Midterm
(You may, but need not, leave any
complete question or questions blank and receive 20% credit)
I.
Briefly explain the idea of rent seeking. [3] (10
points)
Spending resources transferring value from someone else to
oneself.
.
A.
Give two examples, with no more than a sentence or two of explanation for each.
[4]
Burglary--the
benefit to the burglar is at the expense of the victim, and the burglar spends
time and effort burgling.
Lobbying for government policies that favor your interest group
at the expense of others.
B.
Why is rent seeking undesirable? [3]
The gain is cancelled by the other partyÕs loss, and the
resources spent on creating (or preventing!) a transfer are a net loss.
II. Externalities (15
points)
A. What is an external cost?
A cost due to one personÕs action that is born by someone else.
[Could expand this to distinguish costs born by others but only with their
consent]
B. Why are external costs a problem?
Because the actor does not take them into account in making his
decisions. He takes acts where the net benefit to him is positive, but since he
is not including the external costs, some may be acts where the net benefit to
all concerned is negative.
Some students seem to think that the
problem is that the actor doesnÕt know about the external costs. But even if he
knows, he still ignores them, since they are not costs to him.
C. What is the Pigouvian approach to dealing with
externalities? What are its advantages over direct regulation?
The Pigouvian approach is to charge the actor the amount of the
external cost he imposes. Unlike direct regulation, this takes advantage of the
private information of the actor as to ways of reducing the externality and
their costs. If he can reduce it at a cost less than damage done it is in his
interest to do so, if not not. The regulator must be able to measure the damage
but does not have to know how, or if, it can or should be reduced.
D. Give an example of a situation where applying the
Pigouvian approach makes the outcome less efficient than doing nothing.
My factoryÕs air pollution does $100,000 a year of damage to
people living downwind of me. I could eliminate it at a cost of $80,000/year,
and will if charged an effluent fee of $100,000/year. But the people living
downwind from me could move, converting their land to a use not affected by
pollution, at a cost of $60,000/year, and would if there were no effluent fee.
[This assumes that transaction costs make it impractical for me to buy them out
and then point out to the regulator that there is no longer anyone living
downwind, hence no damage to fine me for.] Lot of other examples are possible,
of course.
Mistakes: Thinking that the
advantage of Pigouvian tax is that the company can keep polluting.
III. Coase (10
points)
A. Sketch a Coaseian approach to designing legal rules.
Consider possible rules (who has the right to do what, property
right vs liability right, É), possible outcomes (who ends up doing what), and
how you would get from each rule to each outcome. Choose the rule that
minimizes the combined cost of getting to the efficient outcome plus the cost
of inefficiencies due to not getting to the efficient outcome, given your
estimate of how likely it is for each outcome to be the efficient one.
Mistake: Answering a different
question--explaining CoaseÕs view of externalities, not its implications for
designing legal rules.
B. Give one example of a situation where the move to an
efficient outcome is blocked by a transaction cost.
Factories are allowed to pollute. The efficient outcome is for
them not to--the cost of their preventing the pollution is less than the damage
done or the cost of the victims
preventing it. But the victims canÕt arrange to pay the factories to stop
polluting because of the public good problem faced by the (many) victims in
collecting the money.
IV. Risk. Briefly explain two
of the following ideas, giving one example of its relevance either to designing
an insurance contract or to choosing a legal rule. (10
points)
1. Risk aversion
Due to declining marginal utility of income, an individual
prefers a certain outcome to an uncertain outcome with the same expected value.
This provides a reason to choose an ex ante punishment with a
high probability of a low fine to an ex post punishment with a low probability
of a high fine.
2. Moral Hazard
Insurance transfers at least part of the cost of some adverse
outcome, such as my factory burning down, to someone else, reducing the cost to
me. The result is that I may fail to take some precautions that are worth more
than they cost, since I pay the cost and get at most some of the benefit.
One implication is that the insurance company may choose to
require some precautions, such a an alarm system, as a condition of insurance.
Another is that it may offer to insure for only a fraction of the value of the
factory (co-insurance), leaving the owner with an incentive to at least take
precautions whose benefit is considerably larger than their cost. [One
implication is sufficient]
3. Adverse Selection
One party to a transaction has information relevant to the
terms of the transaction that the other does not have, and cannot believably
share it--for instance the seller of a used car who knows its condition. The
better the condition of the car, the less likely the seller is to accept any
given offer for it, so the average condition of cars sold is lower than that of
cars offered for sale—accepting the offer is a signal that the car is in
poor condition. Buyers, recognizing that, adjust their offers accordingly. The
result is that good cars may not sell, even though they are worth more to the
buyer than to the seller.
One implication is that the parties may try to design the
contract to put those costs affected by the asymmetric information on the party
who has the information, for instance by the seller of the car offering a
guarantee. Another is that one may want to avoid legal rules, such as a rule
forbidding an insurance company from demanding genetic tests, which created
adverse selection. [One implication is sufficient]
V. Pick one of the following and explain it: (10
points)
A. Hawk/Dove equilibrium. How
would you calculate the percentage of hawks (bullies) in equilibrium, assuming
you knew the payoffs to each interaction (dove meets dove, dove meets hawk,
hawk meets hawk). What does this have to do with the question of whether crimes
of passion, such as killing someone in a blind rage, can be deterred?
