Too Many People Are Going to College

That conclusion should be obvious.  Roughly 48 percent of our college graduates are in jobs that the require less than a four-year degree, according to the Bureau of Labor Statistics, and the future looks worse: growth in the number of graduates in this decade is likely to be nearly three times as great as the projected number of jobs requiring such degrees.  Despite incredibly lax standards (the typical full-time student spends about 30 hours a week on academic matters) and rampant grade inflation, well over 40 percent of entering students fail to graduate within six years.

In a market environment with little governmental involvement, problems like this take care of themselves.  With growing “underemployment” of recent college graduates, demand for degrees would fall abruptly, and with that enrollments and fees would decline, and the less strong colleges would close, demonstrating what Joseph Schumpeter aptly called “creative destruction.” Massive government subsidization of students and schools, however, largely prevents that from happening (although there is some evidence, such as falling enrollments, suggesting the disinvesting process is beginning even with the subsidies).

One way to deal with the problem would be to tighten admissions standards of non-selective institutions, but those schools will not voluntarily do that, craving tuition dollars and sometimes state subsidies tied to enrollments. So perhaps we need to speed the process of creative destruction along, by forcing some underperforming universities out of business by cutting off government life support. That rarely happens now. If schools lose their accreditation, they lose access to federal student loan money, a death sentence for most. But very, very few schools lose accreditation, if for no other reasons that the big regional accreditation agencies are largely ultimately controlled by universities themselves. And while the feds have tried to go after for profit schools (“gainful employment” rules), they largely ignore the 90 + percent of higher education that is not-for-profit based.

Hundreds of School Might Go

A new study by Andrew Gillen of Education Sector suggests one possible approach: bar any school from access to federal student financial aid where the default rate on federal student loans exceeds the six-year graduation rate. It turns out there are hundreds of such schools, most of them operating at the two-year level.  Still, over 180 of them are four-year degree institutions. About a hundred of those schools are associated with for-profit institutions like the University of Phoenix, but many of them are state universities. Examples: University of Arkansas at Pine Bluff, University of the District of Columbia, Macon State College (now Macon Middle Georgia College), Chicago State University, University of Maine at Augusta,  Harris-Stowe State University (Missouri), Central State University (Ohio), Murray State University (Oklahoma), Texas Southern University, and Mountain State University (West Virginia).

Some might argue targeting these schools is unjust -many of them serve large economically disadvantaged populations. One might shrug and say, “So what?” If the school’s cost to the taxpayers is high (as measured by default rates) relative to the benefits (percent graduating), the school should be put out of business anyway. But one could use an “adjusted” default rate. Statistically estimate what the school’s default rate should be using a regression model incorporating such factors as the proportion of students receiving Pell Grants and the percent of part-time students. Eliminate federally funding only those universities whose actual default rate measurably exceeds the predicted rate controlling for things like the proportion with Pell Grants.

As Gillen (a former student and co-worker of mine) points out, the default-graduation rate ratio is not without its weaknesses, so a more sophisticated approach would incorporate other factors into the determination of the trigger mechanism.  That said, should any school that has a graduation rate below, say, 15 percent, should be allowed to operate irrespective of its default rate? Similarly, should schools with more than one-third of their loans in default be allowed to operate, regardless of graduation rates?

Another thing that Gillen’s statistical analysis suggests is that default rates on loans received by Pell Grant recipients are extremely high relative to those of other students.  Suppose two schools are otherwise identical, but one has no Pell Grant recipients, the other all Pell Grant recipients. If the actual default rate in the no-Pell school is 3 percent, Gillen’s data predicts the default rate in the all-Pell school would exceed 31 percent.  Despite the fact we spend well over $30 billion annually on Pell Grants, the Education Department resists publishing data on Pell Grant default and graduation rates. If Gillen’s statistical estimates are even remotely correct, the reason probably is that the numbers are politically embarrassing.  Based on some examination of institutional data, I would guess that the national Pell graduation rate is less than 40 percent, whereas the non-Pell rate probably exceeds 60 percent.  Similarly, my guess (partly informed by the Gillen statistical estimation), is that the national Pell Grant default rate is perhaps 25 percent, compared with less well under 5 percent for non-Pell recipients.

If I am even close to right, are Pell Grants a good value proposition for the American people? The default rate is fairly close to the graduation rate. Moreover, if half the graduates are getting jobs little better than what high-school graduates get, under 20 percent of Pell recipients get “good jobs” while 25 percent default on loans.  What are we buying with Pell Grants? Dashed hopes and expectations for many grantees, and a sizable taxpayer burden for little income creation or even redistribution.

Moral of the story: we need to downsize both the supply of higher education services, and the demand (via curtailed federal student aid). Both will work to reduce the overinvestment problem.