The Facts of the Matter: Arizona's Tax Cuts
Tom Rex, Research Manager
Center for Business Research, Arizona State University
In each year since 1991, the Arizona Legislature has reduced state taxes. Some argue that these tax cuts have gone too far while others contend that the cuts have not gone far enough.
This article explores the facts of the matter, quantifying the magnitude of the cuts and identifying other factors that affect government finance. It begins by examining historical revenues and expenditures of state and local governments in Arizona, in total and by type, making comparisons to the national average. This backdrop helps in understanding the changes made to the tax code since 1980, particularly tax cuts passed since 1991. Following this review of government finance in Arizona, the discussion turns to various other factors that have an impact on government finance.
The review of governmental revenues and expenditures focuses on combined finance data of all state and local governments in Arizona, with comparisons to the national average. The data are combined because the level of government responsible for some programs varies over time and by state.
Annual government finance data for all 50 states, including combined figures for state and local governments, are produced and published in the U.S. Bureau of the Census Government Finance series. In order to compare Arizona finance data over time and to other places, these data are adjusted for inflation and population. The latest available data are for fiscal year (FY) 1994 (July 1, 1993 through June 30, 1994).
Since most of the tax cuts of the 1990s have been made at the state government level in Arizona, finances for state government alone also are examined. The source of the state revenue data is the Joint Legislative Budget Committee (JLBC). The latest JLBC data are for FY 1997 (July 1, 1996 through June 30, 1997).
Fiscal years 1994 and 1997 bracket the peak of the current economic cycle. While the absolute peak occurred in FY 1995, economic growth in fiscal years 1994 and 1996 was nearly as strong. Gains during FY 1997 were slower but still solid. Thus, the latest Census Bureau and JLBC data are compared to FY 1985, which was the peak of the prior economic cycle.
Other factors contributed to selecting fiscal year 1985 as the comparison year. Tax cuts and tax increases of the early 1980s had been implemented by then, few changes to the tax code were made at that time, and the tax increases of the late 1980s had not yet occurred. In addition, FY 1985 preceded large increases in expenditures for AHCCCS (Arizona Health Care Cost Containment System), Arizona's equivalent of Medicaid. Expansion of AHCCCS was a significant factor in government finance, especially in the late 1980s and early 1990s.
This analysis emphasizes the tax cuts that began to be made in 1991 and have continued to date. However, to provide historical perspective, some information is provided back to 1980. In that year the first of several significant changes to government finance was passed by the Arizona Legislature. A review of historical tax-code changes shows that because of the economic cycle, tax cuts tend to be made when the economy is strong, while tax increases are most likely when the economy weakens. The countercyclicality of demands on government (for example, expenditures for unemployment insurance rise during recessions) also contributes to this pattern, which holds not only in Arizona, but in most states.
The impact on revenues resulting from changes in the Arizona tax code discussed in this paper should be interpreted as estimates. Such impacts are very difficult to measure, with the uncertainty increasing when projections of the impacts are made into the future. Estimates of the effects of the tax increases of the late 1980s and tax cuts of the 1990s come from the JLBC.
Historical Record of Revenues, Expenditures,
and Tax Changes
In FY 1994, total revenue available to state and local governments in Arizona was $14.9 billion. More than $2.8 billion came from the federal government, while state and local taxes were the source of $8.9 billion. The balance came from user fees, interest earned, and other miscellaneous sources.
Per Capita Revenues. Real (inflation adjusted in FY 1997 dollars) per capita state and local government revenue in Arizona in FY 1994 was $3,880, the highest on record, but only 1 percent more than in FY 1990. In the nine years between the comparison year of FY 1985 and FY 1994, real per capita revenues (in FY 1997 dollars) rose about $529, or 16 percent. This increase was less than the national average, however, causing Arizona's per capita revenues as a ratio to the national average to fall 7 percentage points to 86 (U.S. average = 100). In other words, government revenues in Arizona divided by the state's population were 14 percent less than the national average in FY 1994.
The magnitude of the increase in revenues between fiscal years 1985 and 1994 resulted from a variety of factors, the most important of which are changes to the tax code and variations in economic conditions. Substantial Arizona state tax increases occurred during the late 1980s, but hardly any of the 1990s tax decreases had an effect by FY 1994. During the 1985 to 1994 period, the Arizona economy slid from an economic boom in the mid-1980s to a recession in 1990 91, back to strong growth in FY 1994.
