Corrected Full Text: 'Why Fairness Matters, Progressive Versus Flat Taxes.'

Taxes & The New Economy

Why Fairness Matters

Robert J. Shapiro

April 1996

ABOUT THE AUTHOR

[1] Robert J. Shapiro is the director of economic studies of the Progressive Foundation and vice president of the Progressive Policy Institute. He is also co-chairman of the Committee for Free Trade and Economic Growth, and contributing editor at The New Republic and International Economy. Previously, he served as a principal economic advisor to Bill Clinton in the 1992 presidential campaign, chief economic policy adviser in the Dukakis-Bentsen campaign, and legislative director and tax counsel for Senator Daniel Patrick Moynihan. Dr. Shapiro also has been a Fellow of the National Bureau of Economic Research and of Harvard University. He earned his doctorate at Harvard University and also holds degrees from the University of Chicago and the London School of Economics and Political Science.

CONTENTS

Preface

Summary Of Findings

I. The Case for Progressive Taxation

II. Equity and Economic Growth

III. Elements Of Proportionality and Progressivity in Federal

IV. The Distribution Of Federal Taxation

V. A Close Look at the Leading Tax Proposals

PREFACE

[2] This monograph is the fifth in a series of tax policy studies sponsored by the Progressive Foundation's Project on Tax Reform and Economic Growth and by the Progressive Policy Institute. The Project's mission is to design and assemble a tax reform program that is consistent with a progressive distribution of the tax burden, and can help promote stronger job and business formation, greater productivity, and higher family incomes. This is the first of two new reports outlining the essential features of such a tax reform program. Here we focus on the distributional issues in tax policy: We examine the standards for tax equity or fairness that should guide decisions about who should bear the tax burden, and apply these standards to the current federal tax system and the leading proposals to reform it.

[3] Throughout this analysis, we maintain a clear preference for progressive taxation. In our view, progressive taxes are particularly appropriate for a society that cares about free markets. The more free markets there are, the larger the rewards people can secure by leveraging their talents, resources, or just good fortune. America's markets are generally more free than those in other advanced countries. The happy result is that Americans who start with more talent, resources, or luck than others can prosper more here than elsewhere, and to a greater degree than those who start with less. Progressive taxes are a reasonable price to pay for the privilege of prospering more in such free markets, and a way of limiting the burden on the vast majority who have relatively less to leverage. And this is especially so in the present period, when America's markets are producing growing economic inequality.

[4] Previous publications in this series have included working papers on various aspects of the current federal tax system and certain alternatives to it. The first volume in the series presented two approaches to reforming the most complex area of federal taxation, business taxes: "Alternatives for Corporate Tax Reform" by Alvin C. Warren, Jr., and "U.S. Taxation of International Business Income: Agenda for the 1990s" by Gary Clyde Hufbauer and Joanna M. van Rooij. Volume 2 of the series offered an extensive analysis in comparative taxation by Perry Quick and Tom Neubig, "Comparing Tax Burdens in the United States and the Other G-7 Countries." The third volume examined the primary federal taxes paid directly by individuals, payroll, and income taxes: "Payroll Taxation in the United States: Assessing the Alternatives" by Jonathan Gruber, and "Tax Policy in the Second Clinton Administration: A Fantasy" by Joel Slemrod. Volume 4 explored the case for and against value-added taxes: "Value-Added Taxes and International Competitiveness" by David Raboy, and "The Economics and Politics of a Value-Added Tax" by M. Jeff Hamond.

[5] Our interest in tax reform is part of a broader commitment by the Progressive Foundation and the Progressive Policy Institute to help develop a new "Enterprise Economics" that can place the valid insights of both traditional liberal and conservative economics in a new global economic context. This approach seeks to focus policy not simply on financial capital or aggregate demand, but on the resources for greater innovation and higher efficiency and productivity by American firms and workers. In addition to tax reform, the policy agenda of Enterprise Economics includes budget reform to restrain the growth of public spending while revitalizing genuine public investment, and strategies to strengthen market competition by phasing out subsidies for particular industries and unproductive economic regulation.

ACKNOWLEDGMENTS

[6] This essay owes much to the insights and efforts of others. First among them is my colleague M. Jeff Hamond, a fellow at the Progressive Foundation and economic policy analyst at the Progressive Policy Institute. This report could not have been written without his splendid research, analysis, and general support.

[7] I also want to thank my other comrades at the Progressive Foundation and the Progressive Policy Institute who provided so much assistance in the preparation of this essay -- Will Marshall, Al From, David Datelle, Chuck Alston, Anne Saunders, Lee Lockwood, Lisa Davis, Eliza Culbertson, and Maureen Rehg.

[8] I want to express my gratitude to many fine friends and associates who have advised this project, including Frank Arnold, David Bradford, Gail Fosler, Joseph Gale, Michael Graetz, Gary Hufbauer, James Kiss, Cliff Massa, Tom Neubig, Rudolf Penner, Perry Quick, David Raboy, Alice Rivlin, Ed Stegemann, Eugene Steuerle, Joseph Stiglitz, Lawrence Summers, Joanna van Rooij and Al Warren. Finally, I gratefully acknowledge the generous support of Mars Inc., which provided an unrestricted grant for this project.

SUMMARY OF FINDINGS

TAX REFORM AND THE GOALS OF TAX POLICY

[9] THREE TESTS OF TAX REFORM. The tax system should promote three basic goals: simplicity, growth, and equity. The leading proposals to reform the current tax system fail to meet these tests.

[10] TAX REFORM AND SIMPLICITY. The Armey and Forbes flat tax plans would simplify the current tax code by eliminating many complicated deductions and credits. These proposals, however, still tilt the tax code to favor high income investors and influential industries because both plans replace current deductions and credits with two broader exemptions covering 1) all personal income from interest, dividends, and capital gains, and 2) all business income used for capital investment.

[11] Proposals to replace personal and corporate income taxes with a national sales tax would replace current complications with new ones and expand the bias towards high income investors and certain industries. The "USA Tax" plan to exempt new net savings and investment from income taxes includes only modest simplification.

[12] TAX REFORM AND GROWTH. Economic evidence demonstrates that these plans' broad, new exemptions for capital income or net savings and investment would do little to improve personal saving or growth. The U.S. personal savings rate remains less than 5 percent despite $120 billion in annual federal tax incentives for savings. In fact, the flat tax plans could reduce personal saving: By shifting the tax burden from capital sources to wages, these plans implicitly shift the burden from elderly households with high propensities to consume to younger working generations who save.

[13] If the point of tax reform is higher growth, all of these plans badly misdirect their incentives: The skills of workers and the pace of technological innovation are six to seven times more powerful in driving growth than increases in business fixed investment.

[14] TAX REFORM AND FAIRNESS. The flat tax would reduce tax fairness by shifting tax burdens from owners of capital to wage earners. The plan's new exemption for capital income would eliminate personal tax on 48 percent of the incomes of the richest 1 percent, as compared to just 6 percent to 10 percent of the incomes of the bottom 90 percent (Table 13, page 33). However, stockholders would be hit with a one-time loss, since the business side of the flat tax would depress the market value of existing assets and increase the taxable income of firms carrying debt. While providing a larger exemption for initial income than today, the plan would also increase taxes on the working poor by repealing the earned income tax credit (EITC).

[15] The national retail sales tax plan would eliminate every progressive feature of the current system, more than doubling the tax burden on the poor and substantially increasing taxes on the middle class. The savings-exempt income tax would maintain roughly the current distribution of the tax burden. It would offset the regressive effects of its exemption for new net savings and investment by providing workers with a new tax credit equal to their payroll tax payments and by applying sharply progressive tax rates to remaining taxable income.

THE CASE FOR PROGRESSIVE TAXATION

[16] TAXES AND ECONOMIC INEQUALITY. Since 1973, the average real income of American families has fallen at the bottom of the income scale, stalled in the middle, and risen at the top (Table 1, page 15). One of the reasons is that the United States maintains freer markets than other advanced countries. Talented, industrious, and lucky people can prosper here more than elsewhere, and the wealthy can enlarge their holdings more. Progressive taxes are a reasonable price for high income people to pay for these greater market opportunities, and a direct means of controlling the tax burden on the majority who are not so fortunate.

[17] As America's markets have become freer over the past 20 years, tax changes have increased the total burden on the middle class and reduced it for the wealthy. The flat tax and sales tax plans would accentuate this trend in relative tax burdens and worsen the general drift towards greater economic inequality. At the same time, the more egalitarian income distributions of a generation ago cannot be restored by raising taxes and increasing transfers -- at least not without imposing higher taxes on the struggling middle class as well as the affluent, or applying such high tax rates that growth could be hampered.

[18] WHAT AMERICANS WANT. Public opinion about what constitutes tax fairness is hard to measure but careful academic studies indicate that nearly two-thirds of Americans support progressive tax rates and tax burdens, while one-third would prefer a system with a truly flat burden on everyone but the poor.

BASIC FEATURES AND BURDENS OF THE CURRENT TAX SYSTEM

[19] PROGRESSIVE ELEMENTS IN FEDERAL TAXATION. Current federal taxation contains six progressive elements: (1) the EITC for working poor families, (2) an income tax exemption for initial income, (3) the phaseout of personal deductions at high income levels, (4) graduated tax rates, (5) corporate taxation, and (6) estate taxes. These elements do not produce tax burdens that rise sharply with income because other elements of the system have offsetting effects -- especially payroll and excise taxes, itemized deductions, and tax preferences for saving and investment.

[20] TAX BURDENS ON POOR AND MIDDLE CLASS FAMILIES. Considering all forms of federal tax, the current system is most progressive at the bottom. Poor families bear much smaller tax burdens than others, although they still pay about 5 percent of their incomes in federal taxes. A family's tax burden rises most sharply as it moves up from poverty to the middle class. Still, most middle class families pay a little less today than they would under a comprehensive flat or proportional tax on all forms of income (Tables 6 and 8, pages 26 and 28, respectively).

