Cato Policy Analysis No. 213 July 25, 1994

Policy Analysis

Taming Leviathan:
Are Tax and Spending Limits the Answer?

by Dean Stansel

Dean Stansel is a fiscal policy analyst at the Cato Institute.


Executive Summary

After a decade of dormancy, the tax revolt is back. Fed up with rapidly rising state budgets, Americans are increasingly taking matters into their own hands, voting into law limits on the ability of state lawmakers to tax and spend. In the past two years, five states approved such populist measures. This year voters in as many as six states will have the opportunity to vote on some type of tax limitation initiative.

The opposition to tax and expenditure limitations (TELs) is enormous. Opponents charge that restraining the growth of taxes and spending is impossible without doing things like taking cops off the beat and firefighters out of the firehouse. Other critics make precisely the opposite complaint about TELs, charging that TELs are ineffective and do not limit the growth of taxes and spending as promised.

This study demonstrates that properly designed TELs can and do limit the growth of state taxes and spending. For example, the growth rate of per capita state spending in TEL states fell from 0.8 percentage points above the U.S. average in the five years preceding TEL enactment to 2.9 percentage points below the U.S. average in the five years after TEL enactment.

Unfortunately, many TELs are designed in a way that minimizes their effectiveness. This study examines that issue and provides a detailed description of how an effective TEL should be designed.

If the citizens of a state wish to limit the growth of Leviathan, they should not abandon TELs; instead, they should ensure that the TELs are properly constructed.

Introduction

The grassroots tax revolt, which began in the stagflationary years of the late 1970s, is brewing again. Though the misery index is not what it was in the days of Proposition 13, there is a rising sentiment among voters that state government has grown too large. That sentiment is groundedin reality: since 1980 total state spending has climbed by 60 percent (after adjusting for inflation).[1]

The movement for tax and expenditure limitations (TELs) is growing in much the same way the term-limits movement is. In 1991 two states--Connecticut and North Carolina--enacted TELs for the first time, and two other states--Colorado and Louisiana--modified their TELs. Since then five states have enacted measures to restrain the growth of taxes and spending.[2]

In March 1992 Oklahomans passed a constitutional amendment requiring all tax increases to pass both houses of the state legislature with a three-fourths majority or be approved by a majority of the voters.

In November 1992 voters in Colorado passed Amendment 1, the Taxpayer Bill of Rights. That measure requires that any increase in taxes--state or local--be approved by the voters. It also limits spending growth to that necessary to keep pace with population growth and inflation.

Also in November 1992 voters in Connecticut passed a constitutional amendment limiting spending growth to the rate of growth of personal income or inflation, whichever is greater.

In 1992 an overwhelming 72 percent of Arizona voters approved a constitutional amendment requiring a two-thirds majority in the legislature for any increase in taxes or fees.

Finally, voters in Washington State approved a constitutional amendment in November 1993 that limits state spending growth to the rate of population growth plus inflation and requires voter approval of any tax increases that would exceed that limit.

All told, 23 states now have TELs. In addition, this year on election day in November, voters in as many as six states will have the chance to vote on some type of tax limitation initiative.[3] (Table A.1 in the appendix summarizes each of the prospective ballot initiatives.)

Although many critics have claimed that TELs are not an effective means of restraining the growth of taxes and spending, this study presents new evidence refuting those contentions. It finds that TELs, when designed properly, can be and have been an effective tool for restraining the growth of both taxes and spending.

The five-year growth rate of per capita state spending in TEL states fell from 0.8 percentage points above the U.S. average in the five years before TEL enactment to 2.9 percentage points below the U.S. average in the five years after enactment.

Per capita spending in TEL states fell from 6.4 percent above the U.S. average in the year of TEL enactment to only 1.7 percent above the U.S. average in 1992.

If the level of per capita spending in TEL states had not declined, the state spending burden per family of four in those states would have been, on average, $450 more in 1992 than it was.

Similar declines in the growth and level of spending in TEL states were found when those states were compared with non-TEL-states, and an examination of state taxes and state and local spending shows the same pattern of change. In sum, TELs appear to have imposed restraint on the growth of state budgets and taxes.

Why TELs Are Needed

The expansion of government over the past several decades has been enormous. Since 1950 the real growth of government spending at all levels has outpaced population growth by a margin of almost eight to one. As a result, the per capita burden of government exploded upward by 254 percent, even after adjusting for inflation.[4]

Though real government spending at all levels surged by 480 percent from 1950 to 1990, real state government spending rose even faster, by 534 percent.[5] Over the last two decades that burden has been growing even larger.[6] As Figure 1 shows, from 1970 to 1980, real state spending grew more than two and a half times faster than population. More recently, from 1980 to 1990, real state spending grew over four times faster than the number of those it serves. Since 1980 many state budgets have nearly doubled in size, even after adjusting for inflation. State taxes have climbed at a similar pace.

Figure 1
State Spending Growth vs. Population Growth

Growing evidence indicates that voters do not want government to be as large as it has become. Exit polls on election day in November 1992 indicated that, given a choice between lower taxes and more government services, 55 percent of voters preferred to keep taxes down, even if that meant fewer government services, while only 36 percent said the opposite. One year later in a similar poll in New Jersey, an even higher 60 percent of voters said they would prefer lower taxes and fewer services, compared to just 33 percent who said the opposite.[7] It is in part such hostility to government that is driving the tax revolt. Not trusting politicians to restrain budget growth and rising tax burdens, Americans are increasingly taking matters into their own hands, voting into law strict limits on the ability of their state governments to tax and spend.

History of the Tax Revolt

In June 1978 voters in California, fed up with skyrocketing property taxes, overwhelmingly passed Proposition 13. Hatched by anti-tax crusader Howard Jarvis, Proposition 13 rolled back local property taxes to 1 percent of assessed valuation, limited assessment increases to the lower of 2 percent or the annual inflation rate, required two-thirds voter approval for new local taxes, and required two-thirds approval of the legislature to raise existing or impose new taxes.

The passage of Proposition 13 led to a push for numerous similar measures in other states. Between 1978 and 1980, 43 states adopted new limitations on local property taxes or new property tax relief plans. Emboldened by their success, the tax revolters took aim at state taxes as well. Between 1978 and 1982, 15 states reduced their general income tax rates, 10 indexed their personal income tax systems, 7 eliminated their gift taxes, and 6 repealed their inheritance taxes.[8]

In addition, explicit limits on spending were imposed. In 1976 New Jersey passed the first state TEL, and by 1982 TELs had been enacted in 20 states.[9] Though TEL enactment slowed along with the fervor of the tax revolt during the prosperous 1980s--only two additional states enacted TELs from 1983 to 1990--voter frustration is now reversing that trend.

What Other Studies Have Found

Many studies have contended that TELs do not effectively restrain the growth of government. For instance, Daphne Kenyon and Karen Benker examined the change in expenditures relative to personal income for TEL versus non-TEL states over the period 1978-83. They found that in some years TEL states saw slower spending growth than non-TEL states, while in other years TEL states saw faster spending growth than non-TEL states. Though Kenyon and Benker say that there is some evidence that TELs may have moderated spending in some states, they conclude that "for most states, TELs have not been a constraint on growth in taxing or spending."[10]

Marcia Howard examined both state tax collections and state general fund expenditures as a share of personal income over the period 1980-87. She found that, while state taxes grew faster than personal income in TEL states, taxes outpaced income growth by slightly more in non-TEL states. Further, Howard found that state general fund expenditures as a share of personal income actually fell in TEL states, from 5.5 percent in 1979 to 5.41 percent in 1987, lower than spending in non-TEL states. She nevertheless concluded that, because the differences were small, there was an "absence of strong evidence that tax and expenditure limitations have been successful."[11]

Dale Bails examined the change in five components of per capita state spending and revenue over the period 1981-85, the "post-tax revolt years." He found that each of the measures grew more slowly in TEL states than in non-TEL states. For instance, per capita total general revenues rose 30 percent in TEL states compared to 36 percent in non-TEL states. However, Bails also found that the average annual growth of spending or revenue (the one to which the TEL applies) from the year of enactment to 1985 was lower than such growth over the period 1970 to the year of enactment in only one-third of the TEL states. Bails thus concluded that states' TELs "resulted in virtually no success in limiting growth in their budgets."[12]

James Cox and David Lowery examined the change in state spending and revenue relative to personal income in three TEL states, Michigan, South Carolina, and Tennessee. Each of those states was paired with a non-TEL state in its region. The authors found that TELs did not have a statistically discernible effect on the growth of government. Cox and Lowery concluded that "by and large, the behavior of the cap states has been similar to that of noncap states."[13]

Each of those studies identifies specific factors that limit the effectiveness of TELs, many of which are related to the faulty design of individual TELs. Since TELs are often designed by politicians--the very people whose behavior TELs are intended to restrain--it should come as nosurprise that they frequently are worded in a way that makes their restraints as weak and easily circumvented as possible.

