Report | Health

An idea whose time has returned: Anti-recession fiscal assistance for state and local governments

Briefing Paper #116

 Download PDF

Share this page:

In times of economic weakness, behavior that is prudent for individuals can be damaging for the economy. Rising unemployment and falling consumer confidence encourage people to pull in their belts – to spend less and save more.

The same is true for state and local governments. A slowing economy will reduce tax revenues and increase expenses. State and local budgets are often subject to fiscal constraints aimed at balancing budgets, and during economic slowdowns there is increased pressure to raise taxes and cut spending, sometimes under the mandate of state law (Regan 1995).

These kinds of balanced budget requirements and other tax and expenditure limitations reduce economic growth and employment at a time when the nation can ill-afford such losses. A timely, substantial dose of federal aid can offset these effects. Such aid should be commensurate with the size of the state/local sector, with $20-30 billion in the next 12 months being a reasonable amount to consider under current circumstances.

Such aid would also help buttress the fiscal health of state and local governments against the longer-term trend of devolution in which the administrative responsibility and financial burden for public programs and services has been transferred from federal to state/local governments (Sawicky 1999). Since 1980, aid from the national government to the states has stagnated, while responsibilities assigned to them have expanded (Kenyon 1995). This trend could be alleviated by creating an anti-recessionary fiscal aid program for the states that incorporated some of these basic principles:

  • Limited restrictions. The fewer the restrictions on how aid is used, the more quickly a proposal can become law and the money distributed and spent;
  • Targeted aid. At a minimum, the formula should take into account the effects of recession, population, and fiscal capacity;
  • Fairness in allocation. Formulas should be neutral with respect to the actual fiscal preferences and the budget decisions of state and local governments.

An alternative to the initiation of a new program would be to simply increase funding for a selected set of existing grants-in-aid with broad purposes and well-known, established distribution procedures. A second option would be to make the existing Medicaid formula more generous.

The likely effects of such federal aid, aside from helping the nation out of recession, would be to preclude the need for untimely service cuts or tax increases at the state level. It is also possible that states would use such aid for otherwise-deferred capital expenditures.

Reports of fiscal pressure in the state and local sectors have been surfacing for a year or more (National Governors’ Association and National Association of State Budget Officers 2001). Spending for these sectors was a combined $1,151 billion, or almost 12% of gross domestic product in 2000 (Bureau of Economic Analysis 2001). In September 2001, the National Association of State Budget Officers (2001) reported budget cuts in seven states, ranging from 1.0% to 3.8% of their total budgets (2001). Three states have cut their fiscal year 2002 revenue estimates by 1.6% to 8.2%. Spending in Medicaid, the largest state/local government program aside from education, is now projected to rise above previously forecast levels by an average of 10.8%, possibly forcing unwarranted cuts in other programs.

The simplest way to leverage spending by state and local governments is through grants of cash assistance with relatively few conditions attached. During economic slowdowns, such aid has the potential to reduce the duration and intensity of recessions. With sufficient aid the state/local sector could have a positive effect on the national economy, rather than contribute to slower growth. Since the state/local sector spends about $1,150 billion a year, aid that can replace several percentage points of lost spending – roughly $20-30 billion – could be enough to shield state and local governments from the recession. For the sake of timeliness, the allocations should be determined under a simple, impartial formula, rather than be hampered by annual legislative wrangling or a cumbersome application and review process.

This paper discusses some of the trade-offs associated with unrestricted federal aid and illustrates how such a program could work. It is not intended as a definitive proposal, since technical issues remain in the design of such a program that are beyond this report’s scope. Historical experience with the General Revenue Sharing program provides some insight into these technical and political issues.

A brief history of general fiscal assistance
General Revenue Sharing (GRS) was implemented in 1972 as part of President Nixon’s “new federalism” agenda. This program first began to take shape during President Johnson’s Administration in the work of White House economists Walter Heller and Joseph Pechman (Wallin 1998). Heller was a macroeconomist who wanted to use the program to reduce the federal budget surplus in the mid-1960s.

The program originally provided aid to both state and local governments. At its peak in 1979, the program’s expenditures reached $6.8 billion, which equaled 1.4% of total federal spending at the time. A program of equivalent size today would require $25 billion. In 1980, however, the state component of GRS was terminated. The program died altogether in 1986, after Treasury Secretary James Baker, among others, claimed that the federal government had no revenue to share.

