Why Government Fails So Often: And How It Can Do Better (23 page)

Increasing moral hazard can be sound policy when it is part of a coherent risk management strategy. In 1970, for example, Congress imposed a fifty-dollar cap on consumers’ liability for unauthorized credit card use. Although this protection surely makes us somewhat less careful than if we remained at risk for large losses from credit card theft, it also reassures us about the safety of Internet shopping—a huge social gain—and shifts the risk of unauthorized use to credit card issuers, who are in the best position to efficiently monitor it.
²⁷

Unfortunately, this happy outcome is not true of the vast, egregious moral hazard created by programs to benefit financial giants, higher-income individuals, and students. Fannie Mae and Freddie Mac—dubbed the “toxic twins” by critics—engineered home mortgage policies and markets that systematically created vast moral hazard, ostensibly for the popular purpose of expanding homeownership.
²⁸
(For all their efforts, homeownership in the United States is not particularly high when compared with other nations; it ranks twentieth, just behind the United Kingdom.
²⁹
) After exhaustively reviewing the data on housing finance markets, economists Dwight Jaffee and John Quigley find that none of the rationales invoked to justify the agencies’ policies—especially the need to preserve the long-term, fixed-rate mortgage—is valid. Quite the contrary; they were detrimental to that goal.
³⁰
In addition, these agencies helped inflate the housing bubble by pressing lenders to make subprime loans to high-risk borrowers who, when the bubble burst, had irresistible incentives to walk away from their underwater mortgages. This left lenders—themselves often insured or hedged in ways that increased moral hazard in
their
lending decisions—and ultimately the government (moral hazard’s chief victim) to bear the immense losses, with catastrophic damage to the larger economy. The Treasury estimates that Fannie and Freddie bailouts could amount to $200 billion by the end of 2015. This estimate takes into account some paybacks (
actually, reduced loss reserves) generated by the recovering housing market, paybacks often mischaracterized as “profits” that may tempt Congress to preserve the agencies rather than reform them.
³¹
It does not take into account, however, the opportunity costs of these subsidies—an omission that an expert on subsidies says “would not pass Economics 101.”
³²
As
chapter 8
notes, the Federal Housing Administration is now replicating these moral hazard–stimulated losses on a somewhat smaller but still massive scale—despite the searing experience of Fannie and Freddie. In addition, the regulators seem likely to weaken the traditional 20 percent down-payment requirement, a key bulwark against moral hazard in home mortgages.
³³
Finally, quite apart from the moral hazard issue, a policy to subsidize people to own rather than rent homes can actually be harmful.
³⁴

The government seems to have learned the wrong lesson from this fiasco. The Dodd-Frank reform act has actually increased moral hazard by broadening Wall Street’s safety net—and thus taxpayers’ exposure—in at least two ways. First, it classified some of the largest financial institutions as systemically important financial institutions (SIFIs), which many experts think is tantamount to “too big to fail” status and gives them new marketing and credit advantages. Indeed, a recent Bank of England study finds that these advantages currently increase the profits of the twenty-eight SIFI banks around the world by some $500 billion—substantially more than their combined profits!
³⁵
Regulators’ demand for “living wills” for these institutions is unlikely to be effective.
³⁶
In speeches in July 2013, both the Treasury secretary and the Federal Reserve chairman expressed frustration and concern about the government’s failure to end “too big to fail” among SIFIs.
³⁷
Yet at the same time, regulators proposed to
extend
SIFI status (and thus “too big to fail” status) to
nonbank
entities such as insurers. (This raised a related policy issue: whether it makes sense to treat such entities like banks given their very different risk portfolios, incentives, and regulators.
³⁸
)

Second, Dodd-Frank creates a system for “orderly resolution” of insolvencies that may actually make future bailouts more likely.
³⁹
The Federal Reserve’s policy of keeping interest rates so low for so long,
and of not raising margin requirements since 1974, invites excessive borrowing, speculation, asset bubbles, and future inflation (while harming savers).
⁴⁰
Economists will have to puzzle out whether the policy was justified by the slow recovery from the Great Recession, or instead delayed the recovery and set the stage for more insolvencies and crises.

