The states are broke. The economic recession and the attendant collapse of state revenues, the continuing costs of state programs created during the boom years, and the spiraling costs of federal mandates will force most states to enact massive tax increases and budget cuts. Predictably, the states are demanding more money from Washington, D.C. That, however, is the wrong prescription. The correct approach is to segregate state from federal functions. This Outlook provides principles and practical reform proposals for the most critical programs-Medicaid and education.

The Fiscal Crisis of the States

For the current fiscal year, the states face a combined deficit of over $30 billion. Having exhausted gimmicks, easy fixes, and tobacco payments in last year's budget cycle, the states will have to cover that sum and next year's estimated $80 billion deficit through some combination of tax increases, spending cuts, and higher federal transfer payments--the governors' preferred option.

Each economic downturn produces the same fiscal crisis and accompanying clamor. Think tanks round up their usual culprits: the Cato Institute fingers the states' boom-time spending commitments, while Brookings fellows identify a complicated mix of factors (usually including state tax cuts). The National Governors Association (NGA) characterizes the states' crisis as (a) the worst ever; (b) unforeseeable; and (c) a result of unfunded federal mandates. The NGA's press releases appear nearly verbatim in David Broder's syndicated column.[1]

The handwringing over the states' collective pickle masks large differences among them. Wyoming has no fiscal crisis. Other states are hit hard but have only themselves to blame. California, for instance, owes a good chunk of its $34.6 billion deficit (over 35 percent of its total budget) to the lingering costs of its harebrained energy policies. That said, the states' fiscal crisis is in fact systemic. Like the proverbial canary, it signals a grave problem in federalism's subterranean architecture--the lack of transparency, accountability, and responsibility that is endemic to federally funded, state-administered programs. Instead of silencing the bird by stuffing money down its throat, we should disentangle federal responsibilities from state and local ones.

Up, Up, and Away

Pace the NGA, the states' systemic fiscal crisis is not a revenue crisis. It is a federally induced spending crisis.

To a far greater extent than the national government, states must fear that businesses and citizens will flee high-tax jurisdictions. Relative to federal revenues, one would expect state revenues to erode over time. Curiously, though, the opposite is the case. While the federal tax take--as a percentage of gross domestic product (GDP)--is now roughly what it was after World War II (about 17 percent), state and local tax receipts have more than doubled, from five percent of GDP to over 10 percent. The answer to this puzzle lies in "cooperative" federalism. Federal funding of state-administered programs permanently inflates the demand for government.[2]

The clearest examples are the uncapped federal matching grants for state-provided services, such as Medicaid. A fifty-fifty federal match reduces the price for each unit of service by fifty percent. Hence, a recipient-state will provide the service at up to twice the pre-grant level. The state can do so without raising taxes or shifting money from competing government programs.

When the state has to slash expenses, it can save only 50 cents on the dollar by cutting the federally funded program. Such a cut will look more expensive--and will require more draconian steps--than an equivalent cut in a state-funded program. Even if the federally funded program is more generous than what the local citizens in the pre-grant world were willing to pay for, the state will prefer to cut competing state-funded programs or, failing that, will raise taxes.

Not all federal grants are matching grants, and few are uncapped. Education funding, for example, typically takes the form of conditional, capped grants-in-aid. Unlike matching grants, grants-in-aid have an income effect but no price effect. To illustrate, suppose a state provides a basic education for $90 and the feds offer $5 funding for, say, diversity enhancement in driver's education, at an additional cost of $5 to the state. For the same reasons that induce the acceptance of matching grants, the state may take the fifty-fifty bait, bringing its total education spending to $100. The state's cost of providing a basic education remains unaffected.

When budget constraints force a 5 percent cut in education spending, a rational state should send the diverse driving teachers and their federal paymasters packing. (It would be no worse off than in the pre-grant world.) But that will not look like a 5 percent cut. Because of the "loss" of federal dollars, it will look like a brutal 10 percent cut. In other words, states and their citizens will confuse the income effect of the grant-in-aid with a (nonexistent) effect on the price of education. Once that natural fiscal illusion takes hold, a grant-in-aid works the demand-inflating wonders of a matching grant.

Whatever their form, then, federal funding programs tend to expand state budgets. Over time, those programs grow faster than wholly state-owned functions and account for an ever-larger portion of state budgets. By making expansions look cheap and cuts outrageously expensive, the programs tend to exacerbate the states' boom-and-bust budget cycles. All along, citizens get more government than they are willing to pay for.

The Medicaid Monster

The biggest federal funding program by far is Medicaid, created in 1965 as a Great Society program. Medicaid is an open-ended health care entitlement program under which the federal government generously reimburses state health care expenditures for eligible individuals. The federal reimbursement ranges from 50 percent of state expenditures in rich (mostly northeastern) states to 77 percent in the poorest (Mississippi). In good times, state revenues cover the remainder. In bad times (such as now), states demand more federal money.

