Why Government Fails So Often: And How It Can Do Better (34 page)

Informational demands on regulators
. Government cannot influence markets effectively unless it has access to timely, accurate, inexpensive information about market actors, dynamics, and strategies—information that is appropriately tailored to the immense market diversity just described. Yet the government repeatedly experiences systemic difficulties in efficiently managing information flows generated by private market actors at any reasonable cost. Some information management is easier than others. For the Social Security Administration to send pension checks to seniors, for example, it does not need much individualized data about them other than their names, ages, addresses, and the number of quarters during which they have been employed—all readily ascertainable from standardized public records. It is only a slight exaggeration to say that this part of the program is essentially a check-writing operation. Contrast this, however, with the same agency’s disability benefits program, which must make individualized determinations of disability status under vague legal standards and with a complex process and uncertain evidence, while relying to some (uncertain) degree on earlier decisions made by other people and institutions that together largely determine outcomes: federal and state hearing examiners, private physicians, vocational rehabilitation agencies, and other actors beyond close federal supervision or control.
³⁷

The problem is illustrated by the Bush and Obama administrations’ failed efforts to generate accurate clinical information by subsidizing providers to adopt electronic medical records (EMRs). As a
New York Times
article on the $6.5 billion program put it, EMRs “can make health care more efficient and less expensive, and improve the quality of care by making patients’ medical history easily accessible to all who treat them.”
³⁸
A 2005 RAND Corporation study, which digital record vendors strongly promoted, induced policy makers to give billions of dollars in federal stimulus subsidies to induce providers to adopt EMRs, with scant concern for how this would be implemented.
³⁹

Yet implementation has been a costly fiasco. The RAND study was grossly overoptimistic, as the Congressional Budget Office predicted
at the time. The new systems were so “clunky and time-consuming” that it slowed the public health system of a major California county to a crawl, with doctors and nurses seeing only half as many patients as usual. The systems have increased Medicare billings and even facilitated fraud by inviting providers to check boxes opportunistically. In November 2012, the inspector general of the department that administers Medicare issued a scathing report that Medicare’s payment of incentives to providers to digitize their records has inadequate safeguards against fraud and abuse, despite the agency’s promise to control overbilling.
⁴⁰
As one internist put it, “reading the electronic chart has become a game of looking for a small needle of new information in a haystack of falsely comprehensive documentation and outdated, copied text. Why do we doctors to this to ourselves? Largely, it turns out, for the same reason most people do most things: money.”
⁴¹
Years after the subsidy payments, EMR’s ballyhooed promise of lower costs remains wholly hypothetical.
⁴²

Price and substitution effects
. Government programs cannot escape the inexorable market-driving law of supply and demand, whose price effects in turn activates behavioral incentives in ways often unwanted by policy makers. Nevertheless, policy makers often ignore this basic fact of programmatic life in ways that undermine their policy goals.

If government builds more roads to alleviate traffic congestion, the lower congestion costs will attract more drivers to the roads, thus frustrating the policy. If it seeks to reduce the price of gasoline, the lower price will discourage producers from bringing new supply online, thus raising the price; the lower price will also increase demand, which will also push the price back up. Requiring more energy-efficient air conditioners may actually increase power consumption as consumers respond to the lower electricity costs by increasing their utilization. Requiring child-resistant packaging of drugs may reduce safety on balance by making them harder and slower (i.e., more costly) for adults to open.
⁴³
If one must be homeless to receive free housing, then—
ceteris paribus
—more people will claim that they are homeless, and if the shelters are more attractive, they will tend to remain
longer.
⁴⁴
When the National Highway Traffic Safety Administration (NHTSA) mandated safety features for cars, drivers took more risks and the accident rate rose
⁴⁵
—an example of the Peltzman effect, according to which people adjust their behavior to a regulation in ways that counteract the regulation’s intended effect. Other examples of the Peltzman effect appear in
chapter 8
.

A related phenomenon is the substitution effect, in which raising a good’s cost (or banning it outright) will induce its producers to find cheaper (or legally permitted) substitutes. This substitution is often socially undesirable on balance. Thus, when the Environmental Protection Agency (EPA) raised gas mileage standards, auto manufacturers complied by switching to lighter cars that did consume less fuel but were more dangerous to occupants in collisions.
⁴⁶
(The price effect also encourages owners to offset the lower fuel costs by driving more.
⁴⁷
) Banning DDT led users to substitute other insecticides that are even riskier or more costly (or both), like malathion. And if they fail to find and deploy a better substitute, the result may be disastrous, as with the countless malaria-related deaths that scientists attributed to the government’s ban on DDT.
⁴⁸
Some Occupational Safety and Health Administration (OSHA) regulations have also produced “regrettable substitution,” in which employers replace one source of deadly fumes with an even more dangerous one.
⁴⁹
The 2009 stimulus law sought to create jobs, but its provisions raised the cost of hiring workers (or reduced job-seeking) in several ways—for example, with easier access to food stamps, lengthened unemployment benefits, minimum wage increases, and retention of the labor-cost-inflating Davis-Bacon Act. Economist Casey Mulligan estimates that such measures caused a substantial part of the precipitous decline in labor force participation and hours worked from 2007 to 2011.
⁵⁰

The list of policies frustrated by these market effects on behavior is almost endless. All things considered, the new equilibrium that is produced once a policy’s price and substitution effects work their way through the system may improve the situation overall—or it may make the situation much worse, especially for those with fewer choices. Dodd-Frank compliance is imposing vast new costs on banks—an estimated
$34 billion a year—that the market has passed on to low-income consumers in sharply reduced access to credit cards, loans, and other forms of credit.
⁵¹
When products are banned, consumers may turn to more costly and less satisfactory ones, including more dangerous ones. As political scientist Aaron Wildavsky showed, laws that raise consumers’ costs in order to improve their health also produce offsetting health risks by reducing their wealth. As he put it, wealthier is healthier.
⁵²
We can safely predict that policies which fail to consider or underestimate these price and substitution effects will produce outcomes that are less effective, more costly, and sometimes the opposite of what they intended.
Chapter 8
, on policy implementation, presents many more examples.

