Waive of the Future? Federalism and the Next Phase of Welfare Reform

The Problem

The 1996 welfare reform law returned to the states primary responsibility for establishing rules under which impoverished households could receive cash assistance. Aid to Families with Dependent Children (AFDC), the program that gave cash payments to low-income households, was replaced by a block grant to the states called Temporary Assistance for Needy Families (TANF). In exchange for receiving this grant of fixed funding, states were afforded a great deal of flexibility to design their own welfare reform strategies. However, some new federal mandates were introduced, including a requirement that each state gradually put at least half of its welfare recipients to work or into work-related activities.

The welfare-to-work objective was predicated on a simple proposition: poor families are better off employed than on welfare. Jobs are the best antidote to poverty. The work requirements have helped increase the employment rate of single mothers, lowering welfare dependency and child poverty.

Unlike the straightforward provision of welfare payments, successfully putting unskilled people to work is a complicated process. In the late 1990s, states were able to meet the challenge because caseloads fell sharply and states used the savings to further fund child care and other services designed to help individuals and families become more self-sufficient, in some cases enabling them to avoid the welfare rolls altogether. But the shift in investment also threw into sharper focus the importance of providing work-related supports—including food stamps, child support, housing assistance, workforce programs, and more—that extend well beyond traditional income maintenance and instead assist the transition from welfare to work. A number of these programs are subject to disparate standards as to who can qualify, how funds are to be spent, how cases must be managed, and other separate stipulations that, unless aligned, may ultimately diminish the capacity of states to raise the long-term living standards of poor people.

A perceived need to improve program coordination led the Bush administration in 2002 to recommend that states be allowed to combine waivers from federal strictures across a wider range of social programs. This proposal—which came to be called a superwaiver—was incorporated into legislation that emerged from the House of Representatives later that year and again in 2003. A variation of the concept was also included in a bill passed by the Senate Finance Committee in 2003. These measures remain the subject of much debate.


Superwaiver legislation, along the lines of the House’s version, would authorize federal agencies to approve a state’s petition for relief from various statutory and regulatory provisions pertaining primarily to TANF, food stamps, child care, employment and job training services, family services, and public housing. (The Senate bill was less comprehensive.) How would this differ from the system of waiver requests and approvals in the past?

Executive branch agencies have for decades had authority to waive certain requirements of programs they oversee. Amendments to the Social Security Act in 1962 introduced the possibility of waivers to encourage innovation in administering AFDC. Legislation in 1977, in 1996, and in 2002 authorized the Secretary of Agriculture to allow significant discretionary adjustments in dispensing food stamps. State governors seeking more latitude in operating these or other programs, however, ordinarily would have to appeal to each patron agency in Washington separately. Under the new superwaiver regime, a joint interagency waiver board would accept, modify, or reject a single, unified application. The board would have to do so within ninety days, or a state’s request would automatically take effect.

More importantly, the superwaiver mechanism—at least as envisioned by legislators on the House side—would dispense with a number of traditional constraints on state waiver requests. Under the old system, projects undertaken with waivers in the administration of AFDC were frequently subject to independent evaluation. Precautions also were taken to prevent states from sharply cutting food stamp benefits or severely curtailing child care assistance. Most superwaivers (according to the House bill) could be less restrained. Not only would their programmatic content potentially be more extensive; so would the number of possible qualifying states that could duplicate more or less the same projects. Moreover, the states themselves, not an outside group, would assess the results of their own projects. And quite a few of the earlier waiver system’s restrictions—on possible benefit reductions, especially in the food stamp program—could be relaxed.

Still, superwaivers do not give states carte blanche. Restrictions in appropriations bills would be off-limits to superwaiver requests. The same goes for numerous other federal directives, including child support enforcement, immigrant eligibility rules, civil rights regulations, health and safety standards, and environmental protections. Of course, exactly how a future administration in Washington, or the courts, would ultimately interpret some of these restraints remains to be seen.

The Controversy

Several broad criticisms have been leveled at the superwaiver idea, especially as embodied in the House bill. One is that its administrative proceedings could invite closed-doors connivance between governors and cabinet secretaries whose decisions could bind without explicit congressional or public consent. Thus, authorizing superwaivers might signify a shift of control from Congress to the executive branch. The rejoinder to this complaint is that considerable precedent exists for sweeping delegation of decision-making power to the executive. Consider, for example, the authority Congress has repeatedly granted presidents to negotiate trade agreements.

