Tax And Spending Limits: Theory, Analysis, And Policy

Tax and Spending Limits: Theory, Analysis, and Policy by Barry W. Poulson Professor of Economics University of Colorado at Boulder IP-2-2004 January, 2004

14142 Denver West Parkway • Suite 185 • Golden, Colorado 80401-3119 www.IndependenceInstitute.org • 303-279-6536 • 303-279-4176 fax

Executive Summary This study surveys the literature on tax and spending limits (TELs). The recent literature includes more rigorous econometric analysis, and more thorough case studies of TELs in individual states. These studies provide important insights into the design and implementation of TELs.

will redistribute income from those who paid the excess taxes to those benefiting from government spending. Special interest groups are often successful in earmarking revenue that is then exempt from the TEL limit and used finance expenditures benefiting their constituents

Recent empirical studies support the ‘public choice’ view that budget institutions significantly affect fiscal policy. TELs, as well as other budget rules, can significantly reduce state and local spending. However, decision makers must pay attention to the design of TELs if they are to have a significant impact in constraining government spending. The most effective TELs are ones that: a. are constitutional rather than statutory b. limit the growth of government spending to inflation and population growth rather than other aggregate measures of economic activity c. provide for immediate refunds of surplus revenue above the TEL limit d. are linked to other budget rules, most importantly to balanced budget requirements.

Some TELs mandate tax cuts and tax rebates when revenue exceeds the TEL limit. One approach would return the surplus revenue to those who paid the excess taxes. Rarely is this achieved; the tax cuts and tax rebates have the effect of redistributing income from those who paid the excess taxes to others who may have paid little or no taxes.

Budget rules also affect the way that states respond to revenue shortfalls by cutting expenditures and/or raising taxes. States with TELs experience lower tax increases in periods of recession than states without TELs. State response to a deficit is also affected by the general fund balance. The general fund balance is a broad measure of the total reserves available to stabilize the budget. States with low general fund balances must make larger spending cuts in response to budget deficits. In some states these reserves are allocated to a budget stabilization. There is also evidence that tax increases as a fraction of the deficit are larger in states with low general fund balances.

Some economists have explored case studies of TELs in individual states. In each of these cases we find evidence that political and legal institutions have eroded the effectiveness of TELs. In some states this influence was apparent in the original design of a weak and ineffective TEL. In other states erosion in the effectiveness of the TEL was the result of legislative actions and court decisions that weakened various provisions of the TEL. The next generation of TELs is designed to achieve an optimum tradeoff between constraining the growth of government and stabilizing government budgets over the business cycle. These TELs link a stringent tax and spending limit to a budget stabilization fund. New TELs embodying elements of this design have now been proposed in half a dozen states in recent years.1 A model TEL with this design is included as an appendix to this study.

When surplus revenue is generated above the TEL limit some states simply return the revenue to the general fund to finance expenditures. The allocation of surplus revenue to finance government spending

Page 1

I. INTRODUCTION Tax and spending limits (TELs) are budgetary rules that determine how much taxes and/or expenditures can increase from one year to the next. TELs can be statutory or constitutional rules. Statutory TELs can be modified by legislative action, while constitutional TELs can only be modified by a majority vote of citizens. TELs may originate through a legislative statute or referendum, or they may be initiated by citizens in states that provide for this form of direct democracy. TELs are now in place in 26 states (Rafool 1996). Empirical evidence indicates that well designed TELs impose fiscal discipline on elected officials; poorly designed TELs do not. TELs introduced through citizen initiatives tend to be better designed than TELs developed by elected officials. The most effective TELs are ones that are constitutional, that limit the growth of government spending to inflation and populaAt least some tion growth, and that provide for states with TELs immediate refunds of surplus revappear to have enue to the citizens who paid the been more sucexcess taxes. TELs are most effective cessful than other when linked to other budget rules, states without most importantly to balanced budget TELs in conrequirements. However, the available straining the evidence indicates that political and growth of legal institutions can (and have) erodgovernment. ed the effectiveness of TELs through legislation and court decisions. Two recent developments have stimulated renewed interest in TELs in the U.S. The first is divergence in levels and long-term rates of growth in state revenue and spending in the states. At least some states with TELs appear to have been more successful than other states without TELs in constraining the growth of government. Whether or not this fiscal discipline is a result of the impact of TELs has been the subject of a large and growing empirical literature. The second development is the diverging response of the states to the revenue shortfalls that have accompanied the recent recession. States have

responded differently to deficits in their budgets: some states cut spending; other states raised taxes; and a few states, such as California, have accumulated debt of a magnitude greater than that which has occurred in any previous recession. Sorting out the impact of TELs and other budget rules as an explanation for the different response of states to revenue shortfalls has emerged as a major research question. In this study we will survey the literature on TELs. The recent literature includes more rigorous econometric analysis, and more thorough case studies of TELs in individual states. These studies provide important insights into the design and implementation of TELs. The most comprehensive survey of this literature was that conducted by James Poterba (1994). In that survey Poterba explored the fiscal impact of a wide range of budget rules and institutions, including TELs. His survey included international, national, and state experience with budget rules. Poterba concluded that budget rules do affect fiscal policy, however, he reached an agnostic conclusion regarding the impact of specific rules such as TELs: “ The existing evidence is not refined enough, however, to provide detailed advice on how narrowly defined changes in budget rules might affect policy outcomes.”(Poterba 1996: Abstract) Poterba’s conclusion reflected the narrow and limited scope of research on TELs a decade ago. Most empirical studies lumped TELs together as a single homogeneous fiscal constraint, and then proceeded to test whether they had a significant impact in constraining the growth of government. The literature on TELs focused almost exclusively on the impact these budget rules have in constraining the growth of government. However, TELs have much broader impacts about which we knew very little. Research on TELs over the past decade has explored a wider range of issues, including:

Page 2

A. The Political Economy of TELs

D. The Impact of TELs on Governance

Anyone familiar with TELs is aware of the complex issues of political economy that have influenced the design and implementation of these budget rules.2 Beyond the complexities of the TEL legislation itself, are the institutions that have influenced the implementation of TELs. Even the most stringently designed TEL will have little fiscal impact if the legislature and the courts do not enforce the provisions of the TEL.

From the outset, when TELs were first introduced to constrain local property taxes, people worried that the financing of government programs would be shifted to the state. When states began introducing TELs, people worried that the tax burden would be shifted to local governments. Most TELs contain provisions precluding states from mandating local government functions without providing the financing to support those functions. This issue of the impact of TELs on fiscal federalism and the relations of local, state, and federal government remains largely unexplored.

B. The Impact of TELs on State Budgets Over the Business Cycle Most states have mandated growth in state expenditures for education K-12, Medicaid, and prisons. If the TEL ratchets down the revenue the state can spend during a recession, the only way to satisfy these mandates is to drastically reduce spending in other state programs, such as higher education, transportation, and social welfare. The combination of an effective TEL and mandated growth in state expenditures may result in a structural deficit in the budget in the long run. C. The Impact of TELs on the Distribution of Income When TELs do impact revenue and expenditures they will have direct and indirect impacts on the distribution of income. The direct impact occurs when surplus revenue is returned to taxpayers in tax rebates and tax cuts. Some TELs require that these tax rebates and tax cuts be in some proportion to the excess taxes paid. However, some TELs leave wide discretion to the legislature, which opens opportunities for privilege seeking through targeted tax rebates and tax cuts to benefit special interests. Indirect effects of TELs on the distribution of income occur through budgetary changes as some state programs expand and others contract in response to the constraints imposed by the TEL.

As this brief introduction suggests, the reader should not expect a synthesis of a well-developed literature; but rather a discussion of conceptual and theoretical issues, a review of the relevant empirical analysis and case studies, and an exploration of policy issues. We conclude with a summary of recommendations in designing and implementing a TEL, and an appendix with a model TEL that embodies these recommendations.

II. THEORETICAL ISSUES Underlying every study of TELs is a theory of government, whether that theory is made explicit or not. In his literature survey Poterba contrasted the ‘institutional irrelevance’ view with the ‘public choice’ perspective of budget rules. In the ‘institutional irrelevance’ view, TELs, as well as other budget rules, have no impact on fiscal policy. The problem from this perspective is the interdependent nature of budget rules. Riker (1980), for example, maintains that all political institutions result from the preferences of the electorate. ‘Institutional irrelevance’ means that the institutional structure of a nation or state contains no information beyond the aggregate of information of voter preferences. If institutions no longer suit the majority of the electorate they will be overcome. From this ‘institutional irrelevance’ perspective, states with TELs may be ones with a conservative

Page 3

electorate that would have constrained revenue and spending, even in the absence of the TEL. On the other hand, some states with TELs may have a more profligate electorate that will find ways to evade and avoid the constraints imposed by the TEL. In either case TELs, as well as other budget rules, are nothing more than a summary statement of voter preferences. Some econometric studies have attempted to resolve this interdependence problem, as discussed later in this paper. In contrast to the ‘institutional irrelevance’ view, some economists argue that TELs and other budget rules are exogenous. The leading group of economists identified with this view are referred to as ‘public choice’ economists.3 They point to the durability of many of these budgetary institutions. For example, every state but one, Maine, has a written constitution requiring a balanced budget. These balanced budget rules emerged from constitutional conventions in which, as Buchanan and Wagner (1977) argue, citizens believed in the ‘old time fiscal religion’ of balanced budgets. Politicians have often found ways to circumvent these balanced budget rules in the short run, but they have tended not to violate them in the long run. There has been no attempt to rescind balanced budget rules in the way that TELs have been rescinded in some states. TELs are more recent budgetary rules that have emerged from a different decision process. Nonetheless, we now have several decades of experience with TELs, and in some states they are acquiring an acceptance and durability of their own. From this ‘public choice’ perspective, TELs and other budget rules can have significant impacts on fiscal policy. Further, ‘public choice’ economists challenge the view that social planners are capable of designing optimum tax and expenditure policies over the business cycle as well as in the long run. Their assumption is that the democratic decision process may inherently bias fiscal policy. In the long run a bias toward excessive spending may result in too much growth in government. This is especially true at the national level where government can incur deficits and accumulate debt. At the state

level this bias may result in excessive ratcheting up of government revenue and spending over the business cycle. Faced with a revenue shortfall, states frequently increase taxes in order to balance the budget. The higher taxes then generate higher revenue growth as the economy recovers from recession, ratcheting up revenue and spending in the long run. In the ‘public choice’ view, TELs and other budget rules can provide a fiscal discipline that constrains the discretion of decision makers, offsetting the bias toward expansionary government inherent in the democratic process. Different ‘public choice’ models are designed to capture this fiscal decision process in a democratic society. We will explore the design of TELs using several of these ‘public choice’ models. A. Leviathan Model of TELs

In the ‘public choice’ view, TELs and other budget rules can provide a fiscal discipline that constrains the discretion of decision makers, offsetting the bias toward expansionary government inherent in the democratic process.

