Why Government Fails So Often: And How It Can Do Better (29 page)

Regrettably, such instances of social changes that increase government effectiveness are vastly outnumbered by changes that frustrate or impair it. This can happen for several reasons. New developments may render a policy anachronistic by altering the nature of the underlying problem and the means available for dealing with it. Consider two cases in point. The Delaney Clause, enacted by Congress in 1958, prohibited the sale of processed foods with any additive or pesticide—regardless of its concentration—that causes cancer in animals or humans. By the 1970s, analytical chemists armed with powerful new mass spectrometers could detect infinitesimal amounts of residues, which rendered the Delaney Clause unworkable and dangerous to the food supply and public health. Congress finally repealed it in 1996, long after it had become anachronistic.

A second example comes from the financial services industry. In the decade or so leading up to the economic meltdown of 2008, the industry created, traded, collateralized, securitized, and multiplied instruments so exotic, complex, and opaque that federal regulators could barely comprehend the intricacies of these new markets, much less assess their effects and how to control them.
⁵⁸
Indeed, as I write in December 2013, more than three years after Congress enacted the Dodd-Frank law, the financial regulators finally managed, under immense White House pressure, to issue the Volcker Rule, only to retract part of it barely a week later. I discuss the rule in
chapter 8
.

Another obstacle to policy adaptability is what political scientists call “policy inheritance.” Policy innovation is severely limited by the policies adopted earlier, which leave current policy makers with limited freedom of action. This inheritance effect is particularly powerful if the older policies were instituted by members of the same political party, but it is not confined to such partisan continuities.
⁵⁹
Indeed, as discussed in
chapters 3
and
6
, political inertia is built into the constitutional system and the policy process. Because of separation of powers and numerous veto points, new statutory (and even administrative) policies are difficult to enact, but once in place are difficult to dislodge. This inertia likely explains why it took years, and an experiment, to finally move Medicare to adopt competitive bidding for the medical equipment it buys.
⁶⁰

Social change can also confound existing policies by significantly altering the magnitudes of their benefits and costs and undermining the rationale for the earlier policy. Consider the tax subsidy for employer-provided health insurance. Originating as a relatively small temporary adjustment to World War II wage controls, it has become the single largest “tax expenditure,” costing the Treasury more than $175 billion per year. In its current bloated form, this policy causes many large inefficiencies in the health care system while disproportionately favoring workers who already enjoy relatively high wages and fringe benefits. Most health reformers and budgetary hawks favor eliminating or severely limiting this subsidy, but it is so politically entrenched and union-friendly that it cannot be dislodged.
⁶¹
Another example of an old policy that makes far less sense today than it did at its inception is the impacted aid program, enacted in 1950 to provide formula-based federal assistance directly to local school districts serving students whose parents work for the military, are Native Americans, or are otherwise “federally connected.” It now costs about $1.3 billion a year, but its original rationale—that such districts needed special assistance—has been undermined by the passage of numerous federal education and Native American assistance programs starting in the 1960s, and by the fact that the federal connection often is
on balance economically advantageous to these districts rather than burdensome. Again, however, the program is politically bulletproof.

The federal statute books are replete with programs of this kind—programs for which the cost and other parameters have greatly increased while the original rationale for the policies has eroded, and which nevertheless remain, even many decades later, essentially impervious to fundamental reform or repeal. Indeed, they are as impregnable as Fort Knox. As one contemplates these programs, then, one is immediately reminded of the French expression
plus ca change, plus c’est la meme chose
(the more things change, the more they stay the same). The government’s rapid, decisive responses to Pearl Harbor and Sputnik were exceptional. Far more typical is immigration policy, where adaptation to new conditions is far slower than in other countries like Canada and Australia whose more flexible systems compete with us for highly skilled immigrants. Indeed, our legal immigration policy still operates under strictures of the 1965 Hart-Celler law, which was updated only marginally in 1990 and before immense technological and labor market changes transformed the policy challenge.
⁶²
As of June 2013, nearly a half century after Hart-Celler, it was still not clear that Congress would bring our system into the twenty-first century.

Some programs with enduring rationales—for example, entitlements like Social Security (discussed above and in
chapter 11
) and Medicare (also discussed in
chapter 11
)—have over time become fiscally unsustainable in their present forms, necessitating some combination of benefit or eligibility reductions or tax increases that policy rigidity makes difficult. In both cases, Congress and the president have failed the politically difficult test of putting them on a fiscally sound footing for the long term—a task that is much easier for Social Security than for Medicare.

Two final points about inflexibility deserve mention. First, the mobility of capital and people, which
chapter 4
discusses in connection with localism, often constrains policy adaptability. Second, policy rigidity can be desirable to the extent that it fortifies the government’s
credibility and thus its ability to induce reliance and investment by others. I now turn to this subject.

INCREDIBILITY

Even before the Constitution was adopted, James Madison saw the danger of fickle policies, commenting that “no great improvement or laudable enterprise can go forward which requires the auspices of a steady system of national policy.”
⁶³
Government credibility is usually mentioned in analyses of foreign and national security policy, where promises and threats can only be effective if they are believed.

