A student observed in an email from earlier today that moral hazard and adverse selection seem closely related, and was curious whether there is a definitive test to say whether it is one or the other? For what it’s worth, here’s the response that I sent back to him/her:
Moral hazard and adverse selection are related, in the sense that both phenomena exist because it is costly to acquire information. Thus, information is asymmetric – the agent knows things that are important to her relationship with the principal that the principal is either not aware of or harbors suspicions about. Moral hazard is a problem of “hidden” action; action is hidden in the sense that the principal is not able to perfectly monitor the agent, so it’s in the interest of principal and agent alike to write incentive compatible contracts. Incentive compatible contracts motivate the agent to take actions (e.g., work hard and not shirk in labor markets, prevent or mitigate losses in insurance markets, etc.) that are consistent with maximizing the welfare of the principal.
What differentiates adverse selection from moral hazard is that the latter pertains to situations where a contractual relationship between principal and agent is already in place, whereas the former pertains to situations where a contractual relationship between principal and agent is in the process of being formed. Thus what you worry about as a principal is that you end up “getting stuck” with someone who or something that lacks the qualities that you are looking for. This is the “lemons” problem; e.g., you end up buying an inferior used car, hiring a lazy, incompetent and/or corrupt worker, etc. The strategies for mitigating adverse selection include screening on the part of the principal and signaling on the part of the agent. Examples of screening include requiring someone who is applying for health insurance, life insurance, or an annuity to have a medical exam, banks running credit scores on mortgage applicants prior to granting credit, etc. Also, the agent can credibly convey important information by signaling. Signaling typically takes the form of utilizing third party certification (this is what makes the signal credible). So EMBA students take time out of their busy schedules to earn MBA degrees from Baylor University (here Baylor provides credible certification concerning quality of human capital!), used car sellers pay for vehicle inspections, etc. However, even after people have invested resources in screening and signaling, there still may exist some residual information asymmetry, and it’s here that one utilizes contract design as a strategy for incenting people to reveal their “true” types. The example I used in class involved insurance, where we know that some people are good risks whereas others are bad risks; the problem is that we can’t figure out who’s who. Therefore, we limit contract choices so that it will be in the agent’s self interest to effectively reveal truth by the contract choice. In the insurance example, risky drivers are attracted to expensive contracts that offer high levels of coverage, whereas safe drivers self-select into cheaper contracts that offer lower levels of coverage.
In summary, the definitive test for differntiating between moral hazard and adverse selection involves asking the following questions: 1) is there asymmetric information?, and 2) does a contractual relationship already exist? If the answer to 1 is yes and to 2 is no, then you need to be worried about adverse selection. If the answer to 1 is yes and the answer to 2 is yes, then you need to be worried about moral hazard.