Why Government Fails So Often: And How It Can Do Better (48 page)

This nondelegation doctrine, however, is generally misguided.
³⁸
In fact, Congress’s choices about a delegation—about how it distributes
implementation power among the other branches, the locus of expertise, the importance of context-specificity to problem-solving, and competing modes of public participation and lawmaking process—constitutes a fundamental policy decision that will in turn strongly influence the agency’s particular policy choices. Arguably (depending on its terms), Congress’s delegation to an agency represents the most consequential set of policy choices of all. Whether or not Congress gets this choice “right” in any particular delegation, the choice is replete with difficult political, institutional, and functional trade-offs that courts are poorly equipped to second-guess.

My defense of congressional delegations to agencies, however, does not deny the risk of excessively broad delegations. For all discretion’s policy-making virtues, it can threaten the most precious rule-of-law values—legitimacy, accuracy, dignity, consistency, predictability, transparency, and democratic accountability—if it is too broad or uncontrolled. An example is prosecutorial discretion to decide whether or not to initiate a criminal proceeding against a person or entity, and if so how to conduct it and what sanctions to seek. This discretion is probably the greatest power than an official can exercise against an individual, and it is often abused by overzealous, politically ambitious officials. This danger is only magnified by the vast expansion of federal criminal law in recent decades,
³⁹
and also by Supreme Court decisions giving prosecutors very broad immunity from civil liability.
⁴⁰
Analogous forms of prosecutorial discretion exist outside the criminal justice area, although the sanctions for violations are usually less severe. In the myriad federal programs, agency officials are expected, as a practical and sound policy matter, to act against some but not all rule violations. Indeed, they must make choices about whether and how to act in many policy actions other than rule compliance. Agencies’ passivity and inaction partly reflects the relative difficulty of legally challenging their nonactions.

The vast field of administrative law prescribes how almost all of this agency discretion is to be controlled without sacrificing the considerable virtues of delegation. (Under the Administrative Procedure Act and other laws, some agency discretion is protected from judicial
review). Even the purest, most “lawless” grant of discretion—the president’s constitutional power to pardon criminals
⁴¹
—is limited by some procedural rules.
⁴²
Although administrative law contains a number of discretion-controlling techniques, each of them is seriously flawed. Thus, judicial review of agency decisions entails the many disadvantages of litigation described earlier and in the next section, while also requiring reviewing courts to make policy judgments (in the guise of legal ones) on technical issues for which they usually are ill-equipped. Agencies can usually control their own discretion by issuing clarifying or narrowing regulations, but seldom want to limit their future freedom of action—which, after all, is a central aspect (and usually a virtue) of their power. Congress can restrict agencies’ discretion, but this will deprive them of the flexibility and authority they need to identify and resolve the difficult policy issues that it delegated to them. Congress’s oversight, appropriations, and confirmation powers can be potent forms of control over agency policy choices, but also crude, intermittent ones. This pervasive discretion helps to explain why the government responded to the 2008–9 financial crisis through deal making of a kind normally practiced in the private sector, which was conducted by the two federal agencies least constrained by law, the Federal Reserve and Treasury.
⁴³
In the end, the risk that agencies will abuse their discretion and arbitrarily harm members of the public remains formidable and resistant to legal remedies.

PROCEDURAL APPARATUS

It is impossible to think coherently about law without also thinking of the procedures through which law must be created, identified as law, interpreted, elaborated, applied to specific cases, enforced, and revised. This is particularly true of the procedures that surround the decisions of administrative agencies. One cannot begin to assess law’s effectiveness as our predominant policy instrument without appreciating the nature and consequences, both positive and negative, of these procedures.

First, some essential background. In some cases, these procedures are constitutionally required by the Due Process Clauses of the Fifth and (for state governments) Fourteenth Amendments. Briefly, a constitutionally defined process is due whenever the government acts in ways that “deprive” one of a “protected interest” in “life, liberty, or property.” Each of the quoted phrases is a term of legal art with its own dense history and highly complex jurisprudence. The specific procedures that are constitutionally required in a particular situation—for example, the type and timing of a hearing, the kind of evidence that may or must be used, the nature of the official decision maker, the explanation of reasons for the decision, the scope of judicial review, and so forth—depend on what is known as the
Mathews v. Eldridge
test, named for the Supreme Court decision that defined it.
⁴⁴
This judicially applied test balances three factors: (1) the importance of the interest at stake; (2) the risk of an erroneous deprivation of that interest because of the procedures used and the probable value of additional procedural safeguards; and (3) the government’s interest.
⁴⁵
In the vast majority of administrative decisions, however, the decision procedures used by the agency are fairly standard, well-established ones that raise no constitutional issues. In designing these procedures, the agency must balance a number of competing values, in much the same spirit as
Mathews v. Eldridge
. These values focus on the various dignitary and accuracy benefits of particular procedures, and on their corresponding costs, especially to the government.

For present purposes, the important point is that the legal procedures that surround public programs constrain their effectiveness in some important ways. Like the other limits of law discussed in this chapter, procedural forms are inherent in public policy making; they go with the substantive content of law and are inseparable from it. Different procedures may be more or less effective in promoting legality, fairness, accuracy, regularity, transparency, accountability, honesty, and other procedural values that vitally affect the design and implementation of public policies. Civil litigation includes a variety of
procedural techniques that private lawyers have cultivated to both challenge and shape important public policies.

