Revised October 29, 1997 

Are State Taxes Becoming More Regressive?
by Nicholas Johnson and Iris J. Lav  


Many states throughout the 1990s have been making their tax systems less progressive. When states have raised taxes this decade to meet recession-induced shortfalls, they predominantly raised those taxes that fall most heavily on low- and moderate-income households. When a stronger economy has allowed taxes to be reduced, however, much of the benefit has been targeted on higher-income families. As a result, state taxes appear to have become relatively more burdensome to low- and moderate-income families than they were in the late 1980s.

In the last four years, states have lowered personal income taxes, which are the major taxes paid by upper-income families, by $9.9 billion. This is approximately equivalent to the $8.2 billion income taxes were raised in the early 1990s if inflation is taken into account. But states have not reversed the increases in sales and excise taxes that took place in the earlier years. While sales and excise taxes — the most burdensome taxes for lower-income families — were increased $12.0 billion in the early 1990s, there has been a net reduction of only $0.1 billion in sales and excise taxes in the 1994-97 period.

Looking ahead to 1998, it is likely that the strong economy and healthy state fiscal conditions will continue to prevail as policymakers consider budgets for state fiscal year 1998-99, which begins on July 1, 1998 in most states. In addition, states that levy an income tax may anticipate a temporary revenue boost as a result of the new federal tax law; investors may choose to increase their asset sales, and thus their capital gains income, in response to cuts in the federal capital gains tax rate. Because of these factors, many states are likely to consider additional tax reductions as governors propose budgets and legislatures meet in early 1998. Given the trends of the last decade detailed in this report, it will be particularly important for states to assess the impact on different income groups as they consider further tax reductions.

The tax systems of most states already are significantly regressive, that is, they take a larger proportion of the income of lower-income families than the income of more affluent families. This is largely because states derive about half of their revenues — 49 percent in 1995 — from sales and excise taxes. These consumption taxes impose a disproportionate burden on lower-income families that must consume most or all of their income, as contrasted with higher income families that save or invest some of their income. States derive only a third of their revenues from the personal income tax, the major state tax that can have a progressive effect. Regressive taxes not only burden families that can least afford to pay these taxes, but also can hamper other policies states are pursuing to make families self-sufficient and less dependent on state assistance such as welfare.

An analysis of the tax cuts and increases states enacted in the 1990s shows the trend toward still more reliance on regressive taxes.

In the period from 1990 through 1997, about two-thirds of the states changed either their top personal income tax rate, their general sales tax rate, or both. These rate changes confirm the trend toward more regressive taxation.


Table 1
Significant Changes to State Taxes in the 1990s

  Changes enacted 1990 to 1993   Changes enacted 1994 to 1997

Progressive Taxes Regressive Taxes Progressive Taxes Regressive Taxes
Alabama   Increase    
Alaska       Increase
Arizona Increase   Cut  
Arkansas   Increase Cut Increase
California Increase Increase Cut  
Colorado Increase      
Connecticut Increase Cut Cut Increase
Delaware   Increase Cut  
Florida   Increase    
Georgia     Cut Cut
Hawaii Cut Increase Increase  
Idaho   Increase   Increase
Indiana       Cut
Iowa   Increase Cut  
Kansas Increase Increase    
Kentucky Increase Increase    
Louisiana   Increase   Cut
Maine Increase Increase   Increase
Maryland Increase Increase Cut  
Massachusetts Increase Increase Cut  
Michigan     Cut  
Minnesota Increase Increase    
Mississippi   Increase    
Missouri Increase Increase   Cut
Montana Increase Increase Cut  
Nebraska Increase Increase Cut  
Nevada   Increase    
New Hampshire   Increase   Increase
New Jersey Increase Increase Cut  
New Mexico Increase Increase Cut Cut
New York   Increase Cut  
North Carolina   Increase Cut  
North Dakota   Increase    
Ohio   Increase Cut  
Oklahoma Increase Increase    
Oregon Increase   Cut Increase
Pennsylvania Increase Increase    
Rhode Island Increase Increase Cut* Increase
South Carolina        
South Dakota        
Tennessee   Increase    
Texas   Increase    
Utah     Cut  
Vermont Increase Increase   Increase
Virginia Increase      
Washington   Increase    
West Virginia   Increase    
Wisconsin   Increase   Increase
Wyoming   Increase    
Total Increases 21 37 1 10
Total Cuts 1 1 20 5
Significant changes are defined as aggregate net changes over all four years that exceed one percent of previous year's tax revenue. Blank spaces mean that net tax changes were less than one percent of tax revenues.
"Progressive taxes" includes personal income taxes. "Regressive taxes" includes sales and excise taxes. "Other taxes" includes statewide property taxes, corporate income taxes, and other taxes.
"Total taxes" includes the combined effects of all taxes.
* Rhode Island's personal income tax cuts are being phased in and therefore will not have a significant effect on state revenue until fiscal year 1998-99.