Despite their recognition that faulty design is often the root of the problem, most studies that find TELs to be ineffective do not recommend strengthening them by eliminat ing their flaws. Instead, the studies conclude that TELs should be abandoned.

In contrast, there have been several studies that have found TELs to be effective. Stephen Moore examined a variety of fiscal discipline mechanisms and found that average state spending as a share of income in TEL states fell from 4 percent higher than the average for non-TEL states in 1979 to 10 percent lower than the average for non-TEL states by 1987.[14]

Barry Poulson found that the impact of TELs on state spending during their first four years of existence was "negative for all [TEL] states and significant for seven of those states. . . . The implication is that for these seven states the absence of the TEL would have resulted in significantly greater increases in government expenditures in the short run."[15]

So while there is some dispute as to whether TELs are effective at restraining the growth of government, even the staunchest TEL advocates will admit that TELs have not worked as well as proponents would have liked. There is broad agreement, however, on the importance of the design of TELs to their effectiveness.

Methodology

This study focuses primarily on the performance of TELs as a whole, with a few minor exceptions noted below. In later sections, the performance of well-designed and poorly designed TELs is compared.

To measure the effectiveness of TELs in restraining government growth, this study examines how the growth rates and levels of taxes and spending in states with TELs changed after the TELs were enacted.[16] To adjust for the effect of population changes, per capita figures were used.

Of the 23 states with TELs, 21 had enacted them by 1986. While no state enacted a TEL for the first time between 1986 and 1990, two states, Connecticut and North Carolina, have done so since 1990. Those two states are excluded because there is not yet sufficient data to adequately examine the effect of their TELs. In addition, Rhode Island and Nevada are excluded because their TELs are nonbinding, applying only to the governor's recommended budget. Finally, Alaska is also excluded because of peculiarities in its budgetary structure that make comparisons with other states problematic.[17] Unless otherwise noted, "TEL-state average" herein refers to the 18 states with binding TELs that were enacted by 1986.[18] Table 1 lists those states and the years of initial TEL enactment.

Table A.2 gives the growth rate of spending in each of the 18 TEL states over the five years immediately preceding and the five years immediately following enactment of the TEL. The number of percentage points by which each TEL state's spending growth rate differs from the U.S. average growth rate over those periods is then calculated.[19]

Table 1: States with Binding TELs
State
Year of Adoption
Arizona 1978
California 1979
Colorado 1977
Delaware 1980
Hawaii 1978
Idaho 1980
Louisiana 1979
Massachusetts 1986
Michigan 1978
Missouri 1980
Montana 1981
Oklahoma 1985
Oregon 1979
South Carolina 1980
Tennessee 1978
Texas 1978
Utah 1979
Washington 1979

A TEL-state average is then found for the percentage points by which the five-year growth rates differ from the U.S. average. That average summarizes how the growth rate of spending in TEL states changed relative to the growth rate of the U.S. average after enactment of a TEL.

The same methodology is used to compare TEL-state spending growth to the non-TEL-state average and to calculate and compare per capita state taxes and per capita state and local spending.[20] The results of those calculations for TEL states as a group are summarized below, and a completelist can be found in Tables A.3-A.6 in the appendix.[21]

State Spending

Most TELs are limits on spending. So the most direct way to evaluate the effectiveness of TELs is by examining their impact on state expenditures. If a TEL is effective, it should lower the growth rate of state spending. The average growth rate of per capita state spending in TELstates changed, relative to the U.S. average, in the following respects after TEL enactment.

The five-year growth rate of per capita state spending in TEL states fell from 0.8 percentage points above the U.S. average before TEL enactment to 2.9 percentage points below the U.S. average after TEL enactment (Figure 2).

Figure 2
Growth Rate of per Capita State Spending Relative to U.S. Average for the Five Years Immediately Preceeding and the Five Years Immediately Following TEL Enactment

The growth rate of per capita state spending in TEL states from the year of enactment through 1992, as opposed to just five years after enactment, fell even lower, to 12.1 percentage points below the U.S. average.

The five-year real growth rate of per capita state spending in TEL states fell from 7.1 percent before TEL enactment to 1.8 percent after TEL enactment.

If the real growth rate of per capita state spending in TEL states had not slowed, the state spending burden per family of four would have been, on average, $400 higher in those states five years after TEL enactment and $450 higher in 1992 than it was with the lower rate of spending growth.

In addition to the growth rate's falling (relative to the U.S. average), the average level of per capita state spending in TEL states fell from 6.4 percent above the U.S. average in the year of enactment to 1.7 percent above the U.S. average in 1992 (Figure 3).

Figure 3
Per Capita State Spending Relative to U.S. Average, Year of TEL Enactment and 1992

State Taxes

Though most TELs apply to spending--of the 18 TELs examined here, only 5 apply directly to taxes or revenue--an effective TEL should slow the growth of state taxes as well. The average growth rate of per capita state taxes in TEL states changed in the following ways after TEL enactment.

The five-year growth rate of per capita state taxes in TEL states fell from 5.5 percentage points above the non-TEL-state average before TEL enactment to 12.5 percentage points below the non-TEL-state average after TEL enactment (Figure 4).

Figure 4
Growth Rate of per Capita State Taxes Relative to the Non-TEL State Average for the Five Years Immediately Preceding and the Five Years Immediately Following TEL Enactment

The growth rate of per capita state taxes in TEL states from the year of enactment through 1992, as opposed to just five years after enactment, fell even lower, to 13.2 percentage points below the non-TEL-state average.

The five-year real growth rate of per capita state taxes in TEL states was 11.9 percent before TEL enactment. State taxes then fell by 2.8 percent over the five years after TEL enactment.

If real per capita tax growth in TEL states had not been reversed--that is, if real per capita taxes had risen 11.9 percent instead of declining 2.8 percent over the five years following TEL enactment--a family of four in one of the TEL states would have had to face, on average, a state tax burden that was $650 higher five years after TEL enactment than they faced with the lower growth rate of taxes.

As Figure 5 shows, as a result of the slower growth of state taxes, the level of taxes in TEL states fell in relation to taxes in the rest of the country after TEL enactment. Average per capita state taxes in TEL states fell from 8.8 percent above the non-TEL-state average in the year that TELs were enacted to 0.2 percent below it five years later.

Figure 5
Per Capita State Taxes Relative to the Non-TEL-State Average, Year of TEL Enactment and Five Years Later

So in each case in which the growth rates of spending and taxes in TEL states were compared to the U.S. and the non-TEL-state averages, TELs appeared to have been effective in slowing the growth of state government--although probably not as effective as their supporters had wished.

State and Local Spending

Many TEL critics claim that the real impact of TELs is not in restraining the overall tax and spending burden but in shifting it to local governments. That would presumably defeat the purpose of the TEL. While some TELs include a provision prohibiting the state government from shifting costs to local governments without providing the funding to cover those costs, most do not. Thus, some cost shifting is to be expected.