Most studies that test the impact of TELs implicitly assume a Leviathan model of state government. In the Leviathan model, introduced by Brennan and Buchannan (1980) TELs are viewed as part of the social contract. Politicians are assumed to maximize revenue and expenditures within the constraints imposed by TELs.

Even a cursory look at the experience with TELs in different states reveals a more complex decision process than that assumed in the Leviathan model. The design and implementation of TELs reflects interaction between citizens, special interest groups, and politicians. Some economists have attempted to capture this more complex decision process in other ‘public choice’ models. B. Median Voter Model of TELs Some models attempt to capture the response of politicians to different influences. One approach is based upon the median voter principle: the median

Page 4

voter is defined such that he equates his value of an additional public service with the tax price of that service (Flowers 1997; Sjonquist 1981, and Chicoine and Walzer 1986). In a median voter model politicians support TELs when at least a majority of citizens incur net benefits from the TEL. C. Rent Seeking Model of TELs An alternative approach is based on the assumption that a self- interested politician will attempt to minimize the political costs associated with raising a given budget or revenue (Hettich and Winer 1984). In these models taxpayers are on opposite sides of the budget constraint from special interest groups. In the ‘public choice’ literature privilege seeking by these special interest groups is referred to as rent seeking, and models designed to capture this privilege seeking are called rent seeking models. To the extent that taxpayers can impose costs on profligate politicians we would expect TELs to be designed and implemented to impose a stringent budget constraint. Some TELs are constitutional amendments that originate through the citizen initiative process. Such TELs tend to impose more stringent constraints on the fiscal powers of government than statutory TELs that emerge from a legislative decision process. In either case it is unrealistic to assume that rent seeking groups will simply sit back and accept new rules of the game designed to limit their rent seeking activities. We expect rent seeking groups to become actively involved in the design and implementation of TELs in order to protect what they perceive to be their entitlements to such rents. We expect self interested politicians to respond to rent seeking groups in the design of TELs. Politicians may design a TEL which is completely ineffective in constraining government revenue and expenditure. They can appear to be fiscally prudent to appease taxpayer groups, when in fact they introduce a weak and ineffective TEL to preempt a more stringent TEL from being introduced through citizen initiative.

Politicians may also co-opt TELs in order to benefit one interest group at the expense of other interest groups (Spindler 1990). The constraints imposed by TELs rarely impact interest groups uniformly. The TEL limits may be imposed on a subset of state expenditures, but not on other expenditures. Some interest groups may be successful in earmarking revenue used to finance their expenditures, shifting those revenues outside the TEL limit. In effect these interest groups are removed from the rent seeking battles that accompany the budget process. Politicians can appear to be fiscally prudent in supporting TELs, when in fact they are benefiting rent seeking groups. The ability of rent seeking groups to protect what they perceive as their entitlements to rents depends upon the response of politicians. The special interest effect occurs when politicians pursue policies that benefit special interests at the expense of the public interest. Politicians may design TELs that are weak and ineffective in constraining the fiscal powers of government. Even with a well-designed TEL, politicians may successfully evade or erode these fiscal constraints to permit a more rapid growth in revenue and spending.4 Once we recognize that TELs emerge from a complex decision process involving citizens, special interest groups, and politicians, the impact of TELs on fiscal policy is uncertain. If there is a bias toward excessive levels of revenue and spending over the business cycle and in the long run, then TELs have the potential to enhance the welfare of citizens. A TEL can be designed and implemented to constrain the growth of government in the long run. TELs may also be designed to provide for greater stability of government budgets over the business cycle. On the other hand, TELs may be designed and implemented that have little impact on fiscal policy. At worse, poorly designed TELs can result in inefficiencies and inequities in fiscal policies that diminish the welfare of citizens.

Page 5

III. EMPIRICAL ANALYSIS Most empirical studies have focused on the issue of TELs as constraints on government revenues and expenditures. This is consistent with the perception that TELs are designed and implemented to impose fiscal discipline on governments. However, it is not always clear what is meant by fiscal discipline.

get stabilization fund to provide greater stability to revenues and expenditures over the business cycle, and to limit the ratchet down effect in the long run. This combination of a TEL and budget stabilization fund permits citizens to achieve an optimum tradeoff between stabilizing the budget over the business cycle and constraining the growth of government in the long run.

Some people might argue that fiscal discipline means decreasing the absolute size of government. Indeed a recent failed TEL initiative in Colorado called for decreasing the absolute size of local government by a fixed dollar amount each year.

Whether or not we perceive TELs to have been successful in constraining the growth of government depends upon what we mean by fiscal discipline. With this caveat in mind we turn to the empirical literature that has attempted to answer this question.

Others might argue that TELs are designed to constrain the growth of government. Few would disagree that an effective TEL should hold the growth of government below the double digit rates of increase in revenue and spending that occurred during the inflationary period of the 1970’s. It was that inflationary expansion, in which government at all levels increased relative to the private sector, that triggered the tax revolt, with TELs designed to constrain the growth of local and state government.

A. TELs as Constraints on Government Revenue and Spending

If fiscal discipline is interpreted to mean constraining the growth of government relative to the private sector, then tax and spending limits may be linked to a measure of aggregate economic activity such as state income. The problem with these TELs is that they permit a rapid growth in government in periods of prosperity that cannot be sustained when the economy enters a recession with falling revenues. As noted earlier, such TELs may result in a ratcheting up of government spending over time, if governments respond to falling revenues by increasing taxes.

Cross section studies of the impact of TELs on state and local spending tended to find TELs ineffective in constraining government spending. Bails (1983) conducted one of the earliest cross section studies and found that in the majority of states TELs had an insignificant impact on state spending. An extension of this work essentially confirmed this finding (Bails 1990). Cross section studies by Abrams and Dougan (1986) and Cox and Lawry (1990) also found TELs to be ineffective in constraining state spending.

A more stringently designed TEL is linked to inflation and population growth. In the long run this type of TEL has the potential to constrain the growth of government below the growth of the private sector. The criticism of this type of TEL is a ratchet down effect as revenue falls in periods of recession. However, this type of TEL can be linked to a bud-

A number of studies have tested whether TELs constrain government revenue and spending. The wide divergence in the results of these empirical studies reflects the use of different methodologies and data sets. Early empirical studies focused on different political units, and analyzed the impact of TELs over different time periods.

Time series studies have found mixed results for the impact of TELs on state and local spending. A study by Kenyon and Benker (1984) compared the growth of state spending relative to personal income in TEL states with that in non TEL states and found no significant difference. A study by Dougan (1988) examined time series data from 1960 to 1984 for 16 states. He found that in 7 of these states the TEL had a significant negative impact on government spending.

Page 6

Poulson and Kaplan (1994) used time series analysis to examine the impact of Colorado’s first statutory TEL, and found that it significantly reduced state spending in the short run, but not in the long run. Joyce and Mullins (1991) also found evidence that TELs significantly reduce state revenues in the short run, but not the long run. One of the problems in all of these empirical studies is that they do not account for the interdependent nature of TELs. If states that experience rapid increases in spending are more likely to adopt TELs, then the empirical analysis of the impact of TELs on state spending will be biased. A few empirical studies have addressed this interdependence problem. One approach to this problem is to control for some measure of voter preference, such as the political party of elected officials. This can reduce the proclivity that observed correlation between budget rules and fiscal outcomes simply reflects correlation between these variables and an omitted variable for voter preferences. But, if this approach does not fully capture the impact of voter preferences it may still lead to spurious results. An alternative approach to this problem is to solve for interdependence using variables that affect budget rules but not fiscal policy. Reubens (1995) uses this approach to estimate the impact of TELs. In some states, laws provide for the initiative process that allows citizens to directly propose and vote on TELs. These laws should be positively correlated with the passage of TELs, but unrelated to current expenditure levels. Substituting this variable for the TEL variable in her model enables Reubens to separate the effects of the TEL from the effects of changes in voter preferences for TELs. When this model is run with the TEL variable the results show a positive relationship between TELs and government spending. When the model is run using the initiative variable the effect of the TELs is to reduce state spending 1.8%. Another criticism of these empirical studies is that they do not account for other variables that could

impact state spending. More recent empirical studies use panel data, cross sectional time series data, which can produce more reliable estimates because they include other variables besides TELs that can influence government revenue and spending. For example empirical evidence suggests that government expenditure, in both levels and growth rates, is closely related to income, population, and other economic variables. One of the first studies to utilize panel data controlling for the effects of population and income on the public sector is a study by Elders (1992). Elders found that when he controls for the effects of these economic variables, TELs have a significant effect in reducing the growth of government. Shadbegian (1996) also uses panel data to test the impact of TELs on the size and growth of state government. He includes three control variables, population, income, and federal intergovernmental grants. Shadbegian finds that TELs reduce the size of government, and lower government spending growth in states with low income growth. However, he finds that TELs increase government spending growth in states with high income growth. In states with high income growth TELs provide cover for politicians who appear to be fiscally prudent, but who in fact are increasing government spending more than their counterparts in states without TELs. A criticism of all the studies mentioned above is that they treat all TELs alike. The only study to take into account differences in TELs is that of New (2001). New uses panel data to analyze the impact of TELs on the level of state and local spending. As in other panel data studies he controls for other economic and demographic variables that could impact state and local spending. He includes variables indicating whether the TEL was passed by initiative, by the legislature, by constitutional convention, or by referendum. He finds that TELs passed through the initiative process significantly reduce state and local spending; but, TELs enacted by state legislatures actually increase spending. Further, he finds that states in which TELs limit increases in revenue