The general trend toward more regressive taxation is made up of varying combinations of tax changes in specific states. (See Table 1 above.) The following are examples of ways in which states have shifted tax burdens onto lower-income taxpayers.

In most states, tax reductions or increases are considered without much information or debate over the extent to which various income groups would benefit or be harmed by the proposed tax changes. Only a few states have the capacity in either their executive budget offices or legislative fiscal offices to analyze routinely and disseminate information in a timely way during the legislative process on the distribution of the benefits that would result from a tax proposal. Even states that have such capacity do not necessarily produce and disseminate analyses throughout the session, when negotiations become intense, compromises are hammered out, and legislation can undergo substantial change. Nor is it common for states to prepare analyses of the distribution of tax changes that have been enacted over a period of years. Policymakers in most states do not have access to analytic information describing the impact on families at different income levels of decisions they have made or might make.

The types of trends detailed in this report, which may have gone unnoticed in individual states as well as across states, highlight the need for states and legislatures to produce such information in a consistent, timely manner. Minnesota has routinely produced such information and a few states, including Connecticut and Wisconsin, have performed such studies in the past. Arizona, California, Iowa, and Maryland, among other states, have analyzed the distribution of their income taxes, but not all of their taxes together.

Texas is moving in the direction of providing comprehensive information on the impact of its tax system and proposed tax changes. In the most recent legislative session, Texas enacted new legislation requiring distributional impact analyses on all significant tax bills. The availability of this type of information can help the public participate in debates over the type of tax changes that are desirable for the state, and can help policymakers make informed decisions.


The Tax Changes of the 1990s

In aggregate, state tax changes from 1990 through 1997 have mirrored the business cycle. A recession hit the national economy in 1990, depressing tax revenue and increasing the costs of social programs. Largely in response to that national recession and the resulting budget shortfalls in many states, states in 1990, 1991, 1992 and 1993 enacted more than $25 billion in net tax increases.(1)

As the economy recovered from the recession, state tax revenue began to grow at a rate faster than projected, leaving states with surplus revenues. These surpluses prompted states to enact tax cuts in 1994, 1995, 1996 and 1997 totaling about $13 billion.(2)

But this simple picture of taxes — up in the early 1990s, down in the mid-1990s — disguises dramatic differences in the specific taxes that were raised or lowered and thus masks the impact on different population groups.

In the early 1990s, sales and excise taxes were increased by $12 billion, personal income taxes by $8.2 billion, and other taxes by $5.8 billion. In other words, when states needed new revenue in the early 1990s, they turned predominantly to consumption taxes and secondarily to personal income taxes to provide the extra revenue.

When states decided to cut taxes in the middle 1990s, they cut personal income taxes by $9.9 billion. But they only reduced sales and excise taxes by a net of less than $100 million, a figure that includes about $1.1 billion in net sales tax cuts and $1 billion in excise tax increases. Other taxes, principally business taxes such as corporate income taxes, were reduced $3.5 billion.(3)  In other words, when states began cutting taxes, they focused those cuts overwhelmingly on personal income taxes, with minimal cuts in consumption taxes.

Although the dollar figures from the early 1990s are not strictly comparable to those of the mid-1990s because they are not adjusted for inflation or other economic changes during the decade, these data suggest that states in aggregate have reversed the personal income tax increases of the early 1990s, while they have not reversed the sales and excise tax increases.

Many of the tax changes of the 1990s, both increases and decreases, were accomplished by raising or lowering tax rates. From 1990 to 1997, about 24 states changed their top personal income tax rates and 21 states changed their sales tax rates. (About a dozen states did both. The remaining 17 states changed neither their income tax rates or their sales tax rates.) As with the aggregate tax changes of the 1990s, however, the patterns of rate changes varied dramatically by type of tax.


Personal Income Tax Rate Changes

It is possible for states to target the benefits of personal income tax cuts to families with lower incomes. States can raise the income level at which families must first pay income taxes, for example. States can create or increase tax credits for expenses that are particularly burdensome for low- and moderate-income working families, such as child care costs. Or states can provide refundable tax credits for low- and moderate-income families that offset the burden of state and local sales and property taxes. Although some states have gone down this path in the mid-1990s, few states have chosen to provide the bulk of their tax relief in this fashion.