Incidentally, if the only impact of TELs were to shift government responsibility--for both financing and decision-making--to the local level, TELs might still be worthwhile. Since local politicians are closer to the people, they are more accountable and less likely to be able to get away with excessive spending. Also, moving spending to the local level more closely links the costs of public services to the beneficiaries. In fact, many studies have found that decentralized governments provide public services more efficiently than do centralized governments.[22]

Nevertheless, if the purpose of TELs is to limit the burden of government, then cost shifting may mean that those measures are less successful than their proponents intended.For example, if costs are fully shifted to local governments (i.e., local spending growth rises as much as state spending growth falls), then the total state and local tax burden in TEL states will not have fallen. And if that were the case, though state spending growth would have fallen, it would be misleading to conclude that the burden of government had been reduced.

One way to check for such cost shifting is to substitute per capita state and local spending for per capita state spending and perform the same tests that were used earlier. If a decline in state spending growth leads to a proportional rise in local spending growth, then combined state and local spending growth should not decline. Indeed, both the growth and the level of state and local spending in TEL states did decline, relative to those of the rest of the nation, though not as dramatically as state-level spending.

The five-year growth rate of per capita state and local spending in TEL states fell from 2.3 percentage points above the U.S. average before TEL enactment to 1.2 percentage points below the U.S. average after TEL enactment (Figure 6).

Figure 6
Growth Rate of per Captia State and Local Spending Relative to U.S. Average for the Five Years Immediately Preceeding and the Five Years Immediately Following TEL Enactment

The growth rate of per capita state and local spending in TEL states from the year of enactment through 1991, as opposed to just five years after enactment, fell even lower, to 10.2 percentage points below the U.S. average.[23]

The five-year real growth rate of per capita state and local spending in TEL states fell from 6.1 percent before TEL enactment to 2.4 percent after TEL enactment.

Figure 7 shows that, like the growth rate, the level of state and local spending in TEL states fell (relative to the rest of the nation) after TEL enactment. Average per capita state and local spending in TEL states fell from 0.1 percent below the U.S. average in the year that TELs were enacted to 4.4 percent below it in 1991. Thus, in addition to the burden of state government's falling after TEL enactment, the overall burden of state and local government fell.

Figure 7
Per Capita State and Local Spending Relative to U.S. Average, Year of Enactment and 1991

The data suggest that, as expected, some cost shifting to local governments has, indeed, occurred. However, the data also show that in addition to restraining the growth of state government, as indicated earlier, TELs have been effective at restraining the growth of the overall burden of state and local government.

TELs: Why Some Succeed and Others Fail

Some TELs are more effective than others in restraining the growth rate of taxes and spending. In fact, some of the TELs included in this analysis were conclusively ineffective--most notably, those of Louisiana, Montana, and South Carolina (see Table A.2). The reasons some TELs succeed and others fail can be found in their design.

There are several criteria on which TELs may vary that can play a role in their ultimate effectiveness.[24] (Table 2 examines each of the 18 binding TELs included in this study on five of those factors.) The major issues in determining the effectiveness of TELs are discussed in the following subsections.

Table 2: Description of Binding Tax and Expenditure Limitations

State Year Constitutional/
Statutory
Initated by Approved by Limit Applies to
Arizona* 1978 Constitutional Legislature Voters Appropriations of state tax revenues
California 1979 Constitutional Voters Voters Appropriations of state tax revenues
Colorado* 1977 Statutory Legislature Legislature State general fund appropriations
Delaware 1980 Constitutional Legislature Legislature State general fund appropriations
Hawaii 1978 Constitutional Constituional Convention Voters State general fund appropriations
Idaho 1980 Statutory Legislature Legislature State general fund appropriations
Louisiana* 1979 Statutory Legislature Legislature State tax revenue
Massachusetts 1986 Statutory Voters Voters

State revenue

Michigan 1978 Constitutional Voters Voters State revenue
Missouri 1980 Constitutional Voters Voters State revenue
Montana 1981 Statutory Legislature Legislature State appropriations
Oklahoma* 1985 Constitutional Legislature Voters State appropriations
Oregon 1979 Statutory Legislature Legislature State general fund appropriatioins
South Carolina 1980 Constitutional Legislature Legislature State appropriations
Tennessee 1978 Constitutional Constitutional Convention Voters Appropriations of state tax revenues
Texas 1978 Constitutional Legislature Voters Appropriations of state tax revenues
Utah 1979 Statutory Legislature Legislature State appropriations
Washington* 1979 Statutory Voters Voters

State tax revenues

Source: Advisory Commission on Intergovernmental Relations, Significant Features of Fiscal Federalism, 1993, vol. 1, pp. 14-19 (and other years); Kenyon and Benker, p. 437.

*These states have passed new measures or altered existing ones since 1990.

State Limit is Provisions for a Waiver
Arizona* Shall not exceed 7 percent of state personal income 2/3 approval of the legislature on specific additional appropriations
California Yearly growth shall not exceed the percentage increase in population and inflation Declaration of an emergency by a 2/3 vote and compensating reductions in spending over 3 following years
Colorado* Yearly growth shall not exceed 7 percent Statute may be amended by a majority vote of the legislature
Delaware Shall not exceed 98 percent of estimated general fund revenue and prior year's unencumbered funds Declaration of an emergency and 3/5 vote of the legislature
Hawaii Yearly growth shall not exceed the average annual growth rate of state personal income over the preceding 3 calendar years 2/3 approval of the legislature on specific additional appropriations
Idaho Shall not exceed 5.33 percent of state personal income. 2/3 approval of the legislature on specific additional appropriations.
Louisiana* Shall not exceed FY 1978-79 state revenue as a share of 1997 state personal income, multiplied by state personal income in the prior calendar year Statute may be amended by a majority vote of the legislature; certain tax sources (i.e, severance tax revenue) are excluded from computation.
Massachusetts Yearly growth shall not exceed the average annual growth of wages and salaries over the previous 3 years. Statute may be amended by a majority vote of the legislature.
Michigan Shall not exceed FY 1978-79 state revenue of a share of 1977 state personal income in the prior calendar year or average state personal income over the previous 3 calendar years. Declaration of an emergency by governor and 2/3 vote of the legislature.
Missouri Shall not exceed FY1980-81 state revenue as a share of 1979 state personal income in the prior calendar year or average state personal income over the previous 3 calendar years. Declaration of an emergency by governor and 2/3 vote of the legislature.
Montana Biennial growth shall not exceed the percentage difference in the average state personal income over the 3 calendar years immediately preceding the biennium and the average state personal income over the 3 calendar years immediately preceding the current biennium. Declaration of an emergency by governor and 2/3 approval of the legislature on specific additional expenditures.
Oklahoma* Yearly growth shall not exceed 12 percent (adjusted for inflation) or state appropriations shall not exceed 95 percent of certified revenue. None
Oregon Biennial growth shall not exceed the growth rate of state personal income over the preceding 2 calendar years. Statute may be amended by a majority vote of the legislature.
South Carolina Yearly growth shall not exceed the average annual growth rate of state personal income over the preceding 3 calendar years, or state appropriations shall not exceed 9.5 percent of state personal income, whichever is greater. Declaration of an emergency and a 2/3 vote of the legislature;every 5 years the legislature may review the composition of the limit.
Tennessee Yearly growth shall not exceed the projected growth rate of state personal income for the calendar year in which the fiscal year begins. Majority vote of the legislature on a specific additional amount.
Texas Biennial growth shall not exceed the growth rate of state personal income. Declaration of an emergency and a majority vote of the legislature on a specific additional amount.
Utah Yearly growth may not exceed 85 percent of the increase in state personal income. Declaration of an emergency and 2/3 vote of the legislature; legislature must also hold a public hearing.
Washington* Yearly growth shall not exceed the average annual growth rate of state personal income over the preceding 3 calendar years. Declaration of an emergency by a 2/3 vote in the legislature and 2/3 approval of the legislature on specific additional appropriations.

