National Tax Journal Vol. 48, no. 3, (September, 1995), pp. 329-36

BALANCED BUDGET RULES AND FISCAL POLICY: EVIDENCE FROM THE STATES JAMES M. POTERBA* Against this backdrop, the variation in budget practices across states within the United States provides a valuable source of information on the potential effects of fiscal institutions. One must recognize at the outset that it is difficult to interpret correlations between fiscal institutions and budget outcomes; budget institutions may be endogenous. Changes in institutions that affect the prospect of deficit finance may reflect changing political support for budget deficits, and as such cannot be viewed as “natural experiments” in fiscal institutions. Riker (1980) argues that essentially all political institutions reflect the “congealed preferences” of the electorate. Institutions that no longer suit a majority of the electorate will be overturned. This harsh view is probably too extreme, because at least some of the variation in state fiscal institutions is due to historical accidents rather than current fiscal policy tastes.

The persistence of fiscal deficits in many industrial democracies has spawned a vast theoretical literature in political economy on why nations run budget deficits, along with a sharp policy debate concerning fiscal institutions that might reduce these deficits. The debate on fiscal institutions, such as proposed balanced budget amendments for the federal government, involves relatively little empirical evidence on the potential effects of such institutions. The lack of such evidence can be traced to several factors. First, some of the proposed institutions under discussion are budgetary innovations that have not been tried on a national or subnational scale before. Second, there is relatively little intranational variation over time in the nature of budget processes. Therefore, it is difficult to compare fiscal policy before and after significant institutional reforms. Third, while there are differences across countries in fiscal institutions, many analysts are hesitant to draw strong conclusions from cross-national comparisons because it is difficult to hold constant other factors that may affect fiscal policy. *Massachusetts

Institute of Technology

Economic Research, Cambndge,

This paper consrders the nature of balanced budget requirements in the U.S. states and explores what lessons, if any, the state-level experience holds for discussions of a federal balanced budget amendment. It emphasizes that most such requirements are substantially differ-

and Natlonal Bureau of

MA 02139-4307

329

National Tax Journal Vol. 48, no. 3, (September, 1995), pp. 329-36

state borrowing, and thelperception of capital market participants with respect to state fiscal health. This dibcussion also considers the interaction between balanced budget rules and tax and expenditure limits (TELs), because some proposals for balancing the federal budget combine anti-deficit rules with prolisions that resemble state TELs. The thlird section is a bnef conclusion.

ent from those currently being discussed at the federal level. In particular, virtually all states allow some types of borrowing to be used in budget balancing, most states apply the balanced budget rule to only part of their budget, and there are virtually no formal provisions for enforcing state balanced budget rules. These limitations notwithstanding, evidence on the effects of state balanced budget institutions is relevant for the broader question of whether fiscal institutions can affect fiscal policy outcomes. The view of fiscal institutions that is implicit in most empirical public finance research on the demand for state and local government services holds that these institutions are simply “veils” that do not affect spending outc:omes. Studies in this tradition model the demand for spending as a function of median income, the after-tax price of raisinig revenue, dernographic variables, and similar factors, but exclude variables describing the budgeting or legislative environment. Yet the preponderance of evidence from empirical studies of fiscal institutions and budget outcomes suggests that tightly drawn anti-deficit rules, especially when cloupled with limits on government borrowing, induce smaller deficits and more rapid adjustment of taxes and spending to unexpected fiscal shortfalls. This evidence suggests that modifying the federal budget process has the potential to affect federal fiscal policy.