Page 7

and spending to the sum of inflation and population growth, i.e. Colorado and Washington, reduce government spending more than states in which the TEL limit is linked to state personal income. Finally, he finds that states in which the TEL mandates immediate refunds of surplus revenue to taxpayers, i.e. Colorado, Michigan, Missouri, and Oregon, are more successful in reducing government spending. A study by Bails and Tieslau (2000) is unique in using panel data to analyze the impact of a comprehensive set of budget institutions on state and local spending. These include; tax and spending limits, line item veto, balanced budget requirements, and term limits. As in other panel data studies they control for other economic and demographic variables. They find that the following budgetary institutions significantly reduce the rate of growth of state and local spending: tax and spending limits, balanced budget requirements in the presence of tax and spending limits, supermajority vote The most effective requirements in the presence of balTELs are ones anced budget requirements, term that are constitulimits, and the initiative process. Real tional, that limit per capita state and local spending the growth of govin states with TELs, is estimated to ernment spendbe $42 lower than in states without ing to inflation TELs. In states with both TELs and and population balanced budget requirements spendgrowth, and that ing is reduce by nearly $135. In states provide for immewith the comprehensive budgetary diate refunds of institutions tested, they find that real surplus revenue per capita expenditures are reduced above the TEL nearly $473. limit. Thus, recent empirical studies support the public choice view that budget institutions significantly affect fiscal policy. TELs, as well as other budget rules, can significantly reduce state and local spending. Decision makers must pay attention to the design of TELs if they are to have a significant impact in constraining government spending. The most effective TELs are ones that are constitutional, that limit the growth of government spending to inflation and population growth, and that provide for immediate refunds of surplus rev-

enue above the TEL limit. TELs are most effective when linked to other budget rules, most importantly to balanced budget requirements. B. TELS and Budget Stabilization Over the Business Cycle When Governor Owens of Colorado recently encouraged Governor Schwarzenegger of California to adopt a TEL, he argued that this would enable him to stabilize the state budget over the business cycle (Owens 2003). Owens maintained that the TABOR Amendment limited increases in government revenue and spending in the 1990’s, so that it was easier for Colorado to cope with the revenue shortfalls in the current recession. When California essentially abandoned their TEL in the late 1980’s this permitted unconstrained growth in revenue and spending to unsustainable levels in the 1990’s. The question raised by this exchange between Governor Owens and Governor Schwarzenegger is whether states with effective TELs have been subject to less volatility in revenues and spending over the business cycle compared to states without effective TELs. This question remains largely unexplored in the literature. One of the few studies to analyze this question is a study by Poterba (1994) that explores how states adjust their budgets to unexpected shocks. Poterba’s approach is to measure how changes in state spending and taxes are influenced by fiscal institutions. He examines the impact of balanced budget rules, TELs, and general fund balances. Most state constitutions prevent state governments from running deficits. After a budget is passed, however, actual revenues and expenditures may deviate from projections, resulting in unexpected deficits. States differ in the stringency of balanced budget rules that require elimination of the deficit. Some states require elimination of the deficit in the current fiscal year; while others allow the deficit to be carried forward into the next fiscal year; and a few states do not require the deficit to be eliminated in the following fiscal year.

Page 8

To offset deficits states may draw down their general fund balances or rainy day funds. Most states allow borrowing to offset the deficit. In some states the borrowing must be repaid in the current fiscal year; while others allow the debt to be repaid in the following fiscal year, but prevent the use of long term debt to cover the deficit. In the short run, of course, states may rely upon a variety of budgetary gimmicks to satisfy balanced budget rules: deferring expenditures to the following fiscal year, altering the fiscal year, raiding cash and trust funds, etc. Poterba explores how states respond to revenue shortfalls that accompany recessions. He finds that states with TELs experienced lower tax increases in periods of recession ... states with than states without TELs. In TEL TELs experienced states a $1 increase in the budget lower tax increasdeficit results in a $0.47 tax increase, es in periods of compared to a $1.03 tax increase in recession than states without TELs. He finds no states without evidence that spending cuts are any TELs. larger in states with TELs. Poterba also explores how a state response to a deficit is affected by the general fund balance. In some states these reserves are allocated to a budget stabilization or rainy day fund. The general fund balance is a broader measure of the total reserves available to stabilize the budget. Poterba finds evidence that states with low general fund balances must make larger spending cuts in response to budget deficits. In states with general fund balances of more than 2% of spending, spending is reduced by $0.25 per dollar of deficit. In states with lower general fund balances, the spending cuts are more than double, $0.55 per dollar of deficit. There is also some evidence that tax increases as a fraction of the deficit are larger in states with low general fund balances. Poterba finds that fiscal institutions do affect the way that states respond to revenue shortfalls by cutting expenditures and/or raising taxes. However, he concludes that the question of how fiscal institutions

affect the level of government spending remains an open question. C. TELs and the Distribution of Income A fuller understanding of TELs would extend the analysis to their impact to income distribution. When surplus revenue is generated above the TEL limit some states simply return the revenue to the general fund to finance expenditures. In some states the surplus is held in a general fund balance. When that general fund balance is used to finance a budget stabilization fund then the moneys are expended during revenue shortfalls. In some cases the general fund balances are used to finance targeted expenditures such as education or capital expenditures. As we would expect from our rent-seeking model, general fund balances may become a target for different special interest groups. The allocation of surplus revenue to finance government spending will redistribute income from those who paid the excess taxes to those benefiting from government spending. One of the least understood aspects of TEL is the way in which they interact with other statutory and constitutional provisions impacting fiscal policies. As our rent-seeking model suggests, special interest groups are often successful in earmarking revenue that is then exempt from the TEL limit and used finance expenditures benefiting their constituents.

... special interest groups are often successful in earmarking revenue that is then exempt from the TEL limit and used finance expenditures benefiting their constituents.

Some TELs mandate tax cuts and tax rebates when revenue exceeds the TEL limit. One approach would return the surplus revenue to those who paid the excess taxes. Rarely is this achieved; the tax cuts and tax rebates have the effect of redistributing income from those who paid the excess taxes to others who may have paid little or no taxes. One of the few studies to explore how tax cuts and tax rebates impact the distribution of income is Poulson (1999 and 2001). He finds that in Colorado surplus revenue above the TABOR limit was not offset by

Page 9

tax cuts and tax rebates that returned the surplus to those who paid the excess taxes. Most of the surplus revenue was generated by the income tax, and some of the surplus was offset by broad based income and sales tax cuts. But much of the surplus was offset by targeted tax cuts and tax rebates benefiting special interest groups. The impact of this redistribution was egalitarian, shifting surplus revenue from upper income families who paid most of the excess income taxes, to lower income families who paid little or no taxes. D. TELs and Governance A final set of issues involves the impact of TELs on governance. One question is the extent to which TELs may erode fiscal federalism. In some states TELs may be imposed at the local level but not the state level, as was the case with Prop 13 in Califonia. A number of states have introduced TELs at the state level but not the local level. Other states, such as Colorado, introduced TELs designed to constrain fiscal policy at both the state and local level. From the outset a major concern was that TELs imposed at one level of government would simply shift the financing of programs to another level of government. The obvious example is education K12 which is often financed from revenues received by both state and local government. Some TELs address this issue by limiting the power of the state to mandate expenditures at the local level without financing those expenditures. Satisfying the conditions imposed by TELs requires new administrative procedures. For example, when TELs require voter approval for increase in taxes and debt, this requires new accounting, reporting, and election procedures. Litigation surrounding TELs may require additional legal resources. These administrative procedures may be costly, particularly for local governments with limited resources. The costs and other burdens of TELs must be weighed against the benefits to citizens in exercising control over fiscal powers of government directly

rather than leaving fiscal decisions to the discretion of their elected representatives. Surveys often reveal widespread support for TELs; nonetheless, assessing the costs and benefits to citizens of exercising controls over the fiscal powers of government through TELs remains elusive.

IV. CASE STUDIES Empirical studies provide important insights into TELs. However, the effectiveness of TELs also depends upon the unique political and legal institutions found in each state. Our public choice models suggest that the way in which the legislature and the courts choose to interpret and implement the TEL is crucial. Surprisingly little research has been conducted into this question, and as a result it is one of the least understood aspects of TELs, especially by economists. Analysis of the way in which political and legal institutions influence TELs does not lend itself to econometric analysis, and publication in economics journal, and so economists have contributed very little to this question. Some economists have explored case studies of TELs in individual states, which are usually published in state and national think tanks. We will review case studies for several states including: California, Missouri, Florida, Washington, Michigan, and Colorado. In each of these cases we find evidence that political and legal institutions have eroded the effectiveness of TELs. In some states, such as Florida, this influence was apparent in the original design of a weak and ineffective TEL. In other states erosion in the effectiveness of the TEL was the result of legislative actions and court decisions that weakened various provisions of the TEL. A. California5 The origins of the tax revolt can be traced to the Gubernatorial Administration of Ronald Reagan in California in the early 1970s. Reagan’s Prop1 set the standard for all subsequent TELs. Prop 1 set a limit on state spending equal to the current ratio of state expenditures as a share of state income, and

Page 10

required that percentage to decrease by .1% each year until it reached 7%, at which time the legislature (by two thirds vote) could halt the annual reduction. To give some idea how stringent this original TEL was, total spending as a share of state income in California today is approaching 10%. While Prop 1 lost at the polls in 1973, it launched a tax revolt in California that resulted in initiatives to introduce TELs at both the local and state level. Prop 13 to limit local property taxes wan enacted in 1978. Prop 13 continues to be one of the most stringent limitations on local property taxation in the country. Prop 4, the Gann Amendment, was one of the first TELs adopted by any state. The Gann amendment limited increases in state and local appropriations from tax revenues to the sum of population growth and inflation. The Gann Amendment passed by nearly a 3 to 1 margin. Initially the Gann Amendment was effective in constraining government expenditures. In 1987, when revenue exceeded the budget limit, California taxpayers received a $1.1 billion tax rebate.

In 1990 the Gann amendment was dealt a death blow when Prop 111 was enacted mandating increases in expenditures for education K-12, and increasing the Gann limit.