[21] TAX BURDENS ON THE AFFLUENT. High income people provide the bulk of all tax revenues, mainly because they earn the bulk of all income. Only families earning over $75,000 pay more today than they would under a pure and comprehensive flat tax system; beyond that, tax progressivity exacts moderate costs from most affluent families. At the very top, more than 70 percent of the taxes paid by the richest 1 percent of families represent no more than their proportional share of the costs of government, and tax progressivity reduces their incomes by less than 10 percent (Tables 9 and 10, page 29).

[22] TAX BURDENS ON DIFFERENT KINDS OF HOUSEHOLDS. Two groups bear lighter tax burdens than others: The elderly pay significantly less than younger households with comparable income; and, at low and moderate income levels, families with children pay somewhat less than other households with the same income (Table 7, page 27).

[23] IMPACT OF INCOME TAX DEDUCTIONS. The ability to claim certain deductions under the current income tax tends to shift the tax burden down the income scale because these deductions often cover a larger share of the income of affluent families than of middle class households. Also, the tax savings from these deductions increase as a family's tax rate rises (Table 3, page 20).

[24] THE BURDEN OF PAYROLL TAXES. The payroll tax system has generally regressive effects even when the value of retirement and medical benefits is taken into account.

[25] THE BURDEN OF BUSINESS TAXES. Federal taxes on business are generally progressive, because they fall mainly on the high income people who own most business capital. However, among the owners of capital, federal corporate taxes indirectly favor high income investors over middle class investors.

[26] THE BURDEN OF EXCISE TAXES. Excise taxes are generally regressive, but less so than is often assumed.

[27] THE BURDEN OF ESTATE TAXES. Federal estate taxes are progressive, but they affect only the richest 1 percent of families. Moreover, these estate taxes apply the same effective tax burden to estates of $2.5 million and more as all other federal taxes apply to people earning $20,000 to $40,000 a year (Table 4, page 24).

I. THE CASE FOR PROGRESSIVE TAXATION

[28] Tax reform is Washington's latest answer for the discontent Americans feel with national politics and their own economic prospects. Several presidential hopefuls have promised to replace the current system with a new flat tax or a national retail sales tax, and others would substitute a new consumption-based income tax. Among the various proposals, the flat tax in particular has gained a following, probably because it purports to use tax reform to change politics as well as the economy. As we will see, its advocates claim too much. In fact, their analysis of the current tax system is often wrong. Their promise to reform politics is largely empty. And the policy would probably leave the country worse off, both socially and economically, than it is now.

[29] The three basic measures of a sound tax system and sensible reform are simplicity, growth, and equity; none of the current reform strategies meet all of these tests. The key to the initial popularity of a flat tax is the first measure, its radical simplicity. It would abolish almost all tax deductions, credits, and exclusions in both the personal and corporate income taxes. Make no mistake, such simplification is sound economics. The less the tax code influences how people and firms consume, save, earn, and invest their income, the more efficiently America's markets can allocate the economy's resources. And by broadening the tax base, simplification allows us to raise the same revenues with lower tax rates -- and lower tax rates, like lower taxes, are almost always better for an economy.

[30] The public is drawn to drastic tax simplification, however, because it sees it as an attack on the quiet arrangements and tacit corruptions that allow powerful industries and wealthy people to secure special tax treatment. Even so, the flat tax plans offered by Rep. Richard Armey or Steve Forbes would not deliver such political reform. While they WOULD repeal scores of special provisions used today by profitable businesses and wealthy people to shelter part of their income from tax, they WOULDN'T end the privileged tax status of the well-heeled. That's because both plans would replace those special provisions with new and much broader tax exemptions covering ALL personal income derived from interest, dividends, or capital gains, and ALL business income used for capital investment.

[31] Flat tax advocates defend these blanket exemptions by shifting the argument from political reform to the second measure, economic growth. Sharply progressive tax rates, they insist, discourage the most productive people from working and saving more, which in turn reduces growth for everyone. The flat tax's answer has two parts. First, it would eliminate the personal tax on income from savings and replace progressive tax rates with a single flat rate on the income from labor. Second, it would channel new savings to traditional capital investment by allowing firms to deduct the full cost of new purchases of plant and equipment.

[32] This defense relies on faulty data and flawed economics. The data, first, refute the claim that the current federal tax burden is sharply progressive. Taking all forms of federal taxation into account -- personal and corporate income taxes, payroll taxes, and excise and estate taxes -- middle class families already pay roughly what they would under a pure flat or proportional tax, or a little less. The federal tax burden on the most affluent people is only modestly greater. Even for the richest 1 percent of Americans, more than 70 percent of the taxes they pay represent merely their proportional share of the costs of government.

[33] Second, the 1980s demonstrated that tax cuts for various forms of personal savings and business investment have little effect on overall saving and investment rates. In 1995, despite more than $120 billion in direct tax incentives for personal saving, the U.S. personal savings rate remained less than 5 percent. /1/ And even if such tax incentives did work as flat taxers claim, that wouldn't guarantee higher national growth. Studies of what makes the U.S. economy grow have found repeatedly that technological innovation and improvements in the skills of the work force are SIX-TO-SEVEN TIMES more important than business investment in plant and equipment in promoting higher growth and incomes. /2/ As a result, the flat tax growth strategy has a serious problem of scale: To achieve a permanent 1 percentage point increase in economic growth, we would have to nearly TRIPLE our current national savings rate. /3/

[34] In an upcoming report, the Progressive Foundation will offer a growth-oriented strategy for tax reform designed to stimulate not simply savings but economic innovation, efficiency, and productivity. In this report, we will examine and analyze the current system and the principal proposals to change it through the optic of tax equity and fairness.

THE TERMS OF TAX FAIRNESS

[35] Fairness in the tax system matters because tax collection depends vitally on voluntary compliance. Paying taxes is also most Americans' chief point of contact with their government, and probably their closest approximation to a common civic experience. Yet, some analysts today dismiss equity issues and, with increasing boldness, insist that REGARDLESS of their effects on fairness, all tax cuts are desirable because government's right to tax is less than fully legitimate.

[36] Behind this disdain for tax fairness lies a critique of democracy itself. It posits that the economic choices of individuals are morally superior to the political decisions they make as a community because in some important sense people are "natural" while governments are merely "artificial." Those who make this case misunderstand markets as well as democracy. Society probably cannot maintain itself without individual property rights, but the economic activities that produce property occur within a fabric of relationships shaped by the social and political institutions that people create for themselves. Individuals exercise their economic choices, then, within associations and corporations, which are creatures of the law, and so also, of the political decisions of communities.

[37] Without a doubt, most people don't enjoy paying taxes. But in a democracy like ours, people contribute private resources to provide the public goods they deem appropriate as a community, including helping those unable to make their way by themselves. In America, paying taxes embodies a civic relationship of mutual responsibility, and people's obligation to pay them is as legitimate as any other public duty.

[38] Among those who do not question a citizen's obligation to pay taxes, there are two broad views of the meaning of tax fairness. Fairness under a proportional or flat tax emphasizes equality: Everyone should be subject to the same tax rules, and therefore, everyone should pay taxes at the same tax RATE and bear the same relative tax burden. By contrast, fairness under a progressive tax system stresses people's different circumstances: All people should pay taxes according to their ability to do so, and therefore, the tax system should exempt the poor and apply to everyone else tax rates and relative tax burdens that increase with income. /4/

[39] Conservatives generally believe that a flat tax will best protect people's individual liberty, because by burdening everyone equally, IT CREATES THE BROADEST CONSTITUENCY POSSIBLE FOR LIMITING GOVERNMENT'S DEMANDS ON INDIVIDUALS. Moreover, because a truly flat tax imposes an equal economic burden on everyone, it does not affect the market's distribution of income. This provides the basic measure of a genuine proportional tax: Each person and each income group claims the same share of national income after paying taxes as it did before paying them. And this is thought to be just because, from a conservative's vantage, markets distribute income based on how hard and well people work.

[40] On the other side of the debate, liberals generally believe that progressive taxes protect individuals better than a flat tax, by curbing concentrations of economic power that threaten the opportunity of others, and by providing poor families with the resources they need to live independently. From this vantage, people's incomes reflect not only their own efforts but also a universe of circumstances they cannot affect; therefore the market's distribution of income is not the final word on economic justice. True equality of opportunity becomes a social achievement, one that tax progressivity advances by changing the market's distribution of income. This provides the fundamental measure of a progressive tax system: After paying their taxes, those at the top are left with a smaller share of all national income than before, and those at the bottom a larger share than before.

[41] Both approaches to fairness can claim some basis in economic theory. /5/ The economic logic for a pure flat or proportional tax rests on the basic market notion that people produce economic goods and services because they expect to reap economic benefits. ALL taxes reduce the benefits people receive from working and investing but, the argument goes, a flat tax should discourage work and saving the least because a system with only one tax rate will have the lowest possible TOP tax rate. In addition, a market will produce goods and services most efficiently when its resources are allocated through prices that directly reflect everyone's individual preferences -- and taxes should reflect the "prices" people would pay for the public goods provided by government. Since traditional public goods such as national security or public parks benefit everyone equally, efficiency dictates that everyone pay taxes at the same rate to finance them.

[42] The economic theory behind progressive taxation begins by separating the tax system from the economy that provides its resources. By this view, people work and save because they need or want certain benefits. The existence of taxes does not change those needs or desires, and so should not discourage work and saving. Therefore, tax rates could reasonably rise with income without imposing additional economic costs. In addition, just as an efficient economy uses its most productive resources first, until the cost of using more of them equals the cost of using something else, so an efficient tax system should provide that all taxpayers bear an equal sacrifice. From this optic, tax rates should rise with income, so that an additional dollar of tax entails the same sacrifice by people at every income level.

[43] By itself, economic theory cannot choose between the two cases, and hard economic evidence does not fully support either side. As near as we can tell, tax rates DO affect work effort and savings, but ONLY when the rate is very high, and then only to a MODEST degree. Moreover, the impact of high tax rates is even smaller when people can protect their income from these rates by claiming various deductions, exemptions, and exclusions -- and that's almost always the case. Yet, relatively high tax rates and tax burdens also cannot produce economic equality. Progressive taxes apply only to people's annual income, not to their accumulated wealth, and so have only modest effects on concentrations of economic power. Moreover, while high income people don't stop working or saving because their marginal tax rate is high -- they find ways to avoid it or live with it -- the transfers financed by their taxes can affect the work efforts of those receiving them. In any event, lifting tens of millions of low income people into the middle class by direct income transfers would require much higher taxes not just on the wealthy, but also on strapped middle class families.

