FROM ISSUE NUMBER 7 ~ SPRING 2011 GO TO TABLE OF CONTENTS

TEXT SIZE A A A

Government by Waiver

RICHARD A. EPSTEIN

One of the great achievements of Western civilization is what we commonly call "the rule of law." By this we mean the basic principles of fairness and due process that govern the application of power in both the public and the private spheres. The rule of law requires that all disputes — whether among private parties or among the state and private parties — be tried before neutral judges, under rules that are known and articulated in advance. Every party must have notice of the charge against him and an opportunity to be heard in response; each governing rule must be consistent with all the others, so that no person is forced to violate one legal requirement in order to satisfy a second. In the United States, our respect for such principles has made our economy the world's strongest, and our citizens the world's freest.

Though we may take it for granted, the rule of law is no easy thing to create and preserve. Dictators and petty despots of all sorts will rebel against these constraints in order to exercise dominion over the lives and fortunes of their subjects. But anyone, of any political persuasion, who thinks of government as the servant of its citizens — not their master — will recognize that compliance with the rule of law sets a minimum condition for a just legal order.

That, however, is precisely where the difficulties begin — for minimum conditions by themselves are not enough. Law is not just an idealized system of rules: It also involves the public administration of those rules by a wide range of elected and appointed officials in an endless array of particular circumstances. For those who would defend a just legal order, the basic challenge is to strike a proper balance — between limiting the discretion of these officials so that they do not undermine the rule of law, while also allowing them enough leeway to perform their essential roles.

Lately in America, we have done a poor job of preserving this balance. In practice — and, increasingly, in legal theory — government officials have been given unprecedented ability to make exceptions to the law, both in enforcing it and in respecting the rights granted under it. Indeed, the past year has seen two of the most enormous pieces of legislation in U.S. history — the Patient Protection and Affordable Care Act and the Wall Street Reform and Consumer Protection Act — make the imbalance far worse. Both laws seek to dramatically transform vast swaths of the American economy; both give enormous power to the government to bring about these transformations. And yet both laws are stunningly silent on exactly how these overhauls are to take place. The vague language of these statutes delegates much blanket authority to government officials who will, effectively, make the rules up as they go along.

As these officials stumble through how to implement these sprawling new laws, they will inevitably come up against unanticipated obstacles (or powerful interests) that will demand exceptions to the statutes' far-reaching provisions. In some cases, special benefits or permissions releasing companies from government regulations will simply be granted. In others, the releases will be provided only if the regulated parties agree to waive some legal protection to which they would otherwise be entitled.

Neither of these practices — providing waivers or demanding waivers — is necessarily pernicious. Indeed, in some cases, they are part and parcel of the ordinary course of business in the modern administrative state. But both are open to abuse, and that abuse makes for a particularly dangerous form of government power.

After all, people concerned for their freedom and rights are always most alert to threats that arise when governments (or other powerful institutions) force us to do what we don't want to do. The power of coercion is more easy to define, to identify, and to resist. But we are not sufficiently alert to the flip side of this problem: the risks that come with the power to create exceptions and to grant dispensations. Indeed, this is a much more subtle, insidious assault by government: Rather than setting the state and the private sector against each other in a healthy tension, it fuses them, making the private sphere dependent on the government's benevolence. And when currying the favor of capricious government officials is required for a person's well-being or a firm's very existence, government abuse becomes nearly impossible to oppose.

"Government by waiver" is thus among the most serious challenges to the rule of law in our time.

WAIVERS AND THE MODERN STATE

The issue of government by waiver arises in any system of public administration. But as the size of the state has expanded dramatically, the scope of the problem has grown right along with it.

Under the traditional classical-liberal model of limited government, the state has a few critical, but well-defined, objectives — each directed toward controlling the use of force and facilitating voluntary agreements among private parties. Together, they ensure that the rules of the road are clear and knowable to all individuals, and in turn two felicitous consequences follow: First, individuals who know their rights are able to take easy steps to avoid getting enmeshed with the law; second, clear rules make it easier to monitor the conduct of public officials. Thus, the more limited the scope of government, the fewer difficulties there are in controlling the discretion of its officers.