Did the TEL Originate with the Voters or in the Legislature?

Where a TEL originates and who approves it differ from state to state. Whether a TEL originates with the voters or in the legislature can make a significant difference in its ultimate effectiveness. TELs that originate in the legislature, since they are written by politicians--the very people whose behavior they are intended to restrain--tend to be more vague, less restrictive, and more easily circumvented. In contrast, those originating with the voters usually have more teeth. Of the 18 TELs examined, 11 originated with the legislature and only 5 with the voters. The other two were initiated by constitutional convention. (See Table 2 for astate-by-state listing of who originated and who approved the 18 TELs examined herein.)

It should be noted here, however, that in 24 states citizens have no means by which to put an initiative or referendum on the ballot.[25] Such measures must be initiated by the legislature. Thus in those states, TELs that originate with the voters are not an option, although through alegislative referendum voters can at least have the opportunity to vote on a legislature-originated TEL. (Of some hope for those in noninitiative and referendum states is that, in the three states where TELs were initiated by the legislature but approved by the voters, spending growth did fall somewhat relative to the U.S. average.) Although severely handicapped, tax activists in noninitiative states can pressure their legislators to come up with a sound, well-designed TEL, in addition to pressing for a constitutional change granting citizens the right to put initiatives and referendums on the ballot.[26]

Finally, who approves the TEL is also important. Of the 18 TELs examined, 8 were approved by the legislature and 10 by the voters. Thus, while 3 of the 11 legislature-originated TELs were sent to the voters for approval, 8 never faced voter scrutiny, since they both originated in and were approved by the legislature. Only five TELs were both initiated and approved by the voters.

The same methodology used in the earlier sections reveals that in the five states where TELs were initiated and approved by the voters, spending growth slowed, relative to the U.S. average, while in the eight states where they were both initiated and approved by the legislature, spending growth actually rose, relative to the U.S. average. (The remaining five TELs were not initiated and approved by the same people but by some combination of groups including the voters, the legislature, and constitutional conventions.)

The five-year growth rate of per capita state spending in voter-initiated and voter-approved TEL states fell from 6.5 percentage points above the U.S. average before TEL enactment to 2.2 percentage points above the U.S. average after TEL enactment, a fall of 4.3 percentage points. In contrast, the five-year growth rate of spending in states where TELs were both initiated and approved by the legislature rose from 2.1 percentage points below the U.S. average before TEL enactment to 0.5 percentage points below the U.S. average after TEL enactment, an increase of 1.6 percentage points. These findings support Poulson's theory that TELs initiated by the legislature "could actually erode the budget constraint resulting in an increase in the growth of state government."[27]

The ideal TEL would originate with and be approved by the voters, where possible, rather than the legislature.

Is the TEL Constitutional or Statutory?

In addition to origin and approval, whether a TEL is statutory or constitutional is important. Constitutional TELs are difficult to change. On the other hand, statutory TELs leave open the possibility that the legislature will change the definition of the item limited--often by excluding certain areas of spending or revenue--or weaken the restrictiveness of the limit itself whenever legislators see fit. Thus, constitutional TELs are thought to be more effective than statutory TELs. Of the 18 TELs examined herein, 10 are constitutional and 8 are statutory. (See Table 2 for a state-by-state list.)

Again, the same methodology gives results that indicate that spending growth in the 10 constitutional-TEL states fell, relative to the U.S. average, more so than in the 8 statutory-TEL states.

The five-year growth rate of per capita state spending in constitutional-TEL states fell from 0.8 percentage points below the U.S. average before TEL enactment to 5.6 percentage points below the U.S. average after TEL enactment, a fall of 4.8 percentage points. In contrast, the five-year growth rate of spending in statutory-TEL states fell from 2.9 percentage points above the U.S. average before TEL enactment to 0.6 percentage points above the U.S. average after TEL enactment, a fall of only 2.3 percentage points.

The ideal TEL would be constitutional rather than statutory.

How Much of the Budget Is Being Limited?

Even the most stringent TELs do not pertain to the entire budget. Most apply solely to the general fund. Since, on average, about 44 percent of state-appropriated funds are outside the general fund, a substantial portion of the budget is left uncapped.[28] In fact, in some states the uncapped portion of the budget can be as large as 71 percent.[29] Often exempted are expenditures of special funds such as those earmarked for highways, education, and capital construction. Among the frequently exempted revenue items are federal aid, insurance trust funds, user fees, and earmarked funds.

Six of the 18 TELs examined here apply to state tax revenue.[30] Thus, those TELs do nothing to restrict the growth of nontax revenue--charges, user fees, and the like. Predictably, from 1980 to 1991, charges and miscellaneous general revenue rose as a share of total general revenue from 13.8 percent to 17.7 percent.[31]

The ideal TEL would apply a cap to 100 percent of the budget rather than to only certain categories.

Does the TEL Cap Spending or Revenue?

Whether a TEL applies to spending or revenue can also make a difference in its effectiveness. Of the 18 TELs examined here, 9 apply to spending (4 to appropriations and 5 to general-fund appropriations), and 9 apply to revenue in one way or another (3 to revenue, 2 to tax revenue, and 4 to appropriations of tax revenue).

Each fiscal year, states must estimate their revenues for the coming year, in part on the basis of forecasts of how the state's economy will perform. Those economic forecasts, and hence the revenue estimates, are often quite inaccurate. Further, politicians can manipulate the economic forecasts in an effort to get around their TELs' restrictions. Therefore, the use of revenue estimates as the base for a TEL is not ideal.

Those pitfalls can largely be avoided, however, by applying the TEL to spending, since the spending numbers for the coming fiscal year are far more certain.

The ideal TEL would cap spending rather than revenue or taxes.

What Is the Limit?

Another important factor is the definition of the limit itself. Most TELs--including 14 of the 18 TELs examined here--restrict state spending growth to the growth in state personal income. A few, however, use other limits such as population growth, inflation, or a fixed rate of growth. (See Table 2 for more details.)

Many argue that, of the wide variety of limits in use, those that restrict the growth of government to the rate of growth of personal income are the most appropriate since they prohibit the government from growing faster than the private economy. The fact that an overwhelming majority of TELs are so structured highlights the degree to which this argument has been accepted.

However, the evidence suggests that linking spending to population growth plus inflation is much more restrictive. To illustrate the different effects of those two types of limits--personal income linked versus population linked--Figure 8 shows the record of the 1980s. Throughout most of the 1980s the economy was booming, thus personal income rose at abnormally high rates, 30 percent in real terms from 1980 to 1990. In those prosperous times, government revenues were pouring in, allowing spending to skyrocket as well. Real state spending shot up by 42 percent from 1980 to 1990. Meanwhile, the U.S. population grew at only a modest rate of 10 percent.[32]

Figure 8
Growth in Population, Real Personal Income, and Real State Spending, 1980-90

Thus, a TEL limiting spending growth to the growth rate of personal income would have resulted in a slower rate of state spending growth over the last decade than occurred (30 percent compared to 42 percent). However, a TEL limiting the growth of state spending to the growth rate of population plus inflation would have provided a far stricter limit on real state spending growth, holding it to only 10 percent from 1980 to 1990, less than one-fourth the 42 percent rate at which it actually grew and one-third the rate at which a personal income-linked TEL would have allowed spending to grow. Thus, TELs limiting spending growth to the growth rate of population plus inflation are far preferable if the goal is to tame spending.

In fact, one of the reasons many studies have found TELs ineffective is that so many TELs are linked to personal income growth--which is hardly a restrictive limit. Since most TELs were in place by 1982, the evidence on spending growth since TEL enactment comes largely from the period of the 1980s when rapid growth of income accommodated the surge in spending.

For example, South Carolina has a personal income-linked TEL. The five-year growth rate of spending in South Carolina rose from 18.7 percentage points below the U.S. average before TEL enactment to 3.0 percentage points above the U.S. average after TEL enactment. By contrast, thefive-year growth rate of spending in California, which has a population- and inflation-linked TEL, fell from 9.4 percentage points above the U.S. average before TEL enactment to 5.9 percentage points below the U.S. average after TEL enactment.

