In the course of researching my next novel, I happened upon this old paper by Robert H. Frank, Thomas Gilovich, and Dennis T. Regan, "Does Studying Economics Inhibit Cooperation?" Its conclusions: Economics grad students are more likely to free ride than the general public. Economists are less generous than other academics in charitable giving. Economics undergrads are more likely to defect in prisoner's dilemma problems. Students are less likely to return found money after studying economics but not after studying another subject like astronomy. No wonder they call it "the dismal science."
A study by Gerald Marwell and Ruth Ames found that students of economics are indeed much more likely to free-ride in experiments that called for private contributions to public goods. Their basic experiment involved a group of subjects who were given an initial endowment of money, which they were to allocate between two accounts, one "public," the other "private." Money deposited in a subject's private account was returned dollar for dollar to the subject at the end of the experiment. Money deposited in the public account was first pooled, then multiplied by some factor greater than one, and then distributed equally among all subjects.
Under these circumstances, the socially optimal behavior is for each subject to put her entire endowment in the public account. But the individually most advantageous strategy is to put all of it in the private account. The self-interest model predicts that all subjects will follow the latter strategy. Most don't. Across eleven replications of the experiment, the average contribution to the public account was approximately 49 percent.
It was only in a twelfth replication with first-year graduate students in economics as subjects that Marwell and Ames obtained results more nearly consistent with the self-interest model. These subjects contributed an average of only 20 percent of their initial endowments to the public account, a figure significantly less than the corresponding figure for noneconomists (p.05).