Prior to the rise of the modern administrative state, the delineation of property rights and the enforcement of contracts were governed primarily by such simple common-law rules. Each person was expected to forbear against any physical invasion of the person or property of another, which meant that simple conformity to a bright-line rule could keep most people out of mischief. That one rule was easy to understand, and its content did not vary with the number of people in society, their personal characteristics, or their levels of wealth. Exchanges between individuals, meanwhile, could be conducted through voluntary agreements, which tended to clarify rights and preserve flexibility. The risk of abuse was low, because everyone could pick the parties with whom he chose to deal and the terms on which he wished to interact. Thus labor contracts were often written "at will," which meant that an employer could hire and fire as he chose, and a worker could accept employment or quit his job whenever he wanted.

These rules were, of course, subject to exceptions that dealt with duress, fraud, non-disclosure, incompetence, and undue influence. They left open areas such as child labor, in the case of which there can be genuine differences regarding whether legal limits in fact protect children from parental exploitation or deny families the opportunity to work their way out of poverty. But such exceptions wane in importance in any regime that uses voluntary exchange to achieve economic growth.

Most critical to the rule of law, running such a system requires little administrative oversight. When disputes arise, judicial determinations are generally easy to make, because everything turns on readily verifiable public acts. Moreover, the public enforcement of laws — even when it requires some case-by-case discretion — involves relatively modest functions. Issuing building permits, for instance, is limited to straightforward matters of health and safety — touching only such issues as falling objects or traffic interference (as opposed to, say, more vague interests like "neighborhood character," "economic development," and "urban renewal"). The constrained range of tasks assigned to government officials thus leaves more decisions in private hands, where individual choice is driven by individual preference or by competitive market conditions (rather than by government command). Under such circumstances, the harmful pressures exerted on the rule of law are very low.

But the modern state does not limit itself to these defined objectives. Ostensibly desirous of assuring the welfare of its citizens, today's state claims that it must take a far greater role in the life of the nation. This conception of politics comes to see such welfare — measured in terms of material well-being and access to some essential goods like housing, health care, and education — as a right to be guaranteed to the people by their government. Providing for every such right requires resources, which must be obtained by the state through taxation and mandates. Since no regime of positive rights can repeal the iron law of scarcity — which dictates that the provision of goods to some people necessarily imposes correlative burdens on others — a regime of positive rights is a regime of very demanding duties and requirements.

The question, then, is whether these duties and requirements must be enforced in absolutely all circumstances. To any reasonable observer, the answer is surely not: The immense range of circumstances that present themselves in a huge, complex society means that there will always be hard cases calling for exceptions. Sometimes various requirements have to be waived to avoid unreasonable hardship, and sometimes various requirements have to be waived for the system to function at all.

But can such waivers be made compatible with the rule of law?

UNDERSTANDING WAIVERS

In order to assess the legitimacy of government waivers, it is crucial first to understand what exactly waivers are (in private as well as public affairs), how waivers function, and what effects they have on public life. It is thus useful to start with the standard legal definition, which describes a waiver as the voluntary relinquishment of a known right.

The word "known" means that the granting of a waiver must be a deliberate act. A valid waiver cannot be created through inadvertence: The party providing the waiver must, at a minimum, be fully informed of its conditions and consequences for the waiver to be binding. Any weaker rule would put all complex contractual arrangements at risk.

The word "relinquish" signals that waivers are midstream corrections to ongoing contractual (or other) arrangements through which people have acquired rights and duties with respect to one another. Waivers are usually one-sided transactions, and thus cannot be used to create initial contractual obligations, which are usually bargains between strangers in which each side surrenders something of value to the other. Given the sharp transition from stranger to trading partner, initial contracts are generally more formal than waivers, which are often carried out with greater rapidity and ease.

Most important, the word "voluntary" means that waivers may not be obtained under duress. This refers not only to old-fashioned threats (like a gun to the head) or conventional forms of fraud (as when a waiver is requested on the grounds of a made-up illness): Those forms of duress are not unique to the matter of waivers, and can be invoked to nullify any original contract or agreement. Rather, the form of coercion distinctive to waivers arises in those situations in which the waiver of a benefit or condition is given in response to a promisor's threat or warning that he will deprive the other party of some collateral right if the waiver is not forthcoming.

