Quis custodiet?

My latest column in the Wall Street Journal is about the psychology of bureaucracy. just as we need to understand the human proclivities that give rise to booms and busts in markets, so we need to understand the human proclivities that motivate officials. Here are five identified by Slavisa Tasic, starting with `illusions of competence':

Psychologists have shown that we systematically overestimate how much we understand about the causes and mechanisms of things we half understand. The Swedish health economist Hans Rosling once gave students a list of five pairs of countries and asked which nation in each pair had the higher infant-mortality rate. The students got 1.8 right out of 5. Mr. Rosling noted that if he gave the test to chimpanzees they would get 2.5 right. So his students' problem was not ignorance, but that they knew with confidence things that were false.

The issue of action bias is better known in England as the "dangerous dogs act," after a previous government, confronted with a couple of cases in which dogs injured or killed people, felt the need to bring in a major piece of clumsy and bureaucratic legislation that worked poorly. Undoubtedly the rash of legislation following the current financial crisis will include some equivalents of dangerous dogs acts. It takes unusual courage for a regulator to stand up and say "something must not be done," lest "something" makes the problem worse.

Motivated reasoning means that we tend to believe what it is convenient for us to believe. If you run an organization called, say, the Asteroid Retargeting Group for Humanity (ARGH) and you are worried about potential cuts to your budget, we should not be surprised to find you overreacting to every space rock that passes by. Regulators rarely argue for deregulation.

The focusing illusion partly stems from the fact that people tend to see the benefits of a policy but not the hidden costs. As French theorist Frédéric Bastiat argued, it's a fallacy to think that breaking a window creates work, because while the glazier's gain of work is visible, the tailor's loss of work caused by the window-owner's loss of money-and consequent decision to delay purchase of a coat-is not. Recent history is full of government interventions with this characteristic.

"Affect heuristic'" is a fancy name for a pretty obvious concept, namely that we discount the drawbacks of things we are emotionally in favor of. For example, the Deepwater Horizon oil spill certainly killed about 1,300 birds, maybe a few more. Wind turbines in America kill between 75,000 and 275,000 birds every year, generally of rarer species, such as eagles. Yet wind companies receive neither the enforcement, nor the opprobrium, that oil companies do.