Why Government Fails So Often: And How It Can Do Better (39 page)

The Freedom to Farm Act of 1996 addressed many of these problems and shifted American agriculture toward a free market system. It suspended the authority for income support based on market prices and ended planting restrictions. However, these reforms were short-lived. By 1998, commodity prices had fallen substantially, and Congress soon reinitiated direct payments to farmers. These “market loss payments” totaled $21.4 billion between 1999 and 2001.
¹⁷
Moreover, the government began to require farmers to enroll in subsidized insurance programs; by 2012, these crop insurance subsidies grew to $15.8 billion. Over this same period, the conservation requirements and environmental protections of the 1996 law withered away. Politically, much of this reversion to the status quo has been driven by Democrats and agribusiness lobbyists. President George W. Bush resisted the 2003 farm bill and vetoed the 2008 farm bill outright. Despite this opposition, both bills continue the long and damaging tradition of federal support for American agriculture.

International trade policy has produced some consumer gains through free trade agreements on many nonagricultural products,
¹⁸
but the remaining import quotas and tariff protections outweigh those gains. Synthesizing the published empirical studies, Winston shows that “trade protection has mainly generated gains to established U.S. industries that fall far short of the losses to consumers.” (He also notes the large consumer losses created by the Jones Act, discussed in
chapter 6
.) These protectionist policies are bipartisan, as presidents of both parties succumb to them—most recently, George W. Bush on steel and Chinese clothing, and Barack Obama on Chinese tires.
¹⁹

Another area of economic regulation, not mentioned by Winston, is immigration policy, which limits the number and type of people who may come to the United States to live and work and subjects them to a variety of legal conditions. Because immigration policy is designed to serve a number of different purposes—family unification; labor market expansion; economic growth; innovation and entrepreneurship; protection of human rights; source-country diversity; diplomacy; and others—assessment of its effectiveness is difficult. Historically, immigration has greatly enriched American life and promoted economic vitality, but recent policies have failed egregiously—a fact that has long been acknowledged by the leadership of both political parties. Border and interior enforcement, despite large and growing costs, have not prevented the illegal presence of an estimated ten to eleven million undocumented immigrants, nor has Congress enacted any substantial legalization of this population since 1986. A large number of families contain both legal residents and illegal ones, but there is no coherent policy to address the problems that these mixed-status families present. Policy makers have failed to meet the needs of employers in agriculture and firms in other industries seeking workers with skills in science, technology, engineering, and mathematics—workers who are eager to come to the United States but are increasingly being lost to competing nations with more flexible and responsive immigration systems.
²⁰
By almost any measure, then, U.S. immigration policy has abjectly failed, with no remedy in sight.
²¹

Regulation of safety
. Safety standards regulation is a mixed bag. Standards mandated by the Consumer Product Safety Commission have produced little if any reduction in accident rates. The public
health stakes are far greater in the Food and Drug Administration’s premarket approval of drugs and medical devices, its regulation of advertising of health claims, and its prescription requirements. Regulatory effectiveness has been analyzed extensively with respect to the costs and benefits of different regimes of premarket testing, labeling, and postmarketing surveillance. Such assessments have also examined issues of consumer compliance with labels, regulatory delays in approving valuable new drugs, and consumer behavior that offsets potential regulatory benefits.
²²

Long delays in the implementation of firefighting technologies have contributed to huge losses of life and property from recent conflagrations. In July 2012, less than a year before the devastating Colorado fires, the
Denver Post
reported that the responsible federal agencies had failed to put in place the Fire Program Analysis system that they had heralded back in 2002 and had predicted would take five years to implement. This despite several earlier GAO reports criticizing the agencies’ long delay and inadequate program.
²³

The benefits of National Highway Traffic Safety Administration standards have been affected by offsetting behavior, according to a number of studies using different methodologies. These behaviors include riskier driving induced by air bags, antilock brakes, and other mandated safety equipment. Analysts dispute whether such behavior has wholly erased those benefits, as the famous Peltzman study found,
²⁴
or merely reduced them.

Risk expert W. Kip Viscusi found a similar “lulling effect” to explain why FDA-mandated child-resistant packaging of aspirin and other analgesics did not affect poisoning rates but sharply increased child ingestion rates, as parents assumed that the packaging reduced the risk and thus were more careless (e.g., leaving the caps off the bottles).
²⁵
He also found no beneficial effects from Consumer Product Safety Commission regulation.
²⁶
Winston’s review of studies of workplace safety standards issued by the Occupational Safety and Health Administration (OSHA) and the Mine Safety and Health Administration concludes that they create few marginal benefits. Rather than the regulation reducing accident rates, the studies suggest, the decline
mostly reflects workplace mechanization, independent market incentives to reduce accidents (e.g., workers’ compensation premiums, lost work time, morality), weak enforcement, and inspectors’ tendency to focus on easily identifiable and correctable hazards rather than more elusive but important ones such as employee turnover. Also, OSHA has largely neglected long-term occupational disease in favor of its spotty efforts at accident prevention.
²⁷
A leading analyst of the program, John Mendeloff, notes that inspections in manufacturing have proved effective and some serious toxic exposures (e.g., to asbestos, lead, chromium) have been reduced. He observes, however, that the law’s safety impacts have been limited, partly due to few inspectors in the United States compared with other countries, the difficulty of inspecting construction (a particularly risky industry), and poor data collection on causal factors.
²⁸