Transjurisdictional
effects
. Although some markets (particularly those involving personal services) remain local even today, most markets today are broader than that: interstate, regional, national, or global. Widespread evasion of state excise taxes and other regulations—for example, on cigarettes,
⁵³
liquor, and other products—is widespread, just as it is in Europe, where cross-border travel is simple and inexpensive. (Denmark’s tax on fatty foods was so easily evaded by shopping in neighboring countries that it was quickly repealed.
⁵⁴
)

The federal government lacks extraterritorial jurisdiction to regulate many foreign markets, and even when it possesses such jurisdiction, it is constrained to allow or even encourage foreign competition, which is advantageous to domestic consumers and some commercial interests (importers, for example) even as it may disadvantage some domestic firms, at least in the short run. The more telling point about foreign market actors, however, is that their actions may subvert domestic policy goals. Banking regulators, for example, are finding it difficult to subject foreign banks with large U.S. operations to the Dodd-Frank law’s capital, leveraging, and other requirements for fear of inducing evasive tactics that could harm the interests of American banks and consumers.
⁵⁵
(A similar transjurisdictional problem exists among our states—where insurers risk systemic insolvency by exploiting regulatory differences across borders.
⁵⁶
) Informal markets,
discussed in
chapter 8
, can arise precisely to exploit this transjurisdictional problem.

The possibility that government taxes and regulations will encourage domestic firms to move some or all of their jobs, activities, and investments abroad—where those restrictions are less onerous—is a constant, serious constraint on domestic policy, one with political resonance that conservatives often exploit effectively to defeat such impositions. The Sarbanes-Oxley law, passed in the wake of the Enron fraud, made listing shares on U.S. stock markets so costly that firms increasingly looked abroad for capital or remained private. Not coincidently, the U.S. share of initial public offerings fell from 67 percent in 2002 when the law passed, to 16 percent in 2011.
⁵⁷
Another constraint on policy is a concern that foreign firms will use U.S. law to compete against domestic ones. An example is the Jumpstart Our Business Startups Act of 2012, which Congress designed to create job growth in the United States by reducing the costs to new small firms of raising capital and to established firms of going public, partly by reducing their Securities and Exchange Commission (SEC) disclosure obligations. A recent assessment of the new program finds that foreign companies are using the law to compete more cheaply against American firms; this tends to reduce domestic job growth while enhancing growth abroad. Some foreign companies are also using the program to market to American investors with less disclosure than their domestic competitors make.
⁵⁸
Such foreign competition may on balance be beneficial (or not) to the U.S. economy; my point here is that it often undermines Congress’s more insular policy goals.

Political influence of market actors
. For a host of reasons, market actors exert enormous influence over political outcomes and government policies. Indeed, compared with more centralized parliamentary democracies, our political system is relatively porous and responsive to private interests at multiple points in the policy process.
⁵⁹
This is particularly true where a single government agency intensively regulates a market or related markets, such as the SEC, the Food and Drug Administration (FDA), the NHTSA, and the Nuclear Regulatory Commission. In such regulatory contexts, the mutual dependencies
of regulator and regulated—for information, personnel, political support, and sometimes even revenues (as with the FDA)—are great. The hoary notions of agency “capture” by the regulated firms and of an “iron triangle” of agency, regulated interests, and congressional subcommittee beg a host of difficult questions, as noted in
chapter 4
. These notions, moreover, also vastly oversimplify a complex, opaque reality, as political scientists have consistently demonstrated.
⁶⁰
Still, a subtler, less deterministic view of the populist critique is certainly correct: mutuality of influences and interests are fundamental aspects of regulatory politics that cause program performance to deviate, often greatly, from policy goals.

In regulatory contexts such as health and safety, the environment, and civil rights, the agencies regulate a broader, more diverse group of industries, and the industries’ influence over them tends to be more diffuse. Examples include OSHA, the EPA, and the Equal Employment Opportunity Commission (EEOC). Here the interests among the industries and within an agency may conflict, as regulatory units primarily concerned with one industrial sector (as in OSHA) or with one risk medium (as in the EPA’s solid waste program) compete for congressional attention, budget, legal authority, publicity, and political support. At the same time, the influence of “public interest” organizations demanding that the agency regulate more vigorously—labor unions (vis-à-vis OSHA), environmental groups (the EPA), and civil rights organizations (the EEOC)—is often considerable.
⁶¹

This dynamic of mutual influence and support between market actors and government is not confined to regulatory agencies but extends to agencies charged with other kinds of missions. Aerospace companies, for example, exercise strong influence over the relevant government procurement agencies. Veterans groups shape the activities of the Veterans Administration. Scientific communities shape with the agendas and processes of the National Science Foundation and the National Institutes of Health. The Justice Department is attentive to the substantive views of the American Bar Association, with the ABA’s influence waxing or waning with different administrations and different issues. (Republican administrations, for example, have come
to be suspicious of the ABA’s relatively liberal orientation in its assessment of judicial candidates.) Even a seemingly apolitical agency such as the Census Bureau attracts strong interest from industries, demographers, civil rights advocates, and other groups eager to shape its methodology and use its data.
⁶²