Back to Sales Taxation in Kentucky: Problems and ProspectsFrom Financing State and Local Government
p. 39-56, published 2001
The interpretation of fairness is largely subjective. Nevertheless, some broad principles emerge. Kentucky’s state and local tax system carves out significant exemptions on necessities and provides income tax credits for the very poor. The tax burden is fairly proportional over most income ranges except the upper-income levels, where it is somewhat regressive. Addressing inflation, which has robbed Kentucky’s income tax of its progressivity, would enhance vertical equity. However, Kentucky’s skewed income distribution presents unique challenges to significant tax reform. Horizontal equity issues remain complex but Kentucky’s many exemptions and exclusions that riddle its large taxes create opportunities to promote equitable and efficient tax policy.
Traditional studies of tax equity have concentrated on the ability of citizens to pay their share of the tax burden. It is easy to discern differences in ability to pay as one drives through the Commonwealth, passing from gated-estate communities to high-density starter-home neighborhoods, from small community mobile home parks to urban public housing complexes, and from small-town neighborhoods to single-family farms. Drive Route 460 through Georgetown and out into the countryside and you’ll experience the skewed bell curve of Kentucky’s income distribution, which reflects its demographics, economy, and past investments in education.
The question at hand is how state and local taxes affect the distribution of income available for private spending. Most equity studies look at ability to pay as the basis for assessing the equity of taxes. This chapter looks at various studies done in the past and examines the impact of state and local taxes on Kentucky households across various measures of ability to pay. The chapter also discusses the effects of various exemptions and credits and changing technology on tax equity.
The price of equity is also relevant in any discussion of tax modernization. If fairness were cheap, we’d likely see more of it; however, altering the equity of a tax system always requires sacrifices. How much is Kentucky willing to pay in order to improve the equity of its tax system? This chapter looks at the cost of providing tax relief to low-income households in light of Kentucky’s current income distribution.
Finally, the chapter examines the effect of changes in economic factors on the incidence of Kentucky’s taxes. For example, failure to index rate classes in the individual income tax has dramatically altered the size and burden of the tax. The implication is that the current incidence of the income tax is not necessarily the result of consistent attention to and tinkering with Kentucky taxes, but rather, the rigidity built into the taxes as originally enacted.
It has become popular practice in matters of tax reform to attach the principles of fairness and equity, as in, “the resulting tax systems should be fair and equitable.” This practice raises suspicions that fairness and equity are not substitutes. Fair is defined as just, impartial, or being in accordance with rules, logic, and ethics. Equitable is defined as fair, just, or impartial. Fairness in popular usage connotes the latter meaning, in the sense of a fair shake, that somehow, for example, tax agents are not singling you out for bad treatment. This is essentially the meaning of impartiality, although tax agents can impartially give all persons bad treatment, which no one would consider just or fair. Justice suggests blindness to individual differences, but often hinges on the facts of individual cases. From here on, I’ll use equity to embrace all such related concepts of fairness, justice, and impartiality.
The public finance literature has popularized two definitions of equity: vertical and horizontal. Vertical equity concerns how government affects persons who differ in ability to pay and is typically discussed in terms of burdens of government across income levels. Horizontal equity, equal treatment of equals, is generally accepted as an outcome of broad-based tax and expenditure systems, whereby two “equal” taxpayers can’t legally pay different tax liabilities solely because of some differences in their spending, saving, or investment actions. Both concepts present opportunities for comprehending optimal tax systems, but also considerable definitional and operational hurdles.
The current distribution of income among income taxpayers in Kentucky is shown in Figure 13. According to the Census 2000 Supplementary Survey,(23) average household income in Kentucky was $42,852 vs. $55,253 for the United States. The median Kentucky household earned $32,843 compared with $41,343 for the U.S household. These income figures include most sources of income including some transfers such as Temporary Assistance for Needy Families (TANF) and food stamps. Kentucky’s median income was the fifth lowest in the nation.