[44] Ultimately, progressive taxation has the better of the argument. Flat taxers may be right that it would be morally offensive to tax higher income people more heavily if differences in income reflected only how hard different people work. But income differences reflect much more than that, if only because people don't start in the same place. People are born with different talents and come to age in families, neighborhoods, and cultures with different resources to prepare them for market competition. And plain luck often plays a role.

[45] America's wide-open markets accentuate the impact of all these factors, so that those with more ambition, self-discipline, and talent can prosper greatly. Bill Gates and his investors, for example, would not have enjoyed as great a success in other advanced countries because their markets and laws would not have provided so hospitable an environment. And once a person or family's economic success is secured, America's open markets allow them to increase the value of their wealth at a greater rate than in most other places. The economic benefits of free markets are large and obvious. But there are social costs, because our open markets and laws also produce harsher economic inequality than in other advanced countries -- an urgent issue today when economic inequality is increasing rapidly and for reasons that most working people can do little about.

[46] Progressive taxes cannot undo this inequality or restore upward mobility to poor or middle class Americans -- not even progressive taxes to finance transfer programs. A progressive tax system, however, can protect poor and middle class families from bearing the higher tax burdens entailed in a purely flat or proportional system, and in this sense, ameliorate some of the distributional inequalities achieved through our markets but based on factors other than how hard different people work. And the additional burden of progressive taxation is a reasonable price to pay by those who in some respect start with more, for the privilege of prospering relatively more under America's laws and in her markets. Bill Gates and his investors have a responsibility to not merely bear an equal share of the burden, but a GREATER share because they enjoy a larger share of the benefits provided by these laws and markets.

[47] Furthermore, when accidents of birth and luck affect people's ability to succeed through hard work, economics as well as social considerations can dictate that they receive the means and opportunity to participate more fully in the economy. At the very least, the tax burden to finance these efforts should be progressive at the bottom so that the tax system does not further impair the ability of low income people to participate. At the top, people with higher income can contribute, within limits, without harming the economy or their own basic freedom. In the end, progressive taxation can trade off the benefits of ensuring a broader distribution of opportunity against the modest costs of higher tax rates and higher tax burdens on some to finance it.

WHO PAYS THE TAXES

[48] As we will see, the current federal tax burden -- including the personal and corporate income taxes, payroll taxes, excise, and estate taxes -- is most clearly progressive at the bottom of the income ladder. The total burden is also progressive at the top, but only moderately so and much of that reflects income tax changes enacted in 1993. /6/ For the broad American middle class, the tax system is, on balance, more nearly proportional than progressive, barely affecting the share of national income held by most families. Finally, while everyone would like to pay less taxes, evidence suggests that a substantial majority of Americans would prefer a little MORE progressivity, so that low income families could pay a little less than they do today and very affluent families would pay a little more.

[49] The current distribution of all forms of federal taxation can be summarized in the following five findings:

o FIRST, EVERYONE BEARS SOME OF THE BURDEN. Higher income people

o SECOND, A FAMILY'S TAX BURDEN RISES MOST SHARPLY AS IT MOVES

o THIRD, TWO GROUPS BEAR LIGHTER TAX BURDENS THAN OTHERS:

o FOURTH, THE TOTAL FEDERAL TAX BURDEN IS GENUINELY PROGRESSIVE

o FIFTH, HIGHER INCOME PEOPLE BEAR ALL OF THE CURRENT COSTS OF

FAIRNESS AND TAX REFORM

[50] Plans to broadly reform federal taxation, especially the flat tax and national retail sales tax proposals, would leave the system much less progressive and much less fair. Under the current arrangement, basic fairness is provided through six progressive elements. At the bottom, (1) an exemption for initial income protects poor families from income tax, and (2) the EITC, in effect, refunds part of the payroll tax and other tax payments of working poor families. For everyone else, (3) tax rates on personal income rise from 15 percent to nearly 40 percent as income increases. And for those at the top, (4) most of the value of their personal deductions is phased out; (5) federal corporate taxes affect mainly owners of capital, who are predominantly affluent; and (6) estate taxes affect only the very well-to-do.

[51] FLAT TAX. The Armey and Forbes flat tax proposals would repeal four of these six elements -- the EITC, the graduated tax rates, the phaseout of personal deductions at the top, and estate taxes. The repeal of the EITC, in particular, would worsen the poverty of millions of working poor people and their children. However, the proposals WOULD help many moderate income families, by expanding the current tax exemption for initial income. Yet, the benefits of this change would not be felt by many middle class taxpayers who would lose their personal deductions for mortgage interest, pension contributions, and state and local taxes.

[52] There's no easy way for the flat tax to avoid this problem. Simply preserving the mortgage-interest deduction won't work as long as the flat tax also exempts interest income from tax. If flat taxers tried to allow people to deduct both the interest they pay and the interest income they receive, anyone could take out a second mortgage and deduct the interest cost and then invest the money and exclude the income it earns. The result would be pure tax leverage, producing taxpayer-financed transfers to those holding the greatest home equity, a revenue hemorrhage, and financial distortions as the nation's capital was channeled through second mortgages.

[53] A PURE AND COMPREHENSIVE flat tax system wouldn't have this problem because it would tax all income the same, whether it comes from labor or capital. And a pure proportional tax would affect the disposable income of most middle class families very little, one way or the other. The current proposals, however, cannot avoid imposing higher taxes on most middle class Americans -- as the original architects once noted themselves /7/ -- because they would NOT tax capital and labor the same. Instead, the Armey and Forbes plans would shift more of the total tax burden to labor, because capital would be taxed once under a business tax while wages and salaries are to be taxed twice under both the income tax and the payroll tax. /8/ The middle class HAS to pay more under such a system, and the wealthy much less. That's because virtually all of the income of average families come from wages and salaries, with only 6 percent to 10 percent coming from capital, while people at the top derive much less of their income from labor but 35 percent to 48 percent from the interest, dividends, and capital gains receipts exempt from personal tax under these plans. /9/

[54] Initially, however, the flat tax would likely depress the market value of all corporate stock, by strongly favoring NEW investments in plant and equipment over existing business capital. /10/ Still, by one preliminary estimate, these plans would mean at least $30 billion more in taxes paid by families in the bottom half of the income distribution, and $50 billion less in taxed paid by those in the top 20 percent. /11/

[55] NATIONAL SALES TAX. A national retail sales tax has an even more troubling effect on tax fairness. This approach would repeal all six progressive elements in the present tax system: Along with eliminating the EITC, graduated tax rates, the deduction phaseout, and estate taxes, the plan also would repeal business taxes and the exemption for initial income. /12/ Further, it would create an unlimited tax deduction for new saving and investment, favoring those with high income, since poor and middle class families have to consume much larger shares of their income. The only progressive feature of a sales tax approach is an IMPLICIT tax on existing wealth, since people would pay the tax whenever they sold an existing asset and spent the proceeds. By one preliminary estimate, such a proposal would more than double the effective tax burden on the poor and substantially raise the burden on middle class families, while providing enormous tax relief to wealthy families. /13/

[56] Under the flat tax or a national sales tax, for the first time in American history the tax system would redistribute income towards wealthy people. Families at the top of the income scale could claim a larger share of all national income after paying their taxes, so that the tax system would actually reinforce the country1s growing inequalities in income.

[57] USA TAX. Of all the current major reform proposals1 only the Unlimited Savings Allowance (USA) tax plan of Sens. Sam Nunn and Pete Domenici would have little adverse effect on fairness. This plan would exempt new saving and investment from federal income tax, including all inheritances, but it also would preserve the EITC, graduated tax rates, business taxes, and a substantial exemption for initial income. While including some modest simplification, especially for the corporate tax, the plan retains personal deductions for mortgage interest and charitable contributions and creates a new deduction for college tuition. It also would offset some of the regressive effect of its exemption for new net saving and investment by providing workers a new tax credit equal to their payroll tax payments. In addition, it would create the same kind of implicit tax on existing wealth as a sales tax. Nevertheless, in order to maintain roughly the same progressivity as exists today, the proposal requires higher tax rates than those imposed currently on income which is not saved or invested. /14/

II. EQUITY AND ECONOMIC GROWTH

[58] The debate over tax reform involves issues of both fairness and growth because it is unfolding in an environment in which inequality is increasing and overall growth is slowing.

[59] In the first 25 years following World War II, Americans at every income level could achieve strong income gains by their own efforts. From 1947 to 1973, the real annual income of average families in each income group doubled, along with the size of the real economy. In fact, the LOWER a family's income, the FASTER its income grew. As a result, the share of all national income earned by the bottom 80 percent of families increased steadily, while the share earned by the top 20 percent declined (Table 1).

[60] Over the past 20 years, the entire economy has grown much more slowly AND the gains of typical families at different income levels have diverged sharply. In fact, since 1973 the HIGHER a family's income, the faster it has grown. As a result, the rich have become richer, much of the middle class has stagnated economically, and the poor have become poorer. By 1993, the bottom 60 percent of families earned a smaller share of national income than they did 46 years earlier, and the richest 20 percent claimed a significantly larger share of all income than they had nearly five decades before.

[61] These income developments are not fully understood, but it seems clear that most of the increase in inequality reflects changes in the economy itself. In particular, returns on capital have risen, and changes in the wage structure have increased rewards for advanced knowledge and skills while reducing them for workers, especially men, with only basic skills. /15/ Markets today, in short, increasingly reward the efforts and habits of only a small share of those working hard and well. These changes are themselves associated with two larger developments: The rise of global markets has expanded both the demand for capital and advanced skills, and the supply of low skilled labor; and the increasing economic importance of information technologies has further increased demand for capital and higher worker skills.

