Colin F. Camerer
1200 E California Blvd
MC 228-77
Pasadena, CA 91125
(626) 395-4054

Email me

Most scientists specialize in either theorizing, looking for patterns in naturally-occurring data, or collecting their own data through surveys or experimentation. But theory, naturally-occurring data, and experimental data are complementary—each kind of exploration provokes questions which can be answered with the other kinds of data or theory. To this end, I've also done some studies with field data. Field studies are especially important because many of the phenomena behavioral economists have studied are established in the lab (where it is easier to clearly show that people are making a mistake); so it is crucial to explore whether these phenomena occur in field settings too, especially when people are more experienced and the stakes are high.

For example, people are often sensitive to a "reference point" and dislike losing relative to the reference point or falling short of it. To test whether this sensitivity exists in the amount of labor some workers supply, we collected data on how many hours New York City cabdrivers choose to drive on different days, when their daily wage fluctuates (on rainy days, for example, everyone wants a cab, and on weekends in the summer the cab business is slow). The standard theory of labor supply assumes that drivers plan ahead and "intertemporally substitute"—they drive a lot on high-wage days and save up so they can quit early on low-wage days (they "make hay when the sun shines"). We found the opposite pattern: Inexperienced drivers act as if they set a daily income target and quit when they reach it. While this pattern seems like a good idea, it means they drive a lot on low-wage days (because it takes a lot of hours to reach their target) and quit too early on high-wage days. (They could have earned about 15% more if they switdhed their hours around and drove longer on high-wage days.) Experienced drivers, however, drive about the same number of days on good and bad days so they earn more than inexperienced drivers. The long version is "Labor supply of New York City cab drivers: One day at a time," with L. Babcock, G. Loewenstein, and R. Thaler, Quarterly Journal of Economics, 112, May 1997, 407-441. (A shorter version, published in the book Choices, Values and Frames.) A 2003 New York Times column mentions this research, and some follow-up studies, in the context of whether a tax cut will boost productivity.

Another study explores whether sports bettors have a good sense of what a random lucky streak is. Psychologists showed that basketball players seem to have streak "hot hand" shooting, and are likely to hit one shot after they hit a couple in a row, but they actually do not. The hot hand just seems to exist because most people expect that a random series will reverse itself or "mean revert"; when it doesn't, they are surprised and come to believe that the sequence of data has momentum or "heat". We discovered that the basketball market makes the same mistake in setting the odds that different teams will win. The point spread on teams on losing streaks is too pessimistic (as if the bettors believe in a mythical "cold hand") and the point spread on teams with winning streaks is too optimistic. That paper is "Does the Basketball Market Believe in the 'Hot Hand'?" American Economic Review, 79, December 1989, 1257-1261.

Roberto Weber and I became interested in a phenomenon called "escalation of commitment"—do people who have sunk money in an investment that seems to be going badly sink even more money or do they cut their losses (treating their costs as "sunk")? Many psychologists have shown this pattern in laboratory studies, but there are almost no conclusive field studies. We replicated a creative study by Staw and Hoang looking at how often NBA basketball players who were picked high and low in the draft actually played. We found that players who were drafted high (a large investment because draft picks are valuable) tended to play more minutes, even controlling for actual performance. This means coaches—or perhaps owners, or fans—tend to "throw good minutes after bad" and stick to players they had high hopes for even when it is clear their investment was a mistake. But the coaches gradually give up on underperforming draft picks after a couple of seasons. That paper is "The econometrics and behavioral economics of escalation to commitment in NBA draft choices," with Roberto Weber, Journal of Economic Behavior and Organization, 1999.

Division of the Humanities and Social Sciences
Mail Code 228-77
California Institute of Technology
Pasadena, California 91125

Office: Room 101
Phone: (626) 395-4054
Fax: (626) 432-1726
email icon  E-mail: camerer@hss.caltech.edu
Caltech Home Page