Figure 13: Distribution of Kentucky and U.S. Household Income, 2000
These data reveal the distinct nature of Kentucky’s income distribution: 39 percent of Kentucky households earn less than $25,000 compared with 30 percent for the nation as a whole. To put this in perspective, to make the low end of Kentucky’s income distribution like the average state, Kentucky would have to lift 140,000 families above $25,000. Kentucky at the high end is no different: 7 percent of Kentucky households earn more than $100,000, compared to 12 percent across the nation. To match the high end of the distribution of the average state, Kentucky would have to lift 83,000 households over the $100,000 threshold. The middle of the distribution is disturbing as well; if one considers “middle class” to be $35,000 to $75,000, Kentucky’s middle-class is missing 37,000 households.
Figure 14: Distribution of Kentucky Taxpayers by Federal AGI Class, 1998
Recent years have produced several studies of state and local tax burdens. Because of the effort involved in calculating incidence of individual taxes across 50 states, most of these studies have focused on the average burden or tax effort. However, two studies have provided burden estimates by income classes across multiple states. Both studies use income measures that exclude transfers. As a result, effective tax rates are overstated at low-income levels relative to measures of true ability to pay. Similarly, neither study includes the imputed rent to home ownership as income, thereby overstating effective rates across middle- and upper-income classes. Furthermore, the studies focus on families rather than households, thus failing to represent, among others, single households and the elderly. The 1999 Barents study estimated the effect of Kentucky state and local taxes on families of four at various income levels.(24) The results are shown in Figure 15. Kentucky’s tax structure consumed about 8 percent of income no matter where one falls in the income distribution. A moderate amount of progressivity occurs for low-to-middle income families, which diminishes at high-income levels. The income tax is moderately progressive, while sales and excise taxes are regressive. The mix of property taxes is fairly proportional over all but the lowest income classes.
Figure 15: State and Local Tax Burdens on a Two-Parent Family of Four in Kentucky, 1998
In 1996, the Citizens for Tax Justice published a 50-state analysis of state and local tax systems that assessed the burden of the major taxes by income quintile.(25) The study used fairly standard incidence assumptions and methodology in deriving effective tax rates for nonelderly married couples in 1995. The data on property taxes is outdated, failing to reflect the significant exemption of intangible property precipitated by the St. Ledger(26) decision and reductions in motor vehicle property tax valuation since 1995. Taking those changes into account would lower Kentucky’s property tax burdens across the board. The results for Kentucky are shown in Figure 16. The data show that Kentucky’s state and local tax system is essentially proportional to income across all but the highest-income class. Kentucky taxes take approximately 10 percent of household money income, with the progressivity of the income tax offset by regressivity in sales and excise taxes.
Figure 16: State and Local Taxes on Non-elderly Married Couples in Kentucky, 1995
The two studies paint a similar picture for Kentucky state and local taxes. The tax system is fairly proportional over most income ranges except the upper-income levels. A flat income tax, heavy reliance on sales taxes, and low property taxes contribute to this outcome.
While effective tax rates are instructive, they also require some interpretation. The burden of taxes is generally spread proportionally across income, but the same is not true across households. Figure 17 uses the data on effective rates and income by quintile to calculate the taxes paid by quintile. The data show that the richest 20 percent of Kentucky households pay 43.9 percent of state and local taxes, while the poorest 20 percent pay 4.6 percent of taxes. Essentially, the average household in the highest quintile pays as much tax as 10 poor households put together. Again, the question of whether this is too high or too low depends on one’s view of vertical equity.
Figure 17: Percent of Tax Paid by Income Class
These studies are useful as well in assessing how Kentucky stacks up against other states. Using the data from the Citizens for Tax Justice study, Kentucky’s overall tax burden on the lowest quintile ranks 37th highest in the nation.(27) Kentucky’s ranking by tax types is shown in Table 5 (1 = highest burden). The data reveal why advocates for working poor families give Kentucky’s personal income tax significant scrutiny. Taking into account the property tax changes mentioned earlier would lower Kentucky’s property tax rankings even further. Kentucky’s sales tax burden ranks below the median state for nearly all Kentuckians.
Table 5: How Kentucky Ranks in the United States Based on the Citizens for Tax Justice Study
Table 6 shows how Kentucky ranked among 16 competitor states in the Barents study.(28) A pattern similar to the Citizens for Tax Justice study emerges. Kentucky’s exemption-rich sales tax and moderate property taxes soften much of the burden on very low-income earners but fail to compensate for the relatively high income tax on the lower- to upper-middle class. The low-income credit provides relief for the very poor but fails to cushion what might be called the working poor from much of the 6 percent income tax rate.
