«THE IMPACT OF FISCAL CONSTITUTIONS ON STATE AND LOCAL EXPENDITURES Dale Bails and Margie A. Tieslau The United States has witnessed changing trends, ...»
THE IMPACT OF FISCAL CONSTITUTIONS ON
STATE AND LOCAL EXPENDITURES
Dale Bails and Margie A. Tieslau
The United States has witnessed changing trends, over the past few
decades, in the fiscal behavior and responsibilities of federal, state,
and local governments. In particular, two of the more politically
popular trends include a general movement of fiscal responsibility away from the federal level toward the state and local level and an attempt to implement budgetary mechanisms that instill fiscal discipline at the state and local level. Proponents of these fiscal discipline mechanisms argue that these tools have been or will be effective at slowing the rate of spending at all levels of government.1 The general assumption is that these budget rules will provide the citizenry with some protection against the inherent bias toward excess spending in public-sector decisionmaking. This notion is based on the assumption that state and local governments are more responsive to the needs of the citizenry and, therefore, are more efficient at providing the requisite goods and services.
The most interesting aspect of the reasoning discussed above is that it appears to run counter to the increases in spending that have actually been observed at the state and local level. For example, real per capita state and local direct general expenditures have increased approximately 78 percent over the decade ending in 1994. This paper provides an empirical assessment of this phenomenon in order to determine whether or not the imposition of various budgetary instiCato Journal, Vol. 20, No. 2 (Fall 2000). Copyright © Cato Institute. All rights reserved.
Dale Bails is Professor of Economics and Finance at Christian Brothers University, and Margie A. Tieslau is Associate Professor of Economics at the University of North Texas. They thank J. Matsusaka, Karen Conway, Edward López, Paul Trogen, and Charles Rowley for valuable comments and suggestions.
The justification for combining the state and local sectors stems from the observation that the bundle of publicly provided goods and services varies from state to state (Dye and McGuire 1992). For example, one state may choose to delegate responsibility for park maintenance to local governments while another state may perceive this to be a state function.
CATO JOURNALtutions has been effective at restraining the growth of state and local government spending in recent decades. In particular, we consider the effect of tax and expenditure limits, the ability of the governor to use the line-item veto, balanced budget requirements, super-majority voting requirements, term limits, the length of the budget cycle, the initiative procedure, and the state referendum.
Numerous empirical studies have been conducted in past years regarding the factors that determine and influence state and local government spending. In particular, following the passage of Proposition 13 in 1978, many studies were conducted to investigate the impact of specific fiscal discipline mechanisms on the rate of growth of state government spending or revenue. These studies, however, have generally focused on one or a very small number of fiscal discipline mechanisms (Krol 1997). This study differs in several important respects from previous analyses. Specifically, the objective of this paper is to investigate the impact on state and local spending of a comprehensive set of fiscal and administrative mechanisms. In addition, this analysis employs a panel data set focusing on 49 of the 50 U.S. states using data observed at regular intervals over the 26-year period 1969 to 1994.2 As a result, this analysis is able to capture the effects of fiscal discipline measures over a relatively long period of time.
There are two reasons why the objectives of this analysis are important. First, there is presently a movement in the United States toward imposing, at the federal level, some of the very fiscal discipline mechanisms that are analyzed in this paper.3 Second, there are statebased pressures to implement various discipline mechanisms in states in which they presently do not exist. The knowledge of whether or not a given budgetary rule significantly affects state-level expenditures, and if so, by what magnitude and in what direction, could prove to be extremely valuable to policymakers seeking to control expenditure levels in the face of ever-mounting budgetary difficulties.
The plan of this paper is as follows. The next section provides an overview of the fiscal discipline mechanisms that were in place in various states over the time period of this analysis. We then present the theoretical model of state and local government expenditures and also briefly describe the random-effects model employed in this Due to the unusual nature of fiscal discipline practices in Alaska, our analysis does not include inferences on this state.
In particular, there has been a movement toward implementing balanced budget requirements and the line-item veto, although, more recently, the latter has been declared unconstitutional.
analysis. Next, we provide a description of the data and the sources from which they were obtained, and also present the results of our empirical estimation. A brief summary of the results and concluding statements are presented in the final section.
Fiscal Discipline Mechanisms In their seminal work, The Calculus of Consent, Buchanan and Tullock (1962) first introduced the argument that institutions and rules influence the way collective choices are made with the result that they have important effects on policy outcomes. One of the primary implications of public choice theory is that the rules or institutions within which various groups operate are likely to affect the outcomes of decisions made by these groups. Indeed, the work of Poterba (1994, 1995, 1996) has provided evidence that suggests that budgetary rules and institutions can affect fiscal policy outcomes. This “public choice” view is in direct contrast to the “institutional irrelevance view” that suggests that budgetary rules and institutions can be circumvented through various means, thereby rendering these rules ineffective. Under this view, there is little chance that taxpayers can succeed in instituting changes that reduce the level of government spending. However, this argument neglects the possibility of dissatisfied taxpayers using political pressures to produce constitutional changes that reduce the equilibrium budget size (Abrams and Dougan 1986).