To clarify this point, it is useful to consider it in the context of private-sector waivers (which are common in commercial and other contractual arrangements). Imagine that Party A owes Party B a sum of money for services rendered. A tells B that he will pay for part of the labor only if B waives, in writing, all claims to recover the remainder. In the abstract, it is hard to tell whether A has engaged in an improper practice or not. If B's services were unsatisfactory, the waiver makes sense. But if the services were adequate, A is consciously taking advantage of the fact that B will be reluctant to sue for the remainder because the costs of litigation could well exceed the value of the payment in dispute. Given the difficulty of making basic determinations about the facts of the case — were B's services indeed unsatisfactory? — this form of "duress" by litigation is as difficult to counter as it is easy to practice.

The situation gets more complicated if, as often happens, a waiver is sought in an arrangement in which one provider trades with many buyers — as in the case of a utility company that claims an entire region of customers, of any employer with several workers, or of any landlord with multiple tenants. In such circumstances, the decision of the company, employer, or landlord regarding whether or not to waive some requirement in one case affects all the other customers, employees, or tenants.

The difficulty is especially acute if all of these customers, employees, or tenants are, or even perceive themselves to be, entitled to equal treatment by the company, employer, or landlord. Such expectations — rooted in our commitment to distributive justice, under which like cases should be treated alike — can be very powerful: If several employees are in competition for the same promotion, a waiver of one requirement in the selection process for Candidate A gives that person a serious competitive advantage relative to Candidate B. If an employer rigs enough contests, even if it is technically legal to do so, his workers start to think that reward is not related to effort. Eventually, they slack off or quit.

This set of concerns gives rise to one of the essential, if imperfect, principles of any mature legal system: the so-called non-discrimination rule. The rule was first put in place to deal with "common carriers" — businesses that are required to take all customers — in the bygone days when a town might have had just one stagecoach and one inn, so that each had a monopoly position in its respective market. But that principle quickly insinuates itself into competitive markets as well. People may have many choices regarding where to work or live, but once a worker or tenant enters into a contractual relationship with an employer or landlord, there is only one person to whom he can turn to settle grievances. In these many-on-one settings, parity of treatment among individuals is crucial.

In such settings, therefore, the non-discrimination rule essentially forces the provider of services to take an all-or-nothing position on issues of common concern. Management can waive a particular condition with respect to all parties, or it can waive it with respect to none; it cannot waive it only for some. This prevents a kind of partiality that corrodes economic relations: The only way an actor may hurt his enemies is to hurt his friends as well.

But even in private affairs, the matter is hardly this simple. Although all selective waivers may be suspect, there are surely some circumstances under which they are acceptable. Assume, for example, that all the applicants for a new job were supposed to receive their application forms at the same time, but for some reason the instructions were delivered to one applicant a day late. At this point, a waiver extending the deadline by one day would redress an imbalance that arose through no one's fault. Whenever unequal circumstances arise, some corrective action might easily be justified — and perhaps even required.

Ultimately, an appeal to sound discretion is the best formula anyone has devised for dealing with this problem. In most private circumstances, such discretion can be employed with relatively little trouble — and when trouble arises, it can be addressed in court fairly readily. Indeed, this happens often, and the question of waivers infiltrates every area of private law.

But the process is in every way more complicated when the government gets involved. And those complications are only multiplying as the government becomes ever more tied up in relationships and decisions that were once left to individuals.

PUBLIC WAIVERS

By definition, the government has a strong monopoly position when it chooses to regulate. And so, like the common carriers of bygone days, its operations must be subject to the basic norm of equal treatment for like cases. This, in turn, should limit both the ways in which the government dispenses waivers and the ways in which public officials exercise their discretion.

Yet discretion is, and ought to be, a necessary part of a government's work — just as it is with private businesses. Government managers must run public offices and design and execute public projects. The military, the justice system, public schools, and many other public bodies are large institutions with management imperatives resembling those of private businesses. Executive officers in both settings discharge functions set by boards of directors on one hand and legislators on the other; these overseers act on behalf of shareholders and citizens, respectively.

