It’s that time of year again, when students are deciding where they’re going to go to college. And the NYT is on the case. Today Greg Mankiw tells us that "the financial return to education" has never been higher:
In 1980, each year of college raised a person’s wage by 7.6 percent. In 2005, each year of college yielded an additional 12.9 percent. The rate of return from each year of graduate school has risen even more — from 7.3 to 14.2 percent.
So we know that the benefits of going to college are high. Meanwhile, the costs, according to David Leonhardt, are actually going down, at least at the most elite universities, such as Harvard, where Mankiw teaches:
In less than five years, the entire tuition and financial aid system at the nation’s top colleges has been overhauled. Today, it looks a lot like a highly progressive tax code, in which the affluent are bearing an enormous share of the overall tuition burden. No matter how high the published cost — almost $50,000, typically — these top universities have become significantly more affordable for the majority of students.
But the overwhelming majority of college students don’t go to Harvard or even elite state schools like Berkeley. Instead, says MP Dunleavey, they’re making choices like the one facing Taylor Glauser: should he go to the State University of New York at Potsdam, or should he go to Mercyhurst, a small private college in Pennsylvania which will cost him $15,000 a year even after $20,000 in grants and scholarships? Would spending the extra money be worth it, over the long term?
Both Dunleavey and Leonhardt cite Princeton’s Alan Krueger, who’s done the most research into this question. Krueger’s main finding is that it makes very little sense from a financial perspective to go deep into debt just for the sake of attending a private college. Once you control for things like SAT scores, it turns out that attending a more prestigious university does not actually improve your lot in life.
But as Leonhardt notes, there is one exception to this rule:
Recent research also suggests that lower-income students benefit more from an elite education than other students do. Two economists, Alan B. Krueger and Stacy Berg Dale, studied the earnings of college graduates and found that for most, the selectivity of their alma maters had little effect on their incomes once other factors, like SAT scores, were taken into account. To use a hypothetical example, a graduate of North Carolina State who scored a 1200 on the SAT makes as much, on average, as a Duke graduate with a 1200. But there was an exception: poor students. Even controlling for test scores, they made more money if they went to elite colleges. They evidently gained something like closer contact with professors, exposure to new kinds of jobs or connections that they couldn’t get elsewhere.
“Low-income children,” says Mr. Krueger, a Princeton professor, “gain the most from going to an elite school.”
I find this a very intuitive result. Just as we saw with secondary school, middle-class kids tend to do pretty well wherever they’re educated, which makes it ironic that middle-class parents are the ones who stress the most about where their kids go to school. Meanwhile, the poor, who really do benefit from a better education, are the ones overlooked. As Leonhardt notes:
Given all this, some have argued that Harvard, Yale and other universities are mistaken in giving financial aid to upper-middle-class students rather than using the money to recruit more low-income students.
He gives some not-particularly-convincing reasons why it might make sense to concentrate on the middle classes, but the main takeaway is still clear: if you’re the kind of person who reads Market Movers, and you’re facing big choices in terms of school, then you’re probably worrying far too much, it really doesn’t matter all that much in the long term. But if you’re an educational institution and you really want to maximize your value added, or whatever the jargon might be, then by far the best way of doing that is to get as many poor students as possible.