I just read "Are Economists More Selfish than Other 'Social' Scientists?" (Public Choice, 1999, 100: 85-101) by David Laband and Richard Beil. They looked at lying by members of the American Economic Association, the American Sociological Association, and the American Political Science Association. Each association has higher dues for members with higher incomes, so if you lie and say your income is low, you save on your dues. Laband and Beil surveyed members about their incomes, and then compared the income distribution to what members reported when paying their dues.

Page 96 has the result: in the category of incomes above $50,000, it seems that 26% of political scientists underpay dues (15.3/60.6 from Table 2), 33% of economists, and 50% of sociologists. More economists earn high incomes, so the actual numbers of cheating high-income economists and sociologists look about the same. But the sociologists also have more cheating of middle-income members saying they are in the low-income bracket. Laband and Beil have a clever single summary statistic: an estimate of the percentage of dues not collected because of cheating. The amount lost by cheating is 7% for the economists, 9% for the political scientists, and 22% for the sociologists. The implication seems to be that studying economics or politics does not make people more selfish or dishonest, but sociology is bad for one's morals.