Frank loses me

In the previous post I cited Robert Frank's essay on the financial mess in yesterday's New York Times.

While I agree with much of what he wrote I didn't follow him on the non-financial economic examples he gave. He gives conditions for a particular kind of market failure:

This particular type of market failure occurs when two conditions are met. First, people confront a gamble that offers a highly probable small gain with only a very small chance of a significant loss. Second, the rewards received by market participants depend strongly on relative performance.

These conditions have caused the invisible hand to break down in multiple domains. In unregulated housing markets, for example, there are invariably too many dwellings built on flood plains and in earthquake zones. Similarly, in unregulated labor markets, workers typically face greater health and safety risks.

I see the application to speculative market bubbles. But it would take some unpacking to see the relevance of the two conditions to the two examples -- building in a flood plain, and worker safety risks. They both sound bad, but what is the failure as long as the cost and benefits are privately borne?

In the case of building in a flood, what's the relevance of relative performance? Also, to get a failure don't you have to assume the government will bail out flood "victims"? -- this time inconsistency problem is of course relevant to the financial crisis too.

As to worker safety, as there are no informational asymmetries firms give worker the level of safety they are willing to pay for out of the compensation.