Since its inception as a modest safety net for poor families, Medicaid has grown by leaps and bounds:

• Medicaid spending growth has outstripped even the (wholly federal) Medicare program, which is uniformly viewed as unsustainable. In 2002, for the first time, Medicaid spending ($258 billion) exceeded Medicare spending ($230 billion). The Congressional Budget Office (CBO) projects the gap to widen in coming years.[3]
  
• Medicaid expenditures constitute an ever-growing share of state expenditures. In 1987, that share amounted to slightly over 10 percent. In 1992, the number was 17.5 percent; in 2002, 20.5 percent.4 Medicaid has crimped every other category of state spending.

The states attribute Medicaid's manifestly unsustainable growth to three factors: health care inflation (especially the pharmaceutical industry's pricing practices), federal mandates, and--in recent years--rising poverty.

They are wrong. The AEI Federalism Project's analysis of Medicaid spending trends[5] shows the following:

• Medicaid spending did not explode in consequence of the economic downturn. It exploded during 1994-2000, when poverty rates sank and revenues ballooned in all but a handful of states. Even so, no state except New Hampshire realized meaningful Medicaid expenditure reductions.[6]
  
• Health care inflation and federal mandates cannot explain the Medicaid explosion. Rather, independent state policy decisions produced Medicaid's unsustainable growth. In particular, state revenue growth appears to be a reliable predictor of Medicaid expenditure growth.

Medicaid is indeed responsible for the states' fiscal woes--but not for the reasons suggested by the governors. Far from imposing on the states, Medicaid gives them a strong incentive to experiment beyond their means.

Mistakes Were Made

For the most part, Medicaid is not a mandate. True, participating states must cover basic services for certain groups--principally, poor families that, prior to the 1996 welfare reforms, qualified for the federal Aid to Families with Dependent Children (AFDC) program. Those mandatory services, however, constitute a steadily diminishing share of Medicaid. Nearly 70 percent of expenditures are now devoted to services and populations that the states may, but need not, cover under Medicaid.

Throughout the 1990s, the states expanded optional services and Medicaid eligibility. Some states now cover families with incomes of up to 275 percent of the poverty level. Almost all provide optional prescription drug benefits and long-term care for the poor and low-income elderly. In states like Delaware (Delaware!), one-third of the population is now on Medicaid.

Many statutory requirements that might increase costs have been waived. Under the statute, for example, states may not use Medicaid funds to subsidize health insurance premiums for an insured low-income family. They must provide the entire Medicaid package, now running at an eye-popping $6,000 per year per covered individual. Beginning under the Clinton administration, however, the Department of Health and Human Services (HHS) waived such requirements and permitted states to use Medicaid funds for insurance as well as health service purposes. Upon assuming office in 2001, Secretary Tommy Thompson further loosened regulatory requirements and has since issued waivers like they are going out of style--over 2,500, at last count. Overwhelmingly, HHS waivers have further facilitated the states' expansion of Medicaid coverage and services.

Now that staggering budget shortfalls are forcing cutbacks, constituents and officials are howling. In moments of candor, the states acknowledge their complicity in the Medicaid explosion--but still demand a federal bailout, arguing that that's the way the program should work. In a recent begfest technically designated as congressional "testimony," New Mexico governor Bill Richardson argued that states "should be given more flexibility to cover pregnant women, parents, uninsurable adults and to expand coverage to children with disabilities." But--but!--"flexibility should not mean having to cut people from the rolls."[7]

New Mexico is what Mr. Richardson in an earlier career might have called a "state of concern": reckless even in comparison to sister states, Santa Fe staged a raid on the federal Medicaid bank in the hope that expanding revenues would easily finance the state's contributions.[8] It is unfair that Mr. Richardson should have to bear the consequences of that rogue miscalculation. But then, so is his demand that the feds pay for the getaway car.

TOLES © 2003 The Buffalo News. Reprinted with permission of UNIVERSAL PRESS SYNDICATE. All rights reserved.

True Mandates

Although Medicaid is not a mandate, federal impositions do constitute a prominent and troublesome aspect of federal-state relations. Somewhat surprising for an administration that includes many former state governors (including the president), Team Bush has piled new obligations on state and local governments. Federal election reform is on that list, as are homeland security mandates for which Washington will provide little or no funding.