The ideal TEL would limit the growth of spending to the growth rate of the population plus inflation rather than to the growth of personal income.

Can the TEL's Restrictions Be Circumvented?

How easily politicians can get around the limits of a TEL is an important factor in determining how successful that TEL will be at restraining the growth of government. At the very least, the TEL should be binding. In two cases the limit applies only to the governor's recommended budget and thus has virtually no teeth. (Nevada and Rhode Island have such nonbinding TELs, so as mentioned earlier, they were not included as TEL states in this study.)

Of the other TELs, only Oklahoma's does not provide some mechanism by which the legislature can waive the restrictions of the TEL. Nine of the 18 TELs examined here first require the declaration of an "emergency." However, no state has specified what constitutes an emergency.[33] Of the 18, 6 require only a majority vote to waive their TELs; 1 requires three-fifths approval; 9 require two-thirds approval; 1 requires two-thirds approval plus compensatingreductions over the following three years; and 1 has no waiver provisions. (See Table 2 for more details.)

The ideal TEL would require voter approval for its provisions to be circumvented.

Does the TEL Allow Cost Shifting to Local Governments?

Since the goal of a TEL is to restrict the growth of government, if state governments adhere to their TELs by shifting costs to local governments, the overall size of government will not have been restrained. In an effort to foil such attempts to ignore the true intent of TELs, 5 ofthe 18 TEL states examined here have adopted TELs that prohibit state governments from imposing unfunded mandates on local governments.[34]

Since localized government is preferable to centralized government, a prohibition of unfunded mandates, which by itself could prohibit such decentralization, may not be desirable. To address that, four of the five TELs that prohibit unfunded mandates do allow for transfer of responsibility--for both program financing and decisionmaking--to local governments. In those cases, the state spending limit is then adjusted downward.

Since most TELs do not apply to local governments, such decentralization could, through a disproportionate expansion in local government spending, lead to an increase in the overall level of state and local government spending. To prevent such expansion, TELs should apply to local governments as well.

Colorado's Taxpayer Bill of Rights, passed in 1992, is a recent example of a TEL that applies to both state and local governments. Several of the prospective 1994 ballot measures do so as well.

The ideal TEL would apply to both state and local governments. And it would allow for transfer of responsibility to local governments and provide for the appropriate adjustments in each jurisdiction's limit.

Does the TEL Require Additional Action to Be Implemented?

The manner in which a TEL is to be implemented is also of considerable importance. If the legislature must take additional action to make the TEL operative, it can create a major roadblock in the path toward fiscal restraint. Unfortunately, in several instances that has been the case.

For example, in Connecticut, as part of the 1991 deal to enact a new state income tax, Gov. Lowell Weicker agreed to a statutory TEL limiting the growth in general budget expenditures to the percentage increase in personal income or inflation (whichever was higher). A similar measure was passed as a constitutional amendment in 1992. However, the measure required that the general assembly define "general budget expenditures," "increase in personal income," and "increase in inflation." The legislature has thus far refused to define those terms, and the state's attorney general has ruled that until they do, the TEL is inoperative. Therefore, because of poorly designed implementation provisions, a TEL that voters approved by a four-to-one margin remains impotent.[35]

The ideal TEL would require no additional action by the legislature for implementation.

How Is the TEL Enforced?

Finally, enforceability plays a major role in determining the effectiveness of fiscal discipline mechanisms. If a TEL is not adhered to, it ceases to be useful. Thus, a TEL should clearly articulate what is to take place should the government violate the dictates of the TEL. Unfortunately, that is rarely the case. As a result, several states have simply ignored their TELs.

A case in point is Texas, where a constitutional TEL was passed by an overwhelming 84 percent majority in 1978. The TEL was adhered to in the first biennium to which it applied, but it has been ignored ever since.[36] While the Texas constitution is fairly clear on the precise course to be taken in the case of a violation of budgetary laws such as this one, the specific wording of the TEL itself does not mention enforcement mechanisms. In part as a result of that, the Texas Public Policy Foundation reports that "the Texas Reform Act is probably being violated at almost every level of the budget process," leading the legislature's biennial appropriations to exceed their limit by as much as $1.5 billion.[37]

To avoid that sort of enforceability problem, the text of a TEL should clearly state that taxpayers have legal standing to sue to enforce the TEL's provisions.[38] In addition, TELs should require injunctive relief to prevent the illegal collection of taxes or appropriation of expenditures while suit is pending. Missouri's Hancock II, an initiative seeking 1994 ballot status, is an example of a TEL that contains such provisions.[39]

The ideal TEL should give taxpayers standing to sue to enforce its provisions and require injunctive relief to prohibit any illegal taxes or spending while such suits are pending.

The Ideal TEL

The ideal TEL should have the following characteristics.[40]

  1. It should originate with and be approved by the voters, where possible, rather than the legislature.
  2. It should be constitutional rather than statutory.
  3. It should apply a cap to 100 percent of the budget rather than to only certain categories.
  4. It should cap spending rather than revenue or taxes.
  5. It should limit the growth of spending to the growth rate of population plus inflation rather than to the growth of personal income.
  6. It should require voter approval for its provisions to be circumvented.
  7. It should apply to both state and local governments. And it should allow for transfer of responsibility to local governments and provide for the appropriate adjustments in each jurisdiction's limit.
  8. It should not require additional action by the legislature for implementation.
  9. It should give taxpayers standing to sue to enforce its provisions and require injunctive relief to prohibit any illegal taxes or spending while suit is pending.

Do TELs Have a Useful Life Span?

It is often argued that TELs are less effective over the long run than they are when first implemented. Part of the reason for that is that many TELs are statutory and can be amended by the legislature at any time. Thus, since TELs are often weakened over time--through exclusions of additional areas of spending or revenue to which the limit applies or through outright alterations of the limit itself--it is expected that their effectiveness will decline as well.

Examination of the 15 binding TELs in place by 1980 indicates that spending growth is indeed restrained more over the first five years after TEL enactment than over the second five years.[41] The real growth rate of per capita state spending in the 15 states with binding TELs in place by 1980 was 3.5 percentage points below the U.S. average from 1980 to 1985 (6.3 percent vs. 9.8 percent) but 0.4 percentage points above the U.S. average from 1985 to 1990 (16.7 percent vs. 16.3 percent).

California--where a constitutional TEL was enacted in 1979--is one example of a state whose TEL has substantially lost effectiveness over time, as more and more spending has been moved outside the purview of the limit. For example, Proposition 98, passed in 1988, required that a large portion of the state budget go to education, in effect crowding out spending on other things such as streets, highways, sewers, and police protection. The resulting demands for more spending on those areas were then often met with specific earmarked local taxes, thus increasing the overall burden of state and local government. Further, Proposition 99 increased the cigarette tax but excluded that revenue from the general fund, designating it instead for grants to health organizations, thus exempting it from the TEL's cap.[42] As a result, despite the presence of a constitutional TEL, state spending in California continued to rise dramatically throughout the 1980s.

The erosion of TEL effectiveness over time can be linked in part to the culture of spending in state capitals, in which politicians will do whatever it takes to avoid having their spending powers restrained. However, it can also be traced to poor design. A major loophole in every existing TEL is the lack of a cap applying to the entire budget.

For instance, California's TEL applied only to "appropriations of state tax revenue." Others apply only to "state general fund appropriations." As a result, government can be and has been expanded in such states, despite the presence of a TEL, by increasing nontax revenue (e.g., charges and user fees) and excluding certain categories of spending from the general fund. That loophole is probably the single biggest contributor to the limited effectiveness of TELs, especially over time.

Perhaps the most important factor in preventing the erosion of the effectiveness of a TEL is making sure that the language of the TEL clearly states that the cap applies to "100 percent of state expenditures." Inclusion of that language and the other characteristics of the ideal TEL outlined in the preceding section would go a long way toward ensuring that TELs do indeed have a long useful life span.