Table 6: How Kentucky Ranks in the South: Barents Study
Analyzing trends in tax equity can be revealing. The equity of tax systems change over time because of incremental exemptions, rate changes, court-driven exclusions, and exemptions and deductions that weren’t indexed to inflation. In fact, some of the worst criticism of Kentucky’s tax codes concerns its failure to adjust to changing costs of living.(29) Rate classes, low-income credit thresholds, and personal credits have largely remained at their original levels.(30) The current rate brackets, rising from 2 percent on the first $3,000 of taxable income to 6 percent of income above $8,000, are as they were in 1950 when per capita income in Kentucky was $990.(31) The effect has been to remove a large degree of progressivity from the tax. The low-income credit, established in 1990, remedied some of this problem, but the failure to index its thresholds has weakened its effect over time.
In order to discern the magnitude of the effect of nonindexing on tax burdens, we calculated what the current income tax brackets, credits, and standard exemption would be had they been indexed to the Consumer Price Index in the year they were established. Table 7 compares the current law with this hypothetical indexed law. The most notable result is that the 6 percent marginal tax rate would start at $57,000 rather than the current $8,000. Families with taxable income less than $35,000 would face only a 4 percent marginal rate. Furthermore, the standard deduction would be $2,700 larger, and a family of four would receive $380 more in personal tax credits.
Table 7: Kentucky's Income Tax: Current Law (2000) vs. Indexed
The effects of indexing were derived using the Individual Income Tax Microsimulation Model, which uses a large sample of tax return data to calculate tax liabilities of Kentucky taxpayers under the current law and under various tax code changes. As one would expect, one primary effect of indexing is to remove substantial numbers of taxpayers from the tax rolls. Of 1.4 million currently taxable returns, 188,000 would show no tax liability; 170,000 of these returns had less than $20,000 in taxable income.
Other effects include doubling the total amount of credits taken, a $130 million increase, increasing deductions by $3.4 billion, or 42 percent, and reducing the number of itemizers by 200,000. However, the fiscal impact of this degree of indexing is enormous, reducing income tax revenues by $1.2 billion, or 47 percent. Because it is unrealistic to assume the state would have reduced revenues by $1.2 billion, it is necessary to adjust the figures to keep the changes revenue neutral. The state would likely have substituted other tax revenues for at least some of this loss in income tax. The overall effect on tax burdens depends heavily on how this revenue would be raised. For simplicity and because the overall incidence of state taxes tends to be fairly proportional, the analysis makes up the $1.2 billion through a tax that is proportional to income.
Table 8 shows how indexing and proportional taxation affect the average tax paid at various income levels. For Kentuckians in the lowest income class, the new tax system would actually be more onerous than the old one. This results because this income group pays very little income tax, mostly due to the low-income credit. Indexing further reduces their tax, but replacement of the lost revenue through proportional taxation is particularly burdensome. This is not just a theoretical exercise; local governments throughout Kentucky have increasingly turned to occupational taxes to augment their budgets. These local income taxes grew 86 percent from 1992 to 1999 compared with 51 percent for Kentucky state income taxes and 55 percent for local governments across the nation. Many of these taxes are strictly proportional to income and offer no deductions or credits.
Table 8: Effect of Revenue-Neutral Indexing by Income Class
The working poor fare better under indexing and do so even under revenue neutral taxation. The same holds true for households with incomes up to $100,000. This result shows the effect of lower tax rates over more of their income and significant deductions and credits.
Households with income between $100,000 and $200,000 fare worse, but only marginally so, as proportional taxation eats up the benefit of large itemized deductions in the income tax. The highest-income group is made significantly worse off, with their average tax bill rising by more than a third.
This extreme example of full indexing of income tax thresholds and credits is a theoretical exercise to show the effects of nonindexing on the incidence of the tax. Because the indexed tax generates so little revenue, it is unlikely that the current rate structure would have remained. Nonetheless, the same analysis could be extended to various scenarios of indexing, rate structures, and revenue capabilities. The results would be the same: most Kentuckians would have been better off, and the tax system would have been more progressive at the high end, if the income tax had been indexed and augmented with a proportional tax.
