Micro 7: Information Flashcards
What is the difference between adverse selection and moral hazard?
The unobservable action can either take place before or after contracts are made. In adverse selection, the unobservable action (nature’s choice of your type) happens pre-contract. In moral hazard, the unobservable action (your choice of effort) happens post-contract. There are no other possibilities.
Do all markets with asymmetric information fully unravel?
No. Akerlof-style situations can be constructed where there are equilibrium prices, but there are still inefficiencies (ie trades that would be mutually beneficial, but do not occur).
Why can all individuals not be fully insured when there are two types of agents, but only the agents know which type they are?
High-risk agents can make themselves better off by pretending to be low-risk and buying the insurance intended for low-risk agents. This then puts the insurance companies out of business, as they have to pay out more than expected. There is a sustainable equilibrium if we offer only partial insurance to the low-risk types, as then the high-risk agents do not want to buy it, even at a lower price.
What is a pooling equilibrium? Why can there be no pooling equilibrium in the Rothschild-Stiglitz model of insurance?
A pooling equilibrium is where all agents are offered the same contract. Such a contract must lie on the zero-profit line. However, due to the different gradients of high and low risk types, there is always a region where contracts are preferable for the low-risk types only, and would make a profit for the insurance companies. Any pooling equilibrium is therefore not stable.
What is a separating equilibrium? Why is this not first-best efficient in the Rothschild-Stiglitz model?
A separating equilibrium is where different contracts are offered to high- and low-risk types, and each is satisfied with their contract. The best sustainable contract in the RS model has full insurance for the high-risk types, and partial insurance for the low-types. This is not first-best efficient, as fully insuring the low-risk types would make them better off and make the high-risk types no worse off.
How does the relaxation of perfect competition assumptions affect the Rothschild-Stiglitz model?
‘Non-robust’ equilibria can then be sustained, as there can be no undercutting of offers by other firms. For example, a social monopoly can enforce a pooling equilibrium and require the low-risk types to cross-subsidise the high-risk types. A private monopoly can enforce other equilibria that profit maximise.
In the signalling model of education, why is education a good signal?
High-productivity types can gain education more easily than low-productivity types, so possession of a given education level signals your type well. It is, informally speaking, not just ‘cheap talk’.
When would removing the ability to signal be a Pareto improvement?
If the proportion of high-productivity types is high enough, high-productivity workers would prefer not to exert effort getting education and would instead prefer to get the pooling wage.
Why might education (even if not productivity-enhancing) still be socially useful?
In the Stiglitz (1975) model, it can be shown that education helps with labour market signalling and therefore with productivity in jobs.
What is the principal-agent problem, intuitively speaking?
The principal-agent problem is that the principal pays an agent to complete a task, but cannot observe how much effort she puts in to the task. The task generates higher profits if higher effort is put in. Somehow, the principal must induce high effort without being able to observe her effort level.
If effort is observable, how should the agent be paid?
If effort is observable, we simply need to satisfy an Individual Rationality constraint; the contract can require an effort level and pay just enough to compensate her for that level.
What are the Incentive Compatibility and Individual Rationality constraints?
The Individual Rationality constraint is that any contract must be better than reservation utility.
The Incentive Compatibility constraint is that the expected utility from high effort must be greater than that from low effort; otherwise, they will simply put in low effort since it is not observable.
What is agency cost? Why can it generate inefficiencies?
In order to induce high effort, the agent must be exposed to some sort of risk; otherwise, she will always put in low effort. However, given a risk-neutral principal and a risk-averse agent, we must pay the agent a risk premium in order to be willing to undertake this risk. This is the agency cost; it can make it unprofitable to induce high effort, when under full information high effort would be the first-best outcome.