But credibility is at least as important for domestic policy effectiveness. Many programs can succeed only if the federal government induces private and other governmental actors to do things that they will not do if they doubt that it will honor its own commitments.
*
In order to do that, the government must radically reduce the amount of uncertainty that chronically plagues its decisions,
†
including what theorists call “primary ignorance”—when decision makers do not even recognize that they are ignorant of risks and thus court failure without even knowing that they are risking it.
⁶⁴
Uncertainty both increases government’s costs and raises costs to the private actors on whose decisions policy success ultimately depends. These costs are difficult to quantify, of course—three Stanford University and University of Chicago economists find that economic policy uncertainty is much greater today than in the past, estimating its drag on the economy at
$261
billion since 2011 alone
.
⁶⁵
William Galston, a leading Democratic policy expert at Brookings, cites new Federal Reserve research consistent with these estimates; together, these studies show
that policy uncertainty is having dire effects on the economy, especially on hiring.
⁶⁶

This credibility imperative creates a deep and inescapable temporal paradox: for a policy to be effective, people must believe that the government will discharge today’s commitments in the future (usually in the medium or long term), yet the demands of democratic legitimacy and accountability require government to respond to changed conditions in that future in ways that will impair its credibility.
*
Economists Dani Rodrik and Richard Zeckhauser made this important point twenty-five years ago, providing a number of policy examples. The government may have good reasons to adopt an amnesty for tax evaders or undocumented immigrants, but unless it can persuade them that the amnesty is a one-time-only policy, it simply encourages more people to evade taxes or come illegally in hopes of a future amnesty, which is precisely what has happened in both cases. A public utility commission may want to induce a utility to construct a new power plant, but unless it can guarantee the utility that it will grant a future rate increase to cover the necessary investment, the utility will not make the investment.
⁶⁷
Yet in both of these cases (and countless others), the government cannot credibly make the necessary commitments for two related reasons: (1) any rational person will doubt that it will fulfill them because (2) one knows that it will change direction if future conditions demand it. In economists’ terms, the government’s utility preferences will change intertemporally as conditions change. In time B, the government will act according to its new preferences. We must assume first that the government knows that this will undermine its credibility both in time 1 (because of others’ rational expectations about how government will act in the future) and in time 2 (when it actually changes direction) and, second, that it takes this credibility problem into account in deciding what to do. Nevertheless,
a greater necessity—its obligation to voters in time 2—demands that it do what the conditions then require.

This credibility problem affects almost all public policies. Rodrik and Zeckhauser show that the reasons are structural, not just political, and that the problem is far worse for government than for private actors. First, as just noted, government has a moral and political obligation to meet its constituents’ felt needs, which change over time. (Discharge of this obligation can take forms that create severe moral hazard [discussed in
chapter 5
], such as the 2008 bailouts of General Motors, AIG, and more than 700 banks. Five years later, AIG had repaid its debt but the Treasury was likely to lose billions on its investments in GM and the smaller banks.
⁶⁸
) In contrast, a private actor may have an incentive to respond to changes in its market or other domain, but it has no such obligation to do so. Second, the more discretion officials have under the law, the greater their freedom of action to change direction in the future—perhaps by repudiating the existing policy, failing to enforce it, or deciding not to make the promised investment. Even if the change is justified on the merits, it also exacerbates their credibility problem. Third, long-term credibility is even more essential when the policy depends on inducing private firms and other entities and individuals to make long-term, asset-specific investments—that is, economic or other commitments whose value to them would be much lower if redeployed to a different transaction—because investors fear being held hostage to opportunistic governmental behavior. In the public utility case, once its investment in the new power plant is locked in, the commission’s short-term incentive is to reject unpopular rate increases. Anticipating this, of course, the utility may not make the investment in the first place. Similarly, some states have resisted Obamacare’s full-cost federal subsidy for expanding Medicaid, fearing that in three years when the subsidy decreases, they will be stuck with higher costs.
⁶⁹
(One wag likened this to receiving a free baby elephant.)

The greatest structural impediment to government credibility, however, is that it has fewer ways than private actors do to commit itself to a particular course of action over time. It can bind itself by
statute, but statutes can be repealed by future Congresses (think of the Bush-era tax cuts) subject only to constitutional principles that—in recognition of government’s moral duty to respond to changed conditions—are not very constraining. It can bind itself by contract, but public contract law provides government with some special defenses and immunities that make enforcement against the government less certain and more costly than in private contract law. To be sure, future government actions that disappoint these statutory or contractual expectations will make any course changes more costly, which further reduces its credibility, but the government may decide that the advantages of changing course are worth that cost in credibility. And those who might do business with the government know all this, of course, and must factor its lack of credibility into their own decisions about whether to deal with it and under what terms.

Abrupt changes in policy are common. For example, president George W. Bush proposed a $1.2 billion program in research funding to develop a hydrogen-powered car. Only five years later, President Obama cut 80 percent of the funding for this program, proposing instead to place his money on electric-powered cars, leaving scientists and entrepreneurs who had invested in hydrogen-car research in the lurch. One must doubt that they and people like them will stake their own funds and careers on the next round of government promises. Indeed, a 2009 National Academy of Sciences report cited “lack of sustained policies” as one of the three top barriers to promoting renewable energy, and new evidence further demonstrates this effect.
⁷⁰
Congress vastly expanded the National Institutes of Health, whose success depends entirely on attracting and retaining top scientific talent, only to lose much of that talent by cutting appropriations. In real dollar terms, the agency’s funding has decreased by roughly 20 percent over the last decade.
⁷¹