Over time, however, the Gann Amendment became less effective as a constraint on California fiscal policy. Because the limit was imposed only on appropriations from tax revenues, legislators relied increasingly on fees and other non-tax revenues. In 1990 the Gann amendment was dealt a death blow when Prop 111 was enacted mandating increases in expenditures for education K-12, and increasing the Gann limit. Since then, the Gann Amendment has ceased to be an effective constraint on government fiscal policy in California. B. Missouri6 One of the first TELs introduced in the states was the Hancock amendment in Missouri, in 1980. That

TEL limited the growth in state revenue to the rate of increase in personal income. Tax revenue was to be held constant as a share of personal income, with surplus revenue rebated to taxpayers based on state income taxes. The Hancock Amendment was never an effective constraint on the growth of revenue and spending, and no tax rebates were ever made under the Amendment. The reason is that legislators, aided by the courts, were able to successfully evade and avoid the limits imposed by the Amendment. Legislators exempted an increasing share of revenue from the limit. Initially, only 2% of revenue was exempt from the limit; but a decade later 18% was exempt. Over that period state revenue grew far in excess of personal income; in some years the growth in revenue was more than double the growth in personal income. Experience in Missouri demonstrates how the legislature and the courts can erode the effectiveness of a TEL. The Missouri legislature failed to enact enabling legislation to implement the Experience in Hancock Amendment. As a result Missouri demthere was never agreement on what onstrates how constituted total state revenue, nor the legislature what constituted surplus revenue and the courts above the limit. Different state agencan erode the cies came up with their own confliceffectiveness of a ting interpretations. TEL. Soon after the Hancock amendment was passed the voters passed a 1% sales tax increase; half of which was earmarked for property tax reduction, and half of which was earmarked for expenditures for education K-12. Both the Missouri Budget Office, and the State Auditors Office ruled that this earmarked revenue was not subject to the Hancock limit. Amazingly, the Missouri supreme Court agreed, ruling that this statutory tax increase preempted the revenue limit imposed by the Missouri Constitution. Other tax increases soon followed: the legislature enacted a statutory motor fuel tax increase that was also ruled to be exempt from the revenue limit. Other tax increases were approved by voters with language

Page 11

that specifically excluded these revenues from the revenue limit. Over time special interests and politicians became more adept at circumventing the revenue limit, and the courts sanctioned these acts, even when this violated the spirit if not the letter of the constitution. In 1993 the legislature enacted another tax increase earmarked for expenditures for education K-12, without voter approval. Corporate taxes were increased, and corporations, small business, and higher income individuals were for the first time required to pay state income taxes on the federal income taxes they paid.

“This proposal would not affect current state or local funds, future effects would depend upon the actions of the General Assembly and Missouri voters.” (McCarty 2003:20)

Missouri citizens responded to this increased tax burden by attempting to strengthen the constitutional constraints imposed on state fiscal policy. In 1994 citizens placed on the ballot an amendment requiring voter approval for new taxes. The opposition, including the Governor and legislature, were able to include in the ballot language the statement that the proposed Amendment would cause:

In Florida citizens in 1994 gathered enough signatures to place an Amendment on the ballot requiring voter approval for any new tax or tax increase; but, the Florida Supreme Court ruled that the initiative could not be placed on the ballot because it did not pass legal muster as a citizen initiative.

“state and local spending cuts ranging from $1 billion to $5 billion annually. Cuts would affect prisons, schools, colleges, programs for the elderly, job training, highways, public health, and other services.”(McCarty 2003:20) It is not surprising that the proposed Amendment was defeated by a two to one margin. In 1996 citizens were successful in enacting a more stringent Amendment than the one proposed two years earlier. That Amendment requires voter approval for any tax increase that exceeds $50 million, or one percent of state revenues, whichever is less. That Amendment also requires refunds of excess tax revenues if growth in state revenue exceeds the growth of personal income by one percent or more. That Amendment was passed with the support of the new Governor, and the legislature, and the business community. In contrast to the earlier proposed Amendment this new Amendment contained the following language:

In contrast to the original Hancock Amendment, this new TEL has significantly reduced state revenue and spending. Since 1996 Missouri has offset $2.5 billion in surplus revenue with a combination of tax cuts and tax rebates. C. Florida7

The response of the Florida Legislature to this failed citizen initiative was to design an Amendment to preempt another citizen initiative. The legislature placed an Amendment on the ballot to impose a revenue limit linked to the growth of state personal income. The growth rate of state revenue was capped at a five year moving average of state personal income growth. Two Florida legislaprovisions of that TEL would render tors appear to be it completely ineffective as a confiscally prudent straint on the growth of state governwhen in fact they ment in Florida. designed a TEL One provisions excluded several components of revenue from the TEL cap, including: Medicaid matching funds, charges imposed by local and regional governments, and debt service.

sanctioning higher levels of revenue and spending than occurred before the TEL was introduced.

A second provision maintained the growth in the TEL cap regardless whether the cap was reached in the previous year. The revenue cap increases each year with growth in personal income regardless of what is happening to actual revenue. As a result the

Page 12

revenue cap has not and never will constrain the growth of state revenue and spending. Some have described the Florida TEL as a ‘Modern Day Fairy Tale”. They maintain that the Florida Legislature deliberately designed an ineffective TEL to preempt a more stringent TEL from being introduced through citizen initiative. Florida legislators appear to be fiscally prudent when in fact they designed a TEL sanctioning higher levels of revenue and spending than occurred before the TEL was introduced. Clearly the actions of the Florida Legislature and Courts have left TELs as a meaningless exercise in that state. D. Washington8 Washington State joined the tax revolt in 1979 enacting a tax and spending limit through the initiative process, Initiative 62. That Amendment limited increases in state revenue to the rate of growth in personal income. As is often the case with TELs linked to the growth in personal income the Washington Tel was never an effective constraint on the growth of government. In the years after the TEL was enacted the state was able to pass a series of tax increases, and yet remain within the TEL limit. In the early 1990s citizens responded to the increased tax burden by passing a more stringent TEL through the initiative process. Initiative 601 imposed a limit on Statutory TELs the growth in state spending equal are no safer than to inflation plus population growth. the next vote of When the new TEL took effect in the legislature. 1996 legislation was passed to ensure that spending was contained within the TEL limit. Budget cuts of $120 million were enacted reducing expenditures for administration, social services, and prisons. Institutions of higher education were required to trim their benefits by $39 million. In subsequent years, when revenue growth exceeded the limit, tax cuts were enacted to offset the surplus revenue. Washington citizens voted to repeal the motor vehicle tax.

The flaw in the Washington TEL was that it was a statutory rather than a constitutional provision. Statutory TELs are no safer than the next vote of the legislature. In 2000, when the legislature wanted to pass a budget that exceeded the TEL limit, they simply obtained the supermajority vote required to suspend the TEL limit. Since then the Washington TEL has ceased to be an effective constraint on government. E. Michigan9 The Michigan TEL is unique in several respects. The TEL is a constitutional amendment introduced in 1978 that limits the growth of revenue to a fixed ratio of state personal income. That ratio is applied to either personal income in the prior year, or to average personal income in the prior three years, whichever is higher. When actual revenue exceeds the revenue limit by 1% or more, excess revenue is refunded to taxpayers based on the Michigan income tax. What is unique in the Michigan TEL is the formulaic allocation of surplus revenue to and from a budget stabilization or rainy day fund. When annual personal income grows more than 2%, the percentage in excess of 2% is multiplied by total revenue to determine the amount to be automatically transferred from the general fund to the rainy day fund. All general fund surpluses are also transferred into the rainy day fund. Withdrawals from the rainy day fund are also automatic. When personal income growth is less than 0%, the deficiency under 0% is multiplied by total revenue to determine the amount to be transferred from the rainy day fund to the general fund. The legislature can also appropriate money from the rainy day fund if unemployment rises above 8%. In an emergency the legislature can, by supermajority vote, approve emergency transfers from the rainy day fund. Just prior to the recent recession in 2000 the state had accumulated a rainy day fund of $1.2 billion, equal to 13.2% of state spending that year. As you would expect with the accumulation of such a siz-

Page 13

Despite the dissipation of the rainy day fund in response to rent seeking activities by special interest groups, it has served to stabilize the Michigan budget over the business cycle.

able rainy day fund, the fund became a target for special interest groups. There is an ongoing appropriation from the rainy day fund over the period 2000 to 2008 of $32 million for education K-12, and $25 million for water pollution control. Legislators have been able to transfer money from the rainy day fund for a variety of purposes by simply declaring an emergency.

Despite the dissipation of the rainy day fund in response to rent seeking activities by special interest groups, it has served to stabilize the Michigan budget over the business cycle. In the recent recession the state was able to transfer close to $1 billion from the rainy day fund to offset revenue shortfalls. F. Colorado10 Colorado was one of the first states to impose a statutory cap on the growth of state spending. In 1978, a cap of 7% was placed on the growth of general fund expenditures. In the late 1970s some surplus revenue above that limit was rebated to taxpayers. When recession hit in the early 1980’s Colorado, like other states, responded by increasing taxes to balance the budget. That increase in taxes ratcheted up government spending at rates in excess of the growth in state income in subsequent years. The Colorado TEL, like many other statutory TELs, was simply ignored by the legislature. In the late 1980s citizens organized to put a more stringent TEL on the ballot. After several failed attempts the TABOR Amendment was enacted in 1992 through citizen initiative. Tabor is the most stringent TEL introduced in any state. TABOR restricts the growth in state revenue and spending to inflation plus the percentage change in state population. Surplus revenue above that limit must be rebated to taxpayers. The TABOR limit ratchets-down the amount of revenue the state can keep and spend as revenue

falls. The TABOR limit is determined by applying the sum of inflation and population growth to actual TABOR revenues or the TABOR limit, whichever is lower. When revenue falls in a recession, that lower revenue then sets a new base against which the sum of inflation and population growth is applied. The TABOR Amendment also placed a procedural constraint on the power of government to raise taxes. Voter approval is required for any new taxes, tax rate increases, extension of an expiring tax, or tax policy change directly causing a net revenue gain. Voter approval is also required for state and local government to retain and spend revenue in excess of the limit. In a failed attempt to preempt this stringent TEL from being introduced through citizen initiative the legislature enacted a statutory TEL. In 1992 the Arveschaugh-Bird Amendment placed a cap on general fund appropriations equal to the lesser amount of 5% of Colorado personal income in the calendar year two years prior to the start of the fiscal year or 6% over the previous year’s general fund appropriation, with exceptions for federal mandates and court orders. Over time the Legislature has interpreted the 6% statutory cap as a floor rather than a ceiling on the growth in general fund expenditures. The reason is that the general fund spending in a given year determines the base against which the spending cap is applied in determining the amount of general fund spending permitted in the following year. The distortions this introduces in state spending have become very evident in recent years. For the foreseeable future the Arveschaugh-Bird amendment will have little if any impact on state fiscal policies. This is because the 6% limit on the growth in general fund expenditures will not be a binding constraint on the legislature. That limit is significantly above the limit imposed by the Tabor Amendment

Initially the TABOR Amendment was non-binding because the growth in state revenues was less than the TABOR limit. The first year in which TABOR became a binding constraint was 1997. Over the next five years more than $3 billion in surplus revenue was either rebated to taxpayers, or offset by tax reduction. A referendum to spend a portion of the surplus revenue was defeated by taxpayers.

tax cuts. Most of the TABOR surplus has been generated by the income and sales taxes; and some of that surplus has been offset by rebates and reductions in the income and sales taxes. However, the increased use of targeted tax cuts and tax rebates to benefit narrow interest groups means that less of the surplus is refunded to the people who paid the excess taxes.