                  TABLE 1. THE GROWTH IN INEQUALITY

_____________________________________________________________________

                     Annual Real             Shares of

Income Level      1947-73  1973-93       1947   1973   1993

Quintile 1         2.95%    -0.78%        5.0%   5.5%    4.1%

Source: U.S. Bureau of the Census, 1995a, Tables F-1 and F-1a, and

INEQUALITY AND THE TAX SYSTEM

[62] Tax policy has played, at most, a very minor role in these market developments, although tax changes in the 1980s did reinforce the results by generally lessening the system's overall progressivity. From 1980 to 1989, while the total federal tax burden remained stable, changes in the tax code significantly reduced "effective tax rates" on the rich -- the share of their income actually claimed by taxes -- and more modestly increased them on everyone else except the poor (Table 2). The 1993 Tax Act reversed part of this trend at the top and the bottom by raising the top tax rate on wealthy people and expanding EITC benefits for low income families. BY 1994, ONLY MIDDLE CLASS AMERICANS CARRIED A HIGHER TAX BURDEN THAN THEY HAD IN 1980.

[63] The data are not in serious dispute. The disagreements involve what to do about it: Would economic equality increase by shifting more of the tax burden from the middle class to wealthy and highly skilled people? Or, as conservatives argue, would higher taxes on the affluent and the skilled reduce their incentives to invest and work and so ultimately cut everyone's income gains? If the latter argument is correct, the case for a strictly flat tax is strengthened because it could generate a given level of revenues with the lowest top marginal rate. If it's also true that the reductions in work and saving from higher tax rates significantly impair economic growth, it would weaken the case for income transfers financed by progressive taxes.

[64] In truth, economists do not agree about the power of tax rates to determine people's work and savings. BUT THE BULK OF THE EVIDENCE INDICATES THAT WHILE CONSERVATIVES ARE RIGHT ABOUT HIGHER TAXES REDUCING ECONOMIC PRODUCTION, THE TOTAL EFFECTS ARE NOT VERY LARGE. The first studies of this question, conducted in the 1970s and early 1980s, did suggest that higher tax rates to finance income transfers could substantially harm the economy. /16/ One analysis found that raising everyone's income tax rate by 1 percentage point, and shifting the revenues to the poorest two-fifths of families, would discourage work and investment sufficiently to cost the economy nearly $2.50 for every $1 transferred. /17/ Studies also showed that these costs worsen as the tax rate itself rises, so that increasing the tax rate from 35 percent to 36 percent on wealthy people would have greater costs to the economy than raising the rate on middle class people from 15 percent to 16 percent.

[65] Subsequent studies, however, found the economic costs to be much less -- only 50 cents to $1.30 for every $1 transferred. That would mean, for example, that when the federal government in 1995 provided $12.5 billion in direct payments to 15 million children and their parents through Aid to Families with Dependent Children, it cost the economy roughly $11 billion in lost production -- 0.16 percent of GDP -- on top of the funds transferred. The most recent study suggests that the net cost to the economy is as little as 20 cents for every dollar transferred if the transfers are provided in ways that encourage work, such as the EITC's wage subsidy. /19/ Assisting more than 17 million working poor adults and children through the EITC in 1995, therefore, cost the economy as little as $4.4 billion in lower production -- or 0.05 percent of GDP -- on top of the $22.2 billion provided in direct tax benefits.

                TABLE 2. IMPACT OF TAX LAW CHANGES ON

                    Effective Tax Rates (All Federal Taxes)

_______________________________________________________________

Income Level          1980             1989          1994

Quintile 1             8.5%             8.5%          5.0%

Source: Kasten, et a1., 1994, and Congressional Budget Office, 1994.

[66] In the end, MARGINAL TAX RATES DO NOT DETERMINE MOST PEOPLE'S DECISIONS ABOUT HOW MUCH TO WORK OR SAVE BECAUSE THEY USUALLY HAVE MORE PRESSING THINGS TO WORRY ABOUT -- such as how hard they'll have to work to meet the mortgage, or how much they'll have to save to pay their children's college tuition. /20/ And the top marginal rates have little effect on how much even wealthy people work and save because various tax preferences enable them to shift more of their income to forms not subject to those rates.

[67] Similarly, special tax incentives for saving and business investment affect those decisions, but other factors usually matter much more. Various researchers have found that in the 1980s the investment tax credit increased purchases of certain plant and equipment, lower capital gains taxes helped expand certain forms of equity investment, and IRA deductions modestly raised private retirement saving. But OTHER forms of saving and investment declined because economic growth, productivity gains, and income growth were all relatively weak -- and in the end, overall net business investment grew more slowly and the private saving rate declined. Large economic forces overwhelmed the new incentives, and they ended up mainly shifting saving and investment to the particular forms newly favored under the tax law rather than increasing their overall levels.

[68] Ironically, the conservative case for cutting taxes on high income people who do most of the economy's saving and investing is refuted by the overwhelming power of free markets over both these decisions and government tax policies. But the liberal case for sharply raising taxes on high income people, in order to offset growing economic inequality, remains very problematic. Inequality has increased so much that tax and transfer policies could not restore the income distribution of a generation ago without either substantially raising taxes on struggling middle class families or raising tax rates on the wealthy so high that they could undermine growth for everyone.

III. ELEMENTS OF PROPORTIONALITY AND

[69] The current federal tax system is neither truly proportional nor purely progressive; rather, it is a tangled combination of progressive, flat, and regressive elements.

THE PERSONAL INCOME TAX

[70] The personal income tax is the largest part of the federal tax system, accounting for nearly 44 percent of all revenues or roughly $590 billion in 1995. Only two major provisions of the income tax, however, are strictly progressive. At the bottom, the EITC supplements wage income earned by poor and low income people based on need. And for high income families at the top, up to 80 percent of their deductions for mortgage interest, charitable contributions, state and local tax payments, and other expenses are gradually phased out starting when their income, before claiming their deductions, tops $114,700. /21/

[71] Two other elements of the personal income tax are generally progressive: the tax exemption for initial income, and graduated tax rates. Every income tax in U.S. history, along with every current reform proposal except the retail sales tax, has exempted the income of the poor. In 1995, a family of four was not taxed on its first $16,550 of income (four $2,500 personal and dependents' exemptions, plus a $6,550 standard deduction). Yet, even this provision is proportional in one important sense: It applies equally to almost everyone. Its progressivity is based chiefly on the fact that the income exempted represents a larger share of the total income of lower income families than of others. In addition, the personal exemptions are phased out gradually for those at the very top of the income scale. /22/

[72] The other generally progressive feature of the income tax is the schedule of graduated tax rates. A couple with two children filing a joint return, after deducting their first $16,550 of income, has to pay 15 percent on additional taxable income up to $39,000; 28 percent on additional taxable income from that point to $94,250; 31 percent on income from $94,250 up to $143,600; 36 percent on additional income up to $256,500; and 39.6 percent on any taxable income above that amount. However, under these arrangements even people with the highest incomes pay the lowest rates on the first increments of their income.

[73] The high tax rates on higher income levels do not produce a sharply progressive distribution of the total tax burden because virtually every other important feature of the income tax reduces the income subject to those high rates. For example, people who itemize their deductions do not pay taxes on the income they use to finance their home mortgages, to purchase health care coverage through their employers, to pay their state and local taxes, and to support charities. Like the initial income exemption, these deductions are available to almost everyone regardless of income -- so long as they itemize. And since people use deductions to reduce the total income they report for tax purposes, deductions enable higher income persons to reduce the tax rate applied to their remaining or "taxable" income.

[74] Furthermore, the direct tax benefits provided by these deductions INCREASE as a person's income rises (i.e., they are regressive), because the benefits are based on the tax rate that a person would otherwise pay on the income. For example, if an affluent and an average-income family each claims a $5,000 deduction for mortgage interest, the high income family in the 36 percent tax bracket will save $1,800 in taxes by taking the deduction ($5,000 x 0.36 = $1,800), while his middle class neighbor in the 15 percent tax bracket will save only $750 ($5,000 x 0.15 = $750). As a result, the major deductions preserve more post-tax income for well-to-do people than for those who are middle class.

[75] In theory, the major tax deductions should still have generally progressive income effects, because mortgage interest and health care coverage, in particular, usually represent a larger share of the total income of a middle class family than of higher income households. /23/ But in practice, the income effects of the major deductions are often regressive. Most middle class families do not fully benefit from this effect because most of them don't itemize their deductions at all. Roughly 70 percent of all taxpayers claim the standard deduction -- while among those who are better off, more than 75 percent with annual taxable income of $50,000 and above itemize, as do more than 90 percent of those earning more than $100,000 a year (Table 3). As expected, people who itemize their deductions receive larger tax benefits than people with comparable income who claim the standard deduction -- otherwise, they wouldn't bother to itemize.

[76] Furthermore, analysis of the data shows that the itemized deductions claimed by many higher income people protect from tax a LARGER share of their total income than the standard deduction claimed by most average-income families. For example, the standard deduction claimed by nearly two-thirds of all families with taxable income of $30,000 to $40,000 excludes from tax, on average, about 15 percent of their income, while the itemized deductions claimed by more than 90 percent of families with incomes of $75,000 to $200,000 shelter, on average, about 20 percent of their income. The standard deduction has clearly progressive income effects only at low income levels, where people can shelter much of their income by claiming the deduction as part of their exemption for initial income, and at extremely high incomes, where itemized deductions phase out.

      TABLE 3. VALUE OF STANDARD AND ITEMIZED DEDUCTIONS, 1992

Income Level      Standard  itemized        Standard  Itemized

$10-15K             92.8%      7.1%           38.2%     77.6%

[77] Other important provisions of the personal income tax have no progressive income effects at all; notably the deductions for retirement saving, the 28 percent top tax rate on capital gains income, and the right to defer paying a capital gains tax until an investment is finally sold. These income tax preferences not only provide larger tax benefits as a person's income increases, they also involve forms of income that represent much larger shares of the income of wealthy people than of everyone else. The richest 1 percent of American families derive more than 47 percent of their annual income from capital investments, compared to 18 percent for everyone else, and 22.1 percent from capital gains alone, as compared to 1.8 percent for other families. /24/ Moreover, these preferences are NOT phased out for very wealthy taxpayers. The result is a classic regressive effect: These provisions increase the share of national income held by high income families after they pay taxes, compared to their before-tax share.