The larger the fraction of hawks, the more likely a hawk is to
meet another hawk, which is a cost of being a hawk, and the less likely to meet
a dove and get the food without fighting, which is a benefit. As long as
benefit is larger than cost, the number of hawks increases, so equilibrium is
reached when they are equal--when the average payoff to being a hawk is the
same as to being a dove. In the human case, this implies that increasing the
penalty for crimes of passion, thus increasing the cost of the hawk-hawk
interaction, will push down the fraction of hawks--fewer people will commit to
an aggressive personality.
B. Prisoner's Dilemma: Explain
the argument for why the existence of plea bargaining might make criminals
worse off. Would the argument still hold if transaction costs among defendants
were zero? Explain.
[Answer I was looking for] By getting some defendants to cop a
plea, the prosecutor frees up resources that let him do a better job of
prosecuting those who donÕt, raising the chance of convicting them and thus the
risk to those who donÕt cop a plea. The result is that the plea accepted, although
more attractive than the result for that defendant of going to trial (which is
why he accepts), may be less attractive than what the result of trial would be
if all cases went to trial and the prosecutor thus had fewer resources for
each. With zero transaction costs, the defendants would all agree not to accept
a plea bargain, so the argument would no longer hold.
[Alternative legitimate answer] The standard prisonerÕs dilemma
can be interpreted as involving plea bargaining. The prosecutor offers each prisoner
the choice of going to trial or accepting a lower sentence by pleading guilty
and confessing--thus providing evidence against the other prisoner. If neither
prisoner confesses, the prosecutor can still convict them of some lesser
offense and punish it; if only one confesses, the prosecutor offers him a
sentence lower than that but will press for a severe sentence for the other. If
both confess, both get a sentence somewhat lower than that imposed on one if
only the other confesses. Hence either prisoner is better off confessing,
whatever the other one does, making them worse off than if neither confessed.
If transaction costs were zero, the two prisoners would agree that neither
would confess.
Mistake:
Some students thought the problem is that each prisoner does not know what the
other will do, or does not trust the other to not confess. But if the other is
not going to confess, you are better off confessing.
C. My neighbor sues me because noise from my factory
makes it hard for him to sleep. The court agrees that my factory is a nuisance
and must decide whether to award damages or grant injunctive
relief—forbid the factory from operating unless it either eliminates the
noise or has my permission. Discuss the advantages of each. What circumstances
might be relevant to deciding which is preferable?
The advantage of damages is that no negotiation is necessary
between me and my neighbor--I shut down the factory or install sound insulation
if doing so is less than the cost of the damages, which is the efficient choice
(provided the court has the damage right). Otherwise I just pay. The advantage
of the injunction is that it does not depend on the court getting the damages
right. Absent transaction costs, you offer to let me continue making noise if I
pay you enough, and I end up shutting down or continuing to operate according
to whether the cost of preventing the nuisance to me is less or more than the
damage it imposes on you, which is the lowest payment you will accept.
So without transaction costs and with a correct court decision
we get the same outcome either way. Which is better then depends on the
transaction cost and the ability of the court to estimate the damage done by
the nuisance. The easier it is for the court to estimate and the harder it is
for us to bargain successfully, the more attractive the damages rule.
In particular, if the cost to me of shutting down is much
larger than the damage I am doing to you, our bargaining is a bilateral
monopoly bargain with a very large bargaining range; you can try to hold me up
for almost the entire cost of shutting down, while I try to get you to accept a
little more than the damage my nuisance does. Such a bargain is likely to have
high costs, so a large difference is an argument for liability and against
injunction.
Mistakes: Some students ignored the
possibility of bargaining, and so thought the advantage of damages was that the
factory could keep operating, the advantage of injunction was that it would
have to shut down or control the noise.
No. See Chapter 9, starting on page 98, which is on this
subject.
Set the fine for tortious loss of life at a level corresponding
to the value of the life in question, including its value to the
victim—the amount he would pay to reduce his chance of dying by some
small probability divided by that probability—plus its value to others.
Allow him to adjust the amount his heirs will actually receive to the optimal
insurance level by buying or selling insurance—where selling insurance
means selling the right to collect part or all of the damages if he is tortiously
killed.
Mistakes: Almost all students who answered this apparently
believe that peopleÕs lives are worthless to themselves—at least, the
elements that were proposed to go into value of life were entirely values to
other people.
Many students missed the fact that the question was about
tortious loss of life and treated it as a question about murder; murder is a
crime not a tort. They thus argued that the penalty should be set at a level
that would deter all such cases. As we will see later, the calculation of the
optimal criminal penalty does indeed depend on both how much deterrence you get
at different levels and the damage done. But in our discussion so far, damage
payments have been treated as Pigouvian taxes, set equal to damage done.
Many students seemed to think that the optimal level of
insurance was equal to damage done—the level at which the heirs would be
indifferent between the victim being alive, and his being dead but their
receiving compensation. The explanation of why this is not true starts on page
100, in the context of damages for injury. In the context of loss of life,
consider that once the family has been compensated for the financial cost to
them, their marginal utility of income will be back at about its previous
level. Additional compensation would transfer to them additional money to make
up for non-pecuniary costs, pushing the marginal utility of money lower. If
they were buying insurance, they would not choose to transfer money from a
higher to a lower valued use.