Is there a student debt crisis? The figures seem to speak for themselves. The total amount of outstanding student debt in the United States is $1.3 trillion, distributed among 42 million borrowers. That makes for an average balance among borrowers of over $30,000—in fact, $37,000 for the class of 2016. All these numbers have never been higher, the rate of repayment has never been lower, and the average default rate, though down from its peak a few years ago, is still far higher than it was before the Great Recession.

Two new books argue that this state of affairs is not as bad as it looks. Indeed, to Sandy Baum and to Beth Akers and Matt Chingos, the “crisis” is only a media myth—stoked by sensational profiles of low-end service-sector workers with multiple graduate degrees, six-figure debt loads, and no hope of ever paying off their loans. That is the “rhetoric” of student debt, according to both books, while the “reality” is that debt-financed investment in “human capital” pays off for the vast majority.

Two new books argue that the student debt crisis is only a media myth. They are wrong.

To support this claim, the authors point to the tight correlation between total loan balance and income, emphasizing that delinquency and default are concentrated among those with low balances and low incomes. Further evidence, they argue, can be found in the average “debt burden”: the percentage of annual income a borrower spends on student loan payments. This ratio has not changed very much over the last two decades, at least as calculated for borrowers with positive student loan balances, even though the total amount of debt has skyrocketed. Akers and Chingos take this relatively stable average to mean that while the numerator has increased for younger cohorts, the denominator—income—has too. For them as well as for Baum, that is no coincidence: they assume that more education causes higher income.

Baum, Akers, and Chingos—as well as Adam Looney and Constantine Yannelis, authors of a Brookings Institution study both books rely on—are all correct that crisis is felt most acutely by the worst-off borrowers with low incomes, who also tend to have small loan balances. But that pattern does not imply that the policy is a success for everyone else. If there is a myth at work in discussions of student debt, it is not talk of a crisis, but precisely the flawed theory of human capital behind this conclusion.

The federal government’s central labor market policy of the last three decades has been the explosion of lending for higher education. This has occurred alongside cuts in state funding for public institutions to the tune of 11 percent per full-time student. Shifting the financial burden to individuals and increasing the number and types of people getting college and graduate degrees was supposed to be a win-win, for individuals and for the economy as a whole. The debt would be paid off by the higher earnings to be had from more education.

The crisis is felt most acutely by the worst-off borrowers with low incomes.

But this “New Economy” rationale was based on a basic social science error: confusing correlation for causation. Richer people have more education, therefore more education makes people richer—that is the mistaken inference. The same fallacy lay behind the housing bubble in the 2000s: since homeownership is positively correlated with household wealth—and since housing prices would keep rising, most people assumed—taking on debt to buy homes would make people richer. When the bubble finally burst, the result was a massive sell-off that destroyed an enormous percentage of middle-class wealth. The only reason we have not seen a similar result in higher education is that student debt is unsecured by collateral. What we are witnessing instead is a massive policy failure gradually unraveling as the value of the debt-funded asset sinks.

The explosion in government lending for higher education was premised on decades of economic research documenting “Skills-Biased Technological Change”—more plainly, a skills gap. The idea was simple: big earners make big bucks because they have scarce skills. That theory is reassuring for the elite, but it is wrong. Since 2000 wages have hardly risen outside the top 1 percent of income earners, and they haven't risen at all among those workers whose skills are supposedly in the greatest demand. Even among top earners, most of the astronomical increase has been in capital income, not labor income. Rising income inequality has been paralleled by the devaluation of higher education credentials. Achieving a given level of earnings now requires more degrees, and hence much more debt, than it once did.

Student debt functions as a toll—increasingly costly and exclusionary but also increasingly difficult to avoid paying.

Wage stagnation for skilled workers and the credentialization of the labor market are very hard to square with the skills-based explanation of wage inequality. Competing research now suggests that the real culprit is wage segregation by firms, not workers’ skills. Highly paid firms increasingly employ only highly paid workers, relegating everyone else to low-paying firms. In effect, this prevents low-paid workers from accessing the profits of the economy’s leading firms, bidding down those workers’ wages through the threat—often realized—of outsourcing. This disproportionate power employers exercise over workers is far more consequential for the modern labor market than the alleged skills gap, and it is only exacerbated by credentializing the workforce. Indeed the student debt crisis is itself further evidence against the skills gap. Employers have transferred the costs of job training to the workforce, imposing a de facto wage reduction. In this climate, student debt functions as a toll—one that is increasingly costly and exclusionary but also increasingly difficult to avoid paying.