Over the course of its life, the revenue sharing program gained friends and enemies at every point in the political spectrum, with support and opposition never following party lines. GRS drew support from legislators representing minority groups in the inner city, such as Rep. John Conyers; conservatives from low-income rural areas; and “good government” centrists such as Republican Senator David Durenberger from Minnesota. It was opposed by conservatives who sought to reduce the growth of the federal government and by liberals who preferred such funds to be devoted to programs aimed at specific types of spending.

The political context for any program of anti-recession assistance today is different from that characterizing the old battles over revenue sharing. In the 1970s, fears abounded that President Nixon would use GRS to replace programs that targeted funds to relatively specific purposes. Presently, it is more likely that anti-recession aid would supplement such programs.

In the 1970s, there was some centralization of responsibilities for anti-poverty programs in the federal government. But the recent inauguration of the Temporary Assistance for Needy Families program reflects some devolution of anti-poverty policy, so anti-recession aid could augment state government efforts to secure adequate financing for TANF programs (Sawicky 1999).

Some feared that federal aid with few strings was subject to misuse by state and local governments. But now, most recognize that the capabilities of state and local governments have progressed significantly (Conlan 1999). In any case, contr
ol over grants-in-aid from the federal level is exceedingly difficult to establish. Stipulations or mandates aimed at controlling such aid are more like costly window-dressing, and their effectiveness in real practice is open to considerable doubt.

The arguments for and against general assistance
Grants would be biased in favor of “big government” at the state level.
Any “big government” bias depends on the formula used to distribute funds. It is possible to provide aid based on a state government’s ability to finance services, irrespective of its size or the structure of its tax system (Sawicky 1986; U.S. ACIR 1986). In this sense, there need be no “matching” feature aimed at increasing state spending “on the margin,” as is thought to be the case for the Medicaid grants to states.

Unrestricted aid does not force governments to spend more or prevent them from spending less.
This is true, unless the grant formula is configured to encourage spending. The primary purpose of counter-cyclical aid is not to move state spending in any particular direction, but to alleviate pressures to reduce spending or raise taxes due to episodic economic adversity. A grant that appears to be used to cut state or local taxes could in fact preclude budget disputes that might result in service cuts.

Unrestricted aid permits state or local governments to substitute federal revenue for their own taxes.
This, too, is true, though it is also true that the alternative to substituting federal revenue for own-source revenue could be cuts in state and local services. Another justification for this feature is that the substitution of federal for state or local revenue is generally a good practice. State and local jurisdictions are more vulnerable to tax competition than the national government, and state and local taxes are more regressive than federal taxes. Federal aid in this light provides a modest centralization of public-sector finance for the U.S. as a whole. To some extent, federal taxes on income and payroll reduce flexibility for state and local governments, which are under some pressure to resort to less-equitable taxes on property, retail sales, and excises, or to rely more on user fees and other charges. Federal aid compensates for this.

Federal aid redistributes resources among geographic regions in the U.S., leaving some losers and others winners.
Every federal program has the potential to move resources across states and regions. Economic adversity is often unevenly distributed, whether it affects chronically poor regions or those hit hard by cyclical downturns. For the sake of the national economy, redistributing these burdens with federal funds is precisely the outcome desired in order to reduce fiscal drag from state and local government fiscal behavior.

Anti-recession aid is not well timed to respond to the business cycle.
It is possible to design grant allocations that are sensitive to economic conditions in a state, though it is more difficult to do this for local jurisdictions. A pre-determined formula can respond more rapidly to new economic information than a legislative process that might be charged with the responsibility of determining new allocations each year. Built-in, automatic stabilizers are preferable to the vagaries of the appropriations process in Congress.

A unique problem faced today is that a grant geared to the economic conditions in the tri-state New York/New Jersey/Connecticut area would fail to reflect the full impact of the terrorist attacks of September 11. It would only recognize the loss of income and jobs, not the literal destruction of property. Compensation for events of this sort would properly supplement any general assistance program. No broad, national program of aid is a substitute for disaster relief or reconstruction aid.

State governments ought to provide rainy-day funds to tide them over in bad times, rather than rely on the federal government to bail them out.
Indeed, they should, but people, business firms, organizations, and governments often fail to prepare in this way. Past state and local fiscal decisions are secondary to the national priority to address recessions in a timely manner. An aid program need not reward imprudent fiscal behavior or penalize sound fiscal management. Aid can and should be neutral with respect to whatever fiscal decisions – good or bad – a state or local government is responsible for. The aid formula should take into account the economic conditions in each state, not the fiscal condition of its governments.