STATE BALANCED BUDFET RULES Most state constitutions prevent state governments from running deficits in their general operating budgets, but the nature and scope of these limits varies widely across states. Only one state, Vermont, has no balanced budget requirement. This section drawsupon recent summaries of state budgeting rules by the National Association of State Budget Officers (NASBO) (1992) and the U.S. General Accounting Office (GAO) (1993) to highlight several key features of balanced budget requirements in the U.S. states. What Does Budget Ba/@nce Mean? The balanced budget requirements in the 49 states with such requ’rements can be I broadly categorized into three groups, depending on the stage in the budget process at which balance is required. This categorization draws on the survey of budget practices by the NASBO (1992). First, in 44 states, the governor must submit a balanced budget. This is the weakest of the various balanced budget requirernents. Second, 37 states impose a stricter standard, requiring that the legislature enact a balanced budget. Balanced budget rules of this type nevertheless allow for actual revenues and expenditures to diverge from balance if realizations differ from expectations. In many states that require passage of a balanced budget, the actual budget for the year may be in

This paper is divided into three parts. The first describes the nature of current state balanced budget rules. It devotes particular attention to mechanisms that states use to satisfy their balanced budget requirements, such as borrowing, transfer elf funds across accounts, and retiming various income and expenditure flows. The second section summarizes previous empirical evidence oIn how balanced bludget institutions affect state deficits, 330

I

BALANCED BUDGET RULES AND FISCAL POLICY

National Tax Journal Vol. 48, no. 3, (September, 1995), pp. 329-36

deficit and the state can borrow to carry this deficit forward to future years. The third and strictest type of balanced budget rule combines a requirement that the legislature enact a balanced budget with a prohibition on deficit carry-forward. This is the situation in 24 of the 37 states that require the legislature to enact a balanced budget. Such stringent antideficit rules are more common in small than in large states; seven of the ten largest states allow deficits to be carried forward to subsequent years.

of their spending was affected by these rules, nine states reported that 50 to 75 percent of spending was affected, and the remaining states with balanced budget rules indicated that these rules applied to at least 75 percent of their state spending. New Taxes and Spending Budget Gimmicks?

Cuts, or

States that face budget deficits at some stage of their budget process have three options for closing such deficits. They can raise taxes, reduce spending, or change “budget execution” to close the apparent deficit. Proponents of balanced budget rules typically assume that these rules will lead to tax increases or spending cuts, while skeptics argue that balanced budget targets will be met through accounting changes or various budgeting gimmicks. The state experience suggests that while some cosmetic changes are used to meet balanced budget requirements, these changes are quantitatively less important than tax increases and spending cuts.

How Much of the Budget Must Be Balanced? An important difference between existing state balanced budget rules and recent proposals at the federal level is that state rules frequently apply to only part of the budget. The general fund, or state operating budget, is almost always subject to a balanced budget rule. In 48 of the 49 states classified by NASBO (1992) as having balanced budget rules, these rules apply to the general fund. Such rules are less likely to apply to special funds (34 states), such as those with earmarked tax receipts or those used to fund particular programs such as intergovernmental aid, capital spending funds (33 states), and trust funds (30 states) such as those for highways or some social insurance programs. In some cases, particularly with respect to the capital account, funds raised by issuing long-term debt can be included in the revenue flow that balances the budget.

States can use a wide range of accounting changes and related techniques to satisfy balanced budget rules. The GAO’s (1985) study provides several examples of the changes that were used to satisfy budgetary targets in the early 1980s. California transferred revenues from an oil extraction royalty tax from a trust fund to the general fund; New York State adopted a new payroll system to shift its last payroll payment from fiscal year 1983 into the next fiscal year; Minnesota accelerated tax collections to move receipts across fiscal years.

These statistics on the applicability of state balanced budget rules suggest two conclusions. First, balanced budget rules typically apply to more than just the state operating budget. Second, there is substantial variation across states in the fraction of state spending that is likely to be affected by balanced budget rules. In the NASBO (1992) survey, three states reported that between 25 and 50 percent

The GAO’s (1993) survey of state budgeting collected information on the dollar value of various accounting changes that states used to meet balanced budget targets. Twenty-five states reported that they had faced prospective deficits during 331

