Saturday, June 23, 2012

Self Selection and Diverse Views Over Sea Level Rise in North Carolina

If you are looking for me,  you can soon find me in Salzburg, Austria as I slowly make my way to Singapore.   If I hear some good Mozart, I promise to blog about it but now permit me to discuss this recent blog post at environmental economics.

The extent of Sea Level Rise in North Carolina is a random variable and people disagree about the timing and the extent to which it will occur.  In a diverse society, some people believe that it will be severe and these individuals are unlikely to locate in coastal areas even if land there is cheap and these people love the coastal nature experience.  There are other individuals who either have special talents that allow them to reduce their risk exposure or they discount the possibility that they are imposing risks on themselves by living near the coasts. In such a diverse society, the "dare-devils" will be more likely to bear the risk and live in the risky place.  If disaster occurs via climate change, are they victims?

For those who know some labor economics, doesn't this sound like the compensating differentials literature and the value of a statistical life. Kip Viscusi has been the leader in this field as he has built on Thaler and Rosen's original framework.    Hedonic methods have been used to measure the wage "combat pay" that workers receive for working in risky industries where fatality risk is higher.  Dora and I have published a well cited paper on this topic.    The standard hedonic wage paper doesn't examine who self selects into taking the risky jobs.  You don't have to be Heckman to anticipate that these individuals will not be a random sample of the population.  In fact, this 1992 JPE paper argues that the wealthy will avoid these risky jobs while this 2002 JPE paper argues that diverse abilities that protecting one's self against exogenous risks affects who self selects into risky jobs.

If workers are fully informed about the risk of each job, then is OSHA needed to regulate that jobs must become "safer"?  Employers would know that they could pay lower wages (less combat pay) if they could credibly commit that jobs would be safer.  They will invest in such equipment assuming the cost of such equipment is less than the wage premium they must pay if the job remains risky.  In this sense, the free market provides the "optimal" amount of job safety.

The same "assignment problem" arises in the case of climate change.  There are diverse people and there is a set of differentiated geographic areas that differ with respect to their current and future flooding risk. The home price gradient adjust so that one of each of these households lives in each of these homes and so that the gains to trade are exhausted.   Under what circumstances is the ex-ante equilibrium not a pareto optimum?   Why is government zoning needed here?  To justify such zoning, you need to tell a story of "bad learning" and households unable or unwilling to update their probability assessments as "new news" about climate change arises. If people know that they do not know what climate change will do to the NC coast and if they are risk averse, then they will hesitate before buying coastal property.  Only a "Homer Simpson" who doesn't know that the doesn't know about the risk would view the "cheap" coastal property as a steal!  In this case, paternalist zoning regulation could be justified to protect him but the key issue here is what % of the population is Homer Simpson?  As I have argued before, there is a strong "behavioral economic" view lying behind the pessimism about our ability to adapt to climate change.  This view needs to be explored and debated.