The Legislature has also chosen to interpret the TABOR limit as a floor rather than a ceiling on the revenue that the state can keep and spend. This introduces distortion and inefficiency in state finance, and the problem has become especially evident in the current fiscal crises. Even if it might be prudent for the Legislature to hold revenue and spending growth below that permitted by the TABOR limit, the Legislature has often failed to do so.

A major problem emerged when a constitutional provision was introduced requiring constant growth in expenditures for education K-12 from income tax revenue earmarked for that purpose and exempt from the TABOR limit. This Amendment, which requires a constant ratcheting up of expenditures for education K-12, combined with the TABOR Amendment, which ratchets government revenue down, places the legislature in an untenable position. The legislature has discretion over less than one third of the state budget because more than two thirds is mandated by the required expenditures for education K-12, Medicaid, and prisons. When the recession resulted in a revenue shortfall the legislature was required to make draconian cuts in higher education and social services. In the long run these provisions in the fiscal constitution will result in a structural deficit, which is prohibited by the balanced budget provision of the Colorado Constitution. The Colorado legislature is now considering changes in each of these constitutional provisions in order to avoid a future fiscal crises.11 Colorado is currently experiencFinally, in Colorado TELs have been ing the worst more effective in constraining local fiscal crises in government than state government. decades. The result is that financing for edu-

The Legislature has interpreted TABOR so as to erode the constraints imposed by the limit, and to exacerbate the fiscal crises resulting from the current recession. The Legislature chose to retain the surplus revenue generated in the current year in the general fund reserve, and to finance rebates from the surplus revenue generated in the following year. This scheme worked fine when revenues were rising in the boom years; but the flaw in this scheme became apparent in the current fiscal crises. With recession the TABOR surplus disappeared, and the state had to finance a rebate from reductions in current state spending. The decision of the Legislature to finance taxpayer rebates from revenues received in the year after the TABOR surplus is generated has exacerbated the fiscal crises. The Legislature has also chosen to increase the TABOR limit to adjust for undercounting of population growth in the last decade. This permitted the state to retain and spend more revenue over the coming decade. A different set of problems has resulted from the way in which the legislature has chosen to offset surplus revenue More than 20 bills have been passed offsetting the TABOR surplus with tax rebates and

cation K-12 has shifted significantly from local government to the state government in Colorado. The Tabor Amendment requires the state to backfill local governments whenever state legislation has a negative impact on local government revenues. When the state enacted a cut in the business personal property tax it was required to backfill local governments for the loss of revenue resulting from this tax cut. Despite these

provisions, the financing and administration of some programs, such as education K-12, has shifted from local government to the state governWhile tax and ment. spending limits have been effecColorado is currently experiencing tive in slowing the worst fiscal crises in decades. the growth in While tax and spending limits have state government been effective in slowing the growth in the long run, in state government in the long run, they have not they have not been very effective in been very effecsmoothing the growth of state revtive in smoothing enue and spending over the business the growth of cycle. Colorado does not have a true state revenue and budget stabilization fund. The current spending over the fiscal crises reveals the need for such business cycle. revenue and expenditure smoothing in state government during periods of recession and slower economic growth. Legislation has been introduced in Colorado that would link the TABOR limit to both an emergency reserve fund, and a budget stabilization fund. This legislation is similar to the model TEL in the Appendix to this study.

IV. THE NEXT GENERATION OF TELS12 New TELs have been proposed in half a dozen states over the past year.13 This next generation of TELs depart considerably from the TELs now in place. It is possible to introduce a TEL to constrain the growth of government in the long run that does not serve to stabilize the budget over the business cycle. That is in fact the nature of Colorado’s TABOR Amendment. On the other hand, a TEL designed to stabilize the budget over the business cycle, may or may not constrain the growth of government in the long run. For example, Michigan’s TEL has been used to stabilize the budget over the business cycle, but it is not clear that it has constrained the growth of government. If all of the surplus revenue generated in periods of rapid growth is returned to the general fund to finance spending and to offset revenue shortfalls in

periods of recession, then the TEL will do little to constrain government spending in the long run. Thus, we can think of a tradeoff in the allocation of surplus revenue above the TEL limit. The larger the share of that revenue returned to taxpayers through tax cuts and tax rebates, the less of that surplus revenue is available to stabilize the The next genbudget over the business cycle, and eration of TELs the more that TEL will constrain the is designed to growth of government. The larger the achieve an optishare of the surplus revenue set aside mum tradeoff in a Budget Stabilization Fund the between conmore effective the TEL will be in offstraining the setting revenue shortfalls, but, the less growth of governeffective it will be in constraining the ment and stabigrowth of government in the long run. lizing government Understanding this tradeoff is crucial budgets over the in designing a TEL. business cycle. The next generation of TELs is designed to achieve an optimum tradeoff between constraining the growth of government and stabilizing government budgets over the business cycle. One set of rules is required to introduce an effective brake on the revenue the government can spend, i.e. a tax and spending rule (TEL). A second set of rules must be introduced governing how the surplus revenue above the TEL is allocated to the budget stabilization fund. A third set of rules is required for allocating money from the budget stabilization fund to the general fund to offset revenue shortfalls and stabilize the budget during recession. A. Rules Defining the TEL Limit The states have experimented with a variety of limits in designing their TELs. The most widely used limit is a measure of the rate of growth of income over some historical period. There are a number of criticisms of this type of TEL as a constraint on the growth of government. This limit tends to be less stringent compared to other limits, resulting in a more rapid growth of government. Depending upon how the limit is applied it may have little or no impact on the growth of government.

In the present context, the criticism of a TEL limit based on the growth in personal income over an historical time period is that it is less effective in stabilizing the budget over the business cycle. Such a limit permits a very rapid rate of growth in revenue and spending in periods of rapid economic growth. Even when the limit is triggered it does so only with some lag. As a result this type of limit is less effective than one that has no lagged impact; indeed it is possible that such a limit could actually be pro-cyclical. Recent studies have demonstrated that the most stringent TELs define the limit as the sum of inflation and population growth. While TEL limits defined in terms of the growth of population and inflation have proven effective in constraining the growth of government, they have proven less effective in stabilizing the budget over the business cycle. If such TELs ratchet down revenue and spending in periods of recession, they may also be pro-cyclical.

C. Rules Linking the TEL to a Budget Stabilization Fund The basic principle of a rules based budget stabilization fund is straightforward. A good analogy is a hybrid energy vehicle (HEV). An HEV has a braking system such that surplus energy created by the motor when brakes are applied is stored in a battery, and then used to power the motor when the car is operating at a slower speed. Rules governing a budget stabilization fund operate much like an HEV. A brake is applied to limit the amount of revenue the state spends in periods of rapid economic growth. The surplus revenue is accumulated in a budget stabilization fund, and used to offset revenue shortfalls in periods of recession. The rules governing how money is transferred into and out of the budget stabilization fund are extremely important because they determine the tradeoff between constraining the growth of government and stabilizing the budget over the business cycle.

B. Rules Defining the TEL Base. The states have also experimented with different revenue and expenditure bases against which the TEL limit is applied. In general it is fair to say that the broader the base the more effective the TEL. If interest groups are able to exempt portions of the revenue and or expenditures from the TEL limit it becomes less effective both in constraining government and in stabilizing the government budget over the business cycle. A broader base also limits the economic distortions introduced by such privilege seeking. It is also fair to say that the use of actual historical measures of revenue and expenditures has proven more effective than the use of projected measures. Historical data from the most recent budget cycle is easily obtained and does not rely upon projections of revenue or expenditures that are subject to error and bias.

At this point we assume that the budget stabilization fund is clearly distinguished from other funds that might be created. The money set aside in the budget stabilization fund is used for only one purpose and that is to stabilize the budget over the business cycle. The TEL limit is held constant whenever there is a revenue shortfall, and becomes a binding constraint only when revenue recovers to that limit. D. Rules for Allocating Surplus Revenue Into the Budget Stabilization Fund If we assume that an effective TEL is in place, then the opportunity cost of setting aside surplus revenue above the TEL limit in a budget stabilization fund is the return of that surplus revenue to taxpayers through tax rebates and tax cuts. The more of the surplus revenue returned to taxpayers the lower the rate of growth of government relative to the private sector in the long run. Allocating some of that revenue to a budget stabilization fund provides the funds to stabilize the budget, but also reduces the

constraint imposed by the TEL on the growth of government. Rules must establish how much of the surplus revenue to allocate to the budget stabilization fund at a given point in time; and what, if any, cap should be placed on the size of the budget stabilization fund. Wall street and bond rating agencies often suggest an appropriate size for a budget stabilization fund in the range of 6 to 7 percent of the state budget. But, each state is unique in its preferences and requirements for a budget stabilization fund, and it is fair to say that there are no hard and fast rules.. E. Rules for Allocating Money from the Budget Stabilization Fund to the General Fund A rules-based budget stabilization fund would require the allocation of money from the budget stabilization fund to the general fund whenever there is a revenue shortfall. These rules are also important because they determine the tradeoff between constraining the budget over the long run, and stabilizing the budget over the business cycle. If the primary emphasis is on budget stabilization then money would be transferred equal to the revenue shortfall until all of the funds in the budget stabilization fund are exhausted. If the emphasis is on constraining the growth of government in the long run, then only a portion of the budget stabilization fund would be transferred into the general fund at a given point in time. In the latter case a revenue shortfall would also require reductions in government spending. Voter surveys suggest that money should be transferred from the budget stabilization fund to offset some, but not all, of the revenue shortfall. For example, the rule might require that only half of the revenue shortfall be offset by money transferred from the budget stabilization fund in any given fiscal year, with the other half offset by budget cuts. Such a decision rule might mean that all of the budget stabilization fund would not be exhausted in any given recession to offset revenue shortfalls.