Nine of the 12 states with the largest income tax cuts in 1994, 1995, 1996 and 1997 centered those cuts around reductions in tax rates, and a tenth state provided refunds that were similar in effect to rate reductions.(4) When top income tax rates are lowered, or when income tax rates are cut across-the-board, higher-income taxpayers tend to get the largest tax reductions, measured either as dollar amounts or as a percent of income. (See box on page 16.) In the absence of other tax reforms, such rate reductions shift the burden of state taxes onto less well-off taxpayers.

While average top personal income tax rates nationwide remain somewhat higher than they were at the beginning of the decade, the income tax rates applicable to the highest-income taxpayers in many states stand at their pre-1990 level or have sunk even lower.


Table 2
Top State Personal Income Tax Rates in 1989, 1993 and 1997
  1989 1993 1997
Number of states with top rate of less than 4.5% 7 5 6
Number of states with top rate of at least 4.5% but less than 7% 17 17 20
Number of states with top rate of 7% or higher 16 19 15
Note: The figures for 1993 and 1997 include Connecticut, which had no state income tax income tax in 1989; in 1993 and 1997 the top Connecticut rate was 4.5 percent. Nine other states have no personal income tax.


Sales and Excise Tax Rate Changes  

In contrast to the up-and-down movement of personal income tax rates, sales and excise tax rates have risen throughout the 1990s.


Table 3
State General Sales Tax Rates in 1989, 1993 and 1997
  1989 1993 1997
Number of states with rate of less than 4% 4 2 2
Number of states with rate of at least 4% but less than 5% 15 13 13
Number of states with rate of at least 5% but less than 6% 16 14 13
Number of states with top rate of 6% or higher 10 16 17
Note: Five states have no general sales tax.

The changes in tax rates provide further evidence that while states raised both progressive and regressive taxes during the fiscal crisis of the early 1990s, they have cut predominantly the more progressive personal income tax in the middle 1990s.


Examples from Individual States

The national trend of falling personal income taxes but stagnant or rising consumption taxes is reflected in the legislative decisions made by many of the states that were most active in raising taxes in the early 1990s and cutting taxes in the middle 1990s. Table 4 shows the 15 states that enacted the proportionally largest tax changes in the 1990s.

The specific patterns of tax changes vary from state to state.

Table 4
The 15 States with the Largest Tax Changes in the 1990s

  Net Changes 1990-93   Net Changes 1994-97
State Progressive Taxes Regressive Taxes Progressive Taxes Regressive Taxes
Arizona Increase   Cut  
Arkansas   Increase Cut Increase
California Increase Increase Cut  
Connecticut Increase Cut Cut Increase
Iowa   Increase Cut  
Maine Increase Increase   Increase
Massachusetts Increase Increase Cut  
Nebraska Increase Increase Cut  
New Hamphire   Increase   Increase
New Jersey Increase Increase Cut  
New York   Increase Cut  
Oregon Increase   Cut Increase
Pennsylvania Increase Increase    
Rhode Island Increase Increase Cut Increase
Vermont Increase Increase   Increase
Total Increases 11 12 0 7
Total Cuts 0 1 11 0
The "Progressive taxes" column shows changes in personal income taxes. The "Regressive taxes" column includes changes in sales taxes and excise taxes. States shown are those with the largest proportional tax changes (either increases or decreases) relative to the state's tax revenue during the 1990s as a whole. Increases and cuts shown are those that total at least 1 percent of annual tax revenue.


State Tax Systems Are Already Regressive

Most state tax systems already rely more heavily on regressive taxes than on progressive taxes.


Upper-Income Taxpayers Benefit the Most from an Income Tax Rate Reduction

 An "across-the-board" personal income tax cut may appear equitable.  But typically such a tax reduction provides the largest dollar benefits and the largest benefits as a share of income to the state's wealthiest taxpayers.  

 Consider, for example, a hypothetical state in which taxpayers are exempt from income tax on their first $15,000 of income and pay a five percent tax on all income above $15,000.  Under this system, a taxpayer with income of $10,000 pays no income tax, a taxpayer with income of $20,000 pays income tax of $250, a taxpayer with income of $40,000 pays tax of $1,250, and a taxpayer with an income of $100,000 pays tax of $4,250. 