An anti-recession aid program is just another brick in the pyramid of “big government.”
Not necessarily. To some extent, aid precludes the need for state and local taxes. Conversely, the lack of aid puts added upward pressure on state and local tax levels (Kenyon 1995). Aid can be provided in a form that is not biased in favor of big or small government, or in support of or against the use of any particular revenue-raising device or expenditure program. It can, however, provide complete flexibility to the recipient government. Administrative costs for an aid program can be extremely low, if few restrictions are put on the program. The old revenue sharing program required a staff of about 50 people in the Treasury Department. A state program would require fewer still. President Richard Nixon and others justified revenue sharing as a way of “bringing government closer to the people.”

Federal aid is best if targeted to purposes determined by the federal government.
When specific purposes are contemplated, targeting in the form of a categorical grant or a direct federal program is appropriate. Anti-recession aid is not a good tool for such purposes. In the present context, the primary purpose is to reduce fiscal drag in the state/local sector. For this purpose, general assistance is more appropriate. It should be noted that earmarking or targeting in any sense requires more complex legislation and associated regulations, reporting requirements, and administrative personnel. Targeting also reduces the timeliness of aid, which is important if anti-recession effectiveness is the chief objective. Targeting is also difficult to actually accomplish in practice, despite legislative intent and associated regulations. In practice, targeting of funds for some kind of broad purpose runs a high risk of being illusory, resulting in a political appeasement that, in reality, has little impact on the fiscal behavior of the recipient. Unrestricted aid minimizes unnecessary bureaucracy and maximizes opportunities for innovation at the state and local level.

Aid to state and local governments would cause an undesirable expansion of public employment.
Not necessarily. A devotion of federal resources for such purposes means some other use is precluded. The other use might be a categorical grant program or a direct, federally administered program, either of which would have its own employment requirements. Furthermore, any such effect is reduced to the extent that aid is used to substitute for state or local own-source revenues. And if aid is provided purely on a temporary, counter-cyclical basis, then the likelihood is that it would not be used to expand employment. The reason is that state politicians would see such aid as transitory, and they would not want to hire people they would be obliged to dismiss later (Wallin 1998, 1999). The expectation is that the grants would be regarded as “manna from heaven” and used either as a temporary replacement for revenue losses associated with the downturn (which would preclude layoffs of
public employees or postponement of contracts), or for public capital expenditures (Wallin 1998, 1999).

The biggest fiscal inequalities pertain to local governments, not to the states, so federal aid should bypass state governments.
It is true that disparities are greater below the state level. The trade-off here is that a program of grants to states is simpler to construct, cheaper to administer, quicker to implement, and somewhat less effective in terms of targeting. A local program is more complicated and could require the collection of some new data on local governments in a timely manner, since relatively little data on the characteristics of local jurisdictions is available, beyond population and household income. Local government organization in the U.S. is extraordinarily complex. A compromise would be to have state grants with certain general guidelines as to the method of distribution by state governments to their own local governments. State governments are better situated than the federal government to distribute aid to their own local governments.

How to provide anti-recessionary aid for the states1
Grants to states should take into consideration both the need for aid and the ability for a state or local government to raise funds from its own sources.

A government’s need for aid could be determined by population level. Absent any other consideration, a state’s share of a national program’s spending would be its share of the nation’s population. A state’s grant would be some national standard amount multiplied by the state’s resident population. (Measurement of need can be made more sophisticated (see Rafuse 1986; U.S. ACIR 1990), but such refinements are beyond the scope of this paper.)

A determination of need should take into account the uneven state-by-state effects of recession. Each state’s per capita allocation should be adjusted to reflect its economic condition. The state’s unemployment rate is an obvious choice for gauging a state’s economic condition. A formula could stipulate that, if a state’s unemployment rate is, for example, one percentage point above the national average, its per capita grant should be 5% above the national average.

A state government’s ability to raise funds, also known as its fiscal capacity (Barro 1988), is the other key consideration in determining aid allocations. Fiscal capacity should reflect the extent of taxable resources in a state or local jurisdiction. It is analogous to a tax base for an area, rather than an individual (Sawicky 1986). The aggregate tax base for an area is the value of taxable activities within it.