APPENDIX: THE NEXT GENERATION OF TELS - A MODEL TEL The author has been consulting with a number of states on legislation to introduce, or modify, TELs along the lines of the previous discussion of the next generation of TELs.14 The legislation proposing a new TEL for Kansas is included here as an appendix, because it is the most carefully crafted TEL embodying these principles.15

WORKING DRAFT NO. 5 December 22, 2003

3rs1247 DRAFT CONCURRENT RESOLUTION NO. ____

By

A PROPOSITION to amend the constitution of the state of Kansas by adding a new article thereto, prescribing certain limits upon taxes, revenues and expenditures by the state and local governments.

Be it resolved by the Legislature of the State of Kansas, two-thirds of the members elected (or appointed) and qualified to the House of Representatives and two-thirds of the members elected (or appointed) and qualified to the Senate concurring therein: Section 1

. The following proposition to amend the constitution of the state of Kansas shall be submitted to

the qualified electors of the state for their approval or rejection: The constitution of the state of Kansas is amended by adding a new article thereto to read as follows: 2 ”Article16. TAX, REVENUE AND EXPENDITURE LIMITATIONS ON STATE AND LOCAL GOVERNMENTS.

” § 1. Definitions. As used within this article:

(a) ”State” means the state government including all branches, state offices, authorities, agencies, boards, commissions, institutions, instrumentalities and any division or unit of state government which are directly supported with tax funds, except that ”state” does not include any enterprise;

(b) ”local government” means any county, township, city, education district, other special district and any other taxing district or political subdivision of Kansas which is directly supported by tax funds, except that ”local government” does not include any enterprise;

(c) ”enterprise” means a state-owned or local government-owned business authorized to issue its own revenue bonds and receiving less than 10% of annual revenue in grants or other direct cash benefit from the state and local governments combined;

(d) ”bond” means any bond, note, debenture, interim certificate, grant and revenue anticipation note, lease-purchase agreement, lease certificate of participation or other evidence of indebtedness which, in any such case, is entered into or establishes a debt obligation for longer than one fiscal year, whether or not the interest on which is subject to federal income taxation;

(e) ”fiscal year” means the twelve-month fiscal period prescribed by law for the state or the local government;

(f) ”fiscal year spending” means all expenditures and reserve increases except, as to both, expenditures for refunds of any kind or expenditures of moneys received from the federal government, moneys received as grants, gifts or donations which are to be expended for purposes specified by the donor, moneys that are collections for another government, moneys received for pension contributions by employees and pension fund earnings, reserve transfers or expenditures;

(g) ”inflation” means the change expressed as a percentage in the consumer price index for the Kansas City metropolitan area, all goods, all urban consumers, as officially reported by the bureau of

labor statistics of the United States department of labor, or its successor index; and

(h) ”population” means the more recent of either the periodic census conducted by the United States department of commerce or its successor agency or the annual update of such census as prescribed by the legislature by law, which shall be adjusted every decade to match the federal decennial census.

(i) ”education district” means each school district, vocational or technical school, community college, technical college, municipal university, and any other public educational entity established as provided by law, except that ”education district” does not include any state educational institution under the control and supervision of the state board of regents.

(j) ”total state revenue” means all moneys received by the state from any source except any of the following:

(1) Moneys received as grants, gifts or donations which are to be expended for purposes specified by the donor;

(2) moneys received from the federal government; and

(3) moneys which are income earned on moneys in permanent endowment funds, trust funds, deferred compensation funds or pension funds and which are credited to such funds;

(k) ”total local government revenue” means all moneys received by the local government from any source except any of the following:

(1) Moneys received as grants, gifts or donations which are to be expended for purposes specified by the donor;

(2) moneys received from the federal government; and

(3) moneys which are income earned on moneys in permanent endowment funds, trust funds, deferred compensation funds or pension funds and which are credited to such funds.

§ 2. Prior Elector Approval for Tax Increases or Issuance of Certain Bonds. (a) For any fiscal year that commences on or after July 1, 2005, the state must have approval of the electors in advance (1) for any new state income, sales or other excise tax rate increase before the state tax rate increase can take effect, (2) for any state mill levy ad valorem property tax rate increase above the state mill levy ad valorem property tax rate for the prior year before the state mill levy ad valorem property tax rate increase can take effect, (3) for any extension of any expiring state income, sales or other excise tax or expiring state ad valorem property tax before the extension can take effect, or (4) for any state tax policy change enacted into law by the state which would directly cause a net tax revenue gain to the state or local government, before such tax policy change can take effect.

(b) For any fiscal year that commences on or after July 1, 2005, each local government must have approval of the electors in advance (1) for any new local government income, sales or other excise tax rate increase, which is authorized by law, before the local government tax rate increase can take effect, (2) for any local government mill levy ad valorem property tax rate increase, which is authorized by law, before the local government mill levy ad valorem property tax rate increase can take effect, (3) for any extension of any expiring local government income, sales or other excise tax or local government ad valorem property tax, which is authorized by law, before the extension can take effect, or (4) for any local government tax policy change authorized by law and adopted in a resolution or ordinance by a local government which would directly cause a net tax revenue gain to the local government, before such tax policy change can take effect.

(c) For any fiscal year that commences on or after July 1, 2005, the state and each local government must have approval of the electors before authorizing any bonds, except for refinancing existing bonded debt at a lower interest rate.

(d) The legislature shall provide by law for the manner of submitting matters subject to approval under this section to the electors.

§ 3. Spending and Revenue Limits. (a) For any state fiscal year that commences on or after July 1, 2005, fiscal year spending by the state shall not increase above the fiscal year spending for the preceding state fiscal year by more than the maximum percentage increase determined pursuant to this section. The maximum percentage increase in fiscal year spending for a state fiscal year shall be equal to the result

obtained by adding the rate of inflation for the calendar year ending during the preceding state fiscal year, plus the percentage change in state population during the calendar year ending during the preceding state fiscal year if a positive number, adjusted for revenue changes approved by electors under section 2 of this article.

(b) If the amount of the total state revenue for the preceding state fiscal year exceeds the amount of total state revenue for the second preceding state fiscal year, the total state revenue limitation for a state fiscal year shall be the result obtained by adding (1) the lesser of (A) the amount of total state revenue for the preceding state fiscal year or (B) the amount of the total state revenue limitation for the preceding state fiscal year, and (2) the product of (A) the amount determined under clause (1) of this subsection, and (B) the sum of (i) the rate of inflation for the calendar year ending during the preceding state fiscal year, plus (ii) the percentage change in state population during the calendar year ending during the preceding state fiscal year if a positive number.

(c) If the amount of the total state revenue for the preceding state fiscal year is less than the amount of total state revenue for the second preceding state fiscal year, the amount of the total state revenue limitation for a state fiscal year shall be the lesser of (1) the amount of total state revenue for such state fiscal year, or (2) the amount of the total state revenue limitation for the most recent state fiscal year for which the amount of total state revenue exceeded the amount of total state revenue for the preceding state fiscal year.

(d) For any local government, other than an education district, for any local government fiscal year that commences on or after July 1, 2005, fiscal year spending by the local government, other than an education district, shall not increase above the fiscal year spending for the preceding local government fiscal year by more than the maximum percentage increase determined pursuant to this section. For any fiscal year that commences on or after July 1, 2005, the maximum percentage increase in fiscal year spending by a local government, other than an education district, equals the sum of inflation during the preceding calendar year plus the percentage change in the local government population during the preceding calendar year if a positive number, adjusted for revenue changes approved by the electors under section 2 of this article.

(e) For any education district fiscal year that commences on or after July 1, 2005, fiscal year spending by the education district shall not increase above the fiscal year spending for the preceding education district fiscal year by more than the maximum percentage increase determined pursuant to this section.

For any fiscal year that commences on or after July 1, 2005, the maximum percentage increase in fiscal year spending by an education district equals inflation during the preceding calendar year, plus the percentage change in its pupil or student enrollment for the school year commencing during the preceding calendar year as compared to the school year preceding that school year if a positive number, adjusted for revenue changes approved by the electors under section 2 of this article.

(f) If the amount of the total local government revenue for the preceding local government fiscal year exceeds the amount of total local government revenue for the second preceding local government fiscal year, the total local government revenue limitation for a local government fiscal year shall be the result obtained by adding (1) the lesser of (A) the amount of total local government revenue for the preceding local government fiscal year or (B) the amount of the total local government revenue limitation for the preceding local government fiscal year, and (2) the product of (A) the amount determined under clause (1) of this subsection, and (B) the sum of (i) the rate of inflation for the calendar year ending during the preceding local government fiscal year, plus (ii) in the case of a local government other than an education district, the percentage change in local government population during the calendar year ending during the preceding local government fiscal year if a positive number, or (iii) in the case of an education district, the percentage change in its pupil or student enrollment for the school year commencing during the preceding calendar year as compared to the school year preceding that school year if a positive number.

(g) If the amount of the total local government revenue for the preceding local government fiscal year is less than the amount of total local government revenue for the second preceding local government fiscal year, the amount of the total local government revenue limitation for a local government fiscal year shall be the lesser of (1) the amount of total local government revenue for such local government fiscal year, or (2) the amount of the total local government revenue limitation for the most recent local government fiscal year for which the amount of total local government revenue exceeded the amount of total local government revenue for the preceding local government fiscal year.

(h) The legislature, by law, shall provide a mechanism to adjust the amount of a limitation under this section to reflect any subsequent transfer of all or any part of the cost of providing a governmental function. The mechanism shall adjust the amount of a limitation so that total costs are not increased as a result of the transfer. The adjustment mechanism provided for in this subsection shall be used in determining a limitation under this section beginning with the fiscal year immediately following the transfer.

(i) The legislature, by law, shall provide a mechanism to adjust the amount of a limitation under this section to reflect the cost of providing a governmental function on account of any subsequent annexation, creation of a new governmental unit, or any consolidation or change in the boundaries of a governmental unit. The mechanism shall adjust the amount of limitation so that total costs are not increased as a result of the annexation, creation, consolidation or change in boundaries. The adjustment mechanism provided in accordance with this subsection shall be used in determining a limitation under this section beginning with the fiscal year immediately following the annexation, creation of a new governmental unit, or consolidation or change in the boundaries of a governmental unit.

