A Puzzle on College Graduate Earnings

It is well known that on average, college graduates earn more than non-college graduates. The two dominant theories to explain this are the human capital and the signaling theories.

Here’s the most extreme form of the human capital theory:  students enter college as blank slates and the education they receive makes them more productive. This higher productivity leads to higher wages, so colleges get all of the credit for increasing the wages of their students.

The extreme form of the signaling theory is that the students who go to college are more productive than those who don’t, and colleges do absolutely nothing to improve their students. If this is the case, than the higher earnings of college graduates is not attributable to college at all, but is entirely due to preexisting differences.

Most people don’t believe in either extreme theory and acknowledge that there is some truth to both. Then they proceed largely to ignore the signaling theory in favor of the human capital theory. This is not entirely inappropriate as empirical tests often find little evidence in support of the signaling model (this great paper by Fabian Lange and Robert Topel has a good overview).

But…

Findings keep popping up that are hard to explain if the signaling model really only plays a trivial role. A case in point is this finding in College Participation, Persistence, Graduation and Labor Market Outcomes by David L. Wright, William F. Fox, Matthew N. Murray, Celeste K. Carruthers, and Grant Thrall:

Conditional on degree receipt, additional semesters in college reduce earnings later on. This latter finding is consistent with recent research demonstrating that excessive years of schooling can be a negative signal to employers (Flores-Lagunes and Light, 2010).

In other words, students who take, say, 6 years to graduate tend to earn less than those who graduate in 4 years. In the human capital model, these findings would be quite surprising, because additional years of schooling should continue to increase, rather than reduce, earnings. But while the human capital model cannot easily explain how additional schooling leads to lower earnings, the signaling model has no difficultly—those who take 6 years to graduate signal to employers that they are less motivated/disciplined than those students that graduate in 4 years. And since employers value motivation and discipline, the 6 year graduates earn less as a result.

This significantly increases the weight people should attach to the signaling model in explaining why college graduates earn more. Because if you buy the argument that students who take longer to graduate earn less for signaling reasons, it is difficult to then argue that the signaling theory plays a trivial role in explaining the higher earnings of college graduates relative to high school graduates.

Photo Credit: iStock Photo