BUSINESS TAXES

[78] Most economists agree that corporate income taxes, which generated $150 billion in revenues in 1995 or 11 percent of all federal revenues, fall primarily on owners of capital, offsetting the personal income tax benefits for capital income. To be sure, the distribution of the corporate tax burden is not a settled question. Some analysts argue that the burden of federal business taxes is passed on to workers in smaller wage gains or lower wages, to consumers through higher prices, and to shareholders in the form of lower profits. For some industries, at least to some degree, this is probably correct.

[79] Even if federal business taxes are borne entirely by those who own capital, however, its distribution AMONG those capital owners is not progressive. First, the tax burden on the capital of unincorporated firms is much lower than that carried by corporations: One recent study estimated that in 1989, the marginal effective tax rate on capital was 43 percent for corporations, and 22 percent for other firms. /25/ While most unincorporated businesses are small firms with relatively low capital earnings, it is also probably the case that the bulk of this class of business capital income is earned by highly profitable professional partnerships and large family firms, owned chiefly by high income people. As a result, among owners of capital, high income persons may well benefit most from the lower capital tax burden on unincorporated businesses.

[80] Nor is the burden of the corporate tax distributed progressively among all corporate shareholders. With a few minor exceptions for small companies, the tax applies a flat 35 percent rate to most corporate taxable income; and this rate has a regressive income effect among those who own corporate capital because at 35 percent this rate is lower than the 39.6 percent personal rate on wealthy investors, and higher than the 15 percent and 28 percent personal rates on middle class investors. As a result, moderate and middle income investors pay more taxes on their share of a corporation's earnings under the corporate tax than they would have if they had received the income directly, while wealthy investors pay less tax than they otherwise would have if they had received it as personal income. On balance, however, federal taxes on businesses have an overall progressive effect because businesses are predominantly owned by high income people.

PAYROLL TAXES

[81] The Social Security and Medicare payroll taxes, which generated nearly $500 billion in 1995 or nearly 36 percent of all federal revenues, have clearly regressive effects. These taxes provide neither an initial exemption nor graduated tax rates. Instead, they exempt all forms of capital income and tax only wages and salaries from the first dollar to $61,200 (the cap in 1995) at a flat 15.3 percent rate, at which point the tax rate falls to 2.9 percent (the Medicare portion of the tax). Payroll taxes, therefore, inevitably claim a larger share of both the total income and the wages and salaries of poor and moderate income families than of the well-to-do and the wealthy. /26/ The effects are strictly regressive: Those who earn relatively less are left with a SMALLER SHARE of all national income after paying their payroll taxes than before paying them.

[82] The fairness of payroll taxes, however, depends ultimately on how you conceive the Social Security system, and how closely you relate payroll tax payments to later Social Security benefits. If Social Security is thought of as a government-sponsored annuity system, and payroll tax payments are considered personal investments undertaken to produce specific subsequent benefits, then the entire system's effect may be modestly progressive. From this perspective, the important factor offsetting the regressivity of payroll taxes is the value of Medicare benefits. Medicare hospital coverage, while provided to all elderly people regardless of income, has progressive income effects because they are worth relatively more to low and middle income retirees than to wealthy retirees. /27/ (Nevertheless, so long as medical costs rise more rapidly than wages, the largest transfers under Medicare occur not across income groups, but from younger generations who earn the wages to older people who no longer do.)

[83] The popular view that Social Security retirement benefits are also strongly progressive is much more problematic. It is correct that the system's generational transfers always provide higher benefits for each new cohort of retirees. /28/ As a result, people have always been able to expect that the benefits they will receive when they retire will be larger than those collected by the retirees they supported. But WITHIN each generation, once it retires, low earners receive small benefits and high earners receive large benefits. For example, a low income, two-earner couple -- a husband and wife who both worked for low wages -- retiring in 1995 can expect to collect $155,200 (in 1993 dollars) in Social Security benefits over the course of their retirements, while a highly paid two-earner couple retiring at the same time will receive benefits totalling $312,600. /29/

[84] Of course, highly paid people also pay more payroll taxes into the system; and when the value of those taxes is included in the calculations (adjusted for inflation and an average rate of return), the retirement system appears to be neither consistently progressive nor regressive. /30/ Among two-earner couples retiring in 1995, Social Security on a net basis appears to be regressive for those who earned low incomes relative to average income earners, and progressive for those who earned average incomes relative to high income earners. That is, those who earned average incomes receive larger retirement benefits, net of the taxes they paid, than either those who earned less or those who earned more. /31/ Among all one- earner couples, the distribution of net retirement benefits is more consistently regressive: High earners retiring in 1995 will receive larger net benefits over their retirements than their low or average- earnings counterparts. /32/ From this vantage, however, the retirement system is growing less regressive over time. By 2010, newly retired two-earner couples can expect to collect net benefits valued at $40,300 (in 1993 dollars) if their lifetime earnings were low and $39,800 if their earnings were average -- and bear a net COST of $73,100 if their lifetime earnings (and payroll taxes) were high. /33/

[85] From another perspective, however, the entire Social Security system is very regressive and will remain so. If the benefits are thought of as essentially another government program, and the payroll tax as another way of collecting revenues to support government operations, then the system combines a regressive tax that imposes the greatest burdens on those people with relatively less income, and a regressive benefit schedule that provides the largest retirement support to those who had earned more while working. /34/

[86] How you choose to evaluate the Social Security system may depend on your state of mind, but its operations seem to more closely resemble other public programs than a private annuity system. Social Security does NOT save and invest the payroll taxes it collects, but uses them to provide pensions for current retirees and to help finance other government operations. Furthermore, the benefits that people receive are NOT based on their previous payments, as in a private annuity program, but on the wage income from which those payments were calculated. In the final analysis, the primary redistribution occurring through Social Security is not among income groups, but between generations.

EXCISE AND ESTATE TAXES

[87] Of the remaining 9 percent of federal revenues, almost half -- $57 billion in 1995 -- comes from generally regressive excise taxes, with almost 60 percent of those receipts coming from the taxes on alcohol, tobacco, and gasoline. /35/ The burden of these taxes is usually passed on to consumers -- and with regressive results, since lower income people spend larger shares of their income on these items than do families with higher incomes. /36/

[88] The impact of certain excise taxes, however, may not always be simply regressive. The excise tax on airplane tickets has raised as much revenues as the tobacco tax and is not especially regressive. Furthermore, a study by the Congressional Budget Office found that lower income people devote a smaller share of their TOTAL CONSUMPTION SPENDING to alcohol than do people with higher incomes, and the same share of their spending to gasoline as those who are better-off. From this perspective, the burden of the gasoline tax is roughly proportional, and excise taxes on alcohol are actually progressive. Furthermore, gasoline tax revenues go to maintain the interstate highways used by those paying the tax, as airline ticket taxes support services for air travelers provided by the Federal Aviation Administration. To some extent, these taxes may even be considered voluntary.

[89] Estate and gift taxes, the last major part of the federal tax system, are at once the most purely progressive of federal levies and the least significant as a source of revenues. In 1995, estate and gift taxes raised less than $16 billion, barely 1.2 percent of all federal receipts. Virtually all of this revenue comes from roughly the top 1 percent of estates, /38/ and in most of these cases only a small share of the total estate is taxed at all.

[90] The progressivity of these taxes arises from two provisions. First, the law exempts from taxation the first $600,000 of an estate's net worth, covering all but the largest estates. Second, an estate's additional value is taxed at progressive rates that rise from 18 percent to 30 percent for the next $100,000 of net worth; from 30 percent to 41 percent for the following $900,000 of net value; and from 41 percent to 55 percent for the next $2 million and above. Other estate tax provisions, however, enable most wealthy families to protect most of the value of even the largest estates from these high rates. For example, there is an UNLIMITED deduction for bequests to a surviving spouse. Moreover, many wealthy families also transfer considerable assets to trusts or private foundations that take advantage of numerous tax-reducing strategies.

[91] As a result, estate taxes impose significant burdens on only a very small number of large estates. In 1993, 60,211 Americans died leaving estates valued at $600,000 or more; 32,705 of these estates paid NO estate tax, including 12,805 with assets of $1 million or more, and nearly one-fifth of those valued at $10 million or more. /39/ All told, the 1993 effective tax burden on estates valued at $600,000 to $1 million was barely 2 percent, and the burden on those valued at $1 million to $2.5 million came to less than 9 percent (Table 4). The tax burden rises for even larger estates, but then falls for those valued at more than $20 million. Moreover, all of these calculations overstate the tax burden on substantial estates because they do not include wealth transferred to trusts or private foundations by people before they died.

[92] Federal estate taxes clearly are progressive but in the end, they apply roughly the same tax burden to estates of $2.5 million and more as the rest of federal taxation applies to the incomes of people earning between $20,000 and $40,000 a year.

            TABLE 4. BURDEN OF FEDERAL ESTATE TAXES, 1993

$600,000-$1 million      31,955       $23.8 billion          2.3%

IV. THE DISTRIBUTION OF FEDERAL TAXATION

[93] When the various elements of the federal tax system are combined, the final distribution of its burdens turns out to be dominantly flat but also moderately progressive. Low income families pay substantially less than their proportional share of the costs of government -- that is, substantially less than they would pay under a pure and comprehensive flat tax system -- while most middle class Americans pay nearly their proportional share. Only high income people bear an additional burden from progressivity, and the bulk of the taxes paid by even the richest families represents their proportional share of the costs of government.

[94] A detailed picture of this distribution entails totaling up for each income group the share of its income paid in all federal taxes. The taxes covered include not only the personal income tax, payroll taxes, and corporate income taxes (attributed entirely to owners of capital) but also excise taxes and estate and gift taxes. The income data include not only people's wages and salaries, but also government cash transfers, employer contributions to Social Security and to federal unemployment insurance coverage, and capital income including rents, dividends, realized capital gains, and taxable and non-taxable interest.