 A 10 percent income tax rate reduction would reduce the five percent rate to 4.5 percent; the first $15,000 of income would still be exempt.  With this change, the taxpayer with income of $10,000 would receive no tax reduction at all.  The taxpayer with income of $20,000 would receive a tax cut of $25 or 0.13 percent of income.  The taxpayer with income of $40,000 would receive a tax cut of $125 or 0.31 percent of income.  And the taxpayer with the income of $100,000 would receive a tax cut of $425 or 0.42 percent of income. 

 In this example, the taxpayer with $100,000 income receives a tax cut equal to 0.42 percent of income, more than three times as great as the 0.13 percent of income tax cut received by the taxpayer with the $20,000 income. 

 The difference between the tax cuts for upper-income and lower-income taxpayers is more dramatic if the state has a graduated income tax rate structure.  For instance, consider a state in which taxpayers are exempt from tax on their first $15,000 of income but pay a three percent tax on their next $15,000 of income, a five percent tax on their next $30,000 of income, and a seven percent tax on all income over $60,000.  A 10 percent across-the-board rate reduction would reduce those rates to 2.7 percent, 4.5 percent and 6.3 percent respectively. 

 With that rate reduction, a taxpayer with income of $20,000 would receive a tax break of $15 or 0.08 percent of income.  A taxpayer with income of $100,000 would receive a tax break of $475 or 0.48 percent of income.  As a percentage of income, the higher-income taxpayer's tax reduction is eight times as great. 

How Have Legislated Tax Changes Affected Average Tax Burdens?

 Given the trends described in this report, one might reasonably expect that the state sales tax burden would have increased during the 1990s while the income tax burden would have remained level or declined.  From the preliminary data available through fiscal year 1996, however, that trend is not observable.  Based on the Census bureau's quarterly tax collections data — which are not complete and subject to statistical error but are the most recent information available — state personal income tax revenue was at the same level in 1996 as in 1989, about 2.1 percent of personal income.  Likewise, state general sales and excise taxes took about the same amount of personal income — about 2.7 percent — at both the beginning and the end of the period.  Total state tax burdens from 1989 to 1996 declined slightly, from 6.5 percent to 6.4 percent of personal income, while combined state and local tax burdens went from 10.7 percent to 10.5 percent of personal income. 

 The legislated tax changes do not necessarily translate directly into changes in the average percentage of income absorbed by the different types of taxes, because a number of other variables affect that percentage.  For example, the recent Census bureau report on 1996 income showed that virtually all gains in real income between 1989 and 1996 accrued to families in the highest-income 20 percent of the income distribution.  Since state income taxes are at least mildly progressive, that is, they take a greater share of income from higher-income taxpayers than those at lower incomes, income tax revenues tend to grow more rapidly when income growth is concentrated at the top of the income distribution.  State sales tax revenues, however, may grow more slowly when only the incomes of the relatively wealthy grow than they would with a more even distribution of income growth, since higher-income families save more and consume less of their incomes on taxable items.  

 If the legislated tax changes described in this report had not occurred, the share of income paid in income taxes on average across states would have increased slightly and the share paid in sales and excise tax revenues would have decreased.  State taxes probably would have become modestly more progressive.  The legislated tax changes, however, made the state tax systems more regressive than they would have been if no changes in the tax law had been made.  In addition, when the effect of the 1997 tax changes are accounted for and fully in effect, it is quite possible that state taxes in the late 1990s will be not only relatively more regressive than if 1989 policies had remained in place, but also absolutely more regressive than they were at the beginning of the decade. 

The net effect of these various taxes is generally believed to be regressive. One nationwide study based on the 1995 tax returns of non-elderly married couples estimated that low-income families pay about 12.5 percent of their incomes in state and local taxes, while families at the top of the income scale pay about 7.9 percent of their incomes in state and local taxes.(9)


How Would a State Know If it Is Making its Tax System More Regressive?

As long as current economic trends continue, states are likely to maintain healthy fiscal conditions. State revenue collections are growing because of a combination of low unemployment and strong returns on financial investments; the increased personal income and associated consumption translate into rising revenue for many states. Moreover, strong economies temporarily reduce the demand and the need for some social safety net programs. An additional — although short-lived — boost to state revenue may come from the 1997 reduction in the federal capital gains tax rate. To the extent the federal capital gains tax rate reduction persuades investors to cash in additional accumulated capital gains, a temporary rise in state capital gains tax revenue may occur over the next year or two.