This is the theme of Sara Goldrick-Rab’s excellent Paying the Price, a full-throated call for free public college. The book draws on a longitudinal study of Pell Grant recipients entering the University of Wisconsin system, documenting their diverse socioeconomic backgrounds, their unshakeable belief that college is necessary for financial security, and the many intervening forces that caused them to drop out or defer completion—among them family obligations, sheer lack of income on which to subsist, and the difficulty of managing the scheduling and pay uncertainty of jobs open to college students. Goldrick-Rab locates the root cause of these pathologies in a single factor: the rising price of postsecondary education.

The failings of traditional public education can be measured in part by the explosion of the for-profit sector since the mid-2000s. Exploiting the gap between increased demand (thanks to credentialization) and limited supply in the traditional nonprofit sector (thanks to state cutbacks and poor support for non-traditional students), for-profits have extracted maximum debt-financed tuition from disadvantaged populations in exchange for the minimum in education and marketable skills. In fact, some research shows that, on average, for-profits worsen earnings for their students relative to pre-enrollment—on top of the monstrous added debt load.

For-profit colleges have extracted maximum debt-financed tuition from disadvantaged populations in exchange for the minimum in education and marketable skills.

Goldrick-Rab’s study also underscores a problem with the inference Akers and Chingos draw from the fact that student debt burdens—the percentage of income spent paying down debt—has remained relatively stable over the last three decades. They point out that the median debt burden has not changed very much, but they ignore enormous changes in the median borrower. Since it used to be possible to both get a bachelor’s degree and to access the labor market without taking on debt, the debt burden distribution for older cohorts included far fewer individuals, and the ones it did include came from much more advantaged backgrounds—both compared to their debt-free peers without degrees and to borrowers today, when the labor market is far more credentialized. Now it is difficult to make it into the labor market without paying the student debt toll, and, relative to family income, the toll is highest for those from the least advantaged backgrounds.

What is to be done? First we must get clear about the problem. It is not that people have the wrong degrees for today’s jobs, nor is it that the government has been too generous with federal student loans. Yet conservatives continue to make these assumptions. They have argued that federal loans should come with underwriting conditions, but this would only exacerbate inequality of opportunity, reinforce labor market stratification, and push less-qualified borrowers into more exploitative segments of the credit market. They have also proposed different financing terms depending on a student’s major—even though channeling more people into STEM degrees would only further erode their value. And they have argued for rolling back federal Income-Driven Repayment, but doing so would both increase student debt burdens and direct a larger share of graduates to opt to work in highly-paid fields. As Mike Konczal has written, labor market policies ought to be judged by the degree to which they promote workers’ choices, not inhibit them.

Meanwhile, given the result of the election, conservative legislators are likely, very soon, to roll back regulations the Obama administration has put in place to tame the for-profit sector. The conservative argument is that regulations restrict competition and that access to some form of higher education should be made available to all. What this misses, of course, is that regulation is necessary to enforce minimum standards in a predatory industry. This is a familiar pattern among Republican policy thinking, especially regarding the financial sector: use the rhetoric of free and equal access to argue for less regulation on lenders but do nothing to protect borrowers from exploitative credit arrangements like payday lending and mortgage discrimination.

Striking racial disparities are at work in higher education. What we need is a Brown vs. Board of Education for college.

Goldrick-Rab is on the right track in calling for a high-quality, low-price public option to level the playing field. But even she does not go far enough. Her own data reveal striking racial disparities at work within the traditional public sector, as students are sorted among flagship state universities, minority-serving institutions, and community colleges. To tame this problem, we need a Brown vs. Board of Education for higher education. It is too late to de-credentialize the labor market; good jobs that can be had without a degree are not going to proliferate any time soon, at least as long as effective marginal tax rates on the rich remain low and incentives to reduce wages and outsource labor remain rampant. When a college degree has become the minimum credential for decent jobs, we cannot continue to dole out access on a discriminatory basis. Gatekeepers and segmented markets have always been an effective mechanism for maintaining social hierarchy on the basis of race, class, and gender, and higher education is no different.