WHO PAYS THE TAXES (TABLE 5)

[95] These data show, as conservatives have long insisted, that the relatively small number of Americans who earn high incomes provide a large share of all federal revenues. In 1994, the 20 percent of families with the highest incomes paid more than 60 percent of federal taxes, with the richest 5 percent contributing one-third of all revenues, and the top 1 percent providing more than 18 percent (Table 5). BUT THESE DATA SAY LITTLE ABOUT HOW PROGRESSIVE THOSE TAXES ARE BECAUSE THE PEOPLE PAYING MOST OF THE TAXES ALSO EARN MOST OF THE INCOME. For example, the top 20 percent earned nearly 52 percent of all national income produced in 1994, and the richest 1 percent claimed close to 13 percent of the nation's income that year. Under a strictly flat or proportional tax, they would already contribute those shares of revenues, so the bulk of taxes paid by high income Americans seem to represent their proportional share of the cost of government.

                 TABLE 5. SHARES OF NATIONAL INCOME

_____________________________________________________________________

                     Share of         Share of       Share of

     Under $10K        14.0%             1.9%          0.5%

Source: Congressional Budget Office, 1994, Table 6, p. 32; CBO

              TABLE 6. AVERAGE INCOME BEFORE AND AFTER

                         Average Income       Effective

     Under $10K       $6,032        $5,645       6.4%

THE SHARE OF PEOPLES' INCOME PAID IN TAXES (TABLE 6)

[96] The degree of progressivity in the current tax system can be gauged generally by comparing the effective tax rates of families at different income levels -- the share of each group's total income claimed by federal taxes. As expected, these data show that as a family's income increases, a larger share of it goes to federal taxes. But contrary to popular assumption, the greatest increases in effective tax rates occur at relatively moderate income levels (Table 6). FEDERAL TAXES REDUCE A FAMILY'S INCOME FASTER AS IT MOVES OUT OF POVERTY AND INTO THE MIDDLE CLASS, WHILE MOVING UP FROM THE MIDDLE CLASS TO GREATER AFFLUENCE ENTAILS ONLY A MODESTLY GREATER RELATIVE TAX BURDEN. However, the tax burden rises significantly again as a family reaches the very top of the income ladder, principally because the 1993 tax changes raised the top tax rates and increased the effective tax burden on the wealthy from 27.9 percent to 32.7 percent.

[97] These data also show that the system's general progressivity does not exempt anyone from paying taxes, including the very poor. The current federal system cuts the disposable income of the average poor family earning less than $10,000 a year by 6.4 percent, from $6,032 to $5,645. Federal taxes also claim nearly 20 percent of the incomes of middle class families who earn $30,000 to $40,000 a year, reducing their disposable income from an average of $34,860 to $27,919. At the top where families earn an average of nearly $486,000 a year, those taxes claim nearly $159,000.

THE TAX CONSEQUENCES OF GROWING OLD OR HAVING CHILDREN (TABLE 7)

[98] Two personal characteristics, growing old and having children, significantly affect people's tax burdens. /40/ First, households headed by elderly people bear much lower tax burdens -- generally, one-third to two-thirds lower -- than others at the same general income level (Table 7). The relatively low tax burdens borne by older people reflect the tax advantages provided for certain forms of income, which they happen to depend upon. In particular, nearly 80 percent of the income of retired people comes from Social Security and earnings on previous savings, both of which are favored under the income tax and exempt from payroll taxes; while more than 80 percent of the income of everyone else comes from wages and salaries subject fully to both income and payroll taxes. /41/ Moreover, the elderly's tax advantages come on top of the benefits they receive from the major entitlement programs, producing a double-whammy intergenerational transfer: This year, government programs will shift more than $590 billion in cash retirement and in-kind medical benefits from younger working families to older retired people, and then tax those receiving the benefits more lightly than those working to pay for them. /42/

                    TABLE 7. EFFECTIVE TAX RATES

_____________________________________________________________________

     Income          Families with   Elderly     All Other

     Quintile 1            1.6%        2.8%         13.6%

[99] Second, at low and moderate income levels, families with children bear lesser tax burdens than childless families. This result is due to two factors: The EITC was designed to offset most of the tax burden of working poor families with children; and the dependent's exemption, which, while reducing the tax burden of all families with children, has a greater relative effect on those with modest incomes. /43/

HOW FEDERAL TAXES AFFECT THE DISTRIBUTION OF NATIONAL INCOME (TABLE 8)

[100] The clearest measure of how progressive the tax system is comes from comparing the shares of total national income held by each income level, BEFORE and AFTER paying federal taxes, recalling that a strictly flat tax system would leave families at every income level with the same share of national income after paying their taxes as before paying them. These data confirm that current federal taxes are clearly progressive at the bottom and the top (Table 8). HOWEVER, PROGRESSIVITY DOES NOT MUCH AFFECT THE BULK OF MIDDLE CLASS AMERICANS EARNING $30,000 TO $75,000 A YEAR: They claim roughly the same share of national income after paying their taxes as before.

                 TABLE 8. SHARES OF NATIONAL INCOME,

_____________________________________________________________________

                      Share of     Shares of National Income

     Under $10K         14.0%          1.9%           2.3%

HOW PROGRESSIVITY AFFECTS PEOPLES' TAX PAYMENTS (TABLE 9)

[101] FOR THE VAST MAJORITY OF AMERICAN FAMILIES, ALL OF THE FEDERAL TAXES THEY PAY REPRESENT NO MORE THAN THEIR PROPORTIONAL SHARE OF THE COSTS OF GOVERNMENT, AND USUALLY LESS. Virtually all of the burden of progressivity in the current system is borne by families earning more than $75,000 (Table 9). Among those who earn more than $200,000 a year -- with annual incomes averaging nearly $500,000 -- the element of progressivity in the current system increases their taxes by some $45,000, equal to 28 percent of their total tax burden. More than 70 percent of the taxes paid by the richest families, therefore, reflect merely their proportional share of the cost of government.

HOW PROGRESSIVITY AFFECTS PEOPLES' INCOMES (TABLES 10 & 11)

[102] The final two tables in this series show the impact of tax progressivity on the income of Americans. Table 10 shows that the additional tax burdens on affluent families associated with the progressivity of the current system represent a relatively small share of their total, pre-tax income, while the benefits from this progressivity for poor people are substantial. For poor families with income of $10,000 or less, the progressivity in the current tax system provides benefits equal to 17 percent of their total income. For wealthy families with annual incomes of $200,000 or more -- incomes averaging $486,000 a year -- tax progressivity costs the equivalent of 9.3 percent of their income. /44/ Between the two, the benefits received by middle class Americans from tax progressivity, or the additional burdens they bear, are equal to very small shares of their income.

[103] Table 11 shows that RELATIVE TO A PURE AND COMPREHENSIVE FLAT TAX SYSTEM, CURRENT TAXATION DELIVERS SUBSTANTIAL BENEFITS TO LOW INCOME FAMILIES AT LITTLE OR NO COST TO MIDDLE CLASS FAMILIES, AND AT MODERATE COSTS FOR MOST AFFLUENT PEOPLE. Poor families with incomes of less than $10,000 a year have 22 percent greater resources to spend under the current system than they would under a pure flat or proportional tax -- on average, $5,645 as compared to $4,619 under pure flat taxation. Most middle class families would have roughly the same disposable income under a strictly proportional tax, or a little less, as they do today. Moving up the income ladder, the progressivity in the current system costs affluent families earning $100,000 to $200,000 about 4 percent of their disposable income. And wealthy families with incomes averaging nearly $500,000 a year have about 12 percent less to spend or save under the current tax system than they would under a strictly proportional tax.

   TABLE 9. IMPACT OF PROGRESSIVITY ON PEOPLES' TAX PAYMENTS, 1994

_____________________________________________________________________

                       Gains (Costs) under            Share of

     Under $10K              $1,026                      --

     TABLE 10. IMPACT OF PROGRESSIVITY AS A SHARE OF INCOME, 1994

_____________________________________________________________________

                   Average      Gains (Costs) from      Gains (Costs)

     Under $10K     $6,032            $1,026                17.0%

Source: Congressional Budget Office, 1994, Table 6, p. 32; CBO

     TABLE 11. IMPACT OF PROGRESSIVITY ON AFTER-TAX INCOME, 1994

                       Income with a     Income under

     Under $10K           $4,619             $5,645         22.2%

[104] In fact, over the course of most people's lifetimes, the benefits and costs of tax progressivity are even smaller. At the bottom of the income scale, most poor people do not remain poor throughout their lives; rather, their incomes increase over time, and as that occurs the personal benefits they draw from tax progressivity decline. At the other end, most rich people are not born that way, or at least start their careers earning much less, and consequently they may have many years when they bear little of the costs of progressivity or even benefit from it. IN THE END, ONLY A VERY SMALL NUMBER OF AMERICANS WITH GREAT INHERITED WEALTH WILL PAY OVER THEIR LIFETIMES MORE THAN A FEW CENTS ON THE DOLLAR FOR PROGRESSIVITY.

V. A CLOSE LOOK AT THE LEADING TAX PROPOSALS

[105] Both sides of the tax debate usually insist that public opinion is with them. It's hard to say because surveys indicate that most Americans do not really distinguish between a proportional system under which people's tax payments increase with income, and a progressive one under which tax burdens as well as tax payments rise with income.

[106] Everyone would welcome less taxes. Yet, if many Americans feel that their federal taxes impose too heavy a burden, it cannot reflect a surfeit of tax progressivity since most people benefit at least modestly from it. Moreover, the total burden of federal taxes on the economy has not risen over the past 20 years. /45/ However, since tax burdens on average income people have risen to some degree, while falling at the top and the bottom, middle class dissatisfaction with the tax system may well reflect a sense that these increases are unfair.