Thus, while tax increases were responsible for a good part of the increase in revenues, economic weakness held down the magnitude of the increase. Compared to the national average, revenues rose less in Arizona between FY 1985 and FY 1994, probably resulting largely from Arizona's relatively weak economy over much of this period.
About 19 percent of all Arizona state and local government revenues in FY 1994 came from the federal government. Funds from the federal government rose substantially over the nine years (see Table 1). This was due to increased funding of public welfare programs; other federal funding decreased. The overall increase in revenues from the federal government was well in excess of the national average, yet per capita federal aid to Arizona as aratio to the national average had only risen to 83 by FY 1994.
Arizona State and Local Government Revenues
Per Capita Change: FY85-94
Ration to U.S. Avg*
Ratio to U.S. Avg
From federal government
* In per capita dollars. ** Expressed in fiscal year 1997 dollars
Source: Center for Business Research, L. William Seidman Research Institute College of Business, Arizona State University, from the Government Finance series of the U.S. Bureau of the Census. Inflation adjustment by the GDP Implicit Price Deflator of the U.S. Bureau of Economic Analysis. Population estimates from the Center for Business Research and the U.S. Bureau of the Census.
Revenues raised directly by state and local governments in Arizona rose 8 percent over the nine years. This increase was considerably less than the national average, dropping per capita revenues from state and local sources as a ratio to the national average to 86.
Almost three-fourths of Arizona's directly raised revenues came from state and local taxes in FY 1994. The 15 percent inflation-adjusted increase in such taxes per person over the nine years was less than the national average, causing the ratio to the national average to fall to 90. Arizona's per person tax level has been below the national average since state taxes were reduced in 1980.
The general sales tax was the largest single source of tax revenues in FY 1994. Real per person collections from this tax were only a little higher in FY 1994 than in FY 1985, while per capita sales tax collections nationally rose considerably over these nine years. Despite a substantial drop in the ratio to the national average, Arizona's FY 1994 ratio still was far above average at 130. Arizona's tourists and seasonal residents contribute to this high level of sales tax collections. However, even after considering this nonresident spending, Arizona is overly dependent on the general sales tax relative to other states.
The property tax was the second major source of revenue among Arizona state and local governments in FY 1994. Collections from this tax rose substantially over the nine years, and slightly in excess of the national average. Yet per capita collections in FY 1994 as a ratio to the national average was only 87.
The income tax (personal and corporate combined) was the other major source of tax revenues in FY 1994. The individual income tax accounted for $1.41 billion in revenues in FY 1994, while the corporate income tax was responsible for $0.30 billion. Per capita collections from the income tax rose moderately between FY 1985 and FY 1994, at a pace equaling the national average. Both individual and corporate income taxes per capita were more than 30 percent below the national average in FY 1994.
Collections per capita from selective sales taxes (such as on motor fuel) and miscellaneous other taxes (such as motor vehicle licenses) also were far less than the national average in Arizona in FY 1994. The ratios to the national average fell from the FY 1985 levels. Other than taxes, state and local governments receive revenues from current charges (user fees), interest earned, and various other sources. Over the nine years, per capita collections in Arizona from each of these sources rose less than the national average; in FY 1994, the per person level of each was below average.
Thus, of the major sources of state and local government revenues, per person collections in Arizona in FY 1994 were above the national average only for the general sales tax, and were more than 10 percent below the national average in each of the other sources. Except for property and income taxes, growth in per person collections did not keep up with the national average between FY 1985 and FY 1994.
Government-finance data sometimes are adjusted for factors other than population, such as differences in income levels across states. For example, the Census Bureau presents finance data divided by $1,000 of state personal income, which is the total income from all sources received by all residents in a state. The relationship between government finance and personal income is an attempt to get at the concept of "ability to pay." Arizona state and local taxes per $1,000 of personal income were above the national average in each fiscal year from 1984 through 1994 (see Figure 1). However, tax collections as a ratio to personal income presents a misleading picture of ability to pay.
Arizona State and Local Government Taxes
As a Ratio to the National Average
Source: Center for Business Research, L. William Seidman Research Institute College of Business, Arizona State University, from the U.S. Bureau of the Census and the Advisory Commission on Intergovernmental Relations.