Regulation of externalities
. Government has compelling reasons to regulate negative externalities—as with pollution, when one’s activity imposes costs (“spillovers”) on others without paying for them. Unless negative externalities are “internalized” (by making their producer pay for them), the activity will bear less than its full social costs; too much of it will be produced (an inefficiency), and others will have to bear those costs (an inequity). Winston’s review of the assessments finds that some externality controls are “outright failures” in that their costs have exceeded their benefits.
²⁹

The most important category of externality policy is environmental regulation. Environmental conditions and public health have improved significantly since the early 1970s when the basic framework of federal regulation was established.
³⁰
Still, it is hard to know the extent to which that regulation is responsible for this progress.
*
First,
few ex post cost-benefit analyses (CBAs) of these laws exist, and none evaluates the field overall. This dearth of ex post CBAs contrasts with the abundance of ex ante predictions of whether particular environmental regulations will be cost-effective.
³¹
These ex ante evaluations are usually regulatory impact analyses (RIAs) prepared by the promulgating agency, submitted to the Office of Information and Regulatory Affairs (OIRA) for review (assuming that it is a “major” rule), and used for planning and implementation. RIAs are not legally binding, but OIRA would not clear such a regulation unless the RIA passed an ex ante cost-benefit test. Unfortunately, such analyses tend to overpredict both costs
and
benefits.
³²
Indeed, a wide-ranging Office of Management and Budget review found that ex ante analyses often overstate both benefits and costs but especially benefits, which favors predictions of cost-effectiveness.
³³
The most significant and comprehensive ex post analysis was undertaken by the Environmental Protection Agency itself—a 1997 retrospective CBA of the Clean Air Act (CAA) in which the agency estimated that the benefits from 1970 to 1990 vastly exceeded the costs.
³⁴
Two more recent EPA studies reached similar conclusions.
³⁵

A second obstacle to analyzing the effects of environmental regulation is the difficulty of pinpointing which factors caused which outcomes. Before-and-after studies show that air pollution declined at the same rate before 1970, when the states were regulating it, and after 1970 under CAA regulation, suggesting that other factors may have been more causally important.
³⁶
Another causal factor is technological change, which regulation affects in complex ways. For example, newer technologies tend to be safer than older ones,
³⁷
yet regulation may not be responsible for emergence of those technologies. Indeed, although many programs seek to be technology-forcing, they sometimes
retard
, not promote, polluters to adopt cleaner methods by imposing tougher rules on new sources than on existing ones. Thus, many electric utilities continued to use their dirtier power plants because the CAA grandfathered them in, allowing them to operate under the old, lower standards while subjecting new plants to the more stringent, costlier rules.
³⁸
This substitution effect (discussed in
chapter 7
) also explains why the Corporate Average Fuel Economy Standards (CAFE) program, despite changes that gave the auto industry more flexibility than before in meeting the standards, still induces car owners to continue driving their old, higher-pollution cars, significantly offsetting fuel economy savings.
³⁹
Other complicating factors include consumers’ expectations of future fuel prices, their behavioral responses to changes in fuel economy, automakers’ technological and marketing strategies, and changes in used car markets.
⁴⁰
These complications leave analysts uncertain, even after thirty years of studies, whether CAFE has raised or reduced social welfare.
⁴¹

Another kind of causal uncertainty is exemplified by the 1990 amendments to the CAA. Previously, the law required plants to use the best available technology to control sulfur dioxide pollution.
⁴²
The amendments, however, created an emission allowances market, which is detailed at the end of this chapter. While that policy change has generally controlled emissions more efficiently, economists maintain that unrelated factors, including railroad deregulation, have contributed significantly to the program’s apparent success.
⁴³

SUPPLEMENTING MARKETS

The federal government has made an enormous investment in physical infrastructure, typically on the assumption that the private market would not or could not fund such costly projects. (Privately operated toll roads, airports, and communications networks do exist, but such exceptions constitute a very small fraction of the total.) A common theme in public discourse about this infrastructure—public lands and parks, roads, ports, airports, urban mass transit, intercity rail service, and the postal system (see
chapter 6
)—is its deteriorating quality and its inadequacy for a growing and spatially diversifying population.

Winston’s review of the evidence on government performance in these areas leads him to conclude that public financing and management “have been extremely inefficient.”
⁴⁴
He finds vast opportunities for policy improvements in his very recent assessment of the performance
of the U.S. transportation system—where total public and private spending comprised 17 percent of gross domestic product in 2007, equal to our health care spending; if congestion time costs are included, he notes, expenditures were more than $5
trillion
a year.
⁴⁵

Policy on highways, vital social infrastructure with an estimated economic value of almost $3 trillion,
⁴⁶
is replete with huge inefficiencies. Economists repeatedly note the failures by policy makers to remedy these inefficiencies through smarter road pricing for vehicles and assessing capital investments based on cost-benefit analyses. Federal and state gasoline taxes are insensitive to congestion and pavement damage factors, leading policy makers to opt for thinner, faster-deteriorating pavements that require more repair costs over time. Expanding highway capacity, they find, is a poor way to reduce congestion; on average, one dollar of spending in a given year reduces congestion costs to road users by only eleven cents in that year and only a few cents in subsequent years; this expansion strategy is also increasingly constrained by land limitations. Yet the vast inefficiencies persist.
⁴⁷
In a new paper, Winston and his colleague Fred Mannering argue for an approach to circumvent these failures by using available, well-tested technologies to improve the efficiency of highway pricing, investment, and operations, which would improve travel speeds, reliability, and safety and reduce highway expenditures. The speed of private-sector development of driverless car technology sharply contrasts with the government’s failure to exploit technology in the interests of more efficient highway construction, regulation, and financing.
⁴⁸