Advances in information technology have also raised equity concerns for state and local governments. Recent estimates of the avoidance of sales tax and the subsequent evasion of use tax facilitated by the Internet suggest that Kentuckians will reduce their tax burdens by $84 million dollars per year in 2001, rising to $286 million in 2006.(32) Internet purchases require certain assets that are not distributed evenly across the income distribution including computer hardware, software, Internet access, credit cards, and secure points of delivery. The Kentucky Long-Term Policy Research Center found that computer ownership and Internet access were strongly related to income.(33) Among those Kentuckians who have made online purchases, only 4.8 percent have incomes under $20,000 a year. By contrast, one third of online shoppers in the Commonwealth reported annual incomes over $50,000. The equity issue at hand is whether the Internet provides substantially more opportunities to evade taxes for certain taxpayer groups. This so-called “digital divide” makes the sales and use tax more regressive than it has been in the past. Because this change in the burden was largely precipitated by changes in technology rather than public policy, it can be viewed as a deviation from the desired incidence of the tax. Addressing sales and use tax evasion is a complex federalist issue involving court decisions, the Interstate Commerce Clause, and congressional agendas.
The degree of the equity problem is not yet known. Some would argue that these discrepancies in the ability to shop across state lines have existed for years; the recent boom in business-to-consumer (B2C) transactions has merely aggravated a long-standing problem. Furthermore, of the $84 million in 2001 e-commerce use tax evasion, 90 percent is estimated to be from business-to-business (B2B) transactions.(34) As a result, it is clear that having Internet access is not the only means of benefiting from use tax evasion; one merely has to purchase goods from firms that keep costs of inputs low by evading their business use tax liabilities. As such, it is likely that the extent of the equity problem caused by the digital divide is fairly small. However, any tax subject to increasing opportunity for evasion, regardless of equity considerations, should be addressed as a policy concern.
True horizontal inequities can only be discerned when comparing individuals or households in similar situations. If person A and person B start out in the same place, make similar decisions, and yet witness significantly different results (on average), some horizontal inequity exists. In Kentucky, the question of horizontal equity has arisen in several parts of the tax code. Here are but a few examples.
Kentucky sales and use tax currently exempts drugs prescribed by a physician and dispensed by a registered pharmacist, an exemption valued at approximately $170 million. In the case of an individual consuming prescription drugs administered at a doctor’s office or nursing home, there is a sales tax liability. A court decision in the Humana(35) case introduced a second issue: for-profit hospitals with their own pharmacies are now exempt when they administer drugs, while doctor’s offices, and nursing homes, remain taxable. Let us discuss these two issues in turn.
Self- vs. Provider-administered Drugs. Whether this is horizontally inequitable depends on whether the good consumed is the same. If the drug is administered at the doctor’s office, presumably this is the case because the service of a trained professional conveying the drug has some value. It may well be that in the absence of a controlled environment for administering the drug, this treatment may have little value or, in fact, be harmful to the patient, in which case the drug imposes a cost on the patient. Separating the value of the drug itself from the value of administering it properly raises efficiency and compliance questions that would likely swamp equity concerns. Furthermore, if professional administration of drugs is strongly correlated with the type of disease, e.g., diabetes, cancer, renal disease, the tax code discriminates by disease. This outcome is unsavory but will arise whenever certain transactions are exempted from a tax base. The relevant question is whether the efficiency concerns associated with proper determination of the market value of the good vs. the service outweigh the equity concerns.
Hospitals vs. Nonhospitals. It is difficult to defend the concept that the taxability of administered drugs should depend on the nature of the provider institution. Logically, hospitals are collections of provider offices that differ from individual providers’ offices in scale and scope. The nature of the product is the same. Furthermore, the distinction between a hospital and a nursing home is at best a blurry one in terms of the nature of the product provided. As such, it is likely that the current tax law as it pertains to prescription drugs has horizontal equity problems. This example also underscores the difficulties presented by attempts to enhance vertical equity in the tax code as well as the complexity of tax administration.
The exclusion of pension income from the individual income tax also raises equity concerns. First, it treats retirees with different sources of income differently, favoring those households who saved for retirement by means of a pension or IRA relative to those who acquired stocks, bonds, real estate, or other assets, whose income or capital gains realization will be taxed. The farmer or small business owner who re-invested his or her savings into the business receive no such exemption. Furthermore, the exclusion provides current retirees a tax savings with a higher present value than those who will retire in future years, thus creating some intergenerational inequity.
