Uncompahgre

Uncompahgre

Sunday, January 24, 2016

New Mystery Added to the Eleusinian Mysteries of the Education Myth

Two pieces of higher-education-policy-related analysis came out this past week and caught my attention. One, from Beth Akers at the Brookings Institution, calculates the return to completing a four-year college education, by institution and by major, using data from the administration's college scorecard supplemented with the ACS. The other, from Preston Cooper of "Economics21," castigates those politicians and academics promoting "Free College" as a higher education policy.

As I've learned since creating the first map of our Mapping Student Debt series, and also since writing a column on the Education Myth, releasing results that show increasing filtering-down in the labor market, and criticizing the Hershbein-Kearney-Summers paper on higher education attainment, there's a lot of intellectual capital wrapped up in the idea that the problem with the economy, with the labor market, and with rising inequality is that not enough people have high enough educational attainment, or to get more rhetorical, "lack the skills necessary to compete in today's economy." That intellectual capital is depreciating quickly in light of the gathering evidence that Skills Biased Technical Change isn't a serious contender for explaining rising inequality or anything else.

Since that diagnosis is mistaken, the prescription of increasing educational attainment isn't working, but it is creating its own legacy problem: a rising burden of student debt, which is either completely unsupportable thanks to lack of access to the labor market or an increasing economic burden on even those able to service their loans. That student debt crisis, in turn, has generated its own advocates for free college, debt-free college, "pay-as-you-earn," and other reforms to higher education financing. The intellectual dynamic has therefore become more complex. When I've pointed out the intellectual failure of Skills Biased Technical Change and the consequent mistake of increasing educational attainment as a labor market policy, those who advocate for economic justice for those with student loans or for future students get concerned, reasonably, that I'm blaming the victim. In fact, my view is that we need a real, rather than a fake, labor market policy in this country, and separately, we need to deal with that legacy problem of student debt and reform the higher education sector altogether. On those latter issues, I am not expert and remain uncertain what's best, though I lean toward free college from a public-option-for-everything perspective.

But I digress: back to Akers and Cooper. They both cling problematically to the Education Myth, albeit in subtly different ways (which are not all that different on close inspection). The essential idea behind both pieces is that filtering-down is not thanks to a problem in the labor market, but thanks to poor choices and/or obscured data on the higher education side. If only students made the right choice of major and chose, optimally, to remain enrolled and to graduate from a four-year institution, then they would realize the labor market outcomes of today's currently-existing four-year-college grads (needless to say, an older, better-paid, and less indebted population than current students, recent grads, and dropouts).

Starting from there, Akers and Cooper each have different theories of what's wrong. Akers argues that because there's little data following students through education and into the labor market, their successors in subsequent cohorts don't know what to do and make the wrong choices. Thus, she doesn't straightforwardly blame the victim, and more importantly (and in contrast to Cooper), she doesn't blame the higher education institutions either, or at least not all of them. In Akers' world, more or less, there remain good options for today's students, both in terms of institution and major, and she wants to help you find them. That is an admirable self-conception for a public intellectual, but unfortunately it is not based in reality--in part because she makes no attempt to correct her estimates for selection, as she admits, a problem that inherently plagues any normative inference about individual behavior from the College Scorecard data. The subtext of Akers' piece, at least as I read it, is that the Education Department could improve the situation not only by releasing more and more detailed data, but also by regulating institutions more heavily, presumably to either improve the bad institutions that are luring students into their clutches or by pushing them out of business altogether.

In Cooper, the subtext becomes text, but the proposed regulation has a more ideological, and therefore even more problematic cast. First of all, Cooper is notable among conservatives for at least recognizing that filtering-down poses a challenge to the received wisdom of Skills Biased Technical Change. Cooper favors cutting student loan funding to institutions with large numbers of dropouts, an idea he shares with Elizabeth Warren, but he also gestures towards Marco Rubio's awful Income Share Agreements, which its proponents baldly assert would induce people to choose the "right" majors and would incent institutions to lower costs. There's ample empirical evidence the latter contention is incorrect. As for the former, it's at odds with everything conservatives claim to believe about the effect of marginal taxation, since it imagines that skimming some percentage of income off the top would cause people to earn more money, not less. (For more, let me direct your attention to one of my favorite theory papers of all time, the Stiglitz sharecropping model.) Third, it's a disgusting, discriminatory idea that would either do absolutely nothing or violate the 13th Amendment.

The higher education policy more likely than ISAs to see the light of day under a Republican administration would similarly involve heavier regulation of institutions along the Akers line, but with a twist: student loan availability would become contingent on choosing the "right" major, and institutions might be penalized for even offering degrees not on the DoE's approved list. That, in turn, would effectuate at the federal level the ideological purges of higher education that are already underway in many states and departments. And it wouldn't actually solve the student loan crisis or the labor market's problems.

The issues raised by Akers and Cooper get to a larger point: that for decades, in this country, we've looked at the behaviors that distinguish rich people from poor people (get a job, get married, buy a house, save for retirement, go to college) and inferred that if poor people do more rich people things, they too would be rich. Therefore, a good policy would be for the government to create incentives to do more rich people things. First of all, that is wildly regressive in every instance because if you pay people to do things rich people do, you're overwhelmingly paying rich people. Second, estimates of how much those incentives actually change behavior usually turn out low-to-zilch. And when they are not low, that creates its own problems. We have essentially said that if you want money in today's economy, you have to do x, y, and z. When people do x, y, or z, they end up in a precarious position, as was the case with the housing crisis and underwater mortgages, and is the case with the EITC's high pass-through rate of benefits' replacing wages.

But third and most importantly, doing things that rich people do doesn't make you rich. This enormous confusion of correlation with causation and categorical refusal to do anything other than level regressions is maddening, and truly, that is what the voters should be up in arms about this election season.