Students’ choice of college major does not affect their future salaries quite as strongly as many people believe, according to a recent paper flagged today at The New York Times’s Freakonomics blog.

The paper, by the economists Daniel S. Hamermesh and Stephen G. Donald of the University of Texas at Austin, is based on a survey of several thousand Texas alumni. (Only 25.3 percent of the targeted alumni replied to the survey, and most of the paper is devoted to the methodological problem of dealing with “non-ignorable non-response bias.”)

In raw terms, the two scholars found the usual salary disparities: Nursing and social-work majors earn $48,900 per year, on average, while engineering majors bring home an average of $102,290. But Mr. Hamermesh and Mr. Donald found that “a remarkably large fraction” of those gaps appears to be explained by factors such as the students’ SAT scores, family backgrounds, and number of hours worked each week, not their choice of major per se.

They also found that — regardless of their choice of major and regardless of their mathematics SAT scores — students earn significantly more money if they take more upper-level math and science courses. “The importance of access to this information should not be underestimated,” they write. “At this and most other institutions, students have substantial latitude in most majors to supplement required courses with others; and our results suggest that supplementation with more difficult courses pays off economically.” —David Glenn