A man climbs up to the top of a tower and starts shooting into a crowd. If captured alive, even if he demurs, there can be little doubt that someone will try to declare him "insane." A few lame efforts to correlate his behavior with his brain might be made, but ultimately, the issue will be put thusly: "You would HAVE TO BE INSANE to climb up a tower and fire into a crowd."
Now what insight does economic theory give us into this? A few economists, especially in the law & economics literature, occasionally make "throw-away" remarks about insanity (e.g. "of course this does not apply to the insane..."). But aren't there deeper issues involved?
What strikes me is that the popular perception arbitrarily limits the range of rational heterogeneity. We know that people have widely varying tastes: some people sky dive, while others won't take left turns in a car. Is there anything about the extreme tails of the distribution that makes them in any way less "rational" in the economic sense? Why?
One might think that the putatively insane exhibit more intransitivities in their preferences, but I doubt it. "Lunatics" from Manson to Stalin often seem to have been extraordinarily single-minded and decisive. If anything, it seems that when a person is transparently transitive people call them "obsessive" or "megalomaniacs."
Another alternative is that insanity in economic terms is not weird preferences, but systematically biased beliefs about the world. This is somewhat plausible in some cases. But it seems like plenty of people regarded as perfectly sane have systematically biased beliefs. Whatever one's religious views, I think you will have to agree that almost all of mankind has systematically biased religious views. Less controversially, well-established scientific conclusions - from evolution to the age of the earth - are rejected by a majority of the U.S. population. Does this make them insane? (Maybe once a group of people with the same systematic bias reaches a critical mass, people cease to call them insane).
In any case, it seems that many of the most notorious lunatics had a quite sound grasp of the world which enabled them to carry out their plots. As a friend once told me, he might accept an insanity defense from someone who tried to murder his wife with a head of lettuce, but if the guy had the sense to use a knife...
Suppose your metabolism is such that if you eat until you feel full, you will be morbidly obese. Does this make obesity a disease? Or do you just have different biological incentives and constraints than most people? Thinness is still in your opportunity set, it just requires more effort than it does for most people.
"Insanity is a brain disease" is the hard-line psychiatrists like to put forward. At the outset, it should be pointed out that it is extremely rare that insane behavior and brain states can be linked in such a way that everyone who e.g. has brain state X drinks too much. It is much more common to just assume that when someone's preferences are weird enough, it had to be their brain misfiring, as if weirdness were somehow contrary to human nature.
You've probably seen a dozen shows on TV displaying brain scans (or better yet, flashing technicolors representing brain activity) of "healthy" and "sick" people. But I have yet to hear a brain researcher discuss endogeneity problems. What's the endogeneity problem? Well, if the brain state correlates with insane behavior, one may fairly ask: does the brain state cause the insane behavior, or does the insane behavior cause the brain state? Correlation alone does nothing to resolve this problem.
And one might add: (a) If all behavior is caused by brain states, then what is the significance of showing the correlation between purple coloring on a diagram of brain activity and the desire to torture children? Couldn't one just as easily have shown a correlation between yellow coloring and a desire to pet puppies? How does this prove that one is a brain disease and the other isn't? (b) If some behavior causes brain states rather than the other way around, then why assume that weird behavior is less likely to be exogenous than bland behavior?
I'm an economist, not a neurologist or psychiatrist. Still, I think economics has a lot to say about putative insanity that neurologists, psychiatrists, etc. don't grasp. Moreover, I think I know what KIND of evidence empirical researchers need to present to make a valid argument. If they show no appreciation of the difference between correlation and causation, for example, then I don't think I need an MD to know that something is seriously awry.
What exactly does economics have to contribute?
First, a lot of putative insanity is nothing more or less than the extreme tails of a preference distribution with high variance.
Second, "insane" preferences seem to be as transitive as anyone else's.
Third, some putatively insane people may have systematically biased beliefs, but so do lots of people regarded as perfectly sane.
Fourth, a lot of the well-established links between biology and "behavorial disorders" boil down to unusual incentives and constraints.
In sum, while other disciplines regard insanity as a puzzle to be explained, the economic way of thinking inclines me to wonder what the puzzle is.