You may remember this prominent study by a group of economists finding positive effects of teacher quality from Operation STAR, a Tennessee program lowering class sizes in kindergarden.
One of the ancillary goals of that study connected IRS datasets on income with College attendance, enabling the researchers to figure out how going to Colleges affected income at age 28, among people born around 1980. The results point to the importance of elite Colleges in driving the skill premium:
Bear in mind that this reflects earnings at 28, when many people are in graduate school or recently out, providing added earnings that rise over a lifetime.
Still, the results suggest relatively low returns to higher education for all but the most elite schools. The authors write that STAR program graduates had earnings of $16,475 if they did not attend college, $26,920 if they graduated two-year college, and $35,080 if they graduated a four-year college.
The typical human capital view of things is that these higher premiums for education reflects market reward for greater skills. Though college graduates do earn more than high school graduates, they also tend to be very different people; and it’s not clear what the causal impact is of sending more people to college — especially if the marginal college attendee goes to a low-ranked school.
And from above, it looks like much of the reward for college is really a reward for going to a top college. Assuming that instructional quality doesn’t suddenly jump at rank 50, that’s a sign that something special about top schools — the fact that they attract better students, access to stronger peers, connections, signaling – is important, and the returns to those things are strongly increasing in college rank. The market seems to place a pretty high premium on signals, and a fairly low premium on graduating college. The fact that many graduates of top schools attend graduate school or go into the financial industry is probably important here — even the skill premium for people who only have a college education, but no graduate degree, has remained relatively stagnant in the recent past.
These effects make it difficult to judge various programs. The researchers in this study for instance found that better kindergarden teachers made students more likely to attend the good schools, leading to a large social value of better teachers. But to the extent that the new kids simply displaced other students at schools that gain their value from a relative ranking, it’s not as clear how to value that.