Other Measures of Fiscal Restraint

Well-designed TELs can impose significant fiscal restraint on a state's budgetary process. However, other useful measures exist as well. Two such measures are super-majority requirements and voter approval requirements for increasing existing taxes or imposing new taxes. Table 3 is a list of the states that have at least one of those three (including TELs) measures of fiscal restraint.

Table 3: Fiscal Discipline Mechanisms in the States
State
Tax & Expenditure Limit
Supermajority Requirement
Voter Approval Requirement
Alaska
X
 
 
Arizona
X
X
 
Arkansas
 
X
 
California
X
X
 
Colorado
X
 
X
Connecticut
X
 
 
Delaware
X
X
 
Florida
 
X
*
Hawaii
X
 
 
Idaho
X
 
 
Louisiana
X
X
 
Massachusetts
X
 
 
Michigan
X
 
 
Mississippi
 
X
 
Missouri
X,*
 
*
Montana
X
*
*
Nevada
X
*
 
North Carolina
X
 
 
North Dakota
 
 
*
Oklahoma
X
X
X
Oregon
X
 
*
Rhode Island
X
 
 
South Carolina
X
 
 
South Dakota
 
X
 
Tennessee
X
 
 
Texas
X
 
 
Utah
X
 
 
Washington**
X
 
X
Total in 1994
23
9
3
1994 ballot measures
1
2
5

Note: The following states have none of these three fiscal discipline mechanisms currently in place or on the ballot in 1994: Alabama, Georgia, Illinois, Indiana, Iowa, Kansas, Kentucky, Maine, Maryland, Minnesota, Nebraska, New Hampshire, New Jersey, New Mexico, New York, Ohio, Pennsylvania, Vermont, Virginia, West Virginia, Wisconsin, Wyoming.

* Measures seeking 1994 ballot status (source: Americans for Tax Reform)

**Washington State's Initiative 601, passed in 1993, requires voter approval for tax increases that cause spending to exceed the spending limit (population growth plus inflation) but not for all tax hikes.

Supermajority Requirements for New or Increased Taxes

By requiring more than a simple majority (usually a two-thirds majority) to raise existing taxes or impose new ones, supermajority requirements force legislators to reach a broader consensus on the necessity for higher taxes and the wisdom of the spending those taxes will fund.[43] A more sound fiscal policy is the likely result. Nine states have such requirements, two of which were enacted in 1992. In 1994 voters in two states--Montana and Nevada--may have the opportunity to pull the lever in support of supermajority requirements.[44]

Voter Approval Requirements for New or Increased Taxes

Another fairly new fiscal discipline measure requires voter approval to increase existing taxes or impose new ones. In March 1992 Oklahoma kicked off this popular and populist movement by passing a constitutional amendment that requires voter approval of any tax hike that does not pass each house of the legislature by at least a three-fourths vote. Colorado went even further in November 1992 with passage of the Taxpayer Bill of Rights, a constitutional amendment that requires voter approval of any increase in state or local taxes. In 1994 there may be as many as five state ballot initiatives that would require voter approval for new or increased taxes.[45]

Voter approval requirements (VARs) have the further appeal of reconnecting the voters with their governments. Politicians who wish to raise taxes now have to convince only their fellow legislators of the wisdom of such a move. With a VAR, however, would-be tax-hiking politicians would have to take their case to the people. They would be forced to convincingly explain to their constituents why the money is needed and where it will be spent. That is rarely easy.

Of course, VARs are not without their weaknesses. For example, since most VARs merely restrict the growth of tax rates, they would do nothing to prohibit the growth of nontax revenue.[46] As mentioned earlier, charges, fees, and miscellaneous general revenues have been growing as a share of total general revenue. Thus, the restrictions of most VARs could be circumvented by continuing to increase those types of revenues.

Still, measures that require voter approval for increasing existing taxes or imposing new ones provide a powerful method to restrain the growth of state government. Further, VARs are very popular with voters. A recent national poll found that 66 percent of Americans favor requiring tax increases to be approved by a majority vote of the people.[47] The popular appeal of VARs makes it probable that they will follow the path of the term-limits movement. That is, they will most likely pass by wide margins everywhere they are on the ballot in 1994 and be adopted by many more states in the years to come.

Conclusion

TELs, as we know them, are no silver bullet. Even their staunchest supporters will admit that TELs have not slowed the growth of taxes and spending as much as their advocates would have liked. Some TELs, in fact, have been clearly ineffective. However, the ineffective TELs tend tobe the ones that are plagued by specific elements of poor design. For example, they are usually statutory rather than constitutional and thus are designed by the very politicians whose behavior they are intended to restrain. As a result, the cap frequently excludes large areas of the budget, and the TEL is often written in a manner that makes circumventing or even changing its limit quite easy.

Nevertheless, as this study has shown, TELS--despite their substantial flaws--can work. They can slow the rate of growth of taxes and spending, relative to the rest of the nation, and they can do so without shifting massive costs to local governments. In addition, though all existing TELs have notable flaws that limit their effectiveness, the better designed TELs are much more effective than are those that are designed poorly. Furthermore, TELs with the features of the ideal TEL described herein would have an even greater impact.

This fall voters in as many as six states will have the opportunity to pass some type of tax limitation measure. People opposed to TELs--career politicians, lobbyists, and others--will go to great lengths to scare voters into thinking TELs are damaging. The evidence presented here suggests just the opposite. By restraining the ability of politicians to continue to allow taxes and spending to grow unchecked, TELs can finally provide the beleaguered taxpayer some much-needed relief from the crushing burden of state taxes.

Appendix: Detailed Tables
The tables in this appendix provide more detailed information on TELs. Tables A.3 through A.6 give the results of four different measures of changes in taxing and spending.

Table A.1:
Tax Initiatives Seeking Ballot Status in 1994
State
Initiative
Florida Voter approval requirement
Missouri Tax/expenditure limitation and voter approval requirement
Montana Voter approval requirement
Supermajority requirement
Nevada* Supermajority requirement**
North Dakota Voter approval requirement
Oregon Voter approval requirement**

Source: Americans for Tax Reform.

Note: Information is correct as of July 11, 1994.

*State law in Nevada requires that all ballot measures be approved in two consecutive general elections before taking effect. This measure will be seeking its first such approval.

**On ballot.


Table A.2
Per Capita State Government Spending and Five-Year Growth Rates Relative to U.S. Average, before and after TEL Enactment
 
Per Capita Spending
Five-Year Growth Rate
Change in Five-Year Growth Rate Relative to U.S. Average
5 Years before TEL
Year of TEL
5 Years after TEL
5 Years Preceding TEL
% Points above/below U.S. Average
5 Years Following Enactment
% Points above/below U.S. Average
  1972 1977 1982 1972-77 1972-77 1977-82 1977-82  
Colorado
$481
$748
$1,113
+55.5%
-6.3 %
+48.8%
-6.1%
+0.1%
U.S. avg.
$476
$770
$1,193
+61.8%
 
+54.9%
 
 
  1973 1978 1983 1973-78 1973-78 1978-83 1978-83  
Arizona
$500
$811
$1,085
+62.2%
+2.2%
+33.8%
-13.9%
-16.1%
Hawaii
$1,075
$1,520
$1,932
+41.4%
-18.6%
+27.1%
-20.5%
-2.0%
Michigan
$571
$910
$1,340
+59.4%
-0.6%
+47.3%
-0.4%
+0.2%
Tennessee
$386
$653
$867
+69.2%
+69.2%
+9.2%
+32.8%
-14.9%
Texas
$364
$620
$898
+70.
+10.4%
+44.8%
-2.8%
-13.2%
U.S. avg
$517
$827
$1,221
+60.0%
 