Currently, Kentucky provides sales, corporation, and motor vehicle usage tax exemptions to certified businesses located in any of 10 designated zones scattered throughout the state. Furthermore, residents and uncertified businesses located in the zone can receive sales tax exemptions for building materials purchased. Current laws limit the number of zones to the existing ten. Several expansions of the existing zones have occurred in order to provide the tax incentive to more businesses. The result has been that two businesses located one block apart will face different costs of production and expansion because of these exemptions. Also, one family residing over the line from Knox County, for example, will pay 6 percent more for lumber and building supplies for renovating their home. In 1992, an Attorney General’s opinion declared portions of the enterprise zone statute unconstitutional. The opinion declared that limiting the number of zones to ten was unconstitutional on the grounds that it violated the Constitution’s ban against special legislation favoring certain jurisdictions over others. These opinions do not carry the force of law.
Overall, we are left with some clear-cut horizontal equity problems and others that raise questions. However, if the general principle holds that the narrower the tax base, the greater the likelihood and severity of horizontal inequities, a comparison of Kentucky’s tax base with other states should provide some guidance. Kentucky ranks somewhere in the middle of other states and has shown a recent trend toward narrowing the base. According to calculations by Bruce and Fox,(36) Kentucky’s sales tax base in 1995 was 46.5 percent of personal income, which means that of 45 states imposing a sales tax, 19 states have a narrower base than Kentucky. In the decade of the 1990s, Kentucky added or expanded 22 exemptions to the sales tax.
Analysis of equity isn’t straightforward. Even if we can all agree on the effect of taxes in the economy, there are several issues to tackle before discerning the appropriate reform. A few of these issues are addressed here.
Vertical Equity. Difficulties in assessing vertical equity arise in many aspects. First, effective rates of taxes across income classes vary substantially depending on the definition of income used. In a study of 1991 Census data, Browning estimates the ratio of the average income for the highest quintile to the average income of the lowest quintile.(37) The ratio for before-tax and before-transfer money income is 12.1. The ratio after taxes and some in-kind and cash transfers is 8.4. Finally, adjusting for differing household sizes across income levels (more singles in the low quintile) yields a ratio of 5.1. All ratios indicate a substantial level of income inequality; however, degree of inequality varies tremendously. Since vertical equity requires some judgment of how much inequality is preferred, agreement upon the income measure is critical.
The above analysis uses annual income as a measure of ability to pay. Because many households in the low-income quintiles consist of college students, the elderly, and young workers, current annual income will understate their lifetime ability to pay. As a result, their effective tax rates will appear higher than they really are. Students will spend out of future income and the elderly will spend out of past income, thus raising their sales, property, and excise tax burdens in any given year. Correcting for this problem reduces the regressivity (or raises the progressivity) of these taxes.
Even if policymakers can agree on the definition and term of income to be used as a base for the burden study, they may reach vastly different conclusions about appropriate reform, depending on their perception of fairness. Scholars suggest that people’s perception of vertical equity is altered by their current status and that uncertainty about future income increases the call for a more even distribution. Economists theorize that redistribution provides a public good beneficial to all of society, and that public sector redistribution of income can successfully augment private charity. However, none of these arguments guide policymakers in determining the optimal amount of vertical equity.
Horizontal Equity. Discerning whether equals are treated equally, of course, hinges on the definition of equality. Typically, equity studies treat two parties as equals if they have equal incomes. The Smiths and the Joneses may have identical incomes, identical 3-bedroom ranches on identical streets, live equidistant from the elementary school where their children attend, drive the same late-model Toyota cars and vote conservatively in all elections. Horizontal equity requires they pay the same amount in taxes. However, if the Smiths and the Joneses are equal, then any tax system will treat them equally, because they will make identical consumption, labor supply, investment, and savings decisions. Even a state-enacted hodgepodge of loopholes and preferences, as long as it doesn’t include anti-Smith or anti-Jones taxes, will yield equal burdens. No help so far, but this scenario is an unreasonable burden for the concept of horizontal equity to bear.
