Sunday, April 15, 2007

Government and Smart Choices About Natural Disaster Risks

Libertarians like to argue that for most problems people make the best decisions for themselves because they have more knowledge about their own values, circumstances and options than anyone else and because they have productive incentives, so that forcing them to choose specific actions is bad. This is certainly true to an extent, but in some recurring cases, people have unproductive biases or have the poor information about their circumstances and options relative to others. For example, on the issue of natural disasters and insuring assets against damage, asset owners often have both limited knowledge and biases against buying appropriate insurance. To use the example of New Orleans, residents are unlikely to have good information about the risks they face from hurricanes, the costs and likelihood of a hurricane striking. In addition, people have a well known bias against events with small probabilities. When performing conscious or subconscious cost analysis, people tend to discount the importance of low probability events, which can be a serious problem if there are high cost but low probability events, so New Orleans residents will make less than good choices about insuring their homes and other assets against hurricanes.

What is the role of government in evaluating these risks and insuring against them? Producing information about certain common risks is surely a public good and one of the proper roles of government, but should government, as part its social contract, correct for well-known, widspread biases? On the issue of natural disasters this means forcing the purchase (through taxes) of insurance against low-probability, high-cost events.

4 comments:

John said...

Actually it's called selling treasury notes, and that's how the 120 billion dollars or whatever got dumped in Louisiana.

Now funneling that to the actual victims rather than spending 50% on graft? That's a worthy cause.

Chris said...

MR post on soft paternalism (like this gov't disaster insurance idea):

http://www.marginalrevolution.com/marginalrevolution/2005/11/scream_this_fro.html

Wow, "loogel," aren't you maybe just a little concerned about the downside of the government insuring people against natural disasters? I'm skeptical because it seems that the government's approach is usually a total bail-out, not just a little insurance fine-tuned to compensate for consumers' biases. And I just really don't want to pay to insure someone else's house. Why should a member of society get to chooose to live in a hurricane-prone area at my expense? I'm more worried about the pendulum swinging far in the other direction. Also, a small per-unit subsidy on disaster insurance might be a better tactic against bias than the actual purchase or free provision of insurance by the government. Then that marginal person who dismisses buying insurance when they really should will do so.

Of course, we need to reflect carefully on what bias is. If it is the case that someone living in a hurricane-prone area is aware of the danger, then when they choose not to buy insurance, are they making a rational or an irrational (biased) choice? What is a biased choice? Are you thinking of those studies they do in experimental economics? From what I've heard about those studies, risk-aversion is usually the major pattern. So I am more inclined to think that if people are somehow not buying "enough" insurance, it is because they lack good information (perhaps there's a role for government, though it don't see a strong case for it) or because they expect the gov't to bail them out.

loogel said...

You and the paper on the MR post present a good argument, but I have to think about it longer.

I do want to say that producing good information about risks common to large populations is definitely the role of the government because such information is unquestionably a public good, so it won't be produced otherwise.

Chris said...

Although a general sense that the gulf coast is prone to hurricanes might be a public good, up-to-date information on the risks provided in a nice format along with competent, friendly service might be more of a private good. I don't see the need for the government to provide information in this situation.

If homeowners buy hurricane insurance, I would imagine that the insurance company would do its best to estimate the risks involved, as this information is crucial to its bottom line. I suppose even homeowners who do not buy hurricane insurance could use insurance rates as a barometer for risk. But I admit that at some point this becomes a normative issue. If people ignore low-probability high-cost risks, should it be the government's responsibility to compensate for their cognitive biases?