I get so irritated when economists and political scientists try to explain every sort of irrational behavior in life as being part of some utility function.
That’s one reason I love this paper by Erik Snowberg and Justin Wolfers, “Explaining the Favorite-Longshot Bias: Is it Risk-Love or Misperceptions.” They conclude that, yes, it’s misperceptions:
The favorite-longshot bias presents a challenge for theories of decision making under uncertainty. This longstanding empirical regularity is that betting odds provide biased estimates of the probability of a horse winning|longshots are overbet, while favorites are underbet. Neoclassical explanations focus on rational gamblers who overbet longshots due to risk-love. The competing behavioral explanations emphasize the role of misperceptions of probabilities. We provide novel empirical tests that can discriminate between these competing theories by focusing on the pricing of compound bets. We test whether the models that explain gamblers’ choices in one part of their choice set (betting to win) can also rationalize decisions over a wider choice set, including compound bets in the exacta, quinella or trifecta pools. Using a new, large-scale dataset ideally suited to implement these tests we nd evidence in favor of the view that misperceptions of probability drive the favorite-longshot bias, as suggested by Prospect Theory. Along the way we provide more robust evidence on the favorite-longshot bias, falsifying the conventional wisdom that the bias is large enough to yield pro t opportunities (it isn’t) and that it becomes more severe in the last race (it doesn’t).
Also some hot graphs at the end, for example:
I’ve written articles with cameos by Muhammed Ali and Brad Pitt. But will I ever get a chance to sneak the word “quinella” into any of my graphs? Probably not. And that’s too bad.
It is not such a good thing for you to like a paper because you like its conclusions – you should have liked it just as much had it come to the opposite conclusion. If you think they did a good job of analysis, you should be grateful for their info whatever direction it points.
Robin: Yeah, I know what you mean. On the other hand, I'm frustrated that what I consider a silly model (the utility function for money) is taken so seriously by people who should know better–and I hope that this article will help shift people to a deeper understanding.
I have to disagree with "It is not such a good thing for you to like a paper because you like its conclusions."
For example, I am quite confident that there are many times when rationality (however defined) is a bad assumption. If someone writes a paper showing some of those cases, I like that paper because of the conclusions.
500 papers showing examples in which rationality is close to reality do not change the fact that there are some cases in which rationality is a bad assumption. The supporters of certain theories do a sort of 'proof by example' by only publishing papers that support their theories. We should look for papers that show the limitations of the theories instead, because all theories have limits. In my mind, that is one of the most important goals of empirical research, so I usually do like papers based on their conclusions, and don't think there is anything wrong with doing so.
(This is not the same topic Andrew was blogging about, but I think it makes the point.)
Another thanks for a cool article. I'm looking forward to reading it.
Years ago – like two decades – I ran into an analysis of track pay-offs. Bottom line was bet the favorite in every race and you come out ahead. You don't make as much as you could and that's what catches people up.
I think utility analysis is terrific as long as one doesn't fall into the trap of believing that people are actually good at figuring utility or that they're consistent. See Dan Ariely.
Is it silly because the evidence is overwhelmingly against it? If so, then there wouldn't seem to be much reason to celebrate this paper, since the silliness is well established, nor to think that it will shift people away from that theory, since the already overwhelming evidence hasn't done so.
Prof. Gelman, Thanks for sharing this. I know you can't communicate all your thoughts about everything in detail on the blog, but I think this post (especially its title) is likely to cause confusion about when you think utility functions are useful and when you don't.
While it is true that (we) economists may over-use the concept of the utility function it is tremendously useful. But it can be over-done: if you see someone shooting themself in the foot then maybe thats in their utility function. Or maybe they are just crazy. The advantage of the former, I suppose, is that it allows one to get on with modelling the behaviour systematically and to address questions that would not be possible otherwise. But, ok, it might not always make sense. The growth of behavioural economics has introduced some much needed psychology into the framework.
@robin:
Suppose "Hypothesis X" is widely believed to be true.
Paper A examines "example A", finding that A is not a counterexample to Hypothesis X.
Paper B examines "example B", finding that B is a counterexample to Hypothesis X.
Then I predict that Andrew would love paper B more than he loves paper A, and would be correct in doing so.
If by "epicycles" you mean the theory that Fred Hoyle described as no more accurate than the Copernican theory for describing the motion of the planets, and in fact no more complicated, I'd have to agree with you. The proof of utility theory is in its, dare I say it, utility, not in its reality. Economics is not a branch of psychology, nor is psychology a branch of economics, although there are obvious spillovers. I commmend to your attention Milton Friedman's famous example of billiards and "as if" reasoning in economics.
What about the findings contradicts rationality? Broadly defined, rationality merely means that you attempt to maximize your personal net benefits. People can have misperceptions and still be rational – they may have bad information, or (euphemistically) their "information-processing technology" may be poor, but that doesn't rule out an attempt to maximize net gains. If your quest is to strike down rationality, focus on the segment of the population that gambles, drinks, uses drugs, or has sex and which has psychological problems with addiction and other compulsive habits.