Tourists and seasonal residents are unusually numerous in Arizona compared to most states. These nonresidents pay state and local government taxes and user fees while they are here, thus their impact is captured in the revenue portion of the revenue-to-personal-income measure. The incomes of tourists and seasonal residents, however, are not included in calculations of Arizona personal income, biasing the revenue-to-personal-income measure. Thus, Arizona government revenues per $1,000 of personal income can be higher than in most states without placing an undue burden on residents.
A better measure of ability to pay is produced by the Advisory Commission on Intergovernmental Relations (ACIR).1 However, the latest data available are for FY 1991. The ACIR created a measure of tax effort, that is the ratio of a state's actual tax collections to its "tax capacity." Tax capacity is the amount each state would raise if it applied national average tax rates to commonly used tax bases. For example, in the case of the tax on tobacco products, the tax base in Arizona is the actual volume of cigarette sales.
The tax cuts of the early 1980s lowered Arizona's tax effort from well above to well below the national average. The subsequent tax increase in 1984 pushed the tax effort measure close to, but still below, the national average. Then, the combination of a lowered tax capacity due to the weak economy in the late 1980s and early 1990s and the tax increases of the late 1980s pushed Arizona's tax effort index a little above the national average by 1991. Even without the tax cuts of the 1990s, Arizona's tax effort likely would have dropped to the national average, or below, simply from the improved economy raising the state's tax capacity.
Another way of measuring tax burden is to calculate personal taxes for a representative household or set of households in each state. The consensus of various studies conducted in the early-to-mid-1990s using this methodology is that the tax burden in Arizona was a little below average. For example, a 1995 study by Kiplinger's Personal Finance Magazine ranked the tax burden on a retired couple in Arizona as 32nd highest among the 50 states.2 Money magazine in 1995 ranked Arizona 24th on tax burden on a high-income household,3 while a District of Columbia study in 1994 ranked Arizona 32nd on tax burden on a family of four earning $50,000.4 Thus, even before most of the 1990s tax cuts took effect, Arizona ranked as a middle-to-low-tax state.
The most recent study (Kiplinger's Personal Finance Magazine, June 1997) was done by city. Among 106 cities across the country, Tucson ranked 94th, and Phoenix 96th, in tax burden among the major taxes (state and local income, property, and sales).5
In contrast, a 1995 State of Utah study on business taxes indicated that the business tax burden in Arizona was the highest of seven western states.6
State Government Only.
Since so much of the tax cutting in Arizona has occurred at the state government level, it also is useful to look at actual revenues collected just by the state government. These data, which are available on a preliminary basis through the latest complete fiscal year (FY 1997), are produced by the JLBC. Because of different accounting systems, revenues as measured by the JLBC cannot be directly compared to the figures produced by the Census Bureau.
Total state government revenues per person in FY 1997 were $1,124. On an inflation adjusted basis, this was up 3.2 percent from FY 1996 and the highest figure on record. Real per capita revenues were $147, or 15 percent, higher than in the comparison year of 1985. Tax collections accounted for 93 percent of the state's revenues in FY 1997. The real per capita figure of $1,046 was a record. It was up 1.5 percent from FY 1996 and 11 percent from FY 1985.
While real per capita taxes and total revenues rose over the last 12 years, they dropped marginally as a percentage of personal income. State taxes collected in FY 1997 were 4.97 percent of personal income, compared to a figure of 5.27 percent in FY 1985.
Sales and use taxes as a whole are the major source of tax revenues (see Table 2), accounting for 47 percent of all tax revenues in FY 1997. Real per person collections were no higher in FY 1997 than in FY 1985.
Arizona State Government Revenues
Per Capita Change, FY85-97
Sales and Use
*Expressed in fiscal year 1997 dollars.
Source: Center for Business Research, L. William Seidman Research Institute, College of Business, Arizona State University, from the Joint Legislative Budget Committee. Inflation adjustment by the U.S. GDP Implicit Price Index of the U.S. Bureau of Economic Analysis. Population estimates from the Center for Business Research and the U.S. Bureau of the Census.
The individual income tax was the other major source of tax revenues, accounting for 37 percent of all taxes collected. Real per capita collections jumped 9 percent in FY 1997 and were 34 percent higher than in FY 1985, but they were not quite equal to the record set in FY 1991.
Corporate income tax collections surged even more in FY 1997, up 28 percent to a record level. The sharp increase in income tax collections in Arizona in FY 1997 was part of a nationwide phenomenon.
