain an expression for NET BENEFITS
𝑁 𝐵 = [𝐵𝑡 − 𝐶𝑡 ]
𝑡=0
However, this expression neglects an important factor. If we adopt the basic value judgement that
individual preferences ought to count we must incorporate the fact that individuals may prefer present
consumption to future consumption i.e. individuals will prefer a unit of benefit now rather than later even
if the existence of the benefit in the future is certain.
They discount the future. They are said to have a ‘time preference’ and the rate at which they discount
the future is called the ‘marginal time preference rate’ (MRTP)
Thus the expression for N(B) is more accurately written as:
1 1 1 1
N(B)= 𝐵0 − 𝐶0 [ ]+ 𝐵1 − 𝐶1 + 𝐵2 − 𝐶2 + ⋯…+ 𝐵𝑇 − 𝐶𝑇
1+𝑟 0 1+𝑟 1 1+𝑟 2 1+𝑟 𝑇
More compactly,
𝑇
1
𝑁 𝐵 = (𝐵𝑡 −𝐶𝑡 )
(1 + 𝑟)𝑡
𝑡=0
𝑇
or
𝑁 𝐵 = (𝐵𝑡 −𝐶𝑡 )𝑑𝑡
𝑡=0
Hence, net benefits are to be now understood as ‘net benefits as viewed from the present standpoint’
which we call NET PRESENT VALUE
Three contexts arise within which the NPV criterion must be used to evaluate projects:
(i) Accept-Reject: Here the agency must decide whether a given project is to be accepted or not
(ii) Ranking: The agency may have a series of investments all with positive NPVs. How does it
rank them?
(iii) Mutual Exclusion: The agency may have to decide between two projects simply because
undertaking one may mean that the other cannot be undertaken
(i) Accept-Reject: Faced with a single project, the NPV rule dictates that it should be accepted if
NPV > 0, and rejected if NPV < 0. If NPV = 0 the agency would be indifferent.
(ii) Ranking: In the context of unlimited funds all projects with NPV > 0 should be accepted. However,
the more realistic context is always one of rationing. In such a case the agency has to rank projects
in order of desirability and work down the order till the budget is exhausted. However, the ranking
should not be done on the basis of NPV alone
If a capital constraint of 100 units exists NPV ranking of X,Y,Z would result and only X would be chosen
with Net benefits of 100. However, it is apparent that both Y and Z can be chosen within the constraint
of 100 units and this would yield joint benefit of 130.
The final column seems a better way of ranking projects: ranking by benefit-cost ratios
(iii) Mutually Exclusive Projects
Where the choice is between projects, the general rule is to select the project offering the highest
NPV. However, it should be clear whether ‘exclusivity’ between X and Y is strict of some
combinations of X and Y are permissible.
These combined projects should be treated as if they were extra projects. In any case the rule of
selecting the highest NPV applies.
The calculation of NPV requires the use of some social discount rate r which is assumed to be derived
independently.
An alternative approach to investment appraisal is IRR and compare it directly to the social rate of
discount. The IRR is in fact another name for Keynes’ MEC.
It is calculated by setting discounted net benefits stream equal to the initial capital outlay and solving
the equation for the value of the discount rate. That is…
𝑇
1
𝐵𝑡 − 𝐶𝑡 . = 𝐶𝑜
(1 + 𝑟)𝑡
𝑡=1
𝑇 𝑇
1 1
or 𝐵𝑡 . = 𝐶𝑡 .
(1 + 𝑟)𝑡 (1 + 𝑟)𝑡
𝑡=1 𝑡=0
The rule for accepting a project is accept if r* > r where r* is the IRR and r is the
predetermined discount rate. If r* < r reject the project
Multiple Roots Problem
In computing IRR it is quite possible to obtain more than one solution rate.
This is true because the equation to be solved is a polynomial: if the polynomial is of degree n, there will
be n solution rates.
Therefore if a projects’ IRR has two solutions say 10% and 15% while r =12% there appears to be no clear
cut criterion for acceptance or rejection
The only roots of a polynomial equation which are of interest are those with positive and real values.
If the decision formula is expressed as an equation in I, the IRR, the positive roots can be indicated. Thus
in the simple 2 period case:
𝐵1 1+𝑖 +𝐵2
𝐵1 𝐵2 = k((1 + 𝑖)2
(1+𝑖)2
+ =𝑘
(1+𝑖) (1+𝑖)2
𝐵1 1 + 𝑖 + 𝐵2 = k(1+2i+ 𝑖 2)
−𝑘𝑖2 + 𝐵1 − 2𝑘 𝑖 + 𝐵1 + 𝐵2 − 𝑘 = 0
𝐵 1 + 𝐵 1i + 𝐵 2 = k + 2i + ki 2
−𝑘𝑖2 + 𝐵1 − 2𝑘 𝑖 + 𝐵1 + 𝐵2 − 𝑘 = 0
So that the sequence of signs before the terms in i is - , + There is only one change of
sign so there is one positive root.