Critics of superwaivers also assert that, technically, they may be unnecessary. Much of what superwaivers seek to do already can be accomplished under existing law. For example, state governments do not need a superwaiver to unify certain eligibility criteria across three programs: TANF, Medicaid, and now food stamps. In practice, however, the possibilities for action under present law are not always easy to ferret out and implement. State agencies with enough patience to avail themselves of all their supposed options, even within key programs (never mind across programs), are probably few. Superwaivers, if prudently designed, might help lubricate the bureaucratic wheels.

A more telling objection to the superwaiver scheme is that it is susceptible to misuse. Critics worry that some states might use their added flexibility to the detriment of the poor.

That possibility cannot be casually brushed aside. On one hand, dire warnings about state abuses accompanied every stage of the welfare reform process. For the most part, the worst fears were not borne out—at least judging from the record of state experimentation under the AFDC waivers and the actions of the states once the TANF block grant increased their administrative freedom. The states did not stampede to slash benefits. Indeed, under TANF, many states moved to liberalize aspects of their benefit structures, permitting AFDC recipients to keep a larger part of their welfare payment when they took a job, for example.

On the other hand, it would be naïve to rule out any danger that some states, if given free rein, will be tempted to pare programs that have relatively weak political constituencies, particularly during economic downturns. Recently, amid their budget crises, more than thirty-five states squeezed the spending rates of TANF and other programs serving low-income households. Superwaivers, presumably, would give states more room to move funds around and to change the priorities for state contributions. In fact, there is evidence that some states have diverted TANF dollars. In a 2001 study of ten states, the General Accounting Office (GAO) found that nine replaced state monies with federal TANF funds and shifted their own revenues away from cash assistance to service-based assistance like child care. Likewise, a widespread response among states to the block-granting of federal funds for various functions in the 1980s was to redirect outlays for programs like energy assistance to other social services that were deemed more urgent.

But surely such “supplantations” are not always and axiomatically harmful. Is a state that scales back and switches some energy or TANF money into welfare employment programs necessarily imperiling low-income families? Or is the state bettering the odds of providing those families with a potentially bigger payoff—jobs? Moreover, it is not certain that total spending for needy families invariably declines when funds are supplanted. During the opulent 1990s, the overall commitment of state resources (including Medicaid) to low-income families in most of the states surveyed by the GAO either remained about the same or increased. In any event, the gains from government programs for needy families depend on the efficacy of money spent, not just the absolute level of spending.

A final, though mostly unstated, misgiving about introducing superwaivers is that they might eventually pave the way for wholesale block granting of social programs, followed by sharp reductions in total federal funding. The history of block grant funding is mixed. When the Reagan administration consolidated scores of categorical grant programs into nine blocks and then reduced their budgets, the states mostly managed to maintain services for the poor. There have been significant exceptions, however. The Social Service Block Grant went from being an expansive source of money for a variety of services, including child care and training, to a capped program. Subsequently, funding for it was cut substantially in inflation-adjusted dollars. Examples like that disturb advocates for the poor when policymakers contemplate further devolutionary initiatives. Compared to categorical programs, block grants are more vulnerable to budget cuts, especially amid worsening federal deficits. In part, that is simply because the blocks are larger, hence more consequential targets for budgetary saving. Block grants also tend to lack the solid, well-organized political constituencies that protect special purpose programs. Thus, as an example, the block granting of food stamps (which would be permitted in up to five states if the House’s 2003 welfare bill were adopted) could gradually erode the program’s longstanding coalition.

Lawmakers should recognize these realities. Whether the concerns necessarily justify calling a rigid halt to proposed revisions of the status quo is a matter on which thoughtful observers may disagree. But this much seems plain: political arrangements that secure their social objectives by indirection are usually sub-optimal.

The foes of superwaivers have valid reasons to be uneasy. But as practically everyone agrees, none of the states today resemble the antediluvian polities—governed by reactionary governors, malapportioned legislatures, feeble bureaucracies, and passive courts—that existed in some regions of the country through the first half of the twentieth century. Few, if any, American state governments today seem prone to subvert rather than support the nation’s social safety net.

What Should Be Done

The superwaiver issue calls for a reasonable legislative compromise—one that can encourage states to innovate constructively in social policy while preserving adequate safeguards.