These additional revenues and reduced expenditures are likely to spur continued discussion of tax cuts in many states. Although most governors will not submit their budgets and tax proposals until early 1998, at least one governor has already proposed a large reduction in personal income tax rates.

In considering tax changes, policymakers often seek to understand how those changes will affect taxpayers in various economic circumstances. At the federal level, for instance, both the Treasury Department and Congress' Joint Committee on Taxation prepare detailed tables analyzing how major tax proposals will affect taxpayers at a variety of economic levels. The information contained in these "distributional analyses" are frequently cited during debates and often affects the contents of final tax packages. This type of information, however, rarely is available as states debate tax policy.

It is not always clear to state policymakers or to the public whether a state tax proposal will increase or reduce the regressivity of the tax system. For instance, none of the state governments or legislatures in states with the largest tax changes in the last two years have published data that describe how those changes would alter the total tax burdens (including income, sales, excise and other state taxes) borne by various income groups or how the changes would redistribute taxes among taxpayers.(10)  If such information were available, it would allow for more informed debate by policymakers and the public.

There are no insurmountable logistical problems to producing such information. A few states do provide regular, comprehensive "tax incidence" analyses that tell legislators and the public how the state tax burden is distributed among the population and how specific proposals would change that burden. Such analyses require both the technical capacity to conduct such analyses and the procedural requirements that the analyses be conducted and made publicly available in a timely fashion during legislative debates.

Perhaps the best example is Minnesota, where the Department of Revenue issues a biannual "Tax Incidence Study." This study describes the distribution of major state taxes — personal income tax, sales and excise taxes, and property taxes — among taxpayers of various income levels. In addition to the biannual study, the state produces regular analyses of major legislative tax proposals.

A recent example of how these requirements work was provided by a proposal by the governor in the 1997 session to authorize tax-free education savings accounts. A Department of Revenue analysis showed that the bulk of the benefits from such a proposal would go to families with incomes over $100,000, an analysis that may have prompted the legislature to reject that proposal and instead enact tax breaks for education that were more tightly targeted to low- and middle-income taxpayers.

Even with Exemptions, Sales Taxes Remain Regressive

 In the New York Times article about this report, Raymond C. Scheppach, executive director of the National Governors' Association, questioned whether sales taxes fell heavily on low-income households.  He was quoted as saying:  

"You have to be real careful about sales taxes and whether they really are regressive.  A lot of states have exempted food, and many do not tax services, and so their sales taxes are designed to not be as regressive as an across-the-board sales tax."  (New York Times, 10/5/97) 

It is true that states can make their sales taxes modestly less regressive than they otherwise would be by exempting certain items from the tax base, but the sales tax remains significantly regressive.  Exemptions for food purchased for home consumption, prescription or non-prescription medicine, and residential utilities typically provide slightly larger tax breaks to poorer families, relative to their incomes, than to higher-income families.  For that reason, about 25 states exempt food, all states but one exempt prescription medicine, eight states exempt non-prescription medicine, and 20 states exempt residential electricity. 

However, no state — even those with substantial exemptions for consumer necessities — has a sales tax that even comes close to levying as high a tax burden on upper-income families as on lower-income families.  Even families in poverty have spending needs that go beyond food, medicine and utilities; those needs range from lightbulbs to school supplies to a host of other items, all of which are typically subject to sales taxes.  The main reason the sales tax is regressive is that low-income taxpayers expend a higher portion of their total incomes, and save much less, than upper-income taxpayers. 

A study by the Institute on Taxation and Economic Policy of married couples below retirement age found that in each of the 45 states with sales taxes, sales taxes on the poorest 20 percent of taxpayers are at least four times as high, relative to income, as sales taxes on the top 1 percent of taxpayers.  In many states — including some states that exempt food, medicine and other consumer items from the sales tax — the ratio is even higher. In nearly every state, sales taxes are more regressive than any other major state or local tax. 

Taxes Seem to Have Become More Regressive in the 1980s, Too  

 According to the Times, Scheppach also questioned whether inclusion of the tax increases of the 1980s would change the picture described in this report of increasingly regressive taxes.  In fact, it appears that the same pattern occurred in the 1980s as in the 1990s — that is, both regressive and progressive taxes were raised in the wake of the recession of the early years of the decade, but only progressive taxes were cut in the economic expansion or the middle and later years.  

Sales tax rates, and other regressive taxes on consumption, were systematically raised throughout the 1980s.  From 1980 to 1989, the average state sales tax rose from about 3.9 percent to about 4.8 percent.  In 1980, only three states had sales tax rates of at least six percent; this grew to 10 states by 1989.  Nor did states move noticeably in the 1980s to make their sales tax bases less regressive.  The number of states that exempted food for home consumption, for instance, stayed basically level during the decade at 25.  Gasoline and cigarette tax rates also increased during the 1980s. 