The impact of the 1996 legislative action to reduce property taxes was obvious in FY 1997, with collections down $132 million. The real per capita decline was 74 percent in FY 1997, putting the level less than one-half that of any year since 1980.
Arizona state and local government expenditures totaled more than $15.0 billion in FY 1994. In FY 1997 dollars, per person spending was $3,914 in FY 1994. This was 9 percent less than the peak figure of $4,299 in FY 1990. In real per capita terms, spending rose a little more than $550 between FY 1985 and FY 1994, or 16 percent. This increase, however, was considerably less than the national average, pushing Arizona per capita expenditures as a ratio to the national average down to 89 percent. This was the lowest ratio in at least 30 years.
Nearly all types of spending rose less than the national average between FY 1985 and FY 1994. The exception was public welfare (see Table 3), particularly AHCCCS.
Arizona State and Local Government Expenditures
Per Capita Change, FY85-94
Ratio to U.S. Avg*
Ratio to U.S. Avg
Health and Hospitals
* In per capita dollars. ** Expressed in fiscal year 1997 dollars.
Source: Center for Business Research, L. William Seidman Research Institute, College of Business, Arizona State University, from the Government Finance series of the U.S. Bureau of the Census. Inflation adjustment by the U.S. GDP Implicit Price Deflator of the U.S. Bureau of Economic Analysis. Population estimates from the Center for Business Research and the U.S. Bureau of the Census.
Two-thirds of the public-welfare spending was in the category of "vendor payments," which was more than 2.6 times higher in FY 1994 than in FY 1985 on a real per capita basis. Almost all of this category represents spending by AHCCCS. Very large spending increases also occurred nationally, so that per person expenditures on vendor payments in Arizona still were 11 percent less than the national average in FY 1994. In unadjusted total dollars, vendor payments in Arizona rose almost $1.3 billion in the nine years, including nearly $500 million in the three years (FY 1988 through FY 1990) in which taxes were raised by a like amount.
The huge increases in AHCCCS over this time period were mandated, and were a prime reason why taxes were raised in the late 1980s. The nine year real per capita spending increase of 164 percent in vendor payments compares to an increase of only 9 percent in all other types of spending. The real gain in per capita personal income over these nine years was 8 percent.
The total expenditure figures include capital outlays spending on infrastructure (such as new schools and new roads) and for equipment and land. Arizona's rapid population growth is responsible for per person capital outlays being above the national average. Capital outlays fell sharply in the early 1990s, putting the FY 1994 figure lower than that of FY 1985.
Spending on current operations compensation, supplies, materials, operating leases, contractual services was 14 percent less than the national average in FY 1994. The rapid rise in this category, largely due to AHCCCS, still was less than the national average.
Changes to Arizona Tax Code
At the beginning of the 1980s, two changes were made to Arizona state taxes: the sales tax on food to be consumed at home was eliminated and property taxes were reduced. In addition to the substantial reductions in revenues that resulted from these changes, Arizona state and local governmental revenues fell due to reductions in federal-revenue sharing and an economic slump. Even with the strong economic recovery that began in 1983, a substantial tax increase (making permanent a temporary rise in the general sales tax rate) was passed in 1984, putting the state budget back in balance. Even with this tax increase, the tax burden was less than it had been in FY 1980 (see Figure I).
From FY 1985 through FY 1987, few changes were made to the state tax code. After this, the Arizona economy weakened, reducing governmental revenues. At the same time, spending on AHCCCS was skyrocketing. The result was substantial tax increases in three consecutive years from 1988 through 1990. These increases cumulated to nearly $500 million according to estimates made at the time.
Beginning in 1991, Arizona governmental revenues were affected by two circumstances. First, the economy began to recover, boosting revenues. Second, strong philosophical preferences for tax cuts took hold in both the executive branch and the legislature.
These tax reductions began slowly, due to the still-weak economy limiting revenues while AHCCCS spending continued to rise. Very modest cuts, mostly to the individual income tax, were passed in the legislative sessions of 1991 and 1992. More numerous and larger tax cuts were passed in the legislative session of 1993, but most of these changes were phased in over several years. Thus, the net impact on the FY 1993 budget from tax cuts made from 1991 through 1993 was only $9 million, according to estimates by the JLBC.
The tax cuts passed in 1993 affected a number of taxes, most particularly the sales and use tax and the individual income tax. A wide variety of taxes also were cut during 1994, but the individual income tax was especially reduced. This was the first of three consecutive years of tax cuts of about $200 million per year, though the 1994 cuts again were phased in, with most of the impact in FY 1995 and FY 1996.
