Why Government Fails So Often: And How It Can Do Better (64 page)

Many other regulatory agencies are notoriously understaffed in comparison with their vast responsibilities and the growing size of the markets that they regulate. The FDA is a particularly clear example of this, but it is also true of federal banking agencies. As discussed in
chapter 10
, the use of private contractors may or may not be more efficient than staffing functions in-house—this judgment can only be made after agencies conduct detailed function-by-function analyses—but if contractors are to be used, the agency must be far better equipped with staff sufficiently expert to manage the contractual relationship.

THE LIMITS OF LAW

As
chapter 9
explained, the very nature of public law places some severe limits—both constitutional and functional—on the effectiveness of the policies that it communicates and governs. Although most of these limits cannot be avoided, some of them might be eased.

1. In a 1992 essay on which
chapter 9
draws, I proposed four principles for improving the balance between law’s complexity and simplicity.
⁸²
The “cost distribution” principle calls attention to situations in which a law imposes unnecessary complexity costs on diffuse, poorly organized groups to the advantage of politically influential groups that benefit from that complexity, perhaps by burdening smaller competitors and would-be entrants into their industry with additional costs. Reformers can seek to redress the competitive distortions created by this political economy. Under the “audience principle,” a law’s complexity should be tailored to the sophistication and cost-bearing capacities of those who will have to interpret and implement it. The legal scholar Boris Bittker noted that complex tax rules are less problematic if they are addressed to experts who must apply
them to uncommon transactions than if they are addressed instead to numerous lay people who must understand and apply time to common transactions. As it stands, the tax code has almost tripled in size just over the past decade, and 90 percent of tax filers pay for help to complete their returns.
⁸³
Many scholars and policy experts (including Bittker) have proposed tax simplification reforms which could herald vast gains in both efficiency and fairness.
⁸⁴
With taxation as perhaps the most vexing site of citizen-federal interaction, simplification would likely enhance democratic morale and the government’s legitimacy as well. It is clearly a popular theme with voters.
⁸⁵

In addition to knowing
where
to simplify, we need to know
how
to simplify. The “mimicking principle” posits that by studying the ways in which people contract around the formal law, perhaps by selecting simpler competitors, policy makers may obtain good information about its inefficiencies, including excessive complexity, which they can use to make the policy mimic people’s informal simplification strategies. Where feasible, lawmakers should follow a “user fee principle” to tax those who benefit from legal complexity for the extra costs associated with those benefits. Although it is usually hard to isolate, allocate, and quantify complexity costs, and doing so may not always be cost-effective, marginal complexity costs can sometimes be calculated and redistributed. For example, courts may tax litigants that demand excessive, hard-to-provide document searches by their opponents. Some statutes, including the new Affordable Care Act, exempt small businesses from certain complex and costly regulatory requirements.

2. Other things being equal, simpler programs involving fewer levels through which funds and administrative communications must pass are easier to implement effectively.
⁸⁶
Bureaucratically “flatter” programs can reduce the kinds of conflicting interests and incentives that block the long road from program design to implementation, which helped doom the Oakland Project and other failed programs detailed in
chapter 8
. The federal government can use broader grants, rather than detailed federal administrative specifications, to support
state implementation of federal policies. This conflict between detail and delegation is old and probably inevitable,
⁸⁷
but in balancing them Congress should put its thumb on the side of broader delegation. First, states play an enormously important policy role for reasons explained in
chapters 3
and
4
. Second, they have many ways to thwart federal efforts to dictate policy details. Third, they are far more politically accountable and efficient than in the bad old days of Jim Crow and courthouse politics.
⁸⁸
Finally, they are likely to know best how to coordinate and deliver federally funded services through their own administrative apparatus to their own citizenry. This does not tell federal officials precisely how much control is needed to maximize the effectiveness of federal dollars administered by states, but it suggests that less is sometimes more.

3. Like simpler programming, simpler regulation can be more effective—again, other things being equal. This, of course, is a traditional mantra of business firms and other entities subject to ever-expanding governmental requirements in countless policy areas. Most recently, however, the cry for simplification has been taken up by Cass Sunstein, as mentioned earlier. In his book,
Simpler: The Future of Government
, he proposes many simplification measures designed to influence the “architecture” of individual choice. Sunstein proposes regulatory “nudges,” an approach that he and Richard Thaler had advocated earlier and which they call “soft” or “libertarian” paternalism
⁸⁹
—nonmandatory approaches to regulation that can shape better choices.
⁹⁰
Two nudge strategies, for example, are
“smart” disclosure requirements
, which present salient data that people can use with minimal cognitive effort, and
default rules
, which take account of their cognitive patterns, such as the status quo bias, and reduce the likelihood of clearly irrational choices. Sunstein applies this approach to school meals, pensions, health insurance, savings plans, public health, and other policy areas.
⁹¹
New research on health insurance underscores why such simplification is so important: only 14 percent of the insured, it finds, understand the basic concepts they need to shop for policies.
⁹²
Incorporating “nudging” strategies into regulations
, he contends, can make regulations simpler and more effective, while maintaining freedom of choice.
⁹³
Some commentators doubt that he can have it both ways and that regulators can know when they are increasing, rather than reducing, consumer rationality.
⁹⁴