The best way to measure state fiscal capacity is through the use of the Total Taxable Resources (TTR) index (Sawicky 1986), which is presently used in two federal aid programs. An alternative would be to use gross state product (GSP) (Aten 1986), as currently estimated by the Bureau of Economic Analysis. Otherwise the variable customarily employed for fiscal capacity in federal grant formulas is state per capita personal income (PCI). PCI is inferior to both GSP and TTR as a measure of fiscal capacity (Sawicky 1986). Still another alternative is the Representative Tax System, an index developed at the U.S. Advisory Commission on Intergovernmental Relations and employed in Canada’s substantial system of general assistance to provinces (ACIR 1987).

Local fiscal capacity measurement depends on data available to state governments, which is mostly limited to household income and assessed valuation of property. Experience in school funding would inform each state’s decision in this realm.

Population, unemployment, and all of the fiscal capacity indicators mentioned above share the virtue of being neutral with respect to a jurisdiction’s size of government, expenditure decisions, and tax structure. They reflect no incentive for the recipient governments to either spend more or less than they would otherwise, nor to use or avoid any particular type of tax.

The basic procedure for a formula, then, is to simply calculate a state’s per capita fiscal capacity and index it relative to the national average. Each state would then receive aid inversely to its index, and proportional to its need. For instance, suppose a program provides one dollar per capita on a national basis. If a state’s level of TTR is 10% higher than the national average, its index is 110. Its per capita allocation of aid could thus be 91 cents (1.00/1.10). While simple, this type of formula can produce a variation in grant amounts that might not pass political muster. It might be more politically feasible to have a two-part calculation: one component that provides the same per capita amount to all states and a variable component that depends on fiscal capacity. Of course, even more complex formulas can be designed (see Fastrup 1986).

In the U.S. federal system, responsibility for different functions of government varies significantly across the states. Certain patterns do prevail, however. Health care and social services are provided chiefly by state governments, while education is, in large part, a local responsibility. Federal aid that is confined to only one level of government is likely to neglect some of these responsibilities.

Political dynamics often reduce the effectiveness of targeting in grant programs, including anti-recession assistance. A formula that sharply targets aid by jurisdiction is likely to produce a wide variation in allocations per capita across different states or local jurisdictions. The stronger the targeting, the more varied the results. This typically raises objections from representatives of jurisdictions who receive relatively less aid. The usual remedy is to reduce the variation in grants, and hence the target efficiency of the program. A truly targeted program would limit aid to jurisdictions that are below average in terms of program criteria, but politics makes this level of efficiency difficult to achieve in practice.

Two alternatives to unrestricted aid
One alternative to a new aid program would be for Congress to simply increase the allocations of one or more existing grants-in-aid, especially those with a broad scope. There are more than 1,400 federal grant programs, but most of them are small and narrowly targeted. Examples of some larger ones are the Social Services Block Grant and the Community Development Block Grant. An advantage of this approach is that existing programs have allocation procedures that are established and known to all. It might be more feasible to obtain political agreement in this manner, and funds could be distributed more quickly. A disadvantage is that the distribution of funds under any such set of programs would have no particular underlying rationale. It would simply be a way of transferring resources to the states. Greater effect and fairness could be achieved through a formula designed with specific criteria in mind, first among them being to counteract economic recession in the states.

A second alternative is to increase the federal share of funding for the Medicaid program. This is the largest program in the state/local sector aside from education, and it uses a formula designed to counteract lagging fiscal capacity in the states. Medicaid has increased significantly over the past two decades, and, as the population ages, it will increase further. By various means, the Medicaid formula could be adjusted to increase the federal share. A potential shortcoming of this approach is that states have less control over their Medicaid spending, so the effects of recession could be more severe in the non-Med
icaid portions of their budgets. On this account, it would be better to focus aid on the non-Medicaid portion of state and local budgets.