The tax cuts passed in 1995 were almost wholly to the individual income tax. However, an agreement was made to reduce property taxes in 1996. This became a reality during a special session of the legislature during the summer of 1996. Further tax cuts were passed in 1997, amounting to an initial impact of $134 million in FY 1998 and another $20 million in FY 1999. The majority of the cut was $111 million in individual income taxes.
According to the JLBC, the overall annual impact on revenues from tax code changes made between 1991 and 1997 rose from $9 million in FY 1993 to $38 million in FY 1994. After advancing to $215 million in FY 1995, the annual impact jumped to $514 million in FY 1996 and $776 million in FY 1997 (the fiscal year just ended). The projection for FY 1998 is $1,035 million, with a projection of $1,125 million for FY 1999. The impacts by major tax are shown in Table 4. More than one-half of the tax cuts have been made to the individual income tax.
Revenue Adjustments for Recent Legislation
Annual Impact in Millions of Dollars
Sales and Use
Source: Joint Legislative Budget Committee, unpublished data.
In some discussions, the impacts of tax cuts have been expressed on a cumulative basis. For example, over the five years from FY 1993 through FY 1997, the annual tax cuts noted above cumulate to nearly $1.6 billion. Legislative staff estimates that nearly two-thirds of these tax cuts have gone to individuals.
The $1.1 billion in tax cuts projected for FY 1999 probably should be weighed against the $500 million in tax increases passed in the late 1980s. Further, if this $500 million estimate is adjusted for inflation and population growth, the impact grows. The JLBC estimated that by FY 1993, the annual effect had grown to nearly $800 million. No more up-to-date estimates are available, but it seems likely that the $500 million in tax increases between 1988 and 1990 will grow to nearly $1 billion in FY 1999. From this perspective, the net effect of the tax cuts passed between 1991 and 1997 and the tax increases of the late 1980s probably is a loss of a little more than $100 million per year in state revenues. That is, in FY 1999, state government revenues will be approximately $100 million less than they would have been if no tax code changes had occurred in the late 1980s and 1990s.
The Current Situation
State government finished the 1997 fiscal year with a budget surplus of about $550 million. While large in dollar terms, the surplus was not unusual in percentage terms; substantial surpluses typically occur at the peak of economic cycles.
The FY 1997 surplus resulted primarily from unexpectedly high collections of income taxes, both individual and corporate. Strong collections occurred across the nation and seem to have resulted from large capital gains.
The JLBC projects that the surplus will be smaller at the end of FY 1998. If the unexpectedly high collections of income taxes dissipate as quickly as they accumulated, the surplus would be smaller than predicted. Recent stock market volatility makes this possibility more likely.
Eventually, slowing economic growth in combination with growing impacts of tax cuts already passed ($260 million additional impact between FY 1997 and FY 1998 alone) will shrink the surplus. The surplus will turn into a deficit whenever the next economic recession occurs.
Factors Impacting Government Finance
Arizona consistently is one of the fastest growing states in the nation. Percentage-population growth usually is about three times the national average, ranking second or third among all states. This rapid growth puts a considerable strain on the public sector.
This is most easily seen in the category of capital outlays. The state must spend more per person than a moderately growing state on infrastructure such as roads, schools, and sewers. However, taxes levied on newcomers are not high enough to cover the costs of infrastructure and public services required by the new residents. Thus, everything else being equal, Arizona's total public-sector spending per capita should be above the national average, as should per person taxes. This is not the case, however, as both taxes and expenditures in Arizona are below the national per capita average.
National economic-growth rates vary over the period of several years between strong gains and declines. The Arizona economic cycle corresponds closely to the national cycle in timing, but the Arizona economy is more cyclical than the national economy. Gains in Arizona are much stronger at the peak of the cycle, but the last three recessions in Arizona were as deep, or deeper, than the national average.
This extreme cyclicality creates significant variations in revenues realized by state and local governments in Arizona. For example, during periods of economic strength the state government frequently enjoys annual surpluses in excess of $200 million (in FY 1997 dollars). Within just a year or two of such a surplus, however, the state sometimes has had to resort to a mixture of spending cuts and tax increases just to balance the budget. Unlike the federal government, the state constitution prohibits the state government from running more than an insignificant deficit.