Water systems serving more than 3,300 consumers, for example, must henceforth conduct "vulnerability assessments," at an estimated--and wholly unfunded--cost of $900 million. The requirement covers remote rural districts that are threatened by no one except the U.S. Environmental Protection Agency.[9]

All the new mandates, though, pale in comparison to the No Child Left Behind Act (NCLB), the administration's 2001 education reform. At its core, the act requires federally funded schools (meaning effectively all public schools) to demonstrate "adequate yearly progress," as measured by rising student test scores. The administration argues that NCLB authorized, and Congress appropriated, additional funds to implement the statute, including full funding for the development and administration of student tests. State and local officials counter that NCLB requires a wholesale revamping of entire education systems. Preexisting education mandates--especially the chronically underfunded Individuals with Disabilities in Education Act (IDEA), which governs schools' obligations concerning learning-disabled children--already impose inordinate burdens on state and local school budgets. The new NCLB mandates, officials say, could require some additional $35 billion in state and local education spending per year.[10]

Coming as it does from parties with an unquenchable thirst for more federal money, that projection is open to question. Unlike the Medicaid crisis, however, the states' financial difficulties under federally funded education programs are not self-inflicted. To that extent, the states are right. What they have wrong is the remedy.

Bush to States: Drop Dead

The present economic downturn, as all downturns before it, has generated noisy demands for more federal funding. State and local officials wail over the need to cut the few remaining functions (such as firefighting) that are still financed almost entirely from state and local revenues. Nanny state constituencies protest the few cuts that the states are prepared--and allowed--to make in federally funded programs. Economists lament the fiscal drag of state tax hikes on an already soft economy.[11] 

State governments have found open ears on both sides of the congressional aisle. Bipartisan groups of legislators have proposed substantial expansions of federal Medicaid funding, including a temporary upward revision of the federal matching formula. The White House, in contrast, has rejected the states' demands. The administration's budget proposes $398.8 billion in federal aid to state and local governments. That sum represents a nominal increase of 3.8 percent over the current year. After inflation and net of escalating Medicaid payments, federal transfers would in fact decline.[12]

Contrary to conventional wisdom, "drop dead" is precisely the right policy--not simply for fiscal reasons, but as a corrective to the bizarre incentives created by federal funding. A rip-roaring fiscal crisis is the only check on the states' tendency to succumb to federally induced demand inflation. A federal bailout would compound one moral hazard problem with another.

But while tightfistedness is the beginning of good sense, the administration has so far failed to develop a broader strategy that might sustain fiscal restraint against political assaults. It has experimented with "devolution" under Medicaid and with increased centralization and "accountability" in education. Both approaches are political and policy disasters in the making. Spiraling costs are the fiscal face of a much larger problem.

States' Rights Socialism

The administration's Medicaid agenda has been "devolution" under expansive HHS waivers. The purpose is to help states cover the uninsured, which sounds good and, moreover, helps to keep that pesky problem off the congressional agenda (where it might play to the Democrats' advantage). At the same time, HHS has sought to control Medicaid costs by curbing the states' attempts to inflate federal payments. Those policies, however, collide both with each other and with Medicaid's incentive structure.

The states have proven enormously inventive in boosting federal payments. The principal, highly developed technique is the creation of phantom Medicaid expenses for a federal match. For decades, the stratagem has survived periodic congressional and administrative crackdowns. Fifty renegade states will always outfox a single federal cop, and the inventors of phantom expenses have a huge information advantage over the detectives. The credibility of federal controls is measured by the fact that some states (such as Pennsylvania) are currently trying to solve their budget problems by expanding Medicaid eligibility.

Devolution only enhances the room for state shenanigans. HHS waivers are supposedly budget-neutral: states are expected to cover additional services and populations at a cost no greater than what the federal government would have paid under Medicaid-as-usual. A few responsible governors (such as Florida's Jeb Bush) have in fact managed to institute coverage-enhancing, budget-neutral experiments. Most states, however, inflate Medicaid cost projections to maximize federal payments, declare vast groups (in Tennessee, all citizens) eligible for an ever broader array of services, invent some managed-care or price control system that should save money!, and name the beast after the state animal ("BadgerCare"). Invariably, the scheme crowds out private health insurance, while costs go through the roof. During the 1990s, the most monstrous cost increases occurred in the states with the most expansive HHS waivers.[13]

HHS now seems to sense that the problem is not a lack of pilot programs but rather is state experimentation on autopilot. To discipline the states' choices, the department has proposed a minicap for optional Medicaid services: states that develop programs to cover additional eligible populations would get extra funding for a few years but (effectively) less thereafter. HHS has the principle and the timing right. Its proposal, however, is fiscally too modest and, moreover, inadequate to tackle Medicaid's central political problem--the stealth socialization of American health care.

After the spectacular collapse of the early Clinton administration's wholesale attempt to socialize health care, Democrats switched to a piecemeal strategy. They have seized on Medicaid as HillaryCare on federally subsidized wheels. Compare and contrast:

• In 1993, the GOP denounced President Clinton's proposal for a long-term-care entitlement under Medicare as outrageously expensive. The purported beneficiaries turned against the proposal upon learning of its envisioned copayment requirement, and the Clinton proposal collapsed. Or did it? A federally funded long-term care entitlement is now available under Medicaid to the elderly poor, the near-poor, and anyone with an accountant who will render his clients Medicaid-eligible by transforming their assets into annuities. Attempts to curb such widespread sham transactions through federal "granny goes to jail" regulations have proven oddly unpopular.[14]
  
• A Medicare prescription drug benefit has been a matter of intense controversy for over a decade. Most recently, the Bush administration has proposed such coverage on terms that Democrats have denounced as unconscionably stingy. A vastly more generous prescription drug benefit already exists for Medicaid recipients--including the "dual-covered" elderly, who are eligible for both Medicare and Medicaid.