National Tax Journal Vol. 48, no. 3, (September, 1995), pp. 329-36

BALANCED BUDGET R~ULESAND FISCAL POLICY OUTCOMES

a recent budget enactment period and had taken actions to close these deficits. Nearly half (49 percent) of the deficit reduction was achieved through spending cuts, another 32 percent through revenue Increases, and the remaining 19 percent through “other actlons” such as accounting changes. In addition, 32 states re[ported that they had faced prospective deficits after budget enactment and had taken actions to close these deficits. Spending cuts accounted for 60 percent of the within-fiscal-year deficit reduction, revenue increases accounted for 4 percent, and other actions accounted for 36 percent. These other actions included drawing down rainy day funds (32 percent of the total deficit reduction), interfund transfers (22 percent), short-term borrowing (17 percent), deferred payments (13 percent), and several other changes in budget execution. Accounting changes and related actions thus appear to account for a substantial part of fiscal adjustment within the budget cycle, but they are not the primary source of longerterm state deficit reduction.

Existing research has fo used on the effects of balanced budg ,’ t rules on the size and persistence of state’ budget deficits. This section summarizes this work, focusing first on fiscal actionsi, then on state borrowing behavior, an finally on the ef6 feet of fiscal institutions’on the interest rates that states face in t he capital market. Balanced Budget Rule! and State Fiscal Adjustment Two recent studies havq examined the link between state balarhced budget requirements and state fi$al policy. Alt and Lowry (1994) study ho\hi anti-deficit carryover provisions affect th level of state taxes and spending, an my own work, Poterba (1994), tests w i ether anti-deficit rules affect the short-ru tax and spending adjustments that fol ow from unexpected deficits. Each of I hese studies suggest that anti-deficit rulgs have real effects on state fiscal pc/licy. Alt and Lowry (1994) a ‘alyze data from the Census of Governm a nts for the period 1968-87. They mo#el state revenue and expenditures as a fgnction of current state income, current feberal grants, lagged values of state r venue, state expenditures, the lagged ifference between revenues and ex i enditures,, and a set of indicator variablesi for state political circumstances. They conlpare fiscal policy reactions to disparities ‘etween revenues and expenditures, “defi its,” in states with different political a d institutional configurations. It is imp rtant to recall, as the foregoing discussio suggests, that states can exhibit differ 1 rices between revenues and expenditu es while still satisfying their balanced b dget requirements. Alt and Lowry (1:, 94) find that a $1 state deficit in the current year triggers a 77q response, thiough tax increase

Ensuring Compliance Some of the most difficult questions that arise in designing a federal balanced budget rule concern enforcement. On this point, the state experience is not terribly helpful. Gold (1992) notes that most states have no formal enforcement mechanisms for their balanced budget requirements, and the GAO (1993) reports that there have never been lawsuits to challenge state budgeting outcomes, even though there have been instances when budgets failed to balance. The GAO’s (1985) survey suggested that state policy makers view tradition, or a history of balanced budgets, as the pnmary factor encouraging them to maintain budget balance. It is not clear how such experience can be translated to analyze the potential effects of federal anti-deficit rules. 332

I

BALANCED BUDGET RULES AND FISCAL POLICY

National Tax Journal Vol. 48, no. 3, (September, 1995), pp. 329-36

or spending reduction, for states that are Republican-controlled and prohibit deficit carryovers, compared with a 34~ reaction in states that are Democrat-controlled and have such limits. In states that do not restrict deficit carryovers, the adjustments are 31 Q and 404 for Republican and Democratic states, respectively. This empirical evidence suggests that state politics is an important influence on deficit reduction, and that at least in some political configurations, variation in anti-deficit rules is associated with variation in fiscal actions.

and tax increases, surpluses are largely used to fund increases in state rainy day funds. The disparate responses to deficits and surpluses can be viewed as further evidence of the impact of state anti-deficit laws, and it also provides some support for previous empirical work suggesting a “flypaper effect” in state and local fiscal behavior. My investigation of state reactions to unexpected deficits also considers the effect of tax limitation laws. These laws vary widely in their provisions. Some limit annual tax increases to a fixed fraction of previous taxes or of contemporaneous income growth, while others require legislative supermajorities or popular referenda to enact tax increases. States with tax limitation rules enact smaller tax increases in response to unexpected deficits than do states without such limits. These results confirm Crain and Miller’s (1990) finding that taxes grew less between 1979 and 1986 in states with tax limits than in states without such limits. More generally, the finding that state tax limitations matter is consistent with several studies, including those of Dye and McGuire (1995), Elder (1992), lchniowski and Preston (I 991>, and Poterba and Rueben (1995), which conclude that local property tax limits have real effects in reducing the growth rate of local revenues. These studies buttress the central conclusion that fiscal institutions have real effects on fiscal policy outcomes.