47.6%
   
  1974 1979 1984 1974-79 1974-79 1979-84 1979-84  
California
$655
$1,114
$1,537
+70.1%
+9.4%
+38.0%
-5.9%
-15.3%
Louisiana
$594
$939
$1,504
+58.1%
-2.6%
+60.2%
+16.3%
+18.8%
Oregon
$573
$996
$1,388
+73.8%
+13.2%
+39.4%
-4.5%
-17.7%
Utah
$580
$974
$1,350
+67.9%
+7.3%
+38.6%
-5.3%
-12.6%
Washington
$642
$1,064
$1,525
+65.7%
+5.1%
+43.3%
-0.5%
-5.6%
U.S. avg.
$569
$914
$1,315
 
+60.6%
 
+43.9%
 
  1975 1980 1985 1975-80 1975-80 1980-85 1980-85  
Delaware
$889
$1,378
$2,011
+55.0%
-0.1%
+45.9%
+2.5%
+2.6%
Idaho
$654
$971
$1,250
+48.5%
-6.7%
+28.7%
-14.7%
-8.1%
Missouri
$466
$736
$1,082
+57.9%
+2.8%
+47.0%
+3.5%
+0.8%
South Carolina
$658
$898
$1,315
+36.5%
-18.7%
+46.4%
+3.0%
+21.6%
U.S. avg.
$651
$1,010
$1,449
+55.1%
 
+43.5%
   
  1976 1981 1986 1976-81 1976-81 1981-86 1981-86  
Montana
$770
$1,181
$1,705
+53.4%
-3.0%
+44.4%
+4.9%
+8.0%
U.S. avg
$718
$1,123
$$1,566
+56.4%
 
+39.4%
   
1980 1985 1990 1980-85 1980-85 1985-90 1985-90  
Oklahoma
$948
$1,323
$1,784
+39.6%
-3.9%
+34.8%
-6.4%
-2.5%
U.S. avg.
$1,010
$1,449
$2,047
+43.5%
 
+41.3%
   
1981 1986 1991 1981-86 1981-86 1986-91 1986-91  
Massachusetts
$1,265
$1,963
$3,046
+55.2%
+15.7%
+55.2%
+14.4%
-1.3%
U.S. avg.
$1,123
$1,566
$2,204
+39.4%
 
+40.7%
   
18-state avg.        
+0.8%
 
-2.9%
-3.7%
Note: Spending is per capita state "total general expenditure." The "% points above/below" and the "change" numbers were calculated using unrounded values for the various growth rates.

Table A.3
Growth Rates of Taxes and Spending for the Five Years Immediately Preceding and the Five Years Immediately Following TEL Enactment
 
Preceeding Enactment
Following Enactment
Change
 
Real Five-Year Change
Per capita state spending
7.1%
1.8%
-5.3% pts
Per capita state taxes
11.9%
-2.8%
-14.6% pts
Per capita state and local spending
6.1%
2.4%
-3.7% pts
 
Five-Year Change Relative to Change in U.S. Average
Per capita state spending
+0.8% pts
-2.9% pts
-3.7% pts
Per capita state taxes
+4.0% pts
-8.1% pts
-12.1% pts
Per capita state and local spending
+2.3% pts
-1.2% pts
-3.5% pts
 
Five-Year Change Relative to Change in Non-TEL Average
Per capita state spending
-1.5% pts
-4.6% pts
-3.1% pts
Per capita state taxes
+5.5% pts
-12.5% pts
-18.0% pts
Per capita state and local spending
-1.8% pts
-3.5% pts
-1.7% pts
Note: All figures refer to the average for the 18 states with binding TELs examined herein. The "change" numbers were calculated using unrounded values for the various growth rates and levels of taxes and spending.

Table A.4
Growth Rates of Taxes and Spending for the Five Years Immediately Preceding and the Years Since TEL Enactment
 
5 Years preceding Enactment
Since Enactment
Change
 
Change Relative to Change in U.S. Average
Per capita state spending
+0.8% pts
-12.1% pts
-12.9% pts
Per capita state taxes
+4.0% pts
-7.8% pts
-11.8% pts
Per capita state and local spending
+2.3% pts
-10.2% pts
-12.5% pts
 
Change Relative to Change in Non-TEL Average
Per capita state spending
-1.5% pts
-13.9% pts
-12.4% pts
Per capita state taxes
+5.5% pts
-13.2% pts
-18.7% pts
Per capita state and local spending
-1.8% pts
-11.3% pts
-9.5% pts
Note: All figures refer to the average for the 18 states with binding TELs examined herein. The "change" numbers were calculated using unrounded values for the various growth rates and levels of taxes and spending.
*Most recent state spending and tax figures are for 1992. State and local spending figures for 1992 are not yet available, so 1991 numbers were used.

Table A.5
Levels of Taxes and Spending in the Year of TEL Enactment and Five Years Later
 
Year of Enactment
5 Years after Enactment
Change
 
Percent above/below U.S. Average
Per capita state spending
+6.4% pts
+4.2% pts
-2.3% pts
Per capita state taxes
+3.9% pts
-1.8% pts
-5.8% pts
Per capita state and local spending
-0.1% pts
-1.0% pts
-1.0% pts
 
Percent above/below Non-TEL Average
Per capita state spending
+4.7% pts
+1.3% pts
-3.4% pts
Per capita state taxes
+8.8% pts
-0.2% pts
-8.9% pts
Per capita state and local spending
+3.2% pts
+0.5% pts
-2.6% pts
Note: All figures refer to the average for the 18 states with binding TELs examined herein. The "change" numbers were calculated using unrounded values for the various growth rates and levels of taxes and spending.

Table A.6
Levels of Taxes and Spending in the Year of TEL Enactment and the Most Recent Year
 
Year of Enactment
Most Recent Year
Change
 
Percent above/below U.S. Average
Per capita state spending
+6.4% pts
+1.7% pts
-4.7% pts
Per capita state taxes
+3.9% pts
+0.7% pts
-3.2% pts
Per capita state and local spending
-0.1% pts
-4.4% pts
-4.3% pts
 
Percent above/below Non-TEL Average
Per capita state spending
+4.7% pts
-0.6% pts
-5.3% pts
Per capita state taxes
+8.8% pts
+3.0% pts
-5.8% pts
Per capita state and local spending
+3.2% pts
-1.7% pts
-4.8% pts
Note: All figures refer to the average for the 18 states with binding TELs examined herein. The "change" numbers were calculated using unrounded values for the various growth rates and levels of taxes and spending.
*Most recent state spending and tax figures are for 1992. State and local spending figures for 1992 are not yet available, so 1991 numbers were used.

Notes

[1] U.S. Department of Commerce, Bureau of the Census, State Government Finances, 1980 and 1992 editions.

[2] Several of the measures--the ones that require either a supermajority in the legislature or a majority of the voters to approve new or increased taxes--are not, strictly speaking, TELs. Herein, TELs refers to measures that limit the growth of revenue or expenditures by a specific formula (usually involving the growth of personal income or population).

[3] According to Americans for Tax Reform, information is correct as of July 11, 1994. Those states in which signatures are being gathered to put some form of tax limitationmeasure on the 1994 ballot are Florida, Missouri, Montana, Nevada, North Dakota, and Oregon. At this writing, the Oregon and Nevada initiatives have already achieved ballot status. (Note that three of those statesMissouri, Montana, and Oregonalready have TELs in place.)

[4] U.S. Advisory Commission on Intergovernmental Relations, Significant Features of Fiscal Federalism (Washington: ACIR, 1993), vol. 2, pp. 5354; U.S. Department of Commerce, Bureau of the Census, Historical Statistics of the United States, Colonial Times to 1970, part 2, pp. 111920, 1130; and Office of Management and Budget, Economic Report of the President, February 1994, p. 305.

[5] U.S. Advisory Commission on Intergovernmental Relations, vol. 2, p. 82.

[6] For more details on the recent growth of state government, see Stephen Moore, "State Spending Splurge: The Real Story behind the Fiscal Crisis in State Government," Cato Institute Policy Analysis no. 152, May 23, 1991.