Senator Clinton need not do anything to advance her agenda nor even defend it. The wholly federal Medicare program is a subject of constant national debate--because the Congress that legislates the benefits must at least estimate, and later pay, the attendant costs. Medicaid entitlements, in contrast, are created in Albany and paid mostly by Washington. When the states can no longer pay their own modest share, they send their governors to beg Mrs. Clinton for more. That is not a debate; it is a charade.

Serious Medicaid reform must change the states' incentives and, at the same time, renationalize the debate over health care socialization. The HHS minicap cannot do the job; a bolder approach is needed.

Medicaid Rescue

The way to change the states' incentives is to cap the federal Medicaid share and to convert it into a block grant. The way to force a responsible political debate is to make the federal government or the states, separately, bear 100 percent of the costs of any proposed program expansion.

It is understandable that HHS should eschew such bold proposals: both have been shot down before. In the 1980s, President Reagan proposed to "swap" the major welfare programs: the federal government would assume full financial responsibility for Medicaid, while the states would take full financial and political responsibility for AFDC. The states killed that proposal. House Republicans pushed the Medicaid block grant option in 1995. Their proposal died in the intergovernmental showdown and shutdown that year. Two variables have changed, however. First, both the White House and the Congress are in Republican hands. Second, the states' dire fiscal predicament might induce them to bite on proposals that promise immediate relief. Under these circumstances, a combined swap-cap option might succeed.

Since Medicaid now dwarfs welfare programs, the Reagan swap is no longer an option. It is possible, however, to swap funding and program obligations inside Medicaid. Let Washington take full fiscal responsibility for some Medicaid populations; in exchange, block grant, cap, and phase down federal funding for the remainder.

The states already demand full federal funding for optional but budget-busting Medicaid services for the elderly--long-term care, and prescription drug benefits. The administration and Congress should accede to the states' impetuous demand. They should, however, insist on two conditions: an absolute bar against complementary state Medicaid benefits for the elderly, and the conversion of the rest of Medicaid into a capped block grant.

These proposals are ambitious but not outlandish. As explained, we already have a federally cofunded long-term care entitlement--albeit a very expensive and poorly designed one. (Medicaid covers only institutionalized long-term care but not less expensive--and for many recipients more desirable-assisted living arrangements.)[15] The GOP might as well put the entire cost on the federal budget, take credit for the entitlement, and redesign it in a more sensible fashion.

Federal assumption of the responsibility for Medicaid prescription drugs for the elderly is an urgent necessity. If and when a Medicare drug benefit is enacted, a significant segment of the elderly whose prescriptions are now covered by Medicaid will migrate to the new program. Once that bookkeeping transaction brings Medicaid costs "under control" and state revenues improve, states will resume their devolutionary experimentation and by 2005, Nebraska will put Warren Buffett on Medicaid. The administration should roll the dual-covered elderly into the new Medicare before that happens--and extract a Medicaid cap.

The swap need not be "revenue neutral" in some phony CBO projection. As HHS's minicap proposal recognizes, Medicaid reform must offer current state officeholders a bag of cash and lock the federal vault for their successors. Washington should offer whatever it takes to buy back the states' right to feed local interests with federal dollars. That price is lowest when the states are staring down the barrel of a gun--in other words, now.

No Educator Left Behind[16]

Medicaid devolution is one pole of GOP federalism. The No Child Left Behind Act represents the opposite pole--fiscal restraint (as these things go), coupled with aggressive federal interventionism. That was not the plan: under the administration's original design, Washington was to hold states accountable for results, while giving them more flexibility in deciding how to meet tough federal standards. But while such "results-based accountability" works like a charm in the parallel universe inhabited by experts at the Progressive Policy Institute and the Fordham Foundation, it disintegrates on contact with the real world.

By the time President Bush signed education reform into law, his elegant 30-some-page bill had ballooned into an 1100-page compendium of requirements on testing standards, student assignments and ethnic identification, and other minutiae purportedly calculated to ensure "adequate yearly progress." The price of obtaining Senator Kennedy's crucial support for the bill was to cut the unions in on "reform." Hence, many of the most prescriptive--and, for the states, expensive--provisions govern teacher training, paraprofessionals' certification, and other pay-boosting and leisure-enhancing policies. (It's for the children.)