If the sequence of signs were -,-,+,+ one would still have one positive root
However, the sequence -,+,- has two changes and hence two positive roots.
This sequence is perhaps most appropriate for public investment projects: initial expenditure
(-), followed by positive returns (+) and then negative returns (-) as the project ages, possibly
costing money to dismantle and scrap.
In the case where the sequence is -,+,- some rules have been suggested by which valid IRRs can be
distinguished from ‘invalid’ ones
Suppose we have 2 positive roots i1 and i2 with i2 > i1, formulate the following rules:
(a) i2 < e < i1 accept the project NPV
(b) e < i1 reject the project
(c) e > i2 reject the project
where e is the social discount rate. Rule (a) can be shown diagrammatically NPV
i2 i
i1
SHADOW PRICES
Thus far we have assumed that existing market prices (weighted, where necessary, for
distributive reasons) were the appropriate valuations to use when performing social CBA. This
requires:
(i) Item in question was one to be judged in terms of individual preferences as revealed in
the market
(ii) Individual was free to choose how much to consume of the good at a fixed market price
With respect to inputs one requires that the price of the input in question was equal to the
value of its marginal product.
When these assumptions are not fulfilled market prices will no longer necessarily be
appropriate valuations to use in appraisals.
It is then necessary to replace market prices by ‘’shadow prices’’ which reflect the social value
of outputs and inputs concerned.
A shadow price emerges as a marginal valuation imputed to an input or an output at the optimum in an
optimisation problem. At the optimum activities will only be undertaken if the value of their outputs is at
least as great as the cost of inputs measured in shadow prices.
In the sense just mentioned shadow prices exist for all inputs and outputs whether traded or not and
where some combination of market mechanism and government intervention is functioning well this
shadow price will be reflected in the current market prices.
In CBA the term ‘shadow price’ is reserved for a price that is imputed as opposed to being taken directly
from market transactions, whether this is because no market price exists or because market price is
considered to be inappropriate. Nor is the term restricted to valuations in the neighbourhood of the overall
optimum; where constraints render optimisation infeasible, it is used for ‘second best’ valuations too.
The use of market prices for valuation purposes will be improper if:
(a) Market prices are not equal to MCs
(b) MC does not reflect the true social cost of resources
In an imperfectly competitive world prices will tend to be above MC and external effects will exist, so that
(a) and (b) pertain. The sources of these divergences are:
(i) Imperfect Competition: In product market under market imperfection P> MC. Therefore market
prices will overstate the proper shadow price to be applied to output. With imperfection in factor
markets, factor prices = MRP < VMP Consequently the market price of a factor will not be an
appropriate shadow factor price
(ii) Unemployment of Resources: If a public project employs a resource which would otherwise have
been unemployed, the true cost to society of employment is (virtually) zero. In recession the shadow
cost of labour is closer to zero than it is to the wage actually paid.
(iii) Increasing Returns: The existence of IRS will mean that MC < AC. Therefore if P=MC evaluation
of benefits at market prices will involve losses. But losses have to be financed and methods of
financing them have welfare effects
(iv) Taxation: Market prices will contain elements of indirect taxation or may be below true cost if a
subsidy is involved. To correct for any biases in valuation all output should be valued, net of indirect
taxes and subsidies i.e. at factor cost
(v) External Effects: External benefits and external costs are not priced in the market. Consequently,
ignoring external benefits will lead to the benefits of the project being undervalued; costs will be
undervalued if external costs are present
(vi) Public Goods: The benefits of public goods cannot be restricted to particular groups by pricing
policies. If benefits are yielded, those who receive the benefits as a by-product of the supply to a
particular section of the population will, if they act rationally, understate their true valuation of the benefit
knowing that they will obtain the product free as long as it is supplied to at least one person. Clearly the
market price will not be a valid indicator of the social benefit in this case.
In the presence of imperfection this rule is modified so that benefits are valued at marginal social costs.
It must ne remembered that this rule will be relevant for policies designed to move an economy in the
direction of a Pareto social optimum.
However, a problem arises if prices are not equated with marginal social costs throughout the
economy. If marginal social cost pricing is adopted in the public sector but not in the private sector will
the economy be nearer a Pareto optimum than if neither sector used the appropriate shadow prices?
It has been suggested that marginal social cost pricing in this context may move the economy away
from a Pareto optimum i.e. given that a ‘first best’ is not obtainable, it is not possible to state what
pricing rule in the public sector will take the economy as far as possible towards a Pareto optimum.
This is the essential point of Lipsey and Lancaster’s theory of Second Best.