Other examples of moral hazard abound. Under a 1990 law, Congress
requires
the massive federal credit programs to systematically and significantly underestimate and misrepresent their true costs by accounting for loans through essentially ignoring default risks. This practice, denounced by the Congressional Budget Office (CBO) most recently in March 2012,
⁴¹
creates
two
kinds of moral hazard: it encourages government to lend too much by treating its loans as virtually risk-free, and it encourages high-risk borrowing by those who will default. The distorting effect of’ this combination on policy is evident in the enormous but concealed default-driven costs of the student loan programs detailed in
chapter 8
.

The Pension Benefit Guaranty Corporation (PBGC), which guarantees employers’ unfunded defined-benefit pension debts under certain conditions, sets firms’ premiums far too low to underwrite their true risks of default. This tempts firms, especially those in bankruptcy, to shift their obligations to the PBGC, producing a $26 billion deficit in 2011, which could rise to $35 billion if American Airlines succeeds in offloading its pension obligations onto the agency.
⁴²
Similarly, the National Flood Insurance Program (NFIP) run by the Federal Emergency Management Agency (FEMA) encourages homeowners and businesses to locate and build in the flood plain and beaches initially and then to return and rebuild there once the floodwaters recede—a form of moral hazard that Congress reinforced in its Hurricane Sandy legislation in 2013. Alert to this moral hazard, private insurers have fled the flood insurance field, leaving the government to pay the claims with premiums that are far too low to cover the true risk of loss. Indeed, the Hurricane Sandy aid bill enacted early in 2013 bailed out the NFIP for $9.7 billion. Worse, half of all NFIP policyholders apparently cancel their
coverage after only three or four years, thinking that it isn’t worth even the subsidized cost,
⁴³
and Congress is under great pressure to increase the subsidies now that a 2012 law raising the rates has kicked in.
⁴⁴
Finally, the Affordable Care Act increases moral hazard risks by encouraging the young and healthy to delay buying even subsidized coverage until they are ill, given the very low penalty for not buying insurance, their right to buy it later despite preexisting conditions, and their access to cheap or free (to them) emergency room care in the meanwhile.

Numerous other instances of government policies that encourage excessive risk-taking can be cited. This likely contributed to the savings and loan crisis in the late 1980s, in which the Federal Savings and Loan Insurance Corporation bailout cost taxpayers over $160 billion.
⁴⁵
Large new subsidies to FEMA in 2012—it borrowed $18 billion from the Treasury to cover the 2005 and 2008 hurricanes, and will have to borrow even more to cover the Hurricane Sandy losses—impose huge, growing losses on taxpayers while preserving most of the moral hazard by owners that invited such losses.
⁴⁶

In similar fashion, federal drought insurance programs are replete with moral hazard. A recent study by agricultural economist Bruce Babcock finds that the programs, which are also discussed in
chapters 6
and
8
, are especially costly because they guarantee farmers, regardless of their wealth, a portion of their projected income rather than simply paying them for their damaged crops, so farmers buy more coverage than otherwise. As a result of the subsidies, which cost taxpayers over $6 billion a year, many farmers made more money from insurance payouts during the drought than they would have made from healthy crops! Further multiplying the program’s egregious moral hazard, taxpayers cover about 62 percent of the costs of this insurance, which is sold by fifteen private insurance companies which keep the profits while receiving a federal guarantee against any underwriting losses.
⁴⁷
Henry Olsen, former vice president of the market friendly American Enterprise Institute, calls this “obscene,” utterly inconsistent with free markets.
⁴⁸
To make matters even worse, it also induces excessive development in arid
areas, which in turn produces more subsidized insurance and cost-ineffective water projects.
⁴⁹

Other programs induce moral hazard by encouraging the government to make “bad bets.”
⁵⁰
Again, the politically popular student loan programs, detailed in
chapter 8
, confer entitlements rather than screening for ability to succeed at school or to repay the loan, with the predictable result being high and rising default rates. And as Jack Donahue has noted, the government’s own policy and management failures have spawned a different type of moral hazard, in which large aerospace and other contractors derive almost all of their business from the government, feeding on and enabling its chronic weaknesses.
⁵¹
The obvious psychological analogy is to the codependency of addicts and their enablers.