(j) In the case of the state or any local government having a fiscal year ending on June 30, 2005, for the purposes of determining total revenue limitations under this section for the state or any such local government, the total authorized fiscal year expenditures for the fiscal year ending on June 30, 2004, shall be construed to be the total revenue limitation for that preceding fiscal year and the total authorized fiscal year expenditures for the fiscal year ending on June 30, 2005, shall be construed to be the total revenue limitation for that preceding fiscal year. In the case of any local government having a fiscal year ending on December 31, 2005, for the purposes of determining total local government revenue limitations under this section for any such local government, the total authorized fiscal year expenditures for the fiscal year ending on December 31, 2004, shall be construed to be the total revenue limitation for that preceding fiscal year and the total authorized fiscal year expenditures for the fiscal year ending December 31, 2005, shall be construed to be the total revenue limitation for that preceding fiscal year.

§ 4. Emergency Reserve Funds. (a) For any state fiscal year that commences on or after July 1, 2005, if revenue from sources not excluded from total state revenue exceeds the total state revenue limitation for that state fiscal year and subject to the other provisions of this section, a portion of total state revenue in excess of the total state revenue limitation, determined in accordance with section 3 of this article, shall be transferred in the amount and in the manner prescribed by the legislature by law to the emergency reserve fund, which fund is hereby created in the state treasury, to the extent necessary to ensure that a balance of the emergency reserve fund at the end of the state fiscal year is an amount equal to not more than 3% of the total state revenue limitation for the ensuing state fiscal year. Any amount required to be maintained in the ending balance of the state general fund as provided by law shall be excluded from the amount available for transfer to the emergency reserve fund of the state by this section. Each transfer to the emergen-

cy reserve fund of the state prescribed by this section shall be made before making any transfer to the budget stabilization reserve fund as provided in section 5 of this article or any refunds as required by section 6 of this article. The state shall not be required to transfer any moneys other than any amount of total state revenue in excess of the total state revenue limitation to the emergency reserve fund. The moneys in the emergency reserve fund shall be in addition to, and shall not be used to meet, any other reserve requirement under this constitution or any law. In no case shall additional moneys be transferred to the emergency reserve fund if the balance in the emergency reserve fund is more than 3% of the total state revenue limitation for the ensuing state fiscal year.

(b) Moneys in the emergency reserve fund of the state may be expended only for emergencies declared by law. Two-thirds (2/3) of the members then elected (or appointed) and qualified in each house, voting in the affirmative, shall be necessary to declare an emergency within the state of Kansas as a whole and to pass any bill making an appropriation of money from the emergency reserve fund. Income earned on the emergency reserve fund of the state shall accrue to the fund.

(c) For each local government fiscal year that commences on or after July 1, 2005, each local government shall reserve and maintain in an emergency reserve fund an amount equal to not more than 3% of its total local government revenue limitation for the ensuing local government fiscal year in accordance with this section. For any local government fiscal year that commences on or after July 1, 2005, if total local government revenue from sources not excluded from total local government revenue exceeds the total local government revenue limitation for that local government fiscal year and subject to the other provisions of this section, a portion of total local government revenue in excess of the total local government revenue limitation, determined in accordance with section 3 of this article, shall be transferred in the amount and in the manner prescribed by the legislature by law to the emergency reserve fund of the local government. Any amount required to be maintained in the ending balance of the general fund of the local government as provided by law shall be excluded from the amount available for transfer to the emergency reserve fund of the local government by this section.

(d) Moneys in the emergency reserve fund of a local government may be expended only for emergencies within a local government declared by the local government in the manner prescribed by law. Income earned on the emergency reserve fund of the local government shall accrue to the fund.

(e) As used in this section ”emergency” means an extraordinary event or occurrence that could not have been reasonably foreseen or prevented and that requires immediate expenditures to preserve the health, safety and general welfare of the people within a local government or the state and ”emergency” does not mean a revenue shortfall or budget shortfall.

§ 5. Budget Stabilization Reserve Funds. (a) For any state fiscal year that commences on or after July 1, 2005, if total state revenue exceeds the total state revenue limitation for that state fiscal year, then the remaining excess amount, after making any transfer to the emergency reserve fund as required by section 4 of this article, shall be reserved as prescribed by this section or refunded as prescribed by section 6 of this article, subject to the other provisions of this section. Any amount required to be maintained in the ending balance of the state general fund as provided by law shall be excluded from the amount available for transfer to the budget stabilization reserve fund of the state by this section.

(b) After any amount required to be transferred to the emergency reserve fund of the state pursuant to section 4 of this article has been transferred, an amount of any remaining excess amount of total state revenue shall be transferred in the amount and in the manner prescribed by the legislature by law to the budget stabilization reserve fund, which fund is hereby created in the state treasury. The amount transferred to the budget stabilization reserve fund in accordance with this subsection shall be equal to the lesser of (1) the amount necessary to ensure that the balance in the budget stabilization reserve fund at the end of the state fiscal year is an amount equal to 10% of the total state revenue limitation for the ensuing state fiscal year, or (2) the amount equal to 50% of any such remaining excess amount of total state revenue. Income earned on the budget stabilization reserve fund shall accrue to the fund. In no case shall additional moneys be transferred into the budget stabilization reserve fund if the balance in the fund is equal to or more than 10% of the total state revenue limitation for the ensuing state fiscal year.

(c) For any state fiscal year that commences on or after July 1, 2005, if the amount of the total state revenue is less than the amount of total state revenue for the prior state fiscal year, the legislature shall provide by law for the transfer of moneys from the budget stabilization fund to the state general fund in an amount equal to not more than the difference between the amount of total state revenue for the prior state fiscal year and the amount of total state revenue for the state fiscal year. Under no other circumstances shall moneys be transferred or expended from the budget stabilization fund of the state.

(d) For each local government fiscal year that commences on or after July 1, 2005, each local government shall reserve and maintain in a budget stabilization reserve fund an amount equal to not more than 10% of its fiscal year estimated spending in accordance with this section. For any local government fiscal year that commences on or after July 1, 2005, if total local government revenue from sources not excluded from total local government revenue exceeds the total local government revenue limitation for that local government fiscal year and subject to the other provisions of this section, a portion of total local government revenue in excess of the total local government revenue limitation, determined in accordance with section 3 of this article, shall be transferred in the amount and in the manner prescribed by the legislature by law to the budget stabilization reserve fund of the local government. Any amount required to be maintained in the ending balance of the general fund of the local government as provided by law shall be excluded from the amount available for transfer to the budget stabilization reserve fund of the local government by this section. No amount shall be transferred to the budget stabilization reserve fund of a local government until after any amount required to be transferred to the emergency reserve fund of the local government pursuant to section 4 of this article has been transferred.

(e) For any current local government fiscal year that commences on or after July 1, 2005, if the amount of the total local government revenue is less than the amount of total revenue for the prior fiscal year, the local government shall provide for the transfer of moneys from the budget stabilization fund to the general revenue fund of the local government in an amount equal to not more than the difference between the amount of total local government revenue for the prior local government fiscal year and the amount of total revenue for the current local government fiscal year. Under no other circumstances shall moneys be transferred or expended from the budget stabilization fund of the local government.

§ 6. Disposition of Excess Revenues. (a) Any excess amount of total state revenues for a state fiscal year that remains after the transfers to the emergency reserve fund and budget stabilization reserve fund pursuant to section 4 or section 5 of this article, if any, shall be reserved in the current state fiscal year and shall be refunded as provided by law during the next ensuing state fiscal year to the taxpayers who paid the state ad valorem property taxes or state income, sales or other excise taxes for or during the preceding state fiscal year, in a manner that is proportional, on a pro rata basis, to the manner in which such taxes were collected from such taxpayers. Any amount required to be maintained in the ending balance of the state general fund as provided by law shall be excluded from the amount available to be reserved and refunded by the state as prescribed by this section.

(b) Any excess amount of local government revenues for a local government fiscal year that remains after transfers to the emergency reserve fund and budget stabilization reserve fund of the local government pursuant to section 4 or section 5 of this article, if any, shall be reserved in the current local government fiscal year and shall be refunded during the next ensuing local government fiscal year to the taxpayers who paid the ad valorem property taxes or income, sales or other excise taxes of the local government for or during the preceding local government fiscal year, in a manner that is proportional, on a pro rata basis, to the manner in which such taxes were collected from such taxpayers. Any amount required to be maintained in the ending balance of the general fund of the local government as provided by law shall be excluded from the amount available to be reserved and refunded to taxpayers by the local government as prescribed by this section.

(c) In a case of any amount that is received pursuant to any tax and required to be reserved and refunded to taxpayers by the state or a local government pursuant to this section and that is determined by the state or local government in the manner prescribed by law to be insufficient for refunds to be made during the ensuing state or local government fiscal year, as the case may be, such amount shall be reserved for refunds to be made thereafter when the amount reserved is sufficient therefor.

§ 7. State Temporary Borrowing. On or after July 1, 2005, during any state fiscal year, transfers which are temporary and are to be repaid, or any other temporary borrowing, through certificates of indebtedness or any other device or manner, of any moneys in the state treasury to be credited to the state general fund, are prohibited unless the moneys so transferred or otherwise borrowed are restored or repaid to the original funds or accounts of the state treasury from the state general fund within the same state fiscal year. The provisions of this section do not apply to transfers from the emergency reserve fund or the budget stabilization reserve fund to the state general fund in accordance with this article.

§ 8. General Revenue Supplanting. (a) On or after July 1, 2005, any appropriation of moneys in the state treasury that either supplants any appropriation from the state general fund, or that, if not made, would require an appropriation from the state general fund is prohibited. For purposes of this subsection, any appropriation of moneys in the state treasury that is funded by user charges or fees imposed on goods or services that do not exceed the cost of the goods or services provided shall not be deemed to be an appropriation that supplants any appropriation from the state general fund.

(b) For any local government fiscal year that commences on or after July 1, 2005, no local government shall impose or shall increase user charges or fees imposed for goods or services provided by the local government to supplant general revenues of the local government. For the purposes of this subsection, any imposition or increase of user charges or fees imposed on goods or services that do not exceed the cost of the goods or services provided shall not be deemed to be supplanting general revenues of the local government.