Key officers and executives in corporations are protected by a "business-judgment rule," which essentially gives them the benefit of the doubt when they act in good faith and to the best of their abilities while making hard choices. This standard recognizes that no sensible person would fill a position of trust if he could be routinely second-guessed and sanctioned, perhaps with heavy fines, after the fact. By design, however, the business-judgment rule does not offer absolute protection; it leaves key corporate officials exposed to liability for self-dealing or bad-faith transactions. The rule's durability suggests that it strikes the proper balance.

A similar rule applies to government officials. Here, however, the stakes are much higher, given both the necessity and the dangers of government action. The justified fear of frivolous lawsuits against government officials leads most courts to throw out virtually all actions against such officials in their capacities as private citizens, except in cases of extreme malice and misbehavior. Prosecutors and judges routinely receive absolute immunity for their official acts, even though this protection at times leaves ordinary people with no real remedy for serious violations of their rights.

Such protections, together with lawmakers' increasing tendency to grant administrators broad discretion in carrying out immense and vague legislation, leave government officials with enormous freedom of action. This freedom only grows as the scope and range of the government's power to regulate grow — for any law that gives government more to do opens up more opportunities for government abuses and mistakes. Indeed, now that government's reach is broad and deep, we find around us countless examples of abuses of discretion. And prominent among these are abuses of the power to waive requirements and to extract waivers of rights. Administrative discretion thus becomes a real threat to the rule of law.

In recent decades, we have seen several distinct paths by which the broad use of administrative waiver power can compromise our system of government. First, broad discretion and waiver authority can be used as alternatives to legitimate policymaking, with unfortunate results for both the regulated market and the rule of law. The best example may be the history of telecommunications regulation in America — most notably the regulation of the radio-frequency spectrum. The initial (but modest) use of the spectrum at the beginning of the 20th century took place under a decidedly privatized regime: Individuals were given ownership of the portion of the spectrum that they first used to broadcast. In time, that system ran into problems of over-claiming and interference, and so the Radio Act of 1927 — enacted during the Coolidge administration under the leadership of then-secretary of commerce Herbert Hoover — took the sensible step of requiring all parties to obtain licenses to their frequencies from the federal government.

Once the government took over the spectrum, it could have provided such licenses by auction or deed rights that would have given each party absolute control over a certain frequency, subject only to prohibitions against broadcast interference with the rights of others. At that point, the spectrum could have been folded back into the general system of property rights. Owners could have decided how to use, sell, mortgage, license, or subdivide the frequencies under their control. As technology became more sophisticated, subdivision into narrower bands, allowing for more intensive use, would have become more feasible; as tastes shifted, the owner of a given portion of the spectrum could have altered programming or shifted to entirely different usages (such as broadband, where shortages are acute).

Instead, Congress established a regime under which the Federal Communications Commission grants renewable licenses, with renewal contingent on the licensee using his rights in a way that best serves the "public interest, convenience, or necessity." In such a system, someone must decide what criteria above and beyond market acceptance should be used to determine who is best serving the "public interest, convenience, or necessity" — and thus who gets a license or renewal. The standard established in the statute was hopelessly vague, but the Supreme Court nonetheless upheld it in 1943 — determining that administrative agencies would be capable of putting meat on the fragile bones of such formulas of delegation. To Justice Felix Frankfurter, an ardent critic of laissez-faire policies, the point of government action was not merely to set the rules of the road but also to determine the composition of the traffic. In the case of broadcasting, however, the government was never able to formulate a consistent standard for determining what that traffic ought to be composed of.

Thus the three-year license-renewal process quickly became vulnerable to the sort of intrigue that always arises when the government is in a position to dole out or deny benefits. Interest groups could easily threaten to block license renewal if the licensee did not include more ethnic programming, more local news coverage, more high-brow art, or more of anything else the groups might desire. Under this arrangement, the FCC's most meaningful power has become the authority to decide whether to insist that an applicant demonstrate how he will meet the vague requirements of the law — in other words, the power to decide whether to enforce the rules or to waive enforcement (as the FCC does most of the time). And in return for waiving its right to enforce the law, the agency can require that broadcasters waive rights of their own — and especially that they give up some control over the content of what they broadcast.

These days, the allocation of the spectrum is mostly a settled matter, and very few license-renewal requests are denied. But thanks to the excessive discretion granted to the FCC — and exercised in large part through the agency's power to waive its own rules — we have gotten to this point through an economically inefficient and administratively chaotic process.

