Tools: Bounded Rationality, Incomplete Contracting, Moral Hazard
To date we have taken apart the firm and examined where institutions offer
no advantage over bargaining by individuals, or relying on the price mechanism.
We have seen two early explanations for institutions we call firms.
Coase: characterized as the absence of the price mechanism. Firm exists
when it is less costly to direct economic activity by authority.
Alchian and Demsetz: classical firm arises because of a problem of metering
production. Firm exists when team production makes it difficult to compensate
Moving forward, we are going to look at how incomplete contracting, bounded
rationality, private information affect things. Theories which explain
institutions as a result of contracting problems.
contract: an agreement through which trade is mediated. (does not
have to written down)
complete contract: agreement which specifies the actions parties
must take and transfers which take place under each possible contingency.
Arrow-Debreu environment is one of complete contracts -- all markets exist
for all possibly state-contingent goods.
Imagine an employment contract that is complete...
Think about a rental agreement.
conditional on things breaking down, noisyness of neighbors, construction
in neighborhood, how many sunny days, specifies what maintenance can be
done and how fast.
bounded rationality: because contracting parties are unaware of
all possible states of nature, and know that they unaware of them, they
are assumed to act in a boundedly rational way. They maximize their
utility conditional on their uncertainty about the future, knowing that
events they may not have accounted for may occur.
(When I agreed to become a UCLA professor, I know that I was not accounting
for all possible events that might affect my utility. For example, I did
not know if I could find a nice apartment near campus. Or, I did not know
or account for the extent to which the Simpson trial would affect my life.)
Almost no contracts are complete, both because of the costs of complete
contracting, and because of bounded rationality.
Economic agreements are almost always incomplete contracts, in which
actions and transfers are unspecified for at least some (possibly unforeseen)
Furthermore, enforcement problems can make actions and transfers under
some contingencies indeterminant -- even though they are stated in the
The fact that contracts are incomplete allows for opportunistic behavior,
[Incompleteness leading to "anticipated opportunism" often arises
because it is costly for some party to verify the true state of the world
-- asymmetric information. Rather than devoting resources toward
learning the true state of the world, we often rely on incomplete contracts.]
opportunism: self-interest with guile.
Usually, actions taken (or not taken) which affect the values of the contract
to the other parties adversely.
Opportunism is, of course, accounted for (anticipated) when contract is
agreed upon. It represents welfare losses because parties are unable to
commit not to be oppotunistic.
contractual renegotiation (sports)
fraud (auto repair)
shirking (labor contract)
withholding/manipulating information (applying for insurance)
Opportunism often arises because of the private information held
by one or more parties.
can be less trade than optimum
can be resource misallocation
This can be information about ones own attributes, the true state of the
world (inside info), one's own actions, etc.
The task is to design institutions that are "incentive-efficient"
We know that we can't get "first-best" because of contracting
problems such as bounded rationality and private information.
We want to devise institutions that do better under these conditions than
one-to-one bargaining or relying on the market price.
We saw some of this in Alchian and Demsetz.
Workers have private information about their own effort levels which firms/managers/owners
cannot costlessly observe. Institution arises which mitigates this.
Parties seek institutional arrangements which limit the costs associated
with incomplete contracting.
Moral Hazard: post-contractual opportunism.
contract actions uncertainty payoffs
consider insurance, employment examples
1) you sign a contract with a company that promises to fully compensate
you if your car is damaged or stolen. You, in exchange, pay a fixed fee.
2) You then choose where to park and how to drive your car.
3) uncertainty: accident? car stolen?
4) Payoff: you give insurer x if it is not damaged or stolen (your premium);
insurer gives you y if it is.
Contract does not fully resolve incentive conflict. Contracting on outcomes
rather than (unobservable) effort.
Incentive conflict becomes a problem because you bear the full costs of
any effort toward care in driving, etc., but you do not reap the full benefits.
Therefore, you will take less care.
1) you agree to supply labor/manage a 7-11.
2) you decide how hard to work, and toward what ends.
3) randomness: how much did people spill, how busy was the store, demand
for the goods they suppply.
how often to clean the condiment stand
whether to sell cigarettes to minors
how often to stock shelves
how friendly to be to customers
4) output: sales
If agent's actions are costlessly verifiable to a 3rd party, no moral hazard
problem. (But who monitors the monitor? How can we be sure the monitor
is trustworthy and not colluding?)
Moral hazard is one example of a principal-agent or agency
problem. One individual must induce another to at in their interest when
performance is not perfectly observable.
moral hazard within firms:
incentives of owners and workers are not perfectly aligned because workers
bear the full cost of their effort but to not get full benefits.
since effort is costly, and there is opportunity for workers to divert
firms' resources for private gains, there are moral hazard problems.
ex: workers shirking, taking office equipment for personal use, managers
diverting resources to promote own department, using company planes for
any activity that diverges from profit maximization.
Institutions designed to limit moral hazard:
you also need the consequences you threaten to be credible. For example,
if labor markets are perfectly competitive, firing someone does not affect
them -- they just instantly find another job at the same wage.
monitoring, as in A&D. Monitoring within the firm, and by
the market. Since monitoring is costly, there is a trade-off -- monitoring
will generally not be perfect.
Problem: third-party enforcement, and agreement must be verifiable or else
incentive to renege.
bonding: individuals may agree to post a bond which they forfeit
if they do not abide by the contract's terms. (employment example...corporations
paying for employee's schooling -- employee must pay if early quit)
exposes individuals to risk.
payment on output (observable) rather than labor. Piece rates.
if not done right, could create perverse incentives....
(Why would you not pay a basketball player a function of the amount of
points he or she scores? Misallocation of effort in games and practice.
Incentive for coach not to play him in some circumstances where it would
be efficient for him to play.)
(What about paying golfers a function of their score? Looking narrowly,
the incentive properties are good. Only works, though, if it is predetermined
when or where they are playing. This could cause them not to wish to play
in difficult conditions.)
Analogously, paying workers piece rates can create incentives to lobby
supervisors to give them the best working conditions -- give me the new
tools rather than the old ones, let me work with the best workers, not
with the newer ones who need to be trained.