Unrestricted fiscal assistance is a straightforward way to stimulate spending in a time of economic slowdown. It presents minimal federal intrusion into state/local fiscal decisions, and it benefits from low administrative costs. In the end, any anti-recession fiscal assistance should adhere to the following basic principles:

  • An effective response to recession should include aid to both state and local governments; such aid should be of sufficient magnitude to offset several percentage points of cuts in state spending or increases in taxes, given the size of the state/local sector. Twenty to thirty billion dollars in aid over the next 12 months is a reasonable range to consider.
  • The fewer the restrictions on aid, the more quickly a proposal can become law and the money distributed and spent. Restrictions on the uses of money are difficult to enforce in actual practice. Their political effect is ambiguous, since some constituencies might favor targeting, while state and local public officials would generally prefer complete flexibility.
  • Allocations to states or localities should be determined by a simple, transparent formula.
  • At a minimum, the formula should take into account the effects of recession, population, and fiscal capacity.
  • Formulas should be neutral with respect to the actual fiscal preferences and budget decisions of state and local governments.

October 2001

1. There is a substantial body of literature on the economics of formula grants that is beyond the scope of this paper (see Office of State and Local Finance 1985, 1986). Grant formulas can be crafted to respond to a variety of objectives (Fastrup 1986). Often some objectives conflict with others, so some research must go into determining what the program ought to accomplish.

Aten, Robert H. 1986. “Gross State Product: Concept and Calculation.” In Office of State and Local Finance. Washington, D.C.: U.S. Department of the Treasury.

Barro, Stephen M. 1986. “Improved Measures of State Fiscal Capacity: Short-Term Changes in the PCI and RTS Indices.” In Office of State and Local Finance. Washington, D.C.: U.S. Department of the Treasury.

Bureau of Economic Analysis. Various years. “National Income and Product Accounts Tables.”>

Conlan, Timothy J., and James D. Riggle. 1999. “New Opportunities/Nagging Questions: The Politics of State Policy-Making in the 1990s.” In Sawicky, ed., The End of Welfare. Washington, D.C.: Economic Policy Institute.

Fastrup, Jerry. 1986. “Fiscal Capacity, Fiscal Equalization, and Federal Grant Formulas.” In Office of State and Local Finance. Washington, D.C.: U.S. Department of the Treasury.

Fastrup, Jerry. 1986. “Fiscal Equalization and the Design of Federal Grant Allocation Formulas: An Application to General Revenue Sharing.” In Office of State and Local Finance. Washington, D.C.: U.S. Department of the Treasury.

Kenyon, Daphne A. 1995. The Decade of Declining Federal Aid. Washington, D.C.: Economic Policy Institute.

National Association of State Budget Officers (NASBO). 2001. Fiscal Update: State Budgets Post-Attack. September 25.

National Governors’ Association and National Association of State Budget Officers. 2001. Fiscal Survey of the States.

Office of State and Local Finance. 1985. Federal-State-Local Fiscal Relations: Report to the President and the Congress. Washington, D.C.: Department of the Treasury.

Office of State and Local Finance. 1986. Federal-State-Local Fiscal Relations: Technical Papers. Washington, D.C.: U.S. Department of the Treasury.

Rafuse, Robert W., Jr. 1986. “A Representative-Expenditure Approach to the Cost of the Service Responsibilities of States.” In Office of State and Local Finance. Washington, D.C.: U.S. Department of the Treasury.

Regan, Edward V. 1995. “States’ Experience With Balanced Budget Provisions, Prepared Statement for Hearings of the Senate Judiciary Committee, January 5, 1995.” The Levy Economics Institute of Bard College.

Sawicky, Max B. 1986. “The Total Taxable Resources Definition of State Revenue-Raising Ability.” In Office of State and Local Finance.Washington, D.C.: U.S. Department of the Treasury.

Sawicky, Max B. 1999. The End of Welfare? Consequences of Federal Devolution for the Nation. Armonk, N.Y.: M.E. Sharpe.

U.S. Advisory Commission on Intergovernmental Relations (ACIR). 1986. Measuring State Fiscal Capacity: Alternative Methods and Their Uses (M-150). September.

U.S. Advisory Commission on Intergovernmental Relations. 1990. Representative Expenditures: Addressing the Neglected Dimension of Fiscal Capacity (M-174). December.

Wallin, Bruce A. From Revenue Sharing to Deficit Sharing: General Revenue Sharing and Cities. Washington, D.C.: Georgetown University Press, 1998.

Wallin, Bruce A. 1999. “Revenue Sharing as a Public Investment: Why Unrestricted Aid to Local Governments Will Stimulate Municipal Capital Expenditures.” In Sawicky, ed., The End of Welfare. Washington, D.C.: Economic Policy Institute.

See related work on Health

See more work by Max B. Sawicky