State and local governments function under a wide assortment of fiscal discipline mechanisms ranging from initiative procedures to super-majority voting requirements for tax increases. Variations in these mechanisms and institutions are likely to provide some explanation for the varying policy outcomes as they relate to state and local spending patterns and levels. The fiscal discipline mechanisms under investigation here can be divided into three general categories. The first category is budgetary constraint mechanisms that relate directly to spending or revenue levels. This category includes budget rules such as tax and expenditure limits, line-item veto power, balanced budget requirements, and super-majority voting requirements for tax increases.4 The second category is administrative constraint mechaIt might be argued that the various types of debt restrictions that are placed on state and local governments should also be included in this list of fiscal constraint mechanisms. There are, however, several problems associated with including debt restrictions as a fiscal constraint mechanism. Von Hagen (1991), for example, did not find significant differences in per capita debt between states with and without such limits. Furthermore, Bunch (1991)
nisms that focus on the process by which the budgetary process is carried out or on those who enact the budget. This category includes mechanisms such as term limits for state legislatures, bill introduction limits, and the length of the budget cycle. The third category focuses on direct democracy mechanisms that allow citizens to directly participate in the budgetary process. The initiative procedure and the state referendum are members of this category. The following discussion summarizes each of these budget rules and outlines the theoretical justifications for including each in a model of state and local expenditures.
Tax and expenditure limitations are one of the more recently devised fiscal discipline mechanisms. The great majority of the existing tax and expenditure limits, 21 out of 23, were put in place during the decade ending in 1986. New Jersey, in 1976, was the first state to place a limit on state taxing and spending powers but later, in 1982, allowed this limitation to expire. Colorado and Rhode Island followed suit in 1977, with Tennessee joining the ranks in 1978.5 With the passage of Proposition 13 in California, 15 additional states then enacted some form of tax or expenditure limit. Connecticut and North Carolina are the most recent states to impose such limits. Tax and expenditure limits vary widely from state to state. Some limits are constitutional while others are statutory. Some place limits on spending while others constrain revenues. Furthermore, escape clauses in tax and expenditure limits differ significantly from state to state and significant categories of either expenditures or revenues are excluded from coverage in various states. It is also the case that the appropriate base for tax and expenditure limits is subject to wide variations across states (Bails 1990). If tax and expenditure limitations are successful at capping government spending and taxation, then this variable should be statistically significant and negatively related to per capita government expenditures.
The ability of a governor to use the line-item veto or itemreduction veto as a means of reducing expenditure levels is currently available in all but nine states (Alm and Evers 1991). The theoretical argument for including this measure in a model of state and local expenditures focuses on the notion that individual legislators are responsive to the median voter of their individual district, while govershowed that governments use public authorities to circumvent state constitutional debt limits with the effects being especially prevalent in states that have a debt limit that applies to both general obligation and revenue debt.
Tax and expenditure limits in Rhode Island and Nevada are nonbinding and, therefore, are considered to be nonexistent for purposes of this study (Stansel 1994).
THE IMPACT FISCAL CONSTITUTIONSOF nors are responsive to the median voter in their state. Furthermore, there is no inherent reason to argue that the preferences of these two groups of median voters will be identical. The statewide median voter may prefer the governor to have veto authority to offset the power of the district median voter. A related argument is that the platforms offered to the median voter of the state and those offered to the median voter of the various districts may differ because the political parties of the governor and members of the legislature may differ. If this is the case, the state median voter may prefer a governor who is more likely to use the item veto when the majority of the legislature is of a different party. Both of these arguments suggest that those states where the governor has either a line-item veto or itemreduction power are likely to have lower levels of per capita spending than those states where the governor has neither (Dearden and Husted 1993). Thus, the line-item veto variable should have a negative impact on state and local spending.
Balanced budget requirements, of one type or another, are present in each state except Vermont. In 43 states the governor is required to submit a balanced budget; in 40 states the legislature is required to pass a balanced budget. However, in 11 of the latter states, the government can run a deficit legally simply by carrying it over to the next fiscal year, thereby rendering the balanced budget requirement ineffective. In the remaining 31 states the legislature must pass a balanced budget with no carryover and a proviso that if a deficit occurs during that year, it must be eliminated by either reducing spending or increasing taxes. Niskanen (1975) notes that special interest and monopoly models of government suggest that equilibrium levels of government expenditure will be greater than optimal and, as a result, balanced budget requirements should offer taxpayers some protection against excessive government spending. Given the structural differences in tax and spending limits across all states, and taking into consideration the relationship between spending limits and balanced budget requirements, it is logical to assume that effectiveness of balanced budget requirements will depend on whether or not there is also a tax and expenditure limit in place. In particular, we predict that balanced budget requirements will exert only a negative influence on the level of state and local spending in those states that also impose tax and expenditure limitations.
Super-majority voting requirements are another constitutional restriction on legislative tax powers. This mechanism is aimed at reducing the ability of the legislative body to exploit the voters by requiring that more than a simple majority of voters approve tax increases. The origin of super-majority voting requirements dates back to 1934 when
CATO JOURNALArkansas voters approved a constitutional amendment referred by the legislature. This amendment required a two-thirds vote to increase “the rates for property, excise, privilege, or personal taxes now levied” (Mackey 1993). Following theories of democracy and representative government, it is logical to expect that super-majority voting requirements will be effective at lowering government expenditures only when they are combined with a balanced budget requirement (Farnham 1990).
Of all the administrative constraints placed on decisionmakers in the public sector, none has been more controversial than term limits.