Unemployment insurance, which has been expanded on relatively easy terms, discourages workers from seeking or taking new jobs until their benefits run out, a delay that makes it less likely they will find new employment.
⁵²
(Pending proposals to bar firms from considering applicants’ unemployment in hiring decisions would likely have the same effect.) Government health insurance programs increase moral hazard when they reduce the costs of self-destructive behavior—for example, by subsidizing costly treatments for emphysema, which is almost always caused by smoking, of organ transplants for heavy drinkers, and of obesity-related diseases for reckless overeaters.

Moral hazard is common in government programs targeted at the poor, as one usually can receive benefits only by remaining poor.
⁵³
If one’s income rises too much, benefits either phase out or stop abruptly (called a “notch”), which means a very high, work-discouraging marginal tax rate. Indeed, the effective marginal tax rate on the thirty-eight million households receiving benefits from at least one federal welfare-entitlement program in 2011 averaged from 36 percent to over 50 percent (where additional income made them ineligible for Medicaid).
⁵⁴
But difficult policy trade-offs involving moral hazard are almost inevitable in programs designed to help people in difficult circumstances. Discharging bankrupts’ debts, for example, invites irresponsibility but also gives them a fresh start.
⁵⁵
But some programs
can be structured to reduce moral hazard. The Earned Income Tax Credit, discussed in
chapter 11
, does so by providing wage supplements to low-income workers; those without jobs are not eligible. Benefits for the elderly, the widowed, and the seriously disabled entail little or no moral hazard, either because the beneficiaries have no way to affect either the likelihood or the magnitude of the compensable event (old age, spousal death) or because the compensable event (serious disability) is so debilitating that no rational individual would significantly risk it. The Veterans Administration, however, also increasingly defines disability to include minor conditions, including common age-related ailments such as hearing loss, lower back pain, and arthritis. This helps to account for the nearly 900,000 backlogged disability claims, which are discussed in
chapter 6
.
⁵⁶

Most social programs adopted during the New Deal era hoped to minimize moral hazard by targeting individuals whose misfortunates occurred through no fault of their own. Since these misfortunates were common, the risks of such losses could be reduced or socialized by being spread among a much larger pool of individuals. This helped to justify the transfer of resources to those unfortunates who turn out to be what have been called “bad draws.”
⁵⁷
It also made these programs actuarially feasible, once they matured. Rational risk-averse voters, not knowing whether they would be lucky or not, would likely favor such a system—and the American public has emphatically agreed.

To the extent that most poor people are eager to escape poverty even if it means losing their benefits, programs that help them do so mitigate moral hazard but not completely, due to what Charles Murray calls “the law of unintended rewards: Any social transfer increases the net value of being in the condition that prompted the transfer.”
⁵⁸
Nicholas Kristof, a
New York Times
columnist and ardent supporter of antipoverty programs, notes some tragic examples of this in Appalachia.
⁵⁹
Parents, he finds, often pull their children out of literacy classes to make them more likely to qualify for a monthly check from the federal Supplemental Security Income program for being mentally disabled. Young people don’t join the military, a
traditional escape route from poverty, because they can more easily get food stamps and disability payments (discussed just below). Antipoverty programs’ incentives tend to reduce marriages, which in turn tend to reduce poverty. Kristof fails to mention a far more fateful, socially destructive example: poverty programs have encouraged some unmarried women to drop out of school or the workforce to have children that they could not otherwise afford. (The 1996 welfare reform, discussed in
chapter 11
, reduced this incentive but did not wholly eliminate it.) Cash transfer programs also encourage low-income mothers to leave unhappy marriages, which may improve their own lot but damage the life prospects of their children—especially boys.
⁶⁰