In Poterba (I 994), I study how state balanced budget rules affect the way state fiscal policies responded to unexpected deficits or surpluses. My analysis considers both within fiscal year adjustment, through spending cuts or tax increases, as well as adjustment in the next fiscal year. Unexpected deficits could be the result of tax revenues below projections or of spending in excess of projections. I analyze how an indicator variable for states with “weak antideficit rules,” as classified in the Advisory Commission on Intergovernmental Relations (ACIR) (1987) study on fiscal stringency, is related to state reactions to fiscal shocks. My results, which are based on the 27 continental states with annual budget cycles, suggest that states with weak antideficit rules adjust spending less in response to unexpected deficits than do their counterparts with strict anti-deficit rules. A $100 deficit overrun leads to only a $17 expenditure cut in a state with a weak anti-deficit law, while it leads to a $44 cut in other states. I do not find any evidence that anti-deficit rules affect the magnitude of tax changes in the aftermath of an unexpected deficit.

Balanced Budget Rules and State Borrowing In addition to the evidence on how balanced budget rules affect tax and spending policies, there is also some evidence on how these rules and debt limitation laws affect the level of state indebtedness. Von Hagen (1991) compares the level of state general obligation debt, per capita and relative to state income, in

My analysis of state fiscal adjustment also suggests that states adjust asymmetrically to unexpected deficits and surpluses. While deficits are met with spending cuts 333

National Tax Journal Vol. 48, no. 3, (September, 1995), pp. 329-36

states with and without stringent balanced budget requirements. His index of budgetary stringenc:y is the ACIR (1987) index described above. His findings suggest that general obligation indebtedness is substantially lower in states with stringent balanced budget amendments than in other states. He also explores the effect of stringent balanced budget rules on the ratio of non-full faith and credit to full faith and credit debt across states. States with more stringent deficit limits, as well as states with lower general obligation debt limits, exhibit higher levels of revenue debt and other debt that is not backed by the full faith and credit of the state. These results are consistent with Ejunche’s (1991) demonstration that states with tighter debt limits or balanced budget rules are more likely to use public authorities and other alternatives to statebacked borrowing to finance various projects.

Balanced Budget Rulef and the Municipal Bond Market An ingenious line of res$arch developed by Goldstein and Woglo(m (1992), and subsequently extended t/~yBayoumi and Woglom [(I 995) and Loqry and Alt (1995) tests whether the interegt rates at which states can borrow funds~ are a function of state fiscal institutions. This literature is of Interest for two reasons., First, because capital markets are one of the institutions that may discipline state! when they pursue lax fiscal policies, studying how fiscal Institutions affect borrowing rates can provide evidence on whether this disciplinary Idevice is operative. Second, borrowing rates provide a unique market-based measure of prospective ftate fiscal performanc:e. There are a range of data problems that make it difficu~lt to evaluate the results of empirical analyses of how taxes and spending are relate4 to fiscal institutions, precisely because differences in accounting practices may ffect the reported statistics. lnteres s”rates are subject to no such measuremen/t problems and therefore provide a less Brbitrary measure of fiscal policy outcome$ The drawback of studying interest ratef is that they can be affected by many facjtors other than fiscal Institutions, includilng state fiscal history and economic prbspects.