The current economic cycle began with the end of the last recession in mid 1991. Economic growth was relatively slow until mid-1993. It then accelerated to the very strong growth rates typical of the peak of the cycle in Arizona. Growth rates currently are slowing, but remain moderately strong.
The strong growth of recent years is a cyclical phenomenon, with growth rates not exceeding those of prior economic cycles. For example, annual percentage-growth rates in real personal income less transfer payments (the measure used in calculations for the budget stabilization fund discussed in the next section) in each year of the current cycle have been less than in the comparable year of each of the two prior cycles (see Table 5). Similarly, growth in this measure during the current cycle in Arizona, relative to the national average, has been less than in the two prior cycles. In contrast, numeric-employment growth has increased in each economic cycle over the prior cycle. Thus, the stronger gains in the current cycle fit the historical pattern.
Economic growth in the current cycle peaked in 1994 and 1995. Growth rates are expected to continue to slow in 1997 and 1998. However, no sign of a national recession is yet apparent. If the current cycle lasts as long as the previous cycle, the next national recession would begin at the end of 1998. The Arizona economy always slows at about the same time as the national economy. Unlike the last three recessions, the next downturn in Arizona may not be as severe as the national average. Certain industries, such as commercial real estate and commercial construction, got off to a slow start in this cycle and may moderate the next downturn.
Thus, while the next economic slump may be less severe in Arizona than any of the last three, a substantial slowdown in growth from the rates of the last four years is likely. This will have a depressing effect on governmental revenues.
Economic Growth Rates in the Current Cycle Compared to Two Prior Cycles
Inflation-Adjusted Personal Income Less Transfer Payments (percent change)
Nonagricultural Wage and Salary Employment (change in thousands)
Budget Stabilization Fund
In 1990, the Arizona Legislature created the Budget Stabilization Fund (BSF), also known as the "rainy-day fund." The BSF is designed to set aside revenue during times of strong economic growth to be spent during times of weak growth or recession. It applies only to state government. The change in the BSF during a fiscal year is determined by comparing the inflation-adjusted percent change in Arizona personal income, less transfer payments, for the latest calendar year to its average growth rate over the last seven years. The difference is multiplied by the state government general-fund revenue of the prior fiscal year. When growth is above trend, monies are transferred from the general fund into the BSF. When growth is below trend, the transfer is from the BSF to the general fund.
The first payment into the BSF was made in FY 1994. An additional payment into the BSF was made in FY 1995, resulting in a balance of about $225 million. Since the legislature in 1995 changed the maximum balance in the BSF from 15 to 5 percent of revenue, the fund was already at its maximum level. This legislative change precluded the transfer of about $223 million into the BSF during FY 1996, as called for by the formula. Another $121 million deposit for FY 1997 also was deferred. Tax cuts of similar magnitudes were passed in each year.
Growth in revenues has pushed the BSF cap to about $250 million in FY 1997. The fund continues to grow from interest earned, which pushed the balance past the cap in fiscal years 1995 and 1996, causing small revertments to the general fund to have occurred.
In 1997, the legislature again changed the maximum balance in the Budget Stabilization Fund. It will gradually rise from 5 percent in FY 1997 to 7 percent in FY 2000. The legislature also appropriated an ad hoc contribution to the BSF of $30 million in FY 1998. The JLBC projects that the FY 1998 year-end balance in the BSF will be equal to the limit of $291 million. In FY 1999, a balance of $328 million is projected.
Based on the experience of the last three economic cycles, the 7 percent cap imposed on the BSF as of FY 2000 is too low to ensure that adequate funds will be available at the bottom of an economic cycle to meet the transfer that should occur from the BSF to the general fund.
Had the BSF existed before 1990, the formula would have called for a cumulative transfer from the BSF to the general fund of more than $300 million in fiscal years 1974 and 1975. (This and subsequent dollar figures are expressed in terms of FY 1997 dollars.) However, with a 7 percent cap, the BSF would have had only a little more than $100 million. Even with a 15 percent cap, the fund would have had only about $230 million available.
Since general-fund revenues at that time were only one-third those of today in inflation- adjusted dollars, the equivalent payout today would be nearly $1 billion. General-fund revenues are higher today largely because of the state's population growth. The state's population today is twice that of the mid-1970s.