Sunstein also promotes regulatory “look-backs” to make things simpler. He laments that cost-benefit analysis is often only applied to regulations before they are promulgated—and therefore before agencies have much good, empirical evidence on how they function in practice.
⁹⁵
By reevaluating existing regulations using more recent information, agencies can simplify their rules and ease unnecessary regulatory burdens. During Sunstein’s time at the OIRA, agencies solicited suggestions from the public on which regulations to examine and then streamlined or eliminated many of them based on retrospective analysis.
⁹⁶

4. Another approach to simplification is to alter the mix between legal rules and discretion in favor of the latter for policy implementers. Philip K. Howard has made this case most strongly in his book
The Death of Common Sense
, which provides numerous examples of situations at all levels of government in which law needlessly “suffocates” initiative and expert judgment by public and private decision makers who know better than regulators what needs to be done in particular contexts.
⁹⁷

5. In some situations, policy makers should control risks by relying on existing systems of private law—especially tort, contracts, and insurance—rather than by setting up new, costly, and inevitably over-broad systems of administrative regulation. This is not because private law is invariably effective—indeed, many scholars have made fine careers out of dissecting the tort system’s many flaws
⁹⁸
—but instead because that system may on balance still be more effective than the regulatory alternative. Under certain conditions—the wrongdoer’s identity is known, victims can readily prove causation of harm, their damages are straightforward, defenses to liability are well designed, and contingency fees assure access to lawyers—tort law’s limits are less problematic than where these conditions are absent and thus
justice requires a combination of regulation and compensation. Airline accidents illustrate the first type, and mass asbestos exposures the second.

CIVIL SERVICE BUREAUCRACY

Chapter 10
emphasized that improving bureaucratic performance is essential to greater policy success. Not surprisingly, many carefully designed proposals for doing so have been advanced, including the 1989 report of the National Commission on the Public Service chaired by Paul Volcker. Paul Light, whose empirical work on the civil service was featured in
chapter 10
, has extended and updated the commission’s work in a 2008 book that offered comprehensive proposals to “reverse the decline” in the public service.
⁹⁹
Light noted that almost half the federal workforce would reach retirement age in the next decade, creating a rare opportunity to reshape the federal service.
¹⁰⁰
Among his proposals were to sort government missions based on their importance, difficulty, and past success; thin out the bureaucracy by reducing the number of layers between top and bottom to no more than six; reduce the number of managers by half; accelerate the hiring and appointments processes;
*
improve the incentives to “make a difference”; experiment with an up-or-out promotion system for federal supervisors modeled on the military; streamline the disciplinary process; reserve a quarter of all middle-and senior-level job openings for outside candidates; improve performance measurement and incentives; and strengthen oversight of contracts, grants, and mandates to state and local governments.
¹⁰¹
Our disappointing thirty-five-year experience with the Senior Executive Service raises doubts that Congress will promote or even permit Light’s approaches. I have nothing to add to such proposals.

As we saw in
chapters 6
and
10
, the vast army of private contractors now dwarfs the cadre of federal employees. The political and
other reasons that generated this army, including Congress’s political fear of appearing to enlarge the federal workforce, are most unlikely to change. This makes contract administration by federal employees at least as important a feature of federal policy making and implementation as the services that federal employees render directly. Thus, the OMB and the officials of contracting agencies should promote better contract administration, including control of contractor fraud, waste, and abuse, through the kinds of reforms discussed at the end of
chapter 6
.

*
A classic example is the moratorium on special interest earmarks adopted with much fanfare after the 2010 congressional elections. Urged by President Barack Obama in his State of the Union message and celebrated by good government advocates everywhere, it seems to have impeded the political bargaining and compromise necessary to resolve the 2012–13 “fiscal cliff” and other standoffs. Indeed, the fiscal cliff legislation
itself
contained earmarks—probably for this reason. See Alicia Mundy, “Room for Favors in ‘Cliff’ Deal,”
Wall Street Journal
, January 29, 2013.

*
The post–Civil War amendments and the New Deal are the great exceptions. The Great Society is a more arguable case; most of its programs (other than civil rights) were built on earlier reforms.

*
Chapter 4
discussed this bias. Thirty years ago, I suggested electing some representatives in Congress under OPOV but by
regional or national at-large
constituencies. Peter H. Schuck, “Industrial Policy’s Obstacles,”
New York Times
, September 6, 1983.

*
Such projects do have their political uses; as noted in
chapter 5
, they can lubricate an otherwise stalled legislative process, which may or may not serve the public well. A famous example was the “bridge to nowhere” in Ketchikan, Alaska (population 8,900). Ultimately, however, national protests and ridicule ended the project.

*
For a comparative anaylsis, see R. Kent Weaver & Bert A. Rockman,
Do Institutions Matter? Government Capabilities in the United States and Abroad
(1993), 11–41. A more incremental alternative—requiring voters to choose between each party’s slates of candidates, instead of allowing them to split their votes for president and members of Congress—is thoughtfully discussed in Derek Bok,
The Trouble with Government
(2001), 287–91.

*
Insurers, for example, will not issue a life insurance policy to one who lacks an “insurable interest” in the insured event, which of course is why one may not insure a stranger’s life.

*
Recall that
some
moral hazard may be inevitable, given certain legitimate program goals. See
chapter 6
.

*
Judging from the long delays in filling vacant leadership positions that led to a political crisis in the Senate in July 2013 over such appointments, the much-heralded Presidential Appointment Efficiency and Streamlining Act of 2012 has had little effect.