[7] Paul Gigot, "Voters to Bill: Do You Want to Be Florioed?" Wall Street Journal, November 5, 1993, p. A8.

[8] Terry Schwadron and Paul Richter, California and the American Tax Revolt: Proposition 13 Five Years Later (Berkeley: University of California Press, 1984), p. 6.

[9] Information on the existence of TELs in specific states came from U.S. Advisory Commission on Intergovernmental Relations, 1986, 1991, 1992, 1993, vol. 1.

[10] Daphne Kenyon and Karen Benker, "Fiscal Discipline: Lessons from the State Experience," National Tax Journal 37, no. 3 (September 1984): 438.

[11] Marcia Howard, "State Tax and Expenditure Limitations: There Is No Story," Public Budgeting and Finance, Summer 1989, p. 87.

[12] Dale Bails, "The Effectiveness of TaxExpenditure Limitations: A Reevaluation," American Journal of Economics and Sociology 49, no. 2 (April 1990): 223.

[13] James Cox and David Lowery, "The Impact of the Tax Revolt Era: State Fiscal Caps," Social Science Quarterly 71, no. 3 (September 1990): 507.

[14] Stephen Moore, "What the States Can Teach Congress about Balancing the Budget," Heritage Foundation Backgrounder no. 751, February 1990, p. 10.

[15] Barry Poulson, "The Rules of the Game: Designing a State Fiscal Constitution," in The Crisis in America's State Budgets: A Blueprint for Budget Reform, ed. John Berthoud and Samuel Brunelli (Washington: American Legislative Exchange Council, 1993), p. 133.

[16] Information on TELs is from U.S. Advisory Commission on Intergovernmental Relations, 1986, 1991, 1992, 1993, vol. 1. The state tax and spending figures are from U.S. Department of Commerce, Bureau of the Census, State Government Finances, various editions. The state and local spending figures are from U.S. Department of Commerce, Bureau of the Census, Government Finances, various editions.

[17] Alaska gets a disproportionate share of its revenue from severance taxes, mostly on oil sold to nonAlaskans (22 percent of general revenue in 1992). The amount of revenue from those taxes fluctuates highly; thus, in comparison with other states, Alaska's overall levels of taxes and spending fluctuate highly as well. For instance, from 1986 to 1991 per capita state taxes in Alaska went down (in nominal terms) by 9 percent, while the U.S. average rose 30 percent. Including those numbers would skew the TEL-state group average significantly downward (thus making a stronger case for the effectiveness of TELs). Figures from Alaska are excluded from all "nonTEL-state" group averages as well.

[18] Although New Jersey enacted a TEL in 1976, it was by law temporary, expiring in 1983. Thus, New Jersey is not included as a TEL state, nor is it included as a nonTEL state for any of the years during which its TEL was in effect (1976 to 1983).

[19] Since this involves comparisons between growth rates over the same period of analysis, nominal dollars are used.

[20] "NonTEL states" refers to the group of states (different for each year of enactment) that did not have a binding TEL in place for any portion of the relevant period of analysis. Note that nonTEL states thus include both Rhode Island and Nevada (states with nonbinding TELs) but, as mentioned earlier, not Alaska.

[21] More detailed tables for all calculations for all states (similar to Table A.2), as opposed to just TEL-state average figures (as in Tables A.3-A.6), can be obtained from the author.

[22] A list of some of those studies can be found in Sam Staley, "Bigger Is Not Better: The Virtues of Decentralized Local Government," Cato Institute Policy Analysis no. 166, January 21, 1992. See also Stephen Moore and Dean Stansel, "The Myth of America's Underfunded Cities," Cato Institute Policy Analysis no. 188, February 22, 1993.

[23] The most recent year for which state and local spending data are available is 1991.

[24] For a 10-point checklist for evaluating the potential effectiveness of individual TELs, see Poulson, p. 10. For a more detailed contemporary discussion of the expected difficulties in implementing effective TELs due to their design, see Naomi Caiden, "Problems in Implementing Government Expenditure Limitations," appendix C in Aaron Wildavsky, How to Limit Government Spending (Berkeley: University of California Press, 1980), pp. 14362.

[25] Patrick McGuigan, "The Conservative Activist's Primer on Direct Democracy," in Making Government Work: A Conservative Agenda for the States, ed. Tex Lezar (San Antonio: Texas Public Policy Foundation, 1992), p. 402.

[26] The National Referendum Movement, a newly formed group in Memphis, Tennessee, headed by Barbara Vincent, is seeking to obtain the right to initiative, referendum, and recall in states that do not have it by working with grassroots organizations in those states to petition the courts to put initiative, referendum, and recall on the ballot.

[27] Poulson, p. 35. See also Barry Poulson and Jay Kaplan, "A Rent-Seeking Model of TELs," Public Choice 79 (1994): 117-34.

[28] "Tax and Expenditure Limits," in The Source Book of American State Legislation 199394, A Question of Discipline: A Guide to Fiscal Responsibility, ed. Samuel Brunelli and John Berthoud (Washington: American Legislative Exchange Council), p. 32.

[29] Howard, pp. 87-88.

[30] Note that four of those six apply specifically to "ap propriations of state tax revenue."

[31] U.S. Advisory Commission on Intergovernmental Relations, 1993, vol. 2, p. 71.

[32] U.S. Advisory Commission on Intergovernmental Relations, 1993, vol. 2, pp. 24, 53; and Office of Management and Budget, p. 305.

[33] Dale Bails, "A Critique on the Effectiveness of Tax Expenditure Limitations," Public Choice 38, no. 2 (1982): 12938.

[34] Those states are California, Michigan, Missouri, Tennessee, and Washington.

[35] See Jeffrey Christensen, "Weicker's World," National Review, January 24, 1994, p. 26; and Theodore Olson, "A Tax-and-Spend Legislature Ignores the Voters," Wall Street Journal, November 24, 1993, p. A17.

[36] Michael Weiss, "The Texas Tax Relief Act after 12 Years: Adoption, Implementation, and Enforcement," Texas Public Policy Foundation, San Antonio, August 1991, p. 4.

[37] Ibid., pp. 11, 5.

[38] According to Weiss, "Currently, every state except New Mexico permits [taxpayer suits]" (p. 18n. 73). For a fuller discussion of the matter, see ibid., pp. 810.

[39] Hancock II is an attempt to eliminate some of the loop holes of the original Hancock amendment, passed in 1980, which gave taxpayers standing to sue but did not require injunctive relief.

[40] For a sample tax and expenditure limitation act designed by the American Legislative Exchange Council (ALEC), see "Tax and Expenditure Limits," pp. 3740. ALEC's model TEL does exclude certain areas of spending; nevertheless, it generally shares the spirit of the recommendations made herein. Furthermore, it spells out in detail what the textof a TEL should look like, which could prove useful to grassroots organizers seeking to put a TEL initiative on the ballot in their state.

[41] Those 15 states are Arizona, California, Colorado, Delaware, Hawaii, Idaho, Louisiana, Michigan, Missouri, Oregon, South Carolina, Tennessee, Texas, Utah, and Washington.

[42] Steven Hayward, "Conditions Are Building for Another Tax Revolt," Los Angeles Daily News, October 24, 1993, pp. 3132.

[43] Ironically, on the federal level Congress has adopted a supermajority rule for cutting taxes but only a simple majority rule for raising taxes. However, there is now a proposal by Rep. Jim Saxton (R-N.J.) in Congress to require a three-fifths vote to raise taxes.

[44] Americans for Tax Reform.

[45] Ibid. The five initiatives would be in Florida, Missouri, Montana, North Dakota, and Oregon.

[46] Some of the VARs seeking ballot status in 1994 would specifically require voter approval for increased or new "taxes or fees."

[47] Kevin Merida, "Americans Want a Direct Say in Political DecisionMaking, Pollsters Find," Washington Post, April 20, 1994, p. A19.

1994 The Cato Institute
Please send comments to webmaster