Common Property, Public Goods
Common Property, Public Goods
1
Introduction
• Goods can be classified into different categories based on two different properties.
• A good has the property of rivalry if the consumption by one person precludes consumption by other
people.
• A good has the property of excludability if individuals can be prevented from consuming the good.
Excludable Non-Excludable
Rivalry Private Good: Common Property:
Apple, Computer Park, Fishery
No Rivalry Club Good: Public Good:
Concert, Cable TV Clean Air, National Defense
2
Introduction
• In HE1001, HE2001, we have mainly focused on private goods, which are rivalrous and
excludable.
• In HE3001, externalities of private goods, which we saw last lecture, can often be conceptualized
as working through another non-excludable good.
• E.g. Cigarettes and clean air, Fishing and common sea resources…
3
The Tragedy of
Commons
4
The Tragedy of Commons
• Consider a grazing area owned “in common” by all members of
a village. Villagers each have 1 cow which they can graze on the
common.
5
The Tragedy of Commons
• Production via access to the common is non-excludable, but rivalrous.
• Note: the grazing of an additional cow reduces the milk productivity of other cows grazing.
• because given the previous conditions.
6
The Tragedy of Commons
Why is ?
7
The Tragedy of Commons:
Social Planner’s Problem
How should the village graze their cows so as to maximize their total income?
8
The Tragedy of Commons:
Individual Problem
How should a villager decide whether to graze a cow to maximise personal income?
• As long as , there will be a villager who will want to graze an additional cow.
9
The Tragedy of Commons:
Individual Problem
• Since nobody owns the common, entry is not restricted, and entry occurs until
• Intuitively, this is where the individual economic profit of grazing another cow is 0.
• More cows are grazed in the individual optimum than the social optimum.
• At the social optimum , the economic profit per cow is
• This is because . Hence .
• The individual economic profit from introducing an additional cow is positive at the social
optimum , which implies that the number of cows grazed in equilbrium when villagers are
individually optimizing is greater.
10
The Tragedy of Commons (Graphical
Solution)
11
The Tragedy of Commons
• What is the intuition behind this?
• When a villager adds one more cow, his personal income rises but every other villager’s income falls. I.e.
the villager imposes a negative externality on others, but doesn’t take that into account.
12
Tragedy of Commons, Discussion
• This holds in general: common property leads to a “structure” where externalities are imposed on
others from private use.
13
Review Questions 1
Suppose that there is a wi-fi access point with a capacity of 10 people. Because of bandwidth
limitations, after 10 people, wi-fi speed starts to get slower for each unit of usage. Each connected
person decides how much w-ifi to use.
• Suppose there are 5 people who are sharing the a wi-fi access point. Will there be a tragedy of
commons (in wi-fi usage)?
1. Yes
2. No
• Suppose there are now 15 people who are sharing the a wi-fi access point. Will there be a tragedy
of commons (in wi-fi usage)?
1. Yes
2. No
14
A simple model of
Public Goods
15
Public Goods
• A public good is non-excludable and non-rivalrous in consumption.
• As long as 1 individual chooses to consume (provide) the good, it is not possible to exclude others from
consuming it.
• It is also non-rival in the sense that everyone will consume the same amount.
This causes certain problems for the decentralised market …
• To this end, let us consider consumers’ reservation prices of the public good,
16
Public Goods
• Suppose we have 2 individuals A and B who have to decide whether to contribute to providing a
public good.
• Assume that the cost of provision is The public good will only be provided if their contributions
exceed the cost of provision.
• (Contributions are assumed not to be refunded in the case of failure.)
We will now use game theory to see some problems with privately providing the public good.
17
Scenario 1
• Suppose and . Then A would supply the good even if B made no contribution.
• B then enjoys the good for free. This free riding can lead to inefficient provision.
• Let , , .
Suppose they have to decide whether to individually contribute $50 to providing the public good.
Player B
Contribute Don’t Contribute
Contribute
Player A Don’t
Contribute
• Mixed strategy NE here which implies a positive chance of (Don’t contribute, Don’t contribute) .
• Even though both and , the inefficient outcome of no public good might still result.
18
Scenario 2
• Suppose , but and . Here, they will not individually contribute.
Will sharing the costs help?
• Let , , .
Suppose they have to decide whether to equally share the costs of providing the public good
(The public good is only provided if both of them contribute their share.)
Player B
Contribute half Don’t Contribute
Contribute
Player A half
Don’t
Contribute
• Two pure strategy NE here: (Don’t contribute, Don’t contribute) & (Contribute, Contribute)
• The need for both to contribute may hinder provision of the public good even when it is efficient: .
19
An extended model of
Public Goods
20
Variable Public Good Quantities
• Now, let us consider a case where we can have variable amounts of the public good.
The cost of providing amount of a public good is .
• Each individuals has wealth given by and preferences which are a function of private
consumption and public goods consumption .
21
Social Optimal Outcome
• Any pareto efficient allocation can be shown to satisfy the following condition:
Where is the marginal rate of substitution between the private and public good for .
