With the price of college going up every year faster than inflation, more attention has recently been given to figuring out how students can make the best choices when it comes to picking their school and their major. So much so, that – according to The Atlantic – students now overrate the importance of both of these decisions.
As Andrew G. Biggs of the American Enterprise Institute and Abigail Haddad, a Ph.D. student at Pardee RAND Graduate School, write, it is no secret that conventional wisdom says that the surest way of achieving economic success is to enroll in college after graduating high school and majoring in a tech-related discipline. Several recent studies have been published supporting this point of view. Yet, there are reasons to question these findings because many of them are plagued with simple statistical errors, draw conclusions that the evidence doesn’t warrant, and – what is worse – encourage governments to introduce and pursue bad public policy.
It is a basic tenet of statistics that correlation does not imply causation: simply because two things tend to occur together — such as college attendance and higher incomes — does not necessarily mean that one causes the other. While both college attendance and choice of major do affect earnings, their effects are much smaller than has been reported.
Even straightforward-seeming analysis falls prey to these kinds of errors. For example, a fairly popular recent study – authored by Michael Greenstone and Adam Looney of the Hamilton Project – sought to determine if college was still a good investment despite the rising tuition. They estimated the cost of a four-year degree to be about $100,000 and, finding that college graduates tended to make, on average, $13,000 more per year than their peers who didn’t go to college, estimated the ROI of a college degree to be about 16%.
Since the cost of borrowing money for tuition and other expenses is typically far below that number, they concluded that even going deep into debt is a better economic decision than skipping college altogether. Biggs and Haddad find that reasoning faulty:
There are two problems with this analysis. First, it conflates going to college with graduating from college. They’re definitely not the same. Data from the National Center for Education Statistics show that only 58 percent of new college students who began in 2004 had graduated six years later. Dropout rates are even higher at less selective colleges, whose students are presumably most on the margin between attending college following high school and entering the workforce. Dropouts can end up holding the bag for thousands in college debt, but earn significantly less on average than college graduates. Calculating returns to education only for those who attend college and graduate is like measuring stock returns for Google while ignoring those for General Motors. University of Rochester economist Gonzalo Castex has found that dropout risk accounts for a significant portion of the seemingly high returns to college education.
Furthermore, in their analysis, Greenstone and Looney didn’t account for the fact that high schoolers who go on to enter college are a pretty self-selecting group; they differ from their peers in more than just the choice they make about higher education. In reality, students who go on to college make different choices much earlier in their academic careers. They take tougher courses, apply more effort to their studies, typically score better on standardized exams and have a more advantageous upbringing. Such a wide range of differences doesn’t lend itself well to side-by-side comparison.
It is very likely that should those kids have somehow not been able to enroll in college, they would have still been more likely to out-earn their fellow graduates who never contemplated entering college at all.