Any science is a model of some aspect of the world, and is useful as tool for predicting observable phenomena.
Sciences do not have platonic ideals, there is no most moral path for a particle to travel in physics, for instance. An element in the periodic table is not expected to live up to any ideals like honesty or chastity.
These may seem like obvious and fairly banal point. Observe.
Economists call economics a science, a science intended to predict how human beings allocate finite resources. This is fine, and I have no problem with that as stated. The problem I have is with those who think of economics as a religion. They come to believe that the field of economics does not predict how humans behave, but how they should behave. The Rational Man's decisions are platonic ideals to live up to. If you come up with a beautiful equation to predict how a ball travels through the air, and then you go observe a ball travelling through the air and find your beautiful equation got it about 60% right, you do not conclude that the ball is flawed.
I am not trying to pick on him, but he is a ready example. Look again at Yglesias's statement about the Tennessee fire,
it should be easy to set a pricing scheme that doesn’t entail any substantial adverse selection issues.A science based approach would say that our understanding of how people behave is flawed. Optimal pricing does not predict real world behavior, we should figure out why and update our definition of optimal pricing until our predictions are more in line with reality. The whole point of understanding optimal pricing is to be able to predict how humans are likely to behave under some set of conditions, not to figure out how they should behave.
I think there’s a surprising amount of inefficiency in the world deriving from the fact that people with various competencies—fighting fires, organizing rock concerts, cooking tasty food—don’t really understand optimal pricing.
The religious approach says something different, the people are not behaving as we expect them too, they must be flawed (in this case, ignorant of optimal pricing) and if we could teach them their condition would be improved. This is the approach of the Friedman's, George Wills, Yglesias's and others of their ilk who pick up a few economic terms, learn some of the basics, find some old time religion in that, and go on to tell us not how we are, but how we should be.
Yglesias is not alone in this. When John Nash and fellow economists holed up in the RAND corporation and invented game theory, they ran into a problem when they tested out their hypothesis on anyone that was not either a.) someone working on the project or b.) a diagnosed sociopath, it didn't do very well at predicting human behavior. When they first put their prisoner's dilemma before the secretaries working there, most of them chose to remain silent rather than betray their partner. In spite of this small set-back, however, their theories took a prominent role in the free-market work done at the Chicago School and is still influencing decision making today. You find this 'people are not acting as we expected because they don't understand what they are doing' when it comes to economics all the time. We have made the rational man behavioral model our nation's platonic ideals. Greed is good.
There is an uglier thought often found somewhere behind this first one and that is, "the people are behaving not as we predicted, and must be made to change." Few things are as terrifying as movements that lays claims to intellectual and moral superiority. That is the ugly coupling that inspires the massive blood lettings of the Jacobins, the Killing Fields, Purges and other intellectual revolutions that displaces and existing elite by killing or imprisoning the old standard bearers. It is this loose coupling that led a nitwit like Jonah Goldberg to write a crackpot thesis about liberals leading to fascism. The impulse toward controlling behavior is not liberal, it is borne in any belief system that makes claims to moral authority.