Do Human Arts Really Offer a Lower Return to Education?∗

Using Danish micro data on high school education, tertiary education, and wages, this paper investigates the importance of the field of study for the returns to education. Our OLS results suggest that a degree within the human arts seems considerably less valuable (in terms of earnings) than a degree within other fields, echoing previous result for the UK and US. However, since the choice of which type of education to pursue is likely non-random, these results may reflect selection rather than the causal impact of the field of study on wages. To examine this issue we adopt an instrumental variable approach to the issue at hand. Our identification strategy consists of using observed characteristics of the peer group the individual student were exposed to during high school. In particular, we document that the fraction of female students in a high school impacts on individuals ’ choice of education. Similarly, the educational choices of high school seniors and freshmen are highly related. Conditional on high school fixed effects, and a large number of individual specific student characteristics (e.g., family background, high school test performance etc.), these characteristics should not affect wages directly. We therefore employ them as instruments for the choice of field of study at the tertiary level. In contras