More realistic is the notion that Smith and Jones appear equal, but their different economic choices yield different tax burdens. Jones avoids some sales tax by raising ostriches and Smith, some income taxes by sending his children to pre-school. Jones eludes some property taxes by buying a used car, and Smith, some excise taxes by switching from cigarettes to smokeless tobacco. The myriad exemptions yield unequal taxation of what began as equal incomes, and are thus, inequitable.
Upon closer examination, however, we find that the equality of their incomes ends at the dollar measure. The Smiths’ $50,000 income comprises her $30,000 salary, his $19,000 salary, and $1,000 interest from a savings account, while the Joneses get $55,000 per year from interest on an inheritance minus the $5,000 annual loss from the ratite business that has yet to take off, so to speak. The Joneses play golf and tennis four times a week and enjoy gardening, wiffle ball with their children, and long walks in the park. The Smiths go grocery shopping after picking up their children from day care and after-school and spend weekends catching up on their yard work. Having paid off their mortgage, the Joneses earn the imputed rent on their home, while the Smith’s pay this “rent” to the mortgage company but take a tax deduction. The Joneses estimate the value of their leisure time spent with their children at $40,000 per year. The Smiths pay $7,000 each year for pre-school and after-school programs so that they can enjoy the opportunity to work 40 hours per week. Only a poll tax could secure perfect horizontal equity if the Smiths and Joneses are thought of as equals. A real-world system of taxes with relatively narrow bases will yield less horizontal equity. At best we’re left with the guideline that the narrower the base, the more opportunity for different treatment of so-called equals.
Except when presented with evidence that certain wealthy citizens and corporations paid no income tax, the general public rarely cries for more horizontal equity. This is not surprising; it is unlikely that equals can perceive unequal treatment, except in its extreme. For Smith to realize the inequity of the tax system, he would have to know who his equals are and how much they pay in taxes. There is substantial survey evidence that most taxpayers don’t know how much they themselves pay in taxes.(38)
In essence, horizontal equity and vertical equity are difficult to assess because for the former, it’s hard to agree on the right question, and for the latter, it’s hard to agree on the right answer.
The Trade-off between Vertical Equity and Efficiency. Enhancing equity requires a sacrifice in the tax system’s ability to raise revenues without disturbing the economy. Suppose that a head tax of $2,500 per person were placed on every Kentuckian.(39) Efficiency would require that the tax did not affect the relative prices of goods and services. Here, the only decision to make would be whether to leave the state. But let’s assume that is a costly option. In the absence of any means to avoid the tax, Kentuckians will not avoid the tax. They will pay out of existing cash flow, sell assets, borrow, avoid saving, reduce consumption, or some combination of the above, to pay the $2,500. The tax alters no prices, and therefore does not affect people’s decisions to spend or save, work more or less, rent or buy, or purchase a different set of goods. In essence, other than an income effect, the tax is benign, which adheres to a basic principle of tax policy. Efficient taxes don’t steer economic decisions.
The reason we have so few of these taxes is that they fail miserably on vertical equity grounds. Effective tax rates would naturally be higher for poorer families than for richer ones. Large families would be burdened more than small ones. Single millionaires would pay less than single mothers at the poverty level. So governments build progressivity into their tax codes.
Tax codes generate progressivity in two ways: exemptions and rising marginal tax rates. By far, the most popular means of increasing vertical equity is through exemptions of some kind. The sales tax exempts grocery food, some utilities, and most prescription drugs in order to carve out typical expenditures that constitute a large portion of low-income households’ spending. The property tax exempts part of the value of the homes of elderly citizens to alleviate burdens on their incomes. The income tax allows standard or itemized deductions and personal credits to account for common expenditures households make, and for those with low income, a credit against their tax bill, effectively exempting some income from tax.
While serving to promote progressivity, exemptions require, for a given size of government, higher tax rates. Thus, nonexempt income, purchases, and property exact a higher toll than otherwise. The tax on the incremental activity beyond the exempted amount bears a higher tax. For example, absent all exemptions, the sales tax could shoulder its revenue burden with approximately a 3 percent rate. With a 3 percent rate on more purchases, the tax is more efficient, alters fewer decisions on buying clothes, TVs, lathes, presses, or computers. Herein lies the trade-off of equity and efficiency. Fortunately, for equity’s sake, these altered decisions are difficult to discern, for the alternative choice is not made, and therefore can’t be counted.(40) If the high tax rate discourages Smith from buying more bicycles, the transaction never occurs, no receipt exists, and no neighbors ask about that shiny new bike.