A related analysis of the real effects of antideficit rules and related limits on state borrowing is Kiewiet and Szakaly’s (1992) study. It analyzes whether state constitutional debt limits have any effect on total state indebtedness or on the composition of this debt. The only institution that the authors find to be. highly negatively correlated with state indebtedness is a requirement that state debt be approved by popular referendum. This suggests that a combination of a stringent anti-deficit rule and a requirement that debt be approved by the voters is likely to bring pressure for tax increases or spending cuts, rather than debt finance, in response to state deficits. This evidence is also consistent with the findings reported above on the effect of tax limitation laws. Constitutional or legislative provisions t:hat make it more costly to balance the budget in a given fashion, by raising taxes or by issuing long-term debt, appear to have real effects in discouraging these fiscal actions.

Several recent studies hqve explored the link between fiscal institbtions and borrowing rates, with partidular attention to the role that balanced bkdget rules play in this process. Goldsteiq and Woglom (1992) relate the interesb rate on general obligation debt to the A~CIRindex of state limits on deficits described above. Their results, which also contrlol for the level of state indebtedness and the observed state deficit, suggest th;it a state with the most restrictive set of f&al limits faces an interest rate 0.05 perceitage points lower than a state with an avqrage set of limits. Lowry and Alt (1995) show that the bond 334

I

BALANCED BUDGET RULES AND FISCAL POLICY

National Tax Journal Vol. 48, no. 3, (September, 1995), pp. 329-36

ity view, or to reject it. One argument against this view is that several of the studies described above have controlled for some measure of state voter preferences, such as the political party of the governor or the legislature, so they include at least crude proxies for voter preferences. Other work in a similar vein, such as my study, Poterba (1995>, of how capital budgeting institutions affect state capital spending, adds controls for the fraction of the state voting for each party’s presidential candidate in recent elections, again without changing the measured effect of fiscal institutions.

market reaction to a state deficit projection depends on whether the state has a balanced budget requirement. States with balanced budget rules experience smaller increases in their borrowing costs for a given deficit. These results are important in the current context because they suggest that capital market participants, who have strong incentives to monitor and evaluate state fiscal performance, consider the presence of anti-deficit rules relevant in evaluating state fiscal conditions. They also represent a first step toward a normative analysis of whether states benefit from adopting anti-deficit rules, since such rules involve a trade-off between flexibility in responding to fiscal shocks and a commitment to long-term fiscal responsibility.

A second defense of the exogeneity of budget institutions draws on the history of these institutions, summarized, for example, in Kiewiet and Szakaly (1992). Many of the constitutional limits on state deficits were enacted in the nineteenth century; whether these rules reflect the preferences of current state residents is an open issue. The more difficult it is to make changes in state constitutions, the more valuable the cross-state variation is in identifying the effect of these institutions on fiscal policy.

CONCLUSIONS AND INTERPRETATION The range of budgetary institutions across the states in the United States provides a rich opportunity to study the effects of these institutions on fiscal policy. The studies surveyed in this paper suggest that there are correlations between state balanced budget rules and state fiscal policy. The critical question for policy evaluation is how to interpret these correlations. It is possible that they simply reflect a correlation between fiscal discipline, fiscal institutions, and an omitted third variable, voter tastes for fiscal restraint. Voters in some jurisdictions may be less inclined to borrow to support current state outlays and to use deficits to shift the burden of paying for current state programs to the future. If these voters are also more likely to support legislative or constitutional limits on deficit finance, then the observed link between fiscal rules and fiscal policy could be spurious.

While these difficulties of interpretation prohibit strong conclusions, the available evidence suggests that changes in budget processes and in the rules affecting the dynamics of taxes and expenditures can affect fiscal policy outcomes. The view that these fiscal institutions are simply a veil, pierced by voters and their elected representatives, appears to be dominated by the richer view suggesting that fiscal institutions mediate the link between voter tastes and policy outcomes. While this does not imply that all potential reforms in the federal budgetary process would have real effects, it suggests caution in dismissing them as irrelevant institutional reforms that are likely to change accounting practices more than taxes and expenditures.

