Why Government Fails So Often: And How It Can Do Better (42 page)

But even more disturbing than poor targeting and widespread fraud is the vast level of student debt and loan delinquency. According to the New York Federal Reserve Bank, 35 percent of student loan
recipients under age thirty who were required to make payments were at least 90 days late on them, up from 26 percent in 2008 and 21 percent in 2004.
⁹⁸
For every borrower who defaulted, at least two more were delinquent in their payments.
⁹⁹
The
New York Times
noted that the total of $76 billion in default was “greater than the yearly tuition bill for all students at public two-and four-year colleges and universities.”
¹⁰⁰
The program, in short, is an engine of moral hazard. After all, the borrowers have few assets and the government demands no collateral, has no underwriting requirements, and asks little or nothing about students’ ability to repay or about what sort of education they intend to pursue. Students take on greater debt with evidently little concern about future default. Student debt
rose
by more than 56 percent, adjusted for inflation, from 2007 to 2012—a much faster growth rate than credit card debt.
¹⁰¹
This, in a period when the mortgage crisis should have caused the government to focus on moral hazard problems and in which overall private household debt
fell
by 18 percent.
¹⁰²
Students who borrow average about $40,000 in debt by graduation, almost twice what it was a decade ago; professional and graduate students’ debt is about $56,000.
¹⁰³

Student loans have the highest delinquency rates of
any
federal credit program (reaching 11 percent in 2012) and even higher delinquency rates than for private auto, home equity, and mortgage loans; indeed, they approximate the credit card delinquency rate. And the delinquency steadily worsens. In March 2012, subprime borrowers held an estimated 33 percent of outstanding student loans, up from 31 percent in 2007. Also, the gap between college costs and borrowing limits continues to widen, prefiguring more defaults and dropouts.
¹⁰⁴
As of 2012, nearly one in every six borrowers with a loan balance was in default (which occurs when a loan is delinquent for 270 days), and they were defaulting more quickly; the two-year rate was 9.1 percent, the highest in more than a decade. Moreover, according to the New York Fed, these rates actually
understate
the delinquency problem because many borrowers were not yet required to make payments for one reason or another. And in a grim augury of future defaults, the government’s new direct loan program, instituted in 2010 when it stopped
guaranteeing private loans and then rapidly expanded during the weak job market, seems to have made matters worse. In August 2013, the CFPB found that 22 percent of the 27.8 million borrowers under the new program were
already
in default or in predefault forbearance.
¹⁰⁵

Even as the economy improves, the situation grows more dire. Indeed, this student debt overhang threatens to retard the recovery. The outspoken proborrower director of the CFPB likens the situation to the subprime mortgage disaster.
¹⁰⁶
And to make matters worse, Congress has misled the public by imposing an accounting method on the federal lending agencies that systematically underestimates the risk of default and thus its ultimate costs to taxpayers.
¹⁰⁷

To its credit, the government recently stepped up collection efforts. It estimates recovery of 80 percent of the defaulted debt, or $12 billion—although a 2007 academic study found a recovery rate closer to 50 percent.
¹⁰⁸
But even these efforts are flawed. The private collection agencies, which cost the government $1.4 billion a year, may actually increase defaults by skewing incentives. Colleges with very high debt levels and default rates (mostly for-profits
¹⁰⁹
) continue to participate, while any sanctions will only come in the future. In April 2013, Sallie Mae had to withdraw its bond offering from the market because investors in the secondary market that securitizes these loans showed so little confidence in its ability to collect on these delinquent loans,
¹¹⁰
which forced the agency to hive off the huge losses that it anticipates from these delinquencies in order to protect its other, profitable operations.
¹¹¹

Quite apart from moral hazard, the program has produced some seriously perverse and unforeseen consequences. First, it seems to have contributed to tuition increases, which directly contradicts the program’s goal of increasing student access. Tuition and fees have nearly tripled over the last twenty years, rising much faster than wages, presumably in part because federal aid has increased what students can pay. Another explanation is that federal aid may increase the demand for higher education, and with no offsetting factors, this pushes up the price of tuition.
¹¹²
Second, the program may have encouraged institutions to substitute federal money for their own financial
assistance, thus reducing the program’s net effect.
¹¹³
It may also have contributed to the doubling of the administrative staff-student ratio since 1975, while the faculty-student ratio has changed very little.
¹¹⁴
Third, it may have encouraged many marginal students to over-invest in higher education by foregoing options—cheaper on-the-job training or earlier entry into the labor force, for example—that may be better for them on balance but that are undervalued if they underestimate their risk of future default.
¹¹⁵
They may also be underestimating the large risk—47 percent, if community colleges are included—that they will not complete college, a risk that is even larger for students from poor families.
¹¹⁶
Federal data published in February 2013 indicate that many of these loans go to art, music, and design students who carry a disproportionate debt load while facing limited future income prospects.
¹¹⁷
These debts are not dischargeable in bankruptcy—a rare concession to the moral hazard problem (but one that many in Congress want to change).
¹¹⁸
The CFPB rightly fears that however the bankruptcy discharge issue is resolved, these debts may limit students’ future ability to obtain loans, which will limit their adult opportunities and hobble the economy.
¹¹⁹

Given the program’s sheer size, visibility, and cost, one would expect the government to have conducted comprehensive assessments of its effectiveness, but it has seldom done so—in contrast to, say, the frequent assessments of Head Start. (Recall, however, that Congress continues to lavish funds on Head Start despite many findings of ineffectiveness; see
chapter 6
.) GAO reports in 2002 and 2005 called for more assessments of the effectiveness of student grant and loan programs and tax preferences in promoting students’ attendance, choice, persistence, and completion.
¹²⁰
The few studies that have been done reveal mixed results. Some show small increases for some of these parameters, while others show no effects; no studies show large positive effects.
¹²¹