Federal accountability measures follow the Hans Blix model. The Department of Education conducts site visits, threatens sanctions for infractions ($1 million for Michigan's failure to publish student tests in foreign languages), and from time to time unctuously finds progress towards "adequate yearly progress"--though not enough progress to end the inspections. The only nonintrusive accountability mechanism, the federal defunding of noncomplying school districts, is as credible as the prospect of a French invasion. It has not been used in nearly four decades of federal education funding. To dispel suspicions of future use, NCLB threatens additional funding for failing schools.

While less federal oversight would mean zero local accountability--the education "blob" (in William Bennett's memorable phrase) would simply pocket the money--the NCLB policy of accountability to the federal government is also a losing proposition. The late Daniel Patrick Moynihan compared a similarly inspired education initiative (the Clinton administration's Goals 2000) to a Soviet grain production quota. NCLB's "adequate yearly progress" is a moving target--a quota on a bicycle, if you will.

Recognizing that schools must be accountable to parents, the original Bush proposal offered parents vouchers to escape failing schools. Senator Kennedy and the National Education Association, however, declared vouchers anathema. Hell-bent on a reform bill, the White House crushed conservative think tanks and legislators who insisted on meaningful school choice. Parental choice is dead. The pervasive discontent with NCLB, however, coupled with the states' financial woes, presents a different reform opportunity: let states opt out.

If At First You Don't Secede . . .

The only state to resist the Medicaid lure (see above), New Hampshire may also become the first to reject NCLB. In April 2003, the state House of Representatives voted 310-36 to decline NCLB funds, to permit local school districts to do likewise, and to bar state administrators from implementing NCLB. With any luck, the bill will become law later this year.

In antitax New Hampshire, NCLB sailed into trouble when the state's School Administrators Association (NHSAA), in a widely publicized study, estimated that the acceptance of $17 million in NCLB funds would saddle the state with compliance costs of at least seven times that amount. ("Cost to Attract & Retain High Quality Teachers: $11,716,728.93. Costs to Attract & Retain High Quality ParaProfessionals: $16,558,344. Shift of role, function, and control of local budgets: Priceless.")[17] Pro-NCLB forces--a conservative state think tank and the Business Roundtable, in further evidence of the weirdness of the world-countered that New Hampshire would (naturally!) save money by accepting federal funds. The Completely Great Granite State's representatives have rightly ignored that advice.[18]

At first impression, rejecting a federal dollar that forces a seven-dollar local tax hike is a no-brainer.

New Hampshirites would be better off by not taxing themselves--or, alternatively, by taxing themselves to purchase educational services that suit their own preferences, rather than Ted Kennedy's. If states rarely perform that calculation (let alone act on it), that is because federal grants do not fund education but educators--unionized teachers who benefit from federal class-size goals; education schools that credential teachers; grant writers; compliance personnel; and administrators. Feckless state politicians accommodate those constituencies by accepting federal grants. When the costs hit home in later years, state politicians kvetch about "unfunded mandates" and the fact that "the federal government pays only 10 percent of the cost of education."

NCLB's redeeming feature is to push the intergovernmental conspiracy to the breaking point. The law hits state education systems that are already chafing under the costs of other federal education programs (especially the IDEA special-education program), at a time of general financial distress. In addition, NCLB's champions (such as the Fordham Foundation's Chester Finn) describe it as "probably the most intrusive piece of federal education legislation in history." Confronted with onerous federal demands, the interests that usually lobby state legislatures in support of federal objectives are quite ambivalent about NCLB--except in New Hampshire, where school administrators have urged school districts to post "a warrant article designed to encourage state and federal officials to actively resist" the law.[19]

According to the Department of Education and its supporters, the pervasive discontent proves that NCLB "works." But while a blob-supported education agenda would surely be a racket, it is a logical error to conclude that blopposition ipso facto spells sound policy. Organized education interests care primarily about their members, not about children. Federal bribes cannot change that disposition; neither can federal inspections and harangues. NCLB will have zero impact on learning outcomes--if all goes well.

That dire assessment is naturally disputed both by the Department of Education reformers on the quota bicycle and by the experts who supplied the training wheels and mudguard stickers. ("Give Me Results-Based Accountability, Or Give Me Death!") But while the reformers pedal through forbidding terrain and impale themselves on the NEA's Democratic Guards, states should think hard, as New Hampshire has thought, about the wisdom of joining their campaign. A Republican administration, for its part, should fight its education crusade with volunteers, not conscripts.

. . . Try Again . . .

Federal lunches are expensive meals. When state legislators accept the funds, they implicitly vote for an accompanying stream of state taxes. How does one create more New Hampshires--states that seriously and openly consider the trade-off?

When the state assembly committee for the free lunch convenes its interest group clients to discuss the budget or education policy, citizens cannot monitor the backroom maneuvers. Citizens must have some procedural mechanism--ideally, a constitutional mechanism--that constrains politicians and interests in advance.