§ 9. State Mandates on Local Governments. A local government may not be required to fulfill any mandate imposed by the state unless and until, and may be required to fulfill that mandate only to the extent that, funds are provided to the local government by the state for that purpose. The legislature is not required to appropriate funds for mandates if more than two years have passed since the effective date of the mandate and no claim for funding has been made by local government during that period.

§ 10. Construction and Enforcement. (a) The provisions of this article shall be liberally construed for the purpose of effectuating the purposes thereof, except that nothing in this article shall be construed to authorize any new or increased tax of any kind other than as provided or authorized by law enacted by the legislature in accordance with and subject to the other provisions of this constitution.

(b) In any case of a conflict between any provision of this article and any other provision contained in the constitution, the provisions of this article shall control.

(c) All laws in force at the time of the adoption of this amendment and consistent therewith shall remain in full force and effect until amended or repealed by the legislature. The legislature shall repeal or amend all laws inconsistent with the provisions of this article to conform with the provisions of this article.

(d) Any individual or class of individuals shall have standing to bring a suit to enforce this article. A court of record shall award a successful plaintiff costs and reasonable attorney fees in the suit, but may not allow the state or a local government to recover costs and reasonable attorney fees unless a suit against it is ruled frivolous.” 1 Sec. . The following statement shall be printed on the ballot with the amendment as a whole:

”Explanatory statement. This amendment would establish requirements for prior voter approval of tax increases and certain bond issuances by the state and local governments, would place limitations on increases in state and local government spending and provide for the disposition of excess state revenues by establishing certain reserve funds and prescribing certain tax refunds by the state.

”A vote for this proposition would . . . I. ”A vote against this proposition would . . .”

Sec. . This resolution, if approved

by two-thirds of the members elected (or appointed) and qualified to the House of Representatives, and two-thirds of the members elected (or appointed) and qualified to the Senate shall be entered on the journals, together with the yeas and nays. The secretary of state shall cause this resolution to be published as provided by law and shall cause the proposed amendment to be submitted to the electors of the state at the general election to be held on November 2, 2004, unless a special election is called at a sooner date by concurrent resolution of the legislature, in which case it shall be submitted to the electors of the state at the special election.

BIBLIOGRAPHY Abrams, B.A. and W. R. Dougan (1986). “The Effects of Constitutional Restraints on Government Spending.” Public Choice 49:223-38. Bails, Dale G. (1982). “A Critique on the Effectiveness of Tax-Expenditure Limitations.” Public Choice 38:129-138. Bails, Dale G. (1990). “The Effectiveness of Tax-Expenditure Limitations: A Reevaluation.” American Journal of Economics and Sociology 49: 223-38. Bails, Dale G. and Margie A. Tieslau (2000). “The Impact of Fiscal Constitutions on State and Local Expenditures.” Cato Journal 20:255-277. Brennan, Geoffrey and James M. Buchanan (1980). The Power to Tax: Analytical Foundations of a Fiscal Constitution. Cambridge: Cambridge University Press. Broder, David S. (2000). Democracy Derailed: Initiative Campaigns and the Power of Money. Money New York: Harcourt. Buchanan, James M. and Richard E. Wagner (1977). Democracy in Deficit. Deficit New York: Academic Press. Chicoine, David L. and Norman Walzer (1986). “Factors Affecting Property Tax Reliance: Additional Evidence.” Public Choice 49: 17-28. Cox, J and D. Lowery (1990). “The Impact of the Tax Revolt Era State Fiscal Caps.” Social Science Quarterly 71: 492-509. Dougan, W. (1988). “The Effects of Tax or Expenditure Limits on State Government.” Working Paper No. 54, Center for the Study of the Economy and the State. Elders, Harold W. (1992). “Exploring the Tax Revolt: An Analysis of the Effectiveness of State Tax

and Expenditure Limitation Laws.” Public Finance Quarterly 20:47-63. Flowers, Marilyn R. (1977). “Multiple Tax Sources, Voting Equilibrium, and Budget Size.” Public Finance 32:210-224. Hettich, Walter and Stanley Winer (1984). “A Positive Model of Tax Structure.” Journal of Public Economics 24: 67-87. Holcombe, Randall G. (2001). “Tax and Expenditure Limitations: Issues for Florida.” Backgrounder Policy Report #32, The James Madison Institute. Joyce, Phillip G. and Daniel R. Mullins (1991). “The Changing Fiscal Structure of the State and Local Public Sector: The Impact of Tax and Expenditure Limitations.” Public Administration Review 51:240253. Kenyon, Daphne and Karen M. Benker (1984). “Fiscal Discipline: Lessons From the State Experienced.” National Tax Journal 37:437-446. McCary, Ray (2003). 2003 Fiscal Review of Missouri State Government. Missouri Chamber of Commerce and Industry. New, Michael J. (2001). “Limiting Government Through Direct Democracy: the Case of State Tax and Expenditure Limitations.” Policy Analysis No. 420, Cato Institute. New, Michael J. (2003). “Proposition 13 and State Budget Limitations: Past Successes and Future Options.” Briefing Paper No. 83, Cato Institute. Owens, Bill (2003). “Hasta La Vista, Deficit.” Wall Street Journal, Thursday Oct 16. Poterba, James M. (1994). “State Responses to Fiscal Crises: The Effects of Budgetary Institutions and Politics.” Journal of Political Economy 102: 799821.

Poterba, James M. (1996). “Do Budget Rules Work?” NBER Working Paper No. 5550, Cambridge, Mass., National Bureau of Economic research. Poulson, Barry W. and Jay Kaplan (1994). “A Rent Seeking Model of TELs.” Public Choice 79:117-134. Poulson, Barry W.(1999). “Surplus Expenditures in Colorado.” Issue Paper 10-1999, Independence Institute. Poulson, Barry W. (2001a). “Learning to Live With Colorado’s Tax and Spending Limits.” Issue Paper 9-2001, Independence Institute. Poulson, Barry W. (2001b). “Surplus Expenditures: A Case Study of Colorado.” Public Budgeting and Finance Winter 2001: 18-43. Poulson, Barry W. (2003a). “Fiscal Crises in Colorado.” Issue Paper 2-2003 Independence Institute. Poulson, Barry W. (2003b). “A Constitutional Impasse: Gallagher, Bird Arveschaug, TABOR and Amendment 23.” Article 2003-c, Independence Institute. Poulson, Barry W. (2003c). “Creating a Rainy Day Fund for Colorado.” Mimeo, Treasurer’s Advisory Group on Constitutional Amendments. Poulson, Barry W. 2003d) “Putting State fiscal Policies on Cruise Control.” State Factor American Legislative Exchange Council Washington, D.C. Public Interest Institute (2002) The Constitution of the State of Michigan Article IX.

Reubens, Kim S. (1995). “Tax Limitations and Government Growth: The Effect of Tax and Expenditures Limits on state and Local Government.” Mimeo, MIT Department of Economics. Riker, William (1980). Implications for the Disequilibrium of Majority Rule for the Study of Institutions.” American Political Science Review 74: 432-446. Sjonquist, David L. (1981). “A Median Voter Analysis of Variations in the Use of Property Taxes Among Local Government.” Public Choice 36:273285. Shadbegian, Ronald J. (1996). “Do Tax and Expenditure Limitations Affect the Size and Growth of State Government.” Contemporary Economic Policy 14: 22-35. Spindler, Zane A. (1990). “A rent seeking Perspective on Privatization.” North American Review of Economics and Finance 1: 87-105. Stansel, Dean (1994). “Missouri’s Hancock II Amendment: the Case for Real Reform.” Briefing Paper No. 20, Cato Institute. Treasurer’s Advisory Group on Constitutional Amendments (2003) Reports and Proposals From this Advisory Group can Be found on the State Treasurer’s web site. Uhler Lewis K. and Barry W. Poulson (2003). Tax and Expenditure Limits: From Roots to Current Realities. The National Tax Limitation Foundation, Roseville, California.

Public Interest Institute (2002) The Constitution of the State of Missouri Article X Taxation. Rafool, Mandy (1996). “State Tax and Expenditure Limits,” National Conference of State Legislative Finance Paper No. 104.

ENDNOTES

The author has been consulting with these states as a member of the American Legislative Exchange Council (ALEC) Task Force on Tax and Fiscal Policy. 2 For a discussion of the political economy of TELs see (Broder 2000). 3 For an introduction to this literature see (Buchanan and Wagner 1977) 4 For a discussion of the relevance of rent seeking models to TELs see (Poulson and Kaplan 1994). 5 The discussion in this case study is from (Uhler and Poulson 2003) and (New 2001). 6 The discussion of this case study draws from (Stansel 1994), (Public Interest Institute 2002), and (McCarty 2003). 7 The discussion of this case study draws on (Holcombe 2001). 8 The discussion of this case study draws on (New 2001) and (New 2003). 9 The discussion of this case study draws on (Public Interest Institute 2002) and (Poulson 2002). 10 The discussion of this case study draws on (Poulson 2001), (Poulson 2003a), (Poulson 2003b), and (Poulson 2003c). 11 For insight into this controversy see (Treasurer’s Advisory Group on Constitutional Amendments 2003). 12 Much of this discussion is drawn from (Poulson 2003d) 13 The author has been consulting with these states as a member of the American Legislative Exchange Council (ALEC) Task Force on Tax and Fiscal Policy. 14 Ibid. 15 This Amendment has been introduced in the Kansas Legislature by Representative Brenda Landwehr. 1

Copyright ©2004, Independence Institute INDEPENDENCE INSTITUTE is a non-profit, non-partisan Colorado think tank. It is governed by a statewide board of trustees and holds a 501(c)(3) tax exemption from the IRS. Its public policy research focuses on economic growth, education reform, local government effectiveness, and Constitutional rights. JON CALDARA is President of the Institute. DAVID KOPEL is Research Director of the Institute. BARRY POULSON is a Senior Fellow at the Independence Institute and a professor of economics at the University of Colorado. ADDITIONAL RESOURCES on this subject can be found at: http://www.IndependenceInstitute.org NOTHING WRITTEN here is to be construed as necessarily representing the views of the Independence Institute or as an attempt to influence any election or legislative action. PERMISSION TO REPRINT this paper in whole or in part is hereby granted provided full credit is given to the Independence Institute.