The GAO found that many of these programs and tax preferences had never been studied, and even among those that had, vital aspects of effectiveness remained unexamined. Studies on the tax credits show them to be significantly underutilized by those who would
qualify, probably because of their complexity.
¹²²
No research existed on how the program’s tax credits affected students’ persistence in their studies or the type of schools they decided to attend.
¹²³
According to a 2012 GAO report, no research exists on any aspect of effectiveness for tax-exempt savings programs, the student loan interest deduction, or the parental personal exemption.
¹²⁴
This lack of data on the effectiveness of such popular, long-standing, and costly programs is deplorable and symptomatic of the more general problem of official disinterest in assessment data. Researchers have examined the effects of Title IV aid and federal tax expenditures only “on a limited basis—in particular, only for certain states, types of schools, and groups of students.”
¹²⁵
This leaves extensive gaps in policy makers’ knowledge about the programs’ effectiveness. In 2002, the GAO recommended that the Department of Education study their impact on student attendance, choice, completion, and college costs. Three years later, the GAO found little progress. Its 2012 report concluded that the Department’s most recent assessment efforts were “fairly narrow in scope” and “unlikely to yield broadly applicable lessons about the effects of Title IV aid on student attendance, choice, persistence, or completion.”
¹²⁶
Even more dismaying, Nobel Prize–winning economist James Heckman recently found that college-age people from disadvantaged backgrounds may not be able to benefit from them, absent pre-school interventions.
¹²⁷

In sum, this very costly program reaches a large number of people and enables many students to attend college who could not otherwise afford it. This is no small thing. But on the other side of the ledger, it disproportionately benefits those who need it least, and harms the many students who are encouraged to borrow sums that they cannot repay because they drop out of school, end up in jobs that pay too little, or default for other reasons. Remarkably little rigorous assessment has been done of the program’s effectiveness concerning any of its key policy parameters, and what research there is suggests that it is not only ineffective but causes some serious harm. Not surprisingly, reform proposals abound. Congressman Tim Petri, for example, proposes a plan modeled on those used in the United Kingdom, Australia,
and New Zealand, in which employers deduct student loan payments automatically from paychecks in order to reduce the number of graduates who default on their loans and accumulate large fees owed to collection agencies. His bill proposes shifting all repayment to an income-contingent plan that would cap payments at fifteen percent of discretionary income and would eliminate loan forgiveness. To protect borrowers, interest would accrue instead of compounding, and would stop accruing once it reached 50 percent of the initial balance.
¹²⁸

Green energy
. Federal incentives for domestic energy production go back to the nation’s very founding. In 1789, the first Congress imposed a tariff on the sale of British coal brought here as ship ballast. Throughout the twentieth century, the government promoted energy production and technology through a variety of direct and indirect subsidies including tax preferences, price controls, trade restrictions, provision of subsidized public services, regulation, conservation subsidies, and refusals to tax negative externalities. The great bulk of these subsidies went to oil and gas and, to a lesser extent, nuclear production.
¹²⁹
Coal, which policy has subsidized in a number of ways, creates a large climate change externality.

Today the distribution of energy subsidies is altogether different. In 2011, the Institute for Energy Research, using Department of Energy data, calculated that the subsidies per unit of electricity actually produced was 64¢ for oil, gas, and coal; 82¢ for hydropower; $3.14 for nuclear energy; $56 for wind energy; and $775 for solar energy; In 2009, fossil fuels accounted for 78 percent of energy production but received only 12.6 percent of tax incentives. Renewable energy sources (most of which were hydropower) accounted for 11 percent of production but received 77 percent of the tax subsidies, not including direct subsidies. (On the revenue side, fossil fuel industries paid more than $10 billion in taxes; renewables were a net drain.
¹³⁰
) The subsidy imbalance is even greater on a per-energy-unit basis: renewables receive three times as much as fossil fuels per kilowatt hour.
¹³¹

Since the 1970s, federal policy has increasingly focused on promoting renewable (“clean”) energy alternatives in order to reduce pollution, greenhouse gases, and dependence on foreign suppliers, as
well as foster economic growth.
¹³²
If this policy were achievable in a cost-effective manner, no sensible person could possibly object. Success, however, is extraordinarily elusive. First, some regulations—those seeking to redirect product choices to improve green energy efficiency, for example—arbitrarily assume consumer irrationality and market failures.
¹³³
Second, energy is produced today in more globalized markets shaped in part by opaque political decisions. This makes predicting future market conditions—the unattainable holy grail of all investors and other economic actors throughout history—more difficult than ever.

Unless one views clean energy as a moral imperative that trumps all other considerations, it is far from being cost-effective relative to conventional sources—a fact that as of June 2013 showed little sign of change. The costs of conventionals, particularly relatively clean natural gas, are far lower today despite smaller subsidies, and are declining even more due to new discoveries and technologies. Renewables, moreover, face a host of challenges. Much of the new infrastructure necessary to produce, gather, and distribute renewables to users in sufficient quantities is not in place and will be costly, and some EPA rules are perverse—for example, requiring refiners rather than producers to comply with biofuel mandates.
¹³⁴
Moreover, conventionals benefit from a deeply entrenched political status quo that renewable advocates cannot readily overcome.