THE PUBLIC'S PREFERENCES CONCERNING TAX RATES

[107] Academic analysts have analyzed public preferences about particular tax rate structures. /46/ The most rigorous study asked a sample of taxpayers in 1991 to choose among several recent alternatives and found that a convincing majority preferred substantial progressivity, but very few supported the very sharply progressive rates of the 1950s, 1960s and 1970s. /47/ Roughly one- third chose the progressive rates enacted in 1981: an initial exemption and 12 tax brackets rising from 11 percent to 50 percent in increments of 2 to 5 percentage points. Another 28 percent chose the flatter but progressive rates enacted in 1986: an initial exemption and four tax brackets of 15 percent on moderate income, 28 percent on most middle class income, 35 percent on additional income for the upper middle class, and 38.5 percent on high income. The remaining one-third chose a flat system with a substantial exemption for initial income.

[108] The study also asked people to specify the income tax rates they considered fair at each income level. Their responses indicated a clear preference for rates that would be clearly progressive but more flat than current INCOME TAX rates alone. One intriguing finding was that the income tax rates that these Americans would consider fair roughly tracked the current effective rates for all forms of federal tax, although people seem to prefer slightly lower taxation on low income families and slightly higher taxation on affluent families than there is today (Table 12).

[109] Unfortunately, this rough correspondence may not tell us much, since the sample's stated preferences refer only to the personal income tax, a relatively progressive part of the system, while the total effective rates cover all major federal taxes, which in the aggregate are less progressive than the income tax alone. /48/

[110] On balance, our best judgment is that while all Americans would prefer to pay less taxes, most would consider a fair distribution of the burden to be one that was certainly no less progressive than the current system and probably moderately more so.

THE ARMEY AND FORBES FLAT TAX PROPOSALS

[111] Proposals to replace the current system with a flat tax or a retail sales tax challenge this conclusion by calling for the repeal of many or all of the progressive features in the current tax system. The most prominent flat tax plans purport to apply a strictly proportional 17 percent tax on individual and business income, modified by expanded exemptions for initial income. The actual distributional impact of these proposals is more complicated. They would substantially and fairly systematically reduce the tax burden on the wealthy and raise it for many poor people, while among middle class families, some would receive a measure of tax relief and others would face higher tax bills.

[112] The most progressive element of the flat tax proposal is a large exemption for initial income: The plan offered by Rep. Richard Armey would replace the current $2,500 personal exemption and $6,550 standard deduction (1995 levels) with a new "personal allowance" excluding from taxation $22,700 for a married couple, $11,350 for a single person, and $14,850 for a single parent, plus a dependent's allowance excluding another $5,300 for each dependent. The proposal promoted by Steve Forbes is even more generous. As a result, a couple with two children would pay no income tax on its first $33,300 in income under Armey or $36,000 under Forbes, as compared to $16,550 in 1995 under the current system. Under present arrangements, however, the average working poor family in effect pays no income tax on its first $22,350 in income, because the EITC offsets income tax liability on another $5,800 in income.

[113] Yet, as part of "simplification," both flat tax plans would repeal the EITC. As a result, both plans in effect would RAISE the federal tax burden on most of the 17.4 million low income adults and children who currently receive the credit. DESPITE THEIR GENEROUS INITIAL EXEMPTIONS, THE FLAT TAX PROPOSALS WOULD WORSEN THE POVERTY OR NEAR-POVERTY OF MILLIONS OF WORKING AMERICANS AND THEIR CHILDREN.

                TABLE 12. PREFERRED TAX RATES versus

_____________________________________________________________________

                              Preferred     Current Effective

             Under $10K      2.4% to  4.7%        6.4%

[114] Among the middle class, the expanded initial exemption should reduce the income tax burden on some people earning roughly $25,000 to $50,000 today, especially those who do not itemize their deductions. But among middle class families who do currently itemize, many would see no tax relief because these plans repeal their deductions for mortgage interest, state and local taxes, charitable contributions, and retirement saving. In addition, all of their income above the initial exemption would be taxed at a higher rate than the 15 percent now applied to most middle class families because a revenue-neutral flat tax would require a flat rate of at least 20 percent. Many middle income families would pay higher income taxes under the flat tax than they do today, especially singles and childless couples building pensions and carrying substantial mortgages and health care coverage.

[115] Among middle class families, the flat tax on balance would reshuffle tax burdens. The plan would shift the middle class tax burden, to a moderate degree, from families with children to childless households, from those who do not itemize their deductions to those who currently do, and, in a progressive fashion, from families earning roughly between $25,000 and $50,000 to those earning $50,000 to $100,000. Nevertheless, ALL income groups under $100,000 would on average pay MORE under the flat tax than the current system. One preliminary study has estimated that families with incomes of $30,000 to $50,000 would pay 1.4 percent more, those in the $50,000- to-$75,000 income group would pay 4.6 percent more, and those with incomes of $75,000 to $100,000 would pay 3 percent more. /49/

[116] At the top, the flat tax burden turns distinctly regressive. Unlike a truly proportional tax that applies the same tax rate to all income, the Armey and Forbes flat tax plans would exempt from the personal income tax all capital income, including interest, dividends, and capital gains. SINCE HIGHER INCOME FAMILIES DERIVE ONE-THIRD OR MORE OF THEIR INCOME FROM CAPITAL SOURCES, AS COMPARED TO LESS THAN 10 PERCENT FOR MIDDLE CLASS PEOPLE, THIS FEATURE WOULD SYSTEMATICALLY SHIFT THE PERSONAL INCOME TAX BURDEN DOWN THE INCOME SCALE (Table 13). At the very top, where the richest 1 percent of families derive nearly one-half of their total income from capital, this feature would ensure that many wealthy Americans will pay SMALLER shares of annual income in personal income tax than much less affluent people. /50/ Moreover, both flat tax plans also would indirectly repeal estate taxes, which today fall almost entirely on the richest 1 percent of families.

[117] Both proposals would retain a tax on capital by maintaining a broad business-level tax, taxing capital once. But both also would entail a DOUBLE tax on labor, by retaining payroll taxes on top of the personal income tax. As a result, the two flat tax plans would shift the tax burden toward the middle class, which lives primarily on wages and salaries, and away from both the wealthy who depend on capital for their income and those moderate and lower income people who depend more on transfer benefits.

[118] Unfortunately for investors, the flat tax is likely to REDUCE the income they receive from their investments even as it treats that income as tax free at the personal level: Many provisions of the flat tax on business would be bad news for many large firms and the generally affluent people who own their shares. The first reason is what economists call "capitalization effects," based on the fact that a competitive market works to equalize the returns on all forms of investment. Under the flat tax, however, NEW fixed business investment would be untaxed; therefore THE RELATIVE VALUE OF ALL EXISTING BUSINESS INVESTMENT WOULD FALL. The flat tax, in short, would reduce the current value of firms' existing assets.

              TABLE 13. SOURCES OF FAMILY INCOME, 1988

_____________________________________________________________________

           Income    Capital   Labor   Transfer   Other

          Decile 1     6.7%    23.9%     62.4%     7.0%

[119] In addition, the flat tax would neither tax firms on their interest income nor permit firms to deduct their interest payments. That's good news for banks and other current lenders, but bad news for current borrowers. Since most large public companies today are highly leveraged, the flat tax would force them to pay additional new taxes on billions of dollars in current interest payments. In short, the flat tax would flatten the stock market, reducing the wealth of the higher income people who hold most of the shares, even as it cuts their taxes.

[120] The flat tax also would redistribute the nation's post-tax income across generations, in ways that should generally favor the elderly. By limiting personal income taxes to wages, salaries, and pension benefits, the plan would tend to shift tax burdens away from retired persons who depend principally on Social Security and income from previous savings and investments, and onto younger working families who depend on wages and salaries. As a result, the flat tax would intensify the generational transfers already occurring through the Social Security and Medicare systems. To some degree, this effect should be offset by the higher taxes on business just described and by the plan's expanded exemptions for children, which favor younger households. Nevertheless, the indirect effects on the generational distribution of the tax burden further undermine the proposal's basic economic justification: By shifting tax burdens away from elderly people, who tend to consume all of their income, onto younger families who try to save some of their income, the flat tax could have the perverse effect of REDUCING overall personal savings. /51/

THE ARCHER AND LUGAR SALES TAX PROPOSALS

[121] Proposals to replace current individual and corporate income taxes with a national retail sales tax would have even greater and more systematic regressive effects than the flat tax plans. Like the flat tax, everyone would pay taxes at the same rate under a principle of proportionality, but the share of a family's total income subject to the tax would tend to decline as its income increased.

[122] A national sales tax system, supported by House Ways and Means Committee chairman Bill Archer and Sen. Richard Lugar, would eliminate every progressive element in the current tax system. Like the flat tax proposals, it would repeal the EITC, graduated tax rates, and estate taxes, and the phaseout of tax preferences at high incomes. In addition, a sales tax system also would eliminate any exemption for initial income and all corporate taxes. In fact, the only major aspects of the current system preserved under this plan are the regressive ones -- the payroll and excise taxes.

[123] With no EITC or exemption for initial income, tax burdens on poor and moderate income families would rise dramatically. Without the exemption for initial income or the ability to deduct other expenses, income tax burdens on the middle class also would increase. Moreover, the flat rate applied to everyone's purchases would have to be substantially higher than the 15 percent marginal tax applied today to the taxable income of most middle income people: Various estimates find that a revenue-neutral sales tax would require a rate of between 25 percent and 30 percent if various "necessary" goods are exempt. /52/

[124] A retail sales tax system also would sharply cut tax burdens at the top of the income scale. The essential point of the approach is to tax consumption only, in hopes of encouraging saving. Not only would the plan exempt from any tax all income that a person saves or invests, but the tax rate would be much lower than the marginal rates paid today by wealthy people. Furthermore, by eliminating corporate and estate taxes, a sales tax system would shift those elements of the federal tax burden away from those who own or inherit capital and toward all consumers. As a result, the plan would shift the tax burden toward low, moderate, and average income people forced to consume virtually everything they earn, and away from high income families with greater relative resources to save and invest.