While taxes paid by lower-income taxpayers were rising throughout the 1980s, the personal income tax rates paid by the highest-income taxpayers rose only during the recession of the early 1980s and then returned to pre-recession levels — or lower — by the end of the decade.  The average personal income tax rate rose from 6.3 percent in 1980 to 6.8 percent in 1985, then dropped to 6.2 percent in 1989.  The number of states with personal income tax rates of seven percent or higher rose from 16 in 1980 to 18 in 1985 and then returned to 16 in 1989.  In short, in the 1980s  average sales tax rates were higher, and personal income tax rates were lower, at the end of the decade than at the beginning — a pattern repeated in the 1990s.

 Another state with both the technical capacity to carry out incidence studies and the legal structure to ensure that the studies are actually done is Texas. A new state law patterned after the Minnesota law will require the Texas Comptroller to report before every biannual legislative session on the incidence and distribution of local school property taxes and all state taxes that contribute at least 2.5% of state tax revenue. (The state's next legislative session is scheduled for 1999.) The law also requires the Comptroller to report biannually on the distributional effects of tax breaks, tax exemptions, and other special provisions in the tax code. In addition, during legislative sessions committee chairs may order distributional analyses by the Texas Legislative Budget Board of tax changes of $20 million or more.

An analysis by the Legislative Budget Board in 1997 showed that a sweeping proposal submitted by the governor to cut property taxes and raise some sales and business taxes — House Bill 4 — would have provided the largest benefits, both in dollar terms and percentage terms, to the wealthiest Texans. The proposal did not pass.

A number of other states have produced multi-tax distributional analyses in the past but have not maintained the computer model needed to provide these analyses on a regular basis. Examples include Wisconsin and Connecticut. The Connecticut legislature's Office of Fiscal Analysis contracted with a private firm to create a tax incidence model in the early 1990s as it was considering several dramatic options to cover a large state budget deficit. The information the model provided about the distribution of taxes, showing that sales taxes were particularly burdensome on low-income residents, helped persuade legislators to reject major sales tax hikes and instead institute a modestly progressive income tax.

Unfortunately the Connecticut model has not been maintained. The data underlying the model have become obsolete and the state has not updated the system. By the time new tax proposals were contemplated in the middle 1990s, the model was essentially useless. Information on the combined impact of income tax, consumption tax, and personal property tax changes enacted in recent years is not available.

Some states produce various types of distributional analyses using income tax data alone, largely because income tax data are the easiest to acquire. Examples include Arizona, California, Iowa, and Maryland where legislators and the public have been able to learn how at least some proposed income tax changes would affect taxpayers in various income brackets. Connecticut also can analyze personal income tax change. These data allow voters and legislators to see the extent to which upper-income taxpayers receive the largest tax relief. But by disregarding other major state taxes, such as sales and excise taxes, these limited analyses ignore the failure of income tax reductions to direct tax relief to the many low- and moderate-income taxpayers with little income tax liability but who pay substantial amounts of state sales and excise taxes.

Regressive Tax Systems Can be Counterproductive for Other State Goals

Regressive tax systems pose particular problems for state and local governments as they grapple with their new responsibilities for maintaining the safety net for low-income families. By definition, a regressive tax system imposes its heaviest burdens on lower-income families. These burdens include not only sales and local property taxes but often personal income taxes as well.(11)

With the enactment of federal welfare reform, states across the country have been developing plans to reduce these caseloads and help families move from welfare to work. In most cases, the policy debate at the state level has focused on changing the provisions of the welfare program itself, while the impact of state tax policy on low-wage working families has been overlooked.

The failure to consider the impact of state tax systems on low- and moderate-income families as a part of the welfare reform debate is a serious omission. Low-income families often face high marginal tax rates; that is, as a family's income rises up to and beyond the poverty level, the combination of higher taxes and the loss of means-tested benefits (such as food stamps) consumes a significant share of these increased earnings.(12) In addition, the expenses of working, such as child care and transportation, often absorb a large proportion of the earnings of low-income workers. Thus, as part of a larger strategy to "make work pay" for low-wage workers, it is particularly important that the tax burden on those workers remain low.