In the early 1980s recession, a cumulative transfer of nearly $220 million from the BSF to the general fund would have been called for by the formula. With a 7 percent cap, the BSF would have had only $170 million available. However, considering interest earnings, the fund would have fallen only a few million dollars short of the payout called for by the formula. Since general-fund revenues then were about one-half those of today, the equivalent transfer today would be more than $425 million.
In the prolonged economic slump of the late 1980s and early 1990s, the formula would have called for a cumulative transfer from the BSF to the general fund of about $570 million. (An equivalent payout today would be about $725 million.) With a 7 percent cap, the payout from the BSF would have been about $300 million short.
Thus, with a 7 percent cap, BSF funds in each of the last three economic slowdowns would have been exhausted short of the payout to the general fund called for by the formula. Thus, the rainy-day fund with a 7 percent cap should be viewed as inadequate to offset the loss of revenues during an economic downturn. Even if the next downturn is milder than any of the last three, the projected BSF balance of $328 million in FY 1999 is unlikely to be sufficient to meet the formula's recommended transfers.
Politics influence government finance in a variety of ways. For example, general opposition to government spending and a desire for tax cuts are far stronger in the 1990s than in the 1980s. This has led to spending, at least by state government, having been restrained throughout the 1990s. Rather than by eliminating programs, spending has gradually been reduced by disallowing inflation to be considered in state agency budget requests. Salary adjustments for state government employees have barely kept pace with inflation over the last five years, following inflation-adjusted declines in wages in the late 1980s and early 1990s. Further, new or expanded spending that typically occurs during a strong economy has been less than in the past.
Another example of a political factor is the requirement passed in the early 1990s that two-thirds of each house of the legislature must approve a tax increase. This will make it more difficult for the state budget to be balanced in the next economic downturn since tax increases commonly have been used to increase revenues during times of economic difficulties. This then magnifies the consequences of the tax cuts in the 1990s and of capping the BSF at 7 percent.
Funds transferred from the federal government to state and local governments represented about 19 percent of all general fund revenues in Arizona in the early 1990s (about $2.8 billion in FY 1994). Federal aid to state and local governments in Arizona increased to about $3.3 billion in FY 1996.
These intergovernmental transfers from the federal government will be reduced as Congress tries to reduce federal government spending and decentralizes programs, passing responsibility to state and local governments. According to Fiscal Planning Services, Inc. of Bethesda, Maryland, the Congressional Budget Resolution passed in June 19967 will result in a decline of 9 percent in cumulative-federal aid to Arizona state and local governments in the six years from FY 1997 through FY 2002. The cumulative-dollar reduction will exceed $2.2 billion. This reduction in federal aid will be phased in. By FY 2002, the annual reduction will be nearly 15 percent, or several hundred million dollars.
A decline of similar relative magnitude occurred in the early 1980s as federal revenue-sharing programs were cut soon after President Reagan took office. This drop in federal transfers to Arizona followed substantial cuts in the state's sales tax and property tax. The impact of these revenue losses was intensified by an economic recession.
Thus, the consequences of the tax cuts made in Arizona in the very early 1980s were complicated by both an economic downturn and reductions in federal spending. The same situation could occur in coming years: the full impact of the tax cuts passed during the early and mid 1990s could collide with a significant reduction in federal monies and an economic slowdown.
Shifting Tax Sources
Relative to their shares of state government revenues in the early 1990s, a disproportionately large portion of the tax cuts passed in Arizona in the 1990s have been to income taxes, upon which Arizona already had a low dependence relative to other states. A relatively small part of the tax cuts has been to sales taxes. This means that a shift is occurring with the sales tax becoming an even greater share of overall governmental revenues. Arizona governments already were overly dependent on the sales tax, relative to the national average, before the tax cuts of the 1990s occurred.
This increasing reliance on one tax is contrary to prevailing tax theory that a balance of taxes is better. Heavy reliance on the sales tax raises a number of additional issues.
First, the sales tax collected on goods, excluding food to be consumed at home, is highly variable because consumer expenditures are so cyclical. The downside of a heavy reliance on the sales tax was demonstrated in Maricopa County in the late 1980s and early 1990s. Almost immediately after the public vote in 1985 to increase the sales tax to fund highway construction, the economy slumped. Consumer spending dropped for several years, lowering revenues from the sales tax. Sales tax collections thus were far less than anticipated, a major factor in so few miles of freeway being opened to date, relative to what had been planned.