Traditional measures of incidence use effective tax rates by income class to assess the burdens on households at various points of the income distribution. If the question is “Who pays what, and is this a small or large portion of their ability to pay?” these calculations are appropriate. Judgments can then be made about relative burdens and policymakers can then decide whether these relative burdens are appropriate. However, this is not enough, for it neglects the incremental effect of the tax code, which is an important part of any discussion of how government affects its constituents. Average burdens look at how much a household paid across all dollars they earned. This measure ignores the most relevant dollars to the taxpayer, i.e., the marginal dollars. If we ask the question “How does an increase in income of $100 affect one’s tax bill?” we get at how the government treats decisions to work, save, or invest, across different income classes.
Tax credits can also improve progressivity, but in doing so, raise marginal tax rates. Tax credits are sometimes preferred to straight exemptions because they can be targeted toward intended recipients. The low-income credit cuts the income tax by $57 million per year, but only for those with incomes less than $25,000. However, because the credit targets low-income households, the phase-out of the credit as income rises creates high marginal tax rates (see Figure 18). The credit forces the highest marginal tax rates on those helped by the credit. The following chart shows marginal tax rates by income level. Because of the low-income credit thresholds, an increase in income over that threshold can subject poor families to a marginal tax rate of greater than 14 percent. So, low-income families’ decisions to earn extra income are taxed more than for any other households. A marginal rate of 14 percent on Kentucky’s income tax means that the additional $1,000 is worth only $860. Subtract from this the marginal federal income tax, and the rate at which cash or in-kind transfers are reduced as earnings increase, and government has clearly reduced the incentive to work and invest in human capital.
Figure 18: Kentucky Income Tax: Effective Marginal Rates by Income Class
Tax Equity Is Only Half the Story—Benefits Incidence Matters. Studies of tax incidence usually hint at the logical necessity of including the spending side of government when determining the overall burden of government. Of course, if the question is whether the government budget treats its citizens fairly, it makes sense to include every dollar of tax paid and benefits received by citizens. On net, how much better or worse off are citizens across income classes as a result of the current size and nature of government? Using the economist’s definition of consumer surplus, the difference between what people are willing to pay and what they actually pay would be most appropriate to ascertain how government treats its various citizens. While this is probably the question policy analysts have in mind when they think of equity, they are sadly disappointed by the inability to measure benefits received. The indivisibility of public goods, free-rider problems, and the lack of pricing of public services renders benefits incidence all but impossible. The next best strategy is to ask, “Given the current budget, how does the tax share differ across income classes?” This inquiry has more value the more society adheres to an ability-to-pay principle of taxation but still falls short.
A look at average tax rates by income class reveals a roughly proportional tax burden in Kentucky, with some regressivity at the high end. While Kentucky fails to provide real relief to families at the poverty level through its income tax, it does this largely out of inattention to the effect of inflation on tax burdens. In recent years, the legislature has enacted indexing provisions in the standard deduction and in the pension exclusion, but the impact on the working poor has been minimal. Sales and use tax exemptions carve out a portion of a working family’s consumption spending, and property taxes have exacted a dwindling share of state and local taxes.
With regard to marginal tax rates, Kentucky fares a bit worse, charging the highest rates on working families near the low-income thresholds. And those working families not near the thresholds pay at marginal rates similar to those faced by Kentucky’s highest income families. It would be hard to argue that it is fair for any tax system to discourage the poor from economic activity more than the rich. Had Kentucky’s income tax system been indexed to maintain its original progressivity, the credit that creates these problems would have been unnecessary.
Still, reform necessary to significantly alter tax burdens among income classes carries a high cost. Because Kentucky has so many poor families and so few rich ones, redistribution that would really help working families near the poverty level would impose tax increases on the relatively few higher income families that would be politically burdensome for elected officials. It is hard to imagine a successful political strategy built around mildly helping the majority while significantly dunning the few. The existing process is the same one that created the generous menu of exemptions to do just the opposite.