The impact of state tax systems on poor families has been increasing in importance in recent years because the number of working poor families has been growing across the United States. The increase in the number of families that have earnings from work but remain poor can be seen by comparing poverty rates in 1995 with those in 1979, two years when the economy was growing and unemployment rates were similar. The poverty rate for families with children in which the head-of-household worked climbed from 7.7 percent in 1979 to 11.0 percent in 1996, an increase of about 40 percent.(13)

In 1996, about 9.5 million poor children — two out of every three poor children — lived in a family with a working household member.

The healthy economic condition of the middle 1990s offers states an opportunity to address the problem of taxation of poor families — both in the interest of fairness and in order to further the objective of allowing parents who work to support their families adequately. As the trends detailed in this paper suggest, however, this has not been a major thrust of state tax policy. A first step to refocus that policy could be the development of state capacity to understand the distributional choices that are being made.   


State Tax Changes by Type and Year of Enactment
(in billions of dollars)

  1990 1991 1992 1993 1994 1995 1996 1997
Progressive Taxes
Personal income tax $2.8 $5.1 ($0.3) $0.7 ($1.2) ($4.6) ($2.8) ($1.3)
Regressive Taxes
General sales tax $2.3 $4.9 $0.7 ($0.1) ($0.0) ($0.1) ($0.4) ($0.6)
Motor fuels and vehicle sales taxes $0.8 $0.7 $0.5 $0.2 $0.0 $0.0 ($0.1) $0.5
Cigarette and tobacco taxes $0.5 $0.2 $0.3 $0.5 $0.1 $0.1 $0.2 $0.2
Alcohol taxes $0.2 $0.2 $0.0 $0.0 $0.0 ($0.0) ($0.0) $0.0
Total regressive taxes $3.9 $6.1 $1.6 $0.5 $0.1 $0.0 ($0.3) $0.1
Other Taxes
Corporate income* $1.2 $1.6 ($0.0) ($0.2) ($0.7) ($1.0) ($0.6) ($0.3)
Other taxes $1.3 $1.4 $0.5 $0.1 ($0.1) ($0.4) ($0.2) ($0.2)
Total other taxes $2.5 $3.0 $0.4 ($0.1) ($0.7) ($1.4) ($0.9) ($0.5)
Total $9.2 $14.1 $1.7 $1.1 ($1.8) ($6.0) ($4.0) ($1.7)
* Includes other business taxes.
Totals may not add due to rounding. Figures are not adjusted for inflation or economic growth.
Center on Budget and Policy Priorities analysis of data collected by the National Conference of State Legislatures and information from state fiscal offices.


About the Data in This Report

The primary sources for the aggregate dollar amounts of tax changes in the years 1990 through 1996 are a series of annual reports issued by the National Conference of State Legislatures (NCSL) entitled State Tax Actions (prior to 1993, State Budget and Tax Actions). NCSL collects its estimates of the effects of tax changes from state legislative fiscal offices, and reports these changes by state and by type of tax. The 1997 data in this report are based on preliminary data from NCSL's forthcoming report for this year, combined with information collected directly from state legislative fiscal offices.

The NCSL data generally reflect the effects of tax changes implemented in the fiscal year following the one in which the change was enacted. For instance, the aggregate tax changes reported in 1996 State Tax Actions are based on estimates of revenue impacts for fiscal year 1996-97 (the 12-month period which in most states ended June 30, 1997). In this report, where possible the amount of tax change is shown as an annualized, full-year impact.

The dollar totals in this analysis do not exactly equal the total tax changes reported by NCSL in each of the years covered. Adjustments were made for a variety of reasons. In most cases, the adjustments were based on series of analyses produced by the Center for the Study of the States from 1991 to 1995.(14)

Unemployment insurance taxes and motor vehicle license fees, which are not included in the NCSL data, also were not included in this analysis.

End Notes

1.  Tax increases were not states' only response to the fiscal crisis of the early 1990s; states also reduced spending sharply, in part by cutting programs for low-income residents.

2.  The fact that surplus revenue may create political pressure for tax cuts does not necessarily mean that states could afford the tax reductions they enacted in the mid-1990s. Several states in 1997 — for example, Maryland, New York and Rhode Island — have passed major phased-in tax cuts that, according to some official projections, may not be paid for in future years by budget surpluses and instead probably will necessitate future tax hikes or spending cuts.