We are now witnessing the expansion of this approach through the effort to govern what content is transmitted over the internet — both wired and wireless — under so-called "net neutrality" rules. Under these proposed rules, internet-service providers and data carriers would be prohibited from charging different rates for different types of web content, or from blocking outright the transmission of objectionable content over their networks, without the consent of the FCC. Government bureaucrats — through their ability to grant or withhold such consent — would thus have enormous control over the generation and dissemination of information and speech in the United States. Should such rules go into effect, we will see a replay of the same kind of factional intrigue that marred the early history of radio-spectrum development — a classic example of regulation gone awry.

Another danger inherent in the government's waiver power is that it will be used to extract concessions from private actors. And for an example of how this danger plays out in practice, one need only look at the regulation of construction and real-estate development. These days, to begin any new building project, every developer must obtain a sheaf of permits that go far beyond the relatively mundane functions of avoiding falling bricks or aligning curb cuts to secure entryways for indoor parking. Indeed, today's new norm calls for exhaustive hearings before planning commissions and community boards; these investigations are intended to probe the size of a project, its exterior design, the number and type of apartment units, access for the disabled, the amount of affordable housing (with complex subsidies from both the government and the developer), project financing (with government guarantees), proper hiring practices (with appropriate set-asides for women and minority workers), and multiple inspections for just about everything.

Yet just as all these requirements can be imposed, they can also be waived. The waivers, though, often come at a price — or, more accurately, a land-use exaction. For instance, a cash-strapped local government may be willing to waive the requirement that a developer set aside a certain percentage of apartment units to rent at below-market rates to the poor. The catch, however, is that the developer must agree to provide funding to build or refurbish a public school, a public park, or a nearby train station. The developer almost inevitably yields to the exaction, because he knows that, if he does not, he faces prolonged resistance and constrictive red tape from the government — obstacles that could eventually sink his project. But the requests for exactions may come from many varied groups with different expectations and demands. Parents may want a new school or park, commuters may want a new train station, cyclists may want new bike lanes, the arts community a new public performance space, homeless advocates a new shelter, and so on. It may not be possible for the government or the developer to satisfy all of the groups simultaneously — and the attempt to do so can tie up development for years, or cause projects to be scrapped altogether. This phenomenon drives up the number of project failures, which in turn shrinks the supply of housing, which then drives up housing costs and puts even greater pressure on both the developers and the regulators.

Sometimes government even uses its waiver power to demand that regulated entities make concessions not to the government itself but to other private entities. The foremost example of this practice is in labor relations: We have come very far from the time when employers could simply hire and fire at will. Part of the reason is that today's immensely complex scheme of labor rules ensures that practically every sizable employer will, at some point, violate some portion of the laws governing worker hours or overtime pay. Such violations do not generally give rise to enforcement actions, but as long as the possibility of enforcement remains, government officials with strong pro-labor allegiances can threaten enforcement in order to get employers to waive their rights to resist unionization of their plants.

Indeed, under the Obama administration, union-friendly government officials do use such suasion techniques. Under this arrangement (first reported by the Daily Caller's Matthew Boyle in January), the employer who lets a union organize may not have to pay fines for any of the countless minor technical violations of labor rules he may rack up. The entire process takes place covertly, and part of the deal is that the party that complies with the demand does nothing to make its terms public. With these sorts of schemes, the notion of transparency in government decision-making — so essential to the rule of law — necessarily falls by the wayside.

Finally, administrative agencies that are given too much discretion can end up blurring the line between coercive power and waiver power in a way that grants these agencies an immense amount of informal authority — authority that extends well beyond the powers they are granted by Congress.

Here, the foremost example is surely the Food and Drug Administration. The FDA's endless delays in licensing new drugs result largely from the elaborate clinical trials that must be conducted before a drug can be put on the market. The particular standards and requirements that each drug must meet are not all laid out in law, or even in formal agency rules: Often they are established ad hoc by the regulators charged with each individual drug approval, and they may be waived ad hoc, too. Thus in any particular case the FDA has an enormous level of discretion regarding how to proceed with the approval. It is common for the agency to insist that drug companies accept conditions that are found nowhere in the statutory scheme — conditions that, for instance, can limit the distribution or advertisement of a drug, or govern its recall in case a problem with it should arise — in exchange for the FDA's waiving other requirements or granting early approval.

