This post over at Marginal Revolution is a particularly egregious demonstration of the pitfalls that bedevil most public commentary by economists.
It has recently become known that Wal-Mart is considering strategies to avoid hiring unhealthy people. Commenting on this, Alex Tabarrok writes:
For a company like Wal-Mart, which pays many of its workers modest wages but does offer a reasonable health insurance plan, this is an invitation to adverse selection. As the value of the wage component of the Wal-Mart benefit package has declined relative to the value of the health insurance component Wal-Mart has attracted more workers who want the job for the health benefits, i.e. sicker workers...Wal-Mart will probably be pilloried for this sort of thinking but you can hardly blame them when the workers are engaging in almost the identical actions in reverse.
What I find apalling about this passage (and the entire post) is that no actual evidence is brought for the proposition that Wal-Mart's workers are doing the same thing in reverse. This claim is stated as fact, but really only reflects a set of prejudices on Tabarrok's part.
These prejudices are: that people are rational; that they make informed decisions based on all available information; and that they accurately plan for the future in making decisions today. In short, Tabarrok makes a series of assumptions and then states the consequences of these assumptions as fact.
But the assumptions themselves are clearly, provably, false. Many experiments have been done showing that that people are not rational; that they tend to ignore much available information (quick, do you know the fine print in the contracts you signed with your credit card companies?); and that people tend to discount the future relative to the present when making decisions.
This is not to say that assuming rationality is always bad; though false, it may yield results that are good approximations to real life. But this needs to be demonstrated here, with real data, rather than just assumed.
Marginal Revolution has often argued that economics is a science. But no self-respecting scientist would ever present claims founded on untested approximations as fact.