3.  These figures do not include the 1994 increases in Michigan's sales, excise and other taxes that were enacted as part of a school financing reform package, or income tax credits to offset property tax payments enacted in 1997 as part of Minnesota's property tax reform legislation. Those changes were excluded because they were part of broader tax reforms at both the state and local level; an analysis that included the effects of those changes would properly also include the associated changes at the local level, which are beyond the scope of this analysis. Had this analysis included those state-level changes in Minnesota and Michigan, it would have shown an even larger reduction in personal income taxes and an increase rather than a small decrease in consumption taxes during the middle 1990s.

4. The nine states reducing rates were Arizona, Iowa, Maryland, Michigan, Nebraska, New Jersey, New York, Ohio, and Rhode Island. Oregon is in a special category. Oregon's major tax cuts were refunds in 1995 and 1997 that were required by the state's "kicker" law. Although the refunds were not technically rate reductions, they had a similar distributional effect as rate reductions because the refunds were proportional to the previous year's tax liability. The two states that cut personal income taxes without cutting top rates were Arkansas and Connecticut. Arkansas' personal income tax cut included a variety of new and expanded deductions and exemptions, including an increased standard deduction, a tax credit to offset a portion of federal payroll taxes, an expanded poverty exemption, an increased child care credit, and other changes. Connecticut, which has a two-tier graduated income tax, expanded the amount of income subject to its lower 3 percent bracket so that a smaller portion of families' income is subject to the 4.5 percent rate. Connecticut also increased the amount of personal property taxes tapxyers may deduct from their income tax.

5. These firgures, and those in the accompanying table, reflect personal income tax rates for married couples filing jointly.  In states where federal income taxes are deductible, the state rate were adjusted to reflect the actual marginal state tax ratre on income.  Similarly, state income taxes levied as a percent of federal tax liability are counted at the equivalent state rates.  As a result of the increases in the fedral top rate between 1989 and 1993, the income tax rates in three states rose, including one state (Rhode Island) in whixh the top rate increased from less than seven percent to more than seven percent.  The increase in the federal rate also had the effect of reducing to prates in six states that allowed federal taxes to be deducted, including two states (Montana and Iowa) in which the top rate declined from above seven percent to less than seven percent.

6. One of the states, Connecticut, cut its sales tax significantly when it instituted a new personal income tax.

7. A number of other states have enacted significant low-income tax relief in recent years, usually as a part of tax changes that provide the bulk of their benefits to higher-income taxpayers. These include North Carolina and Louisiana, which reduced their sales taxes on food, and New York, Massachusetts, Minnesota, Oregon and Wisconsin, which enacted or expanded state earned income tax credits and other low-income tax relief in recent years.

8. Statewide property taxes, which are important sources of revenue in some Western states, are often viewed as regressive when levied on residential property but progressive when levied on business property. As for corporate income taxes and other business taxes, which are also major revenue sources in certain states, many analysts believe that the national tax on corporate income is borne primarily by the owners of corporations. Since these owners are disproportionately wealthy, and since the tax has been shown to primarily reduce their profits, the tax is progressive. There are differences of opinion, but little research, on whether the state corporate income tax is as progressive as the national tax. In any case, both statewide property taxes and corporate income taxes in most states are much less significant sources of revenue than other state taxes.

9. Citizens for Tax Justice and the Institute on Taxation & Economic Policy, Who Pays? A Distributional Analysis of the Tax Systems in All 50 States, June 1996.

10. This conclusion is based on an informal survey of practices in Arizona, California, Connecticut, Massachusetts, New Jersey, New York, North Carolina, Ohio and Pennsylvania, each of which has enacted total tax changes in the last three years exceeding $500 million.

11. Although the federal government and some states exempt families with poverty-level incomes from income taxes, other states levy income taxes on families with incomes as low as $4,000 a year.

12. At some income levels, particularly those modestly above the poverty line, workers face marginal tax rates ranging from 67 percent to 79 percent from the combination of federal income and social security taxes, the phaseout of the federal earned income tax credit, and the loss of food stamps.

13. Part of this increase reflects growth in the share of working families with children that are headed by a single female parent, since this group is much more likely to be poor than are two-parent families. Nevertheless, the poverty rate among families with a working parent has grown for both single-parent and two-parent families.

14. Steven D. Gold, "1995 Tax Cuts: Widespread But Not Revolutionary," December 1995; "State Tax Cuts: 1994 as Prelude to 1995," January 1995; "Tax Increases Shriveled in 1993," December 1993; "The Anatomy and Magnitude of State Tax Increases in 1992," January 1993; and "How Much Did State Taxes Really Go Up in 1991?", February 1992. All published by the Center for the Study of the States, Nelson A. Rockefeller Institute of Government, Albany, NY.