Private class actions, which are largely unknown outside the United States, illustrate the point. They affect governmental power and policies in fundamental ways. Sometimes functioning alongside conventional administrative regulation, sometimes as an alternative to it, they are driven and shaped by the private incentives of entrepreneurial lawyers and their clients, not by the public values and priorities of politically accountable policy makers.
⁴⁶
Private class actions may reinforce those public values, as in some securities and antitrust enforcement litigation, or they may undermine them, as in some mass tort cases.
⁴⁷
Thus, legal scholar Samuel Issacharoff refers to class actions as a form of “regulatory pluralism,” showing that they may relate to state power in three distinctive ways: as challenges, as complements, and as rivals.
⁴⁸

In the discussion that follows, I shall assume that legal procedures used do in fact promote these rule-of-law values. My purpose, however, is to explain how at the same time these procedures powerfully and inevitably limit policy effectiveness. I briefly consider five of these limits: (1) budgetary costs; (2) delay; (3) opportunistic behavior; (4) lawyers and courts; and (5) effects on program substance.

Budgetary costs
. Procedures are invariably costly in terms of the lawyers needed to design and assess their legality, and the personnel needed to staff them. Also, procedural manuals must be written and employees trained in how to use the procedures. These are costs to the government only. The private costs imposed by government programs are massive—many examples are discussed in other chapters—but as we saw in
chapter 4
, they do not appear on any public budget and thus may be given less attention than is warranted.

Delay
. Depending on the procedures, they can entail significant delay—not just from the time required for the procedures themselves but also because of the protracted agency and court litigation that may result. Sometimes the delay is caused by the agency itself, as when the immigration enforcement agency (ICE) misplaces files, requests continuances, and must retrace its steps. In other cases, the
delay may be caused by the program beneficiary or target, but although the agency may be able to penalize delay, it may be reluctant to do so if the person lacks resources or is one whom the agency has some duty to help. In any event, delays may compromise the program’s integrity, credibility, and effectiveness; they may also produce harmful regulatory inaction.
⁴⁹
This is certainly the case with the ICE, but it is also true of the courts and many other government programs whose procedures contribute to large case backlogs.

Opportunistic behavior
. All costs and delays benefit someone, and those created by law are no exception. Many actors and interests—criminal defendants out on bail, debtors pressed by their private or governmental creditors, undocumented immigrants hoping to stay as long as possible, those benefitting from policies that may be repealed—welcome the sluggishness that legal procedures induce. But this inertia is just the beginning of procedural opportunism. Procedures present endless possibilities for gaming. Parties’ abuse of the discovery process in litigation is notorious and itself exceedingly costly. Indeed, the ability of parties to impose such costs on their adversaries—including the government—is a crucial consideration in decisions about whether to initiate, continue, or settle lawsuits and on what terms. Thus, opportunism of this kind can be outcome-determinative quite apart from the substantive merits of parties’ claims.

The
New York Times
recently reported a classic example, the much-heralded California Environmental Quality Act of 1970, which environmentalists consider a model for the nation. According to the
Times
, the program has “open[ed] the door to lawsuits—sometimes brought by business competitors or for reasons unrelated to the environment—that regardless of their merit, can delay even green development projects for years or sometimes kill them completely.” San Francisco’s plan to paint bicycle lanes was delayed for four years by a lawsuit claiming that the lanes could cause pollution.
⁵⁰
Federal environmental laws provoke similar “greenmail” litigation.

Lawyers and courts
. If the substance of public policy is the province of politicians, economists, policy analysts, and interest groups,
process
is the playground of lawyers. Formally trained in the law of due process, and of civil, criminal, and administrative procedure, they specialize in the analysis, both legal and functional, of different procedural systems.
⁵¹
For present purposes, the important point is that the procedures that the law invariably attaches to government programs—either because they are constitutionally required or because no orderly program can function without them—are a focus of interest to lawyers whose zealous representation of private clients involves holding agencies to procedural standards that will serve their clients’ interests.

Significantly, a lawyer ordinarily can challenge the adequacy of an agency’s procedures or its compliance with those procedures more easily and successfully than she can attack its policies and decisions on substantive grounds. This incentivizes lawyers to demand from agencies more process, or more punctilious compliance with existing process, largely without regard either to the cost of increased process or to its effect on overall program performance. This is also true—but much less so—of private law rules that operate under the precedent-respecting principle of
stare decisis
. Implicit or explicit in these demands is the threat of litigation. This is the client-protecting lawyers’ job, and they perform it well. But the process does produce excesses, distortions, rigidities, and other social costs.
⁵²

Effects on program substance
. Policy makers and program officials fear litigation. With some reason, they anticipate that the intrusion of courts into their domain will attract adverse publicity, alter officials’ incentives, consume scarce bureaucratic and beneficiary resources, require additional procedures or other programmatic changes, and reduce the officials’ policy and operational autonomy. Indeed, federal judges in certain notorious cases—especially those involving prisons, school and housing integration, and other programs whose operation can raise constitutional issues—may even issue “structural injunctions,” empowering the plaintiffs’ lawyers and courts to monitor and supervise the programs, sometimes to a remarkably detailed extent.
⁵³
Officials facing these threats to their autonomy are tempted to head them off by making programmatic changes along the way. In some
cases, they may even welcome the litigation, hoping that the court will order the program to do what might otherwise be politically infeasible, such as spending more money to raise program standards. Officials may institute these judicially prodded programmatic changes, moreover, even if the lawsuit ultimately fails.