Since legislatures have little incentive to discipline themselves in that fashion, states with a constitutional initiative and referendum process--about half of all states--should enact a constitutional amendment that (1) commands the state government to conduct a "fiscal impact analysis" on the costs of accepting a federal grant over the life of that grant and (2) forbids the acceptance, without an affirmative vote in both houses of the state legislature, of any federal grant with a cost impact above a certain threshold-say, one state dollar for every federal dollar received. (This particular threshold would "save" Medicaid, from which no state can realistically withdraw.)

Such a constitutional rule would preclude legislators from taking the money first and whining about "unfunded mandates" later. It would compress the policy choice into a single, up-or-down vote. The rule could apply prospectively, or it could force affirmative votes on programs in which the state already participates. It could require a simple majority or a supermajority. It could apply to individual programs or entire functional categories. The details matter little. What matters is to construct a mechanism that forces an explicit and accountable policy choice.

. . . and Again

The proclivity of state governments to accept federal funds and the attached strings is exacerbated by an oft-observed financial asymmetry: states can opt out of federal grants-in-aid but not the taxes that pay for those grants. When Congress establishes a conditional grant program to the tune of $1,000, the average state's citizens pay $20 of the program. They can choose to have that amount distributed to the other forty-nine or else have the $20 (or whatever is left, after the dollars have spent an expensive Washington night on the town) revert to their own government.[20] Obviously, they will choose the latter option--even if they would never tax themselves to establish a similar program.

The way to counteract the pro-spending bias is to permit a statutory state opt-out from funding as well as program participation. If a state declines participation in a federal funding program (Congress could and should provide), that state's citizens will receive a tax rebate--proportionate to their home state's federal tax share and their individual payments--on the following year's tax return.[21]

Arguably, such an amendment might be ineffectual. K-12 education spending amounts to slightly over one percent of federal spending. Per capita federal taxes are roughly $6,000 in rich states. Even if a state declined to participate in any federal K-12 program, the average taxpayer's refund would be only $60. That amount seems too small to neutralize the fiscal illusions and interest group pressures that operate at the state level. But while the refund would operate against the federal tax, it would require a state legislative decision to opt out. At that level, a few tax dollars often generate serious, ideologically charged debates. A federal refund option would enable state politicians to present citizens with a choice between paying taxes to buy an odious federal imposition and "a tax cut for working families." So framed, the question has an obvious answer.

The education bureaucracies' fear of a debate on those terms would likely turn a proposal for a funding opt-out into a poison pill. Still, some courageous federal legislator should roll it out in full regalia. (While it is probably too late--and, in light of the administration's political commitments, too dangerous--to tack that amendment onto NCLB, the reauthorization of the widely despised IDEA later this year might provide a good opportunity.) At a minimum, an opt-out proposal would enhance political accountability by forcing the political establishment to choose sides. Governors would have to decide whether they want to be beggars or federalists. David "More-Money-For-The-States" Broder might have to tell us whether he would allow independent state choice and experimentation--or whether his "federalism" is welfare state flackery. NCLB's champions would have to declare whether they believe in local control (as they profess) or in "accountability" to a professional elite. We would learn whether the Department of Education has sufficient confidence in its accountability scheme to permit unbiased, independent state choices about the acceptance.

Prediction: they will all denounce an opt-out as a reactionary assault on public education. Unfair? Let's find out.

Why Now (More Than Ever)

Eyeball-glazing at the most uneventful of times, a disquisition on federalism's dysfunctionalities seems, under present circumstances, exquisitely mistimed. Even so, the effort may not be altogether unseasonable.

At the state level, the fiscal crisis and the attendant call for tax hikes and more federal money highlight a long-term problem that ought to occupy antitax constituencies. Grover Norquist's Americans for Tax Reform has had great success in collecting no-new-tax pledges from state governors and legislators. Those valuable commitments, however, cannot stem the hydraulic pressure of federal funding programs. Even in the thriftiest states, under the most responsible governors, something will have to give. A plausible strategy to enforce state spending and tax restraint must address state participation in federal funding programs, along the lines suggested.

At the national level, foreign commitments are on a collision course with perceived domestic needs. The White House has announced its choice: billions for Baghdad, not a nickel more for Nebraska. Wholly apart from obvious national security imperatives, the administration's position is a timely step toward domestic reform. Our dysfunctional federalism can be reformed only in times of a fiscal crisis in the states. Those crisis opportunities, however, are typically missed--and misused as an occasion for further financial and political entanglements. If the administration can resist state and congressional pressure in that direction, it will have done considerable good.

As noted, though, federal funding programs institutionalize a bias for government growth. That bias survived the Reagan revolution and the Gingrich revolt. It will easily swamp a managerial, nonideological fiscal control effort. A plausible agenda must rest on a clearer distinction between national and state and local responsibilities, with the aim of restoring something closer to a constitutional equilibrium.