Horizontal equity issues remain complex but present opportunities where sound tax reform can improve things. Kentucky’s many exemptions and exclusions that riddle its large taxes create opportunities for unequal treatment of taxpayers that appear similar. While unequal treatment is hard to discern (for taxpayers as well as policy analysts), it is also hard to justify. At best, the argument for eliminating horizontal inequities hangs on efficiency. Expanding the base and lowering the rates of Kentucky’s taxes is a win-win solution, mitigating equity effects while removing distortions caused by unnecessarily high tax rates.
The problem of the digital divide, or effect of new technology on the distribution of opportunities to evade taxes, is a symptom of a larger problem. Governments do best when first imposing rules on the populace over which they have jurisdiction. Optimizing these rules as circumstances change, creating innovative approaches, and instituting multijurisdictional solutions are far more problematic. For what it matters, the problem of use tax evasion attributable to Internet commerce may be more of a business tax rather than a personal tax issue. The resulting incidence of the noncompliance problem is such that the poor and the rich may be affected more or less equally.
In conclusion, it is not a stretch to claim that equitable tax systems are hard to come by and hard to maintain. Even the soundest of tax reform has to play off equity with other policy goals. A state such as Kentucky, long grappling with issues of investment in economic and social capital, may have had to forego a degree of equity to pursue growth and development. Furthermore, time, technology, and the economy are not always allies to well-intentioned legislation. At some point, the current tax structure in Kentucky will work against the goals of the Commonwealth; some may make the case that it already does.
To view a list of all chapters in this book, click here. To read the chapters in sequential order, please follow the arrows below.
Back to Sales Taxation in Kentucky: Problems and Prospects
Ahead to Business Taxes in Kentucky
* The author thanks John Scott, Wendell Butler, Michael Jones, Mary Lassiter, and Charles Schroff for their assistance on this project. The views expressed in this chapter are solely those of the author and do not necessarily represent the position of the Governor’s Office. Errors are, of course, the sole responsibility of the author. Return to text.
23 Available at http://factfinder.census.gov/home/en/c2ss.html. Return to text.
24 Barents Group LLC, Comparative Analysis of Kentucky’s Tax Structure (Washington, DC: Author, 1999) 71-80. Return to text.
25 Michael P. Ettlinger, John F. O’Hare, Robert S. McIntyre, Julie King, Neil Miransky and Elizabeth A. Fray, Who Pays?: A Distributional Analysis of the Tax Systems of All 50 States (Washington, DC: Citizens for Tax Justice, 1996). Return to text.
26 Hershel St. Ledger et al. v. Commonwealth of Kentucky. Return to text.
27 Citizens for Tax Justice. Return to text.
28 Barents. Return to text.
29 For a recent study of the burden of state personal income taxes, see Center on Budget and Policy Priorities, State Income Tax Burdens on Low-income Families in 2000 (Washington, DC: Author, 2001). Return to text.
30 The pension exclusion threshold adopted in 1997 is indexed for inflation as is the standard deduction starting in 2001. Return to text.
31 The Census Bureau estimates Kentucky per capita income in 2000 at $17,324. Return to text.
32 Donald Bruce and William F. Fox, State and Local Tax Revenue Losses from E-Commerce: Updated Estimates (Knoxville, TN: Center for Business and Economic Research, University of Tennessee, 2001) 8-9. Return to text.
33 Michal Smith-Mello, Michael T. Childress, Amy L. Watts and John Watkins, Challenges for the New Century, (Frankfort, KY: Kentucky Long-Term Policy Research Center, 2001). Return to text.
34 Bruce and Fox 6, and author’s calculations. Return to text.
35 Revenue Cabinet v. Humana, Inc., (1997-CA-000568-MR). Return to text.
36 Bruce and Fox 4. Return to text.
37 Edgar K. Browning, Public Finance and the Price System, 4th ed. (Englewood Cliffs, NJ: Prentice-Hall, 1994) 259. Return to text.
38 Income Inequality, Joel Slemrod, ed. (Cambridge University Press, 1994) 309-334. Return to text.
39 Per capita state and local taxes in Kentucky were $2,464 in 1999. Source: Census, State and Local Finances. Return to text.
40 For a good presentation of the problem, see Frederic Bastiat, in Selected Essays in Political Economy, G.B. deHuszar, ed., Foundation for Economic Education, 2000, Chapter 1. Return to text.