Second, collections from Arizona's sales tax, which only applies to certain goods and not to faster-growing services, has not kept pace with overall economic growth. Between 1985 and 1996, real personal income in Arizona rose 57 percent, but retail sales subject to the sales tax increased only 43 percent. Thus, an even stronger dependence on the general sales tax will result in a growing gap between governmental revenues and expenditures.
Third, sales taxes are more regressive than income or property taxes. That is, greater dependence on the sales tax shifts more of the tax burden onto the less affluent.
Economic Impacts of Tax Cuts
The link between an increase or decrease in taxes and subsequent economic performance is unclear. A review of economic research articles published in refereed journals shows no consensus. Most of these studies have focused on tax increases, rather than decreases. Some studies conclude that tax increases harm economic growth, while others reveal no impact or even a small positive effect on the economy.
While disagreements still can be found in the literature, on net the most recent and sophisticated studies suggest that state and local taxes have marginal effects on economic development and growth. Since the magnitude of the effects is small, a large differential between states would be required to influence business location decisions. State and local government taxes are just one of many factors that affect economic growth, and studies have not found them to be one of the more important factors.
A specialized aspect of the economic impact of tax cuts is supply-side effects: that a reduction in taxes will stimulate growth so much that governmental revenues will be greater than if tax rates were higher. While the bulk of the literature suggests that such an effect does not exist to any significant degree, it remains a topic of considerable debate. Whether a supply-side effect will accrue in Arizona, however, is a critical point in assessing the consequences of the state's tax cuts, as discussed in other contributions in this volume.
In the early to mid 1990s, before most of the tax cuts passed during the 1990s had taken effect, Arizona's tax burden already was a little below the national average. Similarly, governmental spending was a little below average, despite upward pressures caused by the state's rapid population growth.
When fully implemented, the tax cuts of the 1990s (through 1997) largely will only offset the tax increases of the late 1980s. Despite an impact of about $1.1 billion in FY 1999 from the tax cuts, the net tax cut probably will be about $100 million (relative to the revenue that would have been generated had the tax code of the mid 1980s not been changed).
In FY 1997, the $776 million impact from the tax cuts had not yet equaled the revenue impact from the tax increases of the late 1980s. Thus, it is not surprising that the tax cuts have not yet had a strong impact on state and local government finances. Further, the tax cuts have been made during a strong economy, with cyclical factors pushing up revenues. Spending also has been cut, contributing to budgetary surpluses.
Economic growth in the current economic cycle is similar to that of prior cycles. Growth rates already are slowing and likely will continue to slow in the next few years due to cyclical factors. This will have a downward influence on governmental revenues.
Governmental revenues in Arizona in the next few years also are likely to be affected by reductions in federal funds. Further, increased relative reliance on the sales tax will gradually cause governmental revenues to grow less rapidly than the overall economy.
When the next economic downturn occurs, the rainy-day fund of $328 million projected for FY 1999 will help to offset the decline in general fund revenues. Even in a mild downturn, however, this balance is unlikely to be adequate to meet the transfers called for in the law that set up the Budget Stabilization Fund. Coupled with the other factors that likely will reduce government revenues, the state government could be faced with a significant imbalance between revenues and expenditures.
In the past, the solution to such a situation typically has been a combination of spending reductions and tax increases. The two-thirds majority now needed in the legislature to increase taxes makes such an option questionable in the next economic slowdown. Thus, substantial spending cuts may be necessary to meet the constitutional mandate of a balanced budget.
1. State Revenue Capacity and Effort, (Washington, DC: Advisory Commission on
Intergovernmental Relations, 1993).
2. Melynda Dovel Wilcox, "The Taxes of Madison County," Kiplinger's Personal Finance
Magazine (August 1995): p. 57.
3. "We Rank the States by Tax Load," Money, (January 1995): p. 92
4. Tax Rates and Tax Burden in the District of Columbia: A Nationwide Comparison,
(Washington, DC: District of Columbia Department of Revenue and Finance, 1994).
5. Marc L. Schulhof, "Taxes: There's No Place Like Home," Kiplinger's Personal Finance
Magazine (June 1997): p. 55.
6. Western States' Tax Burden, FY 1993 94: Research Publication 95 13 (Salt Lake City: Utah
State Tax Commission, 1995).
7. Cuts in Federal Aid Under the Congressional Budget Resolution: A Fiscal Impact Report to
the States, (Bethesda: Fiscal Planning Services, Inc.,1996).
Return to the Table of Contents