• tells us how much one marginal unit of the public good is worth in terms of the private good
22
Social Optimal Outcome
Why is this condition optimal?
1. If ,
• Reducing the public good by 1 unit.
• The decline in utility is worth of the private good.
• The cost savings is .
• Use the cost savings to compensate each individual with of the private good.
• Will have left over!
• Can redistribute left over money to everyone and have a possible pareto improvement!
Too much of the public good here..
23
Social Optimal Outcome
Why is this condition optimal?
2. If ,
• Reduce each individual’s private consumption by .
• Increase the public good by 1 unit which costs . Utility will be exactly the same as before.
• But will have left over!
• Can redistribute left over money to everyone or buy more public goods and have a possible pareto
improvement! Too little of the public good here..
24
Social Optimal Outcome (Quasi-
linear)
• Suppose that preferences are quasi-linear .
s.t.
Why? From HE2001, we know that quasi-linearity implies that the pareto efficient allocation with money
transfers is the one which is utilitarian.
25
Private outcomes
• Suppose that each individual decides on how much public good to purchase
• The total public good is just the sum which each purchases:
• For simplicity, assume fixed marginal costs of contributions:
s.t.
• This implies that and there is too little of the public good.
• This is because investing in public goods generates positive externalities on others which individuals do
not account for. Positive externalities also means that individuals want to free-ride on others!
26
Review Questions 2
Consider a public good with cost function . There are 2 individuals with general preferences . They
each contribute amount of the public good and privately consume the left over .
1. Suppose the level of contributions are such that and . Is this pareto efficient?
1. Yes
2. No
27
Mechanisms for Public
Goods Provision
28
Alternative Mechanisms
• To the extent that public goods remain non-excludable, given the externalities and free riding
incentives involved, private market mechanisms are often inefficient.
How may such a mechanism decide on how much/ whether to supply a public good?
29
Demand Revelation
• The previous discussions imply that we need to know their marginal valuations to efficiently
supply a public good!
• How do we do so? If we naively just ask them, people might have incentives to inflate their
valuations of the good to get more of it.
• A scheme that makes it rational for individuals to reveal truthfully their private valuations of a
public good is a revelation mechanism.
30
The Setup
• N individuals; . All have quasilinear preferences.
For simplicity, we consider the case of a discrete public good:
(don’t provide) vs (provide).
31
The VCG mechanism
When is the good provided?
• Everyone is asked to report their true value . Individual report is called .
• If , then the public good is provided.
32
The VCG mechanism: Taxes
• We call an individual pivotal if
and or
• I.e. with his inclusion, the supply decision changes from what the others would want to do (on
average) based on their reports.
33
The VCG mechanism: Taxes
• Suppose
Then, if was like others on average, then public good would not be provided.
• If including causes the public good to be provided.
This goes against what the others want and generates an aggregate loss on them.
• Suppose
Then, if was like others on average, then public good would be provided.
• If including causes the public good not to be provided.
Again this, goes against what others want and generates a loss on them.
34
The VCG mechanism: Taxes
• Everyone is taxed an amount equal to the loss they inflict on others this happens when
a person is pivotal.
35
The VCG mechanism: Payoffs
• Suppose other agents favour the public good:
36
The VCG mechanism: Payoffs
• Suppose other agents do not favour the public good:
37
The VCG mechanism
• The blue terms are the portion of utility for which agents have control over (through their
report).
Given the messages from other agents , agent ’s message potentially determines the value
of .
38
The VCG mechanism
• Expanding the blue term gives:
• Notice that in order to maximise utility, it would be best to report the truth:
• If, a report would result in provision of the public good resulting in this positive utility.
• If, a report would result in non-provision of the public good resulting in the avoidance of
this negative utility.
39
The VCG mechanism
• Reporting is thus a (weakly) dominant strategy.
• In this Nash Equilibrium, since the public good is provided when , and everyone is
telling the truth, the mechanism will indeed provide the public good if it is efficient to
do so.
• Notice that the VCG mechanism like an auction: messages are like bids and these result
in allocations of the public good.
• Truthfulness arises from a mechanism similar to that of the second price auction.
40
Remarks on the VCG mechanism
• The VCG implicitly requires quasi linear preferences as we can’t have payments influencing
demand for the public good.
• While it guarantees the efficient provision of the public good, there is still resource inefficiency
due to taxes removing wealth from pivotal individuals
• These taxes cannot be fed back into the system so as to maintain incentive compatibility.
• There is thus a drop in private consumption.
• This result is more general: there are no processes which are simultaneously dominant
strategy incentive compatible and have no resource inefficiency over a reasonable range of
possible environments. (even under quasi-linear utility.)
• See the textbook for more discussions of weaknesses: collusion and equity.
41
Review Questions 3
The government is considering providing a public good for 3 users, I, J and K using the VCG
mechanism. The cost of the public good is 9.
2. Suppose they all report their true values. Select all agents who are pivotal.
a) I
b) J
c) K
42
References
• Varian Chapter 35
• Varian Chapter 37
43