Instead of entrusting states with entitlement creation on Washington's nickel, let the national government assume political and fiscal responsibility for some state-created national entitlements. Since we can no longer fight Medicaid entitlements, let us buy them back, declare victory, and--as for future entitlements--make the states play with their own money. Instead of serving up bromides about education as "a local function but a federal responsibility" (the administration's incomprehensible NCLB rationale), let us decide what belongs where. While NCLB was a manifest mistake, the administration need not now renounce a federal role. It can continue to offer help and intergovernmental accountability, while permitting states a free and unbiased decision to accept or decline that assistance.

National and local responsibilities cannot always be segregated with theoretical clarity or political plausibility. In the important area of homeland security, compartmentalization is particularly difficult. But while intergovernmentalism in that arena is inevitable, so is its consequence. The first breakdown will be everybody's and nobody's fault, and ensuing attempts to affix blame and responsibility will prove a shell game without a peg. Precisely because we must sometimes tolerate those consequences, we must avoid them when we can.

Notes

1. See, respectively, Stephen Moore and Stephen Slivinski, "Fiscal Policy Report Card on America's Governors: 2002," Cato Institute Policy Analysis No. 454, September 20, 2002; Donald Kettl, "Bush and the 50 Beggars," Governing (February 2003), p. 14; National Governors Association, "The State Fiscal Crisis," Legislative Update, February 22, 2003; David S. Broder, "Stiffing States and Cities," Washington Post, November 24, 2002, p. B7; Broder, "Straight From the Governors," Washington Post, February 26, 2003, p. A23; and yet again Broder, "Bush's Tax Cut Brushoff," Washington Post, March 9, 2003, p. B7.

2. This view is widely held among political scientists (see, for example, Aaron Wildavsky, "Fruitcake Federalism or Birthday Cake Federalism," in Federalism and Political Culture, ed. David Schleicher and Brendon Swedlow [New Brunswick, N.J.: Transaction Publishers, 1998], p. 55) and among economists of the fiscal federalism school (see, for example, Edward M. Gramlich, "Intergovernmental Grants: A Review of the Empirical Literature," in The Political Economy of Fiscal Federalism, ed. Wallace Oates [Lexington: Heath, 1977], p. 219). In contrast to the paragraphs in the text, which discuss state responses to federal grants under changing state budget constraints, the fiscal federalism literature typically models state responses to changing federal grant amounts (under constant state budget constraints). Copious empirical evidence shows that state and local demand for services will rise in response to federal grants but fail to decrease proportionately to a grant decrease. Fiscal federalism theory cannot fully explain the asymmetry; it is usually attributed to interest group dynamics at the state level. See, for example, Gramlich, "Federalism and Federal Deficit Reduction," National Tax Journal 40 (1987): 299. A recent, sophisticated discussion and review of the literature is Shama Gamkhar, Federal Intergovernmental Grants and the States (Northampton, Mass.: Edward Elgar, 2002).

3. For Medicaid and CBO prediction see James Frogue, "The Future of Medicaid: Consumer-Directed Care," Heritage Foundation Backgrounder No. 1618 (http://www.heritage.org/Research/HealthCare/BG1618.cfm) (visited March 11, 2003). For Medicare, see Final Monthly Treasury Statement of the Receipts and Outlays of the United States Government, FY 2002, p. 30 (http://www.fms.treas.gov/mts/mts0902.pdf) (visited March 13, 2003).

4. National Association of State Budget Officers, State Expenditure Report (Washington, D.C.: National Association of State Budget Officers, 1991, 1993, and 2001), pp. 42, 86, and 50 respectively.

5. The analysis, "What Goes Up May Not Go Down: State Medicaid Decisions in Times of Plenty" is available on the project's website at http//www.federalismproject.org/masterpages/publications/index.html.

6. See the study cited in note 5. While Louisiana also experienced lower Medicaid spending, that drop seems largely attributable to a congressional crackdown on certain bookkeeping transactions through which Louisiana had previously inflated federal Medicaid payments. See Senate Budget Committee, Republican Staff, Budget Bulletin No. 6, March 3, 1997 (http://www.senate.gov/~republican/analysis/archive/1997/ bb6.pdf ) (visited April 4, 2003).

7. Testimony of Governor Bill Richardson before the House Committee on Energy and Commerce, March 12, 2003 (visited March 14, 2003).

8. Between 1994 and 2000, New Mexico Medicaid spending rose by 58 percent (constant dollars), despite a 20 percent drop in the state's poverty rate. That growth rate is twice the state average and the fourth highest among all states. Medicaid spending as a percentage of state revenues increased by a mind-blowing 53.8 percent. No other state came close. Data, sources, and further analysis are available in the AEI Federalism Project study cited in note 5.

9. A 2001 EPA rule requires the elimination of trace amounts of arsenic in drinking water, at a cost of over $200 million annually. Considering the necessary shift in state and local resources, an AEI-Brookings Joint Center for Regulatory Studies paper estimated that the rule would result in a net loss of ten lives per year. See Jason K. Burnett and Robert W. Hahn, EPA's Arsenic Rule: The Benefits of the Standard Do Not Justify the Costs, Regulatory Analysis 01-02 (AEI-Brookings Joint Center for Regulatory Studies, January 2001) (http://aei.brookings.org/ admin/pdffiles/phppu.pdf) (visited April 4, 2003).