Because the costs of bringing a new type of drug to market now come to around $1.5 billion, companies are often willing to comply with the FDA's demands, not wanting agency delay (or outright rejection) to cause even more expense (or massive losses). Plus, efforts to sue the FDA for an abuse of discretion only compound the delays — leaving the regulated companies little choice but to follow the unpredictable rules they are given and waive their own rights to object.

THE ERA OF DISCRETION

As corrosive as these consequences of government by waiver have been to the rule of law and the legitimacy of the regulatory state, they are now likely to do much more damage as the reach and power of the state increase even further. The two large legislative achievements of the first half of Barack Obama's presidency — health-care reform and financial regulatory reform — stand to exacerbate all of the problems surrounding regulatory discretion and waiver authority.

The Patient Protection and Affordable Care Act is a 906-page statute that authorizes well over a trillion dollars in spending and taxes, establishes dozens of complex new federal and state bureaucracies, grants the federal government immense new regulatory powers over the health-care and insurance sectors, and grants the Department of Health and Human Services tremendous discretion and freedom to decide precisely how its many hundreds of individual provisions will be carried out. In more than 700 separate instances, the law states that "the Secretary shall" have the power to set specific rules to put into practice the statute's very general requirements.

The major provisions of the law will not take effect until 2014. But even the minor provisions that have already been implemented offer some clear examples of the troubles to come. The most prominent has involved the law's so-called "medical-loss ratio" requirements, which stipulate that only 20% of the total premiums collected by health-insurance plans for individuals (and 15% of those of plans covering small groups) may be devoted to administrative costs. The rest of the premiums collected must pay for actual medical services. Before the passage of the health-care law, however, the market averages for administrative costs were at least 10% higher than these requirements. The law offers no mechanism for bringing down these costs; rather, it simply requires insurers to cap their non-medical spending, even if their legitimate non-medical expenses vastly exceed that limit.

As the Department of Health and Human Services began preparing to enforce this requirement, it became clear that not all insurers would be able to comply. In particular, so-called "mini-med" plans — which provide only very basic coverage, and are offered by some employers to part-time workers so they are not left uninsured — would find it especially difficult to function under the requirement, as the high turnover rate of employees using such plans creates unavoidably high administrative costs. HHS thus had to decide whether it would waive the requirement for the mini-med plans, lest they be forced out of business (thereby pushing even more individuals into the ranks of the uninsured, or onto the public health-care rolls). As it turns out, HHS has been willing to grant exceptions — but only on a case-by-case basis. As of early March, the department has granted 1,040 one-year waivers to organizations offering mini-med plans, including a variety of employers and unions — among them Pepsico, Foot Locker, the Pew Charitable Trusts, many local chapters of the Teamsters, the United Food and Commercial Workers union, and numerous public-employee unions. Together, these organizations and the insurance plans they offer cover more than 2 million Americans.

But what about employers who do not have the resources to navigate the waiver process? What about those lacking the political connections to make their concerns heard in Washington? And what happens when the one-year waivers run out? Will they be renewed? Under what conditions? And what rights will insurers have to waive then in order to avoid going out of business?

The medical-loss-ratio requirement is only the beginning. The remainder of the health-care law — most of which has yet to take effect — contains far more complicated, and in some cases unreasonable, requirements. They will govern participation in the new state-run health-insurance exchanges, the design and pricing of insurance plans, and the various new taxes and subsidies that are essential to the new law's budgeting and financing, among other facets of health care in the United States. An overwhelming number of these new provisions have been enacted with only the vaguest legislative language; the result is that it will be up to HHS bureaucrats to interpret these mandates and turn them into regulations. Inevitably, those rules will come with some waiver authority — and with it great potential for the differential application of generally applicable laws. The danger posed to the basic principles of the rule of law, therefore, is immense.