It is difficult to provide definitive evidence in support of this institutional endogene335

National Tax Journal Vol. 48, no. 3, (September, 1995), pp. 329-36 Masson, 228-69. Cambridbe: Cambridge University Press, 1992. Ichniowski, Casey and Anne E. Preston. “A National Perspective on the Nature and Effects of the Local Property Tax Revolt.” National Tax Journal44 No. 2 (June, 1991): 123-46. Kiewiet, D. Roderick and Kristin Szakaly. “The Efficacy of Constitutional Restrictions on Borrowing, Taxing, and Spending: An Analysis of State Blonded Indebted ess, 1961-90.” California Institute of Technol 1 gy. Manuscript, 1992. Lowry, Robert C. and la es E. Alt. “A Visible Hand? Inter-temporal E?ficiency, Costly Information, and Market-Based Enforcement of Balanced Budget Laws.” Hbrvard University Department of Government. Mimeo, 1995. National Association of State Budget Officers. State Balanced Budget Requirements: Provisions and Practice. Washington, D.C.: NASBO, 1992. Poterba, James M. “State Responses to Fiscal Crises: The Effects of Bud etary Institutions and Politics.” lournal of Pgolitical Economy 102 (August, 1994): 799-82 1. Poterba, James M. “Capital Budgets, Borrowing Rules, and State Capital Spending.” Journal of Public fconomics 56 (January, 1995): 16587. Poterba, James M. and Kim 5. Rueben. “The Effect of Property Tax Limits on Wages and Employment in the Local Public Sector.” American Economic Review 85 (May, 1995): 384-9. Riker, William. “Implications for the Disequilibrium of Majority Rule for the Study of Institutions.” American Political Science Review 74 (1980): 432-46. U.S. General Accounting1 Office. Budget Issues: State Balanced Budget Practices. GAO/ AFMD-86-22BR. Washington, D.C., 1985. U.S. General Accounting1 Office. Balanced Budget Requirements: Stare Experiences and Implications for the Federal Government. GAO/ AFMD-93..58BR. Washington, D.C., 1993. von Hagen, Jurgen. “A Note on the Empirical Effectiveness of Formal Fiscal Restraints.” Journal of Public Economics 44 (March, 1991): 199-210.

ENDNOTES l I am grateful to the National Science Foundation for research support, and to Steve Gold, Sunny Ladd, Matthew McCubbrns, and Kim Rueben for helpful discussions. REFERENCES Advisory Commission on Intergovernmental Relations. Fiscal Disciplrne in the Federal System: National Reform and the Experience of the States. Washington, D C.: ACIR, 1987. Alt, James E. and Robert C. Lowry. “Divided Government and Budget Deficits: Evidence from the States.” American Political Science Review 88 (December, 1994): 81 l-28. Bayoumi, Tamin and Geoffrey Woglom. “Do Credit Markets Discipline Sovereign Borrowers: Evidence from U.S. States.” burnal of Money, Credit and Bankin!g, 1995. Bunche, Beverly 5. “The Effect of Constitutional Debt Limits on State Governments’ Use of Public Authorities.” Public Choice 68 (January, 1991): 57-69. Crain, W. Mark and James C. Miller III. “Budget Process and Spending Growth.” William and Mary Law Review .31 (Spring, 1990): 1021-46. Dye, Richard F. and Therese 1. McGuire. “Impact of Property Tax Ltmitations on Local Governments: Evidence from the Chicago Metropolitan Area.” University of Illinois Institute of Government and Public Affairs. Mimeo, 1995. Elder, Harold W. “Exploring the Tax Revolt: An Analysis of the Effects of Tax and Expenditure Limitations.” Public Finance Quarter/y 20 (June, 1992): 47-63. Gold, Steven D. “State Government Experience with Balanced Budget Requirements: Relevance to Federal Proposals.” Testimony before U.S. House of Representatrves, Budget Committee. Washington, ID.C., May 13, 1992. Goldstein, Morris and Geoffrey Woglom. “Market-Based Fiscal Discipline in Monetary Unions: Evidence fronn the U.S. Municipal Bond Market.” In Establishjng a Central Bank: issues in Europe and Lessons from the United States, edited by M. B. Canzonen, V. Grilli, and P. R.

336