Irrational Economist_ Making Decisions in a Dangerous World - Erwann Michel-Kerjan [74]
What is to be concluded here? Certainly these data do not imply that all is well with rational economic theories of behavior. Rather, they suggest that, as indicated by Mark Pauly in the previous chapter, even for smaller potential losses the report card for insurance decisions for natural disasters is mixed. Some aspects of the choices are broadly consistent with rational choice, but there is also evidence that the standard economic paradigm fails to reflect the substantial failures in individual risk beliefs and decision making. One can, of course, mount other defenses of expected utility theory in this context, but they tend to be somewhat ad hoc. The fixes represent possible amendments to the basic economic model rather than features already incorporated into it.
BEHAVIORAL CONSIDERATIONS
As discussed in previous chapters, how individuals treat low-probability events has continued to be a recurring concern in the economics and psychology literatures. People tend to both overestimate low-probability events and underestimate the very large risks that they face. Disasters often carry a low probability, so one would expect these risks to be easily overestimated. Moreover, highly publicized, dramatic events are prone to overestimation; natural disasters such as floods and earthquakes garner substantial press attention, which should lead the public to believe the risks are more common than they actually are. Given these biases, highly publicized risks would theoretically lead people to be excessively insured with respect to such hazards. However, insurance decision purchases seem to reflect biases in the opposite direction.
Another characteristic of disaster risks other than the level of risk probability is the precision of our knowledge of the risk. People may not know the actual probability of possible disasters, making the risks highly ambiguous. To the extent that people exhibit aversion to poorly understood risks (a widely documented phenomenon), one might expect them to err on the side of excessive insurance and self-protection. Yet, the apparent inadequacy of insurance purchases suggests that on balance this effect of risk ambiguity is not dominant.
Another possible explanation as to why individuals fail to purchase adequate coverage against disasters is that there are real personal costs to becoming fully informed. The percentage of people who have read their insurance contract in its entirety is likely to be very low. Learning about the terms of available insurance contracts involves transactions costs (e.g., making calls to different insurers to compare what they would offer, or carefully reading the entire car or home insurance contract; many consumers just don’t), so people may not know all of the conditions under which they are covered or not.
However, given the opportunity to purchase subsidized insurance, people may need to know little about the insurance terms as long as they know that it is heavily subsidized by the government. Similarly, if purchase of insurance is a requirement of obtaining a mortgage, then no element of choice is involved. Those who expect the government to bail them out completely after a major disaster may not see a need for insurance either, so here, too, the incentive to learn about insurance and to buy appropriate coverage is diminished.
The main implication of these and other findings is that decisions concerning low-probability events appear to be fraught with error. The low probabilities involved are hard for people to think about, and making reliable probability judgments requires substantial personal experience. Some people overestimate the risks, and others may underestimate the risk. The losses they suffer from the errors that lead to underinsurance and inadequate protection are likely to be greater than the losses from excessive insurance. Regardless of the direction of the errors people make, however,