The new financial-regulation law — the Wall Street Reform and Consumer Protection Act — similarly grants vast discretion. The 848-page law creates a host of new regulatory agencies and powers to oversee the financial industry. Addressed to a sector of the economy in which clear and predictable rules are especially important, the law is astonishingly vague and broad, leaving regulators — including new agencies with no experience or track record — with unprecedented freedom to draw up the rules.

Consider, for example, the process the law uses to establish whether a given financial institution creates a systemic financial risk — a risk that in turn justifies some form of government takeover. In the language of the statute, the Federal Deposit Insurance Corporation is empowered to "determine that such action is necessary for purposes of the financial stability of the United States" — a stunningly broad determination, the terms of which are nowhere clearly defined. If the FDIC does make such a determination about a firm, it can inject into the company funds drawn from a new "orderly liquidation fund" created from fees assessed from all regulated financial firms. The amounts of these assessments are not fixed by the law, or applied universally across all financial institutions: Rather, they are to be determined based on the regulator's judgment of such factors as "economic conditions generally affecting financial companies," "the risks presented by the financial company to the financial system," and, of course, "other risk-related factors as the [regulator] may determine to be appropriate."

And because of the nature of the determinations involved, the regulated companies — whose very existence might be at stake in these decisions — often have no effective shot at judicial review. Indeed, in this case as in many others, the breadth of discretion afforded to regulators creates challenges that today's legal system is not well equipped to address. Most judges evince great faith in the administrative state, so that the abuse of discretion that lies at the heart of the waiver problem is, to them, a matter best sorted out by administrative expertise — a perpetually overestimated pool of wisdom.

Even a court intent on stopping or controlling this form of government discretion has relatively few legal instruments at its disposal. A court can intervene in such cases only if it deems the regulator's actions to be "arbitrary and capricious." But often the particular application of the rules being challenged is so distinctive that regulators can easily justify differential treatment, even to a court that casts a suspicious eye over their work.

Moreover, judicial challenges take time to resolve: Agencies that seek to block time-sensitive private actions are quite happy to run out the clock when they are sued by disappointed supplicants. And even if this obstacle could be overcome, most private parties are not foolhardy enough to challenge a government regulator — they fear, understandably, that the government will retaliate elsewhere. It is no accident that it is often public-interest groups or patient groups that take on the FDA, for instance. It is simply too risky for a pharmaceutical company with multiple applications before the agency to challenge one action if it is vulnerable to a government-induced slowdown on another.

In these cases, the best that judges can do is to make clear that the government waiver process does not allow the state to impose any condition that is "not germane" to the application at hand. Thus an environmental permit may be conditioned on increasing the size of a filtration system, but not on a company's willingness to contribute money to some wilderness fund.

This is an important principle of law, one very much in need of reaffirming. And yet that is small consolation to a company in the grip of a regulator wielding unpredictable requirements and waivers. Nor does it offer much comfort to the rest of us — citizens concerned about the basic fairness, stability, and integrity of our legal regime. Clearly, the problem of government by waiver calls out for some other solution.

LEVIATHAN UNBOUND

But can such a solution be found? Ultimately, the only real fix is to impose sharp constraints on the growth of government, and to adopt a different attitude toward government's role in our lives. After all, if the state is confined to the traditional functions envisioned by classical liberalism, the problem of regulatory discretion shrinks nearly to insignificance. If those boundaries are ignored, a ballooning state will also swell the benefits and burdens conferred upon citizens — which in turn will increase the dangers of government by waiver.

Of course, a total transformation of the role of government will not happen any time soon. Meanwhile, the problem of government by waiver — like the larger danger of excessive discretion — can be limited only by a greater awareness of these perils on the part of judges and administrators. The best we can hope for, then, is enlightened leaders.

And that is precisely the problem. The fate of our rights and liberties is left to the wisdom and discretion of individuals; we are therefore governed by men, not by laws. It was this exact circumstance that our system of government was designed to avoid: As James Madison noted in Federalist No. 10, "enlightened statesmen will not always be at the helm." In this sense, the problem of government by waiver shows us just how far we have strayed from the intentions of those who created our system of government — and how we risk betraying their hope that we might preserve it.

Richard A. Epstein is the Laurence A. Tisch Professor of Law at New York University, the Peter and Kirsten Bedford Senior Fellow at the Hoover Institution, and a senior fellow at the University of Chicago.