10. Testimony of Oklahoma state senator Angela Monson (also president of the National Conference of State Legislatures) before the U.S. Senate Finance Committee, February 11, 2003 (http://www.ncsl.org/standcomm/scbudg/econ_stimulustestimony.htm) (visited March 20, 2003).

11. The fiscal drag is real. A federal grant to the states, however, is the wrong response: a federal tax hike (or forgone tax cut) to finance the grant also drags on the economy and, moreover, will fail to reduce state tax increases dollar for dollar. (Federal money sticks where it hits-with state and local governments.) The correct federalism response is the course that the administration-and my colleague John H. Makin-are already advocating on independent, macroeconomic grounds: slash federal taxes to revive the economy, boost local spending, and allow state budgets to recover. See John H. Makin, "More Tax Cuts, Please," AEI Economic Outlook February 2003.

12. U.S. Budget FY 2004, Aid to State and Local Governments, p. 249 (http://www.whitehouse.gov/omb/budget/fy2004/pdf/ spec.pdf) (visited March 13, 2003).

13. Randall R. Bovbjerg et al., "Medicaid Coverage for the Working Uninsured: The Role of State Policy," Health Affairs 21, no. 6 (2002): 231.

14. Ann Davis, "Insurers Find Way to Help the Elderly Get Medicaid to Pay for Nursing Homes," Wall Street Journal, June 6, 2003, p. A1.

15. These statutory requirements, like many others, have been waived for many states. For amply mentioned reasons, however, states have less-than-perfect incentives to distinguish cases in which in-home care functions as an alternative to institutionalized care from a new entitlement for individuals ineligible for long-term care. Here again, state "flexibility" has meant new entitlements and costs.

16. I had thought of this heading as a mildly clever pun until the AEI Federalism Project's incomparable Kate Crawford located an education association's earnest website service by that title (http://educationworld.com/a_issues/archives/ NELB.shtml) (visited March 20, 2003).

17. The report is available at http://www.nhsaa.org/pdffiles/ESEAcostimpactan.pdf (visited March 11, 2003).

18. Beware of business support for education: where it materializes, schools go to pot. For systematic empirical evidence see Sam Peltzman, "The Political Economy of the Decline of American Public Education," Journal of Law and Economics 36: 333. ("When a state's most politically potent industries demand more highly educated workers, school performance deteriorates.")

In the case at hand, the Business Roundtable financed a study that, for all its good intentions and fastidious number-crunching, translates into awful policy advice. The study (available at http://www.jbartlett.org) purports to identify low-cost, more-than-fully financed compliance options that are available "as long as New Hampshire responds in an efficient manner." Technically necessary compliance costs, however, are only a fraction of the costs that should be reasonably expected. For example, the study criticizes the NHSAA report (cited in note 17) for projecting that NCLB "quality teacher" requirements will require a two-percent pay raise, for a total of $11.7 million. A more efficient, $7.3 million option, says the study, is a 14.4 percent pay hike for science and math teachers, where New Hampshire faces a shortage. That might well work--on Pluto. On this planet, history and physical education teachers would howl and the unions would declare World War III. In estimating the costs of federal grant conditions, politically informed ballpark estimates beat technocratic fantasies any day.

19. Department of Education emissaries, sent to preach the NCLB accountability gospel even to the ends of the earth, have encountered distinctly hostile audiences. See Michael Winerip, "A Pervasive Dismay on a Bush School Law," New York Times, March 19, 2003, p. B7. Rebelling alongside conservative New Hampshire is socialist Vermont. North Dakota and Hawaii are also considering withdrawal from NCLB.

20. For simplicity's sake, the discussion here ignores the massive state-to-state redistribution under federal grant programs. For a documentation of those effects see Herman B. Leonard and Jay H. Walder, The Federal Budget and the States, twenty-fourth edition (Taubman Center for State and Local Government at the John F. Kennedy School of Government, 2000). State-to-state redistribution only strengthens the case for the tax refund option proposed in the text. Federal grant programs are popular even in net loser states (such as New York) because citizens have no way of comparing grant proceeds with their own tax contributions. A tax refund option would ameliorate that information asymmetry.

21. While an individual taxpayer refund may appear unduly cumbersome, the seemingly simpler approach of remitting the tax amount to nonparticipating state governments would exacerbate accountability problems. Because of the aforementioned flypaper effect (the money sticks where it hits), state governments would fail to pass the rebate on to their citizens. Thus, an opt-out would simply convert a grant-in-aid into general revenue sharing, which further reduces transparency and accountability. The only way to restore those values is to return the money directly to the citizens.

Michael S. Greve is the John G. Searle Scholar at AEI.