A first step toward such a compromise would be for all sides to acknowledge that the future of low-income families depends on far more than what state bureaucrats or governors do with “welfare” cases. Their decisions, fettered or unfettered, are but one consideration in a larger equation, the most important parts of which are the health of the national economy and the influence of the nation’s other increasingly work-based antipoverty policies, especially the Earned Income Tax Credit and child care subsidies. While opponents may see superwaivers as the opening wedge in the gradual defunding of various supports for distressed families, a more optimistic interpretation is that superwaivers, if granted and monitored with care, may enable states to use existing funds in ways that further encourage work and thus improve the long-term prospects of disadvantaged people.

Whether, on balance, superwaivers are a good idea is an unsettled question. They carry both risks and potential rewards. The challenge is to tilt the balance toward the latter. To that end, we offer several recommendations.

First, program rules initially should be eased only in a limited number of states to get a better look at what they might produce, both in terms of program innovation and results. This prudent approach was envisioned in the Senate Finance Committee’s 2003 welfare reauthorization bill, which permitted up to ten such state experiments. States would apply for superwaivers from a centralized, interagency board, as proposed by the Bush administration. The criteria for approving a state’s superwaiver application would include the extent to which its plan of action was designed to use existing funds in new and useful ways, the state’s track record in managing its programs, acceptance by the state bureaucracies involved, and evidence that the plan’s central aim is to relieve poverty and promote the well-being of children.

Second, if the promise of superwaivers is to be fulfilled, states have to enjoy a suitable measure of trust. Vigilance against their possible misuses of federal funds is necessary, but it has practical limits. If policed too fastidiously, some states—even eminently deserving ones—may not bother to apply. Those that do may wind up taking only marginal steps of little lasting significance. And chances are, the few intent on misusing funds will find ways to get around many of the federal prohibitions. Specifically, we question the practicality, if not the merits, of rigid proscriptions against reallocating funds across programs serving similar target populations. If, for example, a state wants to use housing funds to pay for child care or place a client in a substance abuse treatment program, that decision should not be barred. Too frequently, federal regulations and funding limitations that forbid such choices prove unhelpful at the local level, where the focus needs to be on what will assist the client most.

Third, truly bold reforms are not always cost-neutral. The existing legislative proposals would not allow any waiver to increase outlays by the federal government. While a budget constraint of this sort obviously makes sense over time (say, a five-year period), imposing it rigidly from the outset can be counterproductive. To begin with, realigning some program rules will cost money if existing beneficiaries are to be adequately protected. In addition, there are start-up costs in any new enterprise. People need to be retrained, computer systems upgraded, forms and websites redesigned, and coordination issues addressed. Finally, from a political perspective, underwriting state demonstrations, even if only modestly, could help dispel the suspicion among critics that superwaivers might be a budget cutting exercise in disguise. Just as promoting marriage may be a worthy objective, but one we do not yet know how to achieve, the next generation of antipoverty policies may require an extra investment in ventures that can teach us what really works. For all these reasons, a new waiver system, to be successful, probably calls for some seed money.

Fourth, a strong and enforceable maintenance-of-effort requirement is a must. As the 2001 GAO study indicated, states move money around, and do so in ways that may be hard to trace. GAO suggests that the best way to oversee the situation is not by trying to track separate funding streams, but by ensuring that states maintain their own funding on relevant programs. Such maintenance-of-effort provisions are standard fare in many programs and are easier to enforce than anti-supplantation provisions.

Fifth, the demonstrations should be evaluated rigorously by third parties using scientific methods. At a minimum, these ought to include systematic comparisons of policy outcomes in states or sub-state areas that have demonstrations with those in other states that do not, controlling, to the degree possible, for a variety of unique state or area characteristics. In addition, case studies ought to be conducted to examine what differences superwaivers actually make, both to identify best practices and to better understand liabilities. By setting broad objectives, such as reducing poverty and improving child well-being, and then assessing the states’ ability to achieve these goals through superwaivers, much also may be learned about the pros and cons of performance-based block grants. The lessons from this experience could inform and shape the future of the federal-state partnership for years to come.

In the end, Congress may deem that the superwaiver notion entails unacceptable hazards, and so may refuse to sanction this novel administrative instrument. But such a risk-averse stance has its own potential costs, for it could forfeit some socially desirable policy adjustments that would further assist families in raising their skills and income. There is no infallible way of knowing in advance whether the conceivable benefits of experimenting with superwaivers outweigh the possible perils.