Do college students choose their field of study according to expected income? One would hope so, as the price mechanism would then lead to more efficient uses of human resources. This optimality criterion would only hold if there were particular distribution of abilities across fields prior to the start of studies. But leaving this aside, do prices matter?
Magali Beffy, Denis Fougère and Arnaud Maurel use French data and exploit the fact that the expected returns of study majors vary with the business cycle. As the cost of studies in negligible in France, tuition is not a factor, although the duration of studies varies a little by field and thus the opportunity cost of studying, but that is largely unaffected by the business cycle. They conclude that while expected earnings matter in a statistically significant way, it is not significant in economic terms. That means, students take expected earnings into account, but it influences them little. This means, there will always be a lack of nurses, accountants, and engineers, and always a surplus of humanists, whatever the market wage is (within reasonable bounds of course).