[125] The precise dimensions of the tax windfall for the well- to-do are difficult to estimate. While it is relatively simple to track how much capital income is received by people at different income levels, it is much harder to measure people's annual saving or investment rates. Saving or investment often occur in non-financial assets such as homes or works of art, whose changing values are difficult to assess; and the base from which one would measure new saving is problematic. The result is considerable controversy over estimates of savings rates, especially by income levels. However, there is no dispute about the general pattern: These rates tend to rise as people's incomes increase. /53/ Moreover, financial assets still dominate saving and investment, and the purchase of new financial assets closely tracks the distribution of income from those assets, which is dominated by affluent families.

[126] The only progressive aspect of a sales tax system is that it entails an implicit tax on people's existing wealth, because whenever anyone would liquidate an asset in order to consume, it would fall subject to the sales tax. Unlike current estate taxes, however, this implicit wealth tax would affect everyone. As a consequence, a national sales tax would redistribute the nation's post-tax income across generations in ways that would DISADVANTAGE the elderly -- the opposite generational effects of a flat tax. By exempting from taxation all NEW saving and investment, and implicitly taxing all existing wealth when it is consumed, the plan would shift tax burdens away from families still saving for retirement and toward those who already are retired and consuming out of their previous savings. On this basis alone, a national sales tax could increase overall national savings. /54/

[127] In practice, these saving benefits are unlikely to materialize because long-term, international experience indicates that BASIC COMPLIANCE WITH SALES TAXES DECLINES SHARPLY WHEN THE SALES TAX RATE EXCEEDS 13 PERCENT TO 15 PERCENT. A national sales tax purporting to replace the current individual and corporate income taxes would require a tax rate twice as high as that, creating enormous compliance problems and probably resulting in huge budget deficits that would drain national savings. /55/

[128] On balance, a sales tax system would replace the moderately progressive effects of current federal taxes with a new, highly regressive dynamic. The effective tax burden on poor people could more than double and the burden on most middle class families would rise substantially as well. As a result, poor and middle income Americans would claim SMALLER shares of national income after paying taxes than before paying them. People earning roughly $100,000 to $200,000 a year would no longer bear any burden of progressivity, but rather retain roughly the same share of national income after taxes as before taxes. Wealthy families, whose incomes have soared over the last generation, would avoid any tax on much of their income and claim an even LARGER share of national income after taxes than before taxes.

[129] Sales tax advocates argue that, over time, these regressive effects should recede; and from a certain theoretical perspective, they are correct. Over the course of most people's lifetimes, they will consume all of the income they earn or receive (except what they bequeath to their heirs). Therefore, over a person's lifetime, he or she should bear the same tax burden under a consumption tax as under an income tax. However, this theory holds ONLY if the income and consumption taxes apply the same tax RATES. Instead, the sales tax plan would impose a tax rate higher than the current income tax rates on poor and middle class families, and lower than the rates for affluent people. The theory also assumes that bequests either are relatively insignificant, or supplement everyone's earnings to roughly the same degree. In practice, inheritances are relatively trivial for most people, but are a highly significant source of resources for wealthy people. /56/

[130] Finally, the theoretical equivalence between consumption and income taxes assumes that the tax system doesn't change over the course of people's lifetimes. If Congress adopted a sales tax system next year and later shifted back to an income tax, the initial reduction in the tax burdens of people most able to save and invest would never be offset by a sales tax when they consumed those savings later in life. If the next round of tax reform is not the final word, tax fairness is as important in the present as it is over the course of generations. Over any period shorter than generations, a consumption tax can achieve the same progressivity as an income tax only by applying more sharply progressive rates. /57/

THE NUNN-DOMENICI SAVINGS-EXEMPT INCOME TAX PROPOSAL

[131] One consumption tax reform proposal would maintain significant tax progressivity: The USA or savings-exempt income tax reform proposed by Sens. Sam Nunn and Pete Domenici. Instead of replacing the current system, the USA plan would modify it to focus the tax burden on the income people use to consume.

[132] Unlike the sales tax, this proposal would preserve the most essential progressive features of current federal taxes -- including the EITC, a substantial exemption for initial income, graduated tax rates, and a broad-based business tax. It also would preserve the principal personal deductions that provide both regressive direct tax benefits and some progressive income effects, notably mortgage interest and charitable contributions, and add a new capped deduction for college tuition. And in an important progressive modification of the current system, the proposal also would end the double tax on wages by providing a new income tax credit equal to the payroll taxes paid by workers.

[133] The plan's other major innovation comes in its replacing current tax preferences for capital income with an unlimited personal tax exemption for all new NET saving and investment. All dividends, interest, and capital gains would be taxed at the same rates as wages and salaries; but people could fully deduct any new saving or investment, less the proceeds from previous saving or investment which they liquidate. Since the current system provides only a partial tax preference for capital income, and the USA tax plan's deduction for new net saving and investment would be unlimited, this central element of the proposal would favor higher income people.

[134] Over time, some of the regressive effect from this part of the plan would be offset by its implicit tax on existing wealth, since whenever anyone withdrew savings or sold an investment and used the proceeds to consume, those proceeds would be taxed. The USA tax also adopts the flat tax's approach to business taxation, favoring new investment in plant and equipment. As noted earlier, this change could well depress the value of financial assets on a one-time basis and thus offset some of the benefits higher income people might expect from a tax system that rewards their ability to save and invest.

[135] Nevertheless, in order to maintain roughly the same degree of progressivity as there is today, the USA tax applies MORE steeply progressive tax rates than the current system to the wages, capital income, and wealth that people use for consumption. Even so, by favoring those within each income tax bracket who save and invest more, the proposal would tend to shift the tax burden downward WITHIN each income group and tax bracket. Finally, very wealthy people capable of saving large shares of their incomes would always enjoy a substantial tax cut under this approach.

[136] This kind of consumption-based income tax also would tend to redistribute the tax burden across generations in much the same way as a national sales tax. Elderly people consuming out of their previous savings would, on average, face higher tax burdens than today, while younger people still saving and investing would receive a measure of tax relief -- once again, offsetting some of the generational transfer now occurring through Social Security and Medicare. In fact, most of the plan's economic power depends on this new tax on existing wealth when it is consumed. /58/ Yet, in order to protect most elderly people from higher taxes, the plan includes a special transition provision allowing everyone to withdraw part of their current savings and consume it without tax.

[137] As this suggests, the greatest challenge for the USA tax lies in its transition. Since all new net saving and investment would be tax deductible, the new system requires an accurate public accounting of everyone's existing assets. Without it, federal revenues could collapse, since everyone could liquidate or hide all existing financial assets before the new system took effect and then, after it was in place, claim them as NEW tax deductible saving. Yet, no one knows how to identify and track all of the economy's trillions of dollars of existing savings and investments.

[138] In the final analysis, the plan's real weakness lies not in its political concessions, its technical problems, or even its impact on the distribution of the tax burden. The USA tax is clearly the most carefully conceived and designed of the major reform proposals, and it is the only one that would NOT worsen conditions for poor people or otherwise dramatically shift federal tax burdens down the income scale. Rather, its defect lies in its basic premise: Like the flat tax and the retail sales tax, the USA tax assumes that tax policy can drive economic growth by requiring people to pay taxes based on their propensities to consume. Yet, there is no hard evidence that domestic saving and investment rates are very responsive to new tax incentives or, even if they were, that more business investment in plant and equipment would significantly increase growth. In order to test once again the contrary view, these proposals would upend or nullify a market full of economic arrangements produced by responses to the current tax system. Such an experiment is likely to involve serious transition costs for the economy, for the most modest potential benefits.

[139] Even if there were grounds for confidence in the economic notions underlying these approaches to reform, the flat tax and sales tax proposals in particular would remain profoundly problematic. Both plans falsely assume that sharp progressivity in the current tax system is distorting economic behavior. As we have seen, the current system is significantly progressive only at the bottom, where lighter tax burdens presumably promote work. For most middle class families, federal taxes already are roughly proportional, and the burdens of progressivity on more affluent families are modest. Even among the wealthiest Americans, the vast bulk of the taxes they pay represent their proportional share of the costs of government, and the additional taxes associated with progressivity claims a small share of their total income.

[140] The flat tax and the sales tax proposals would leave the tax system substantially less progressive than it is today and, in certain respects, distinctly regressive. Under both proposals, for the first time in American history, the tax system could well redistribute national income in favor of wealthy families who would claim a larger share after taxes than before taxes. As a result these plans would exacerbate the country's growing inequalities in income.

[141] Our next report will analyze fully the economic issues in tax policy, and outline a new approach to tax reform. In it, we will show how tax reform can enhance the economy's basic capacity for innovation and efficiency without sacrificing basic fairness.

NOTES

/1/ These tax incentives include the tax exclusions for employer contributions to employee pension funds; earnings on these funds; individual contributions to IRA, 401-K and Keogh accounts; earnings on life insurance savings; the rollover of home-sale capital gains; and the one-time exemption for $125,000 in home-sale capital gains. See Executive Office of the President, 1995 (Historical Tables, Table 5.2, pp. 43-46). Personal saving rates are from U.S. Department of Commerce, 1995, Table 2, p. 56.

/2/ Solow (1988) has estimated that technology and labor improvement account for more than 85 percent of the increase in U.S. productivity and wealth from 1909 to 1949, while increases in business capital account for about 12 percent. Similarly, Denison (1985) has estimated that increases in business plant and equipment account for 15 percent of real per capita U.S. growth from 1929 to 1982, while technological advances account for 54 percent and worker educational and skill improvements account for 26 percent.

/3/ Denison, 1983.

/4/ The founder of modern economics, Adam Smith, argued in Inquiry into the Causes of the Wealth of Nations that a nation's tax system should contain both elements: "The subjects of every state ought to contribute toward the support of government, as nearly as possible, in proportion to their respective abilities; that is, in proportion to the revenue which they respectively enjoy under the protection of the State."

/5/ For a discussion of the economic basis of different approaches to taxation, see Musgrave, 1994.

/6/ Kasten, et al., 1994.

/7/ Hall and Rabushka (1995) write, "The flat tax . . . is a little higher than the current income tax in the range from $30,000 to $90,000" (pp. 89-91).

/8/ The flat tax also would significantly affect the distribution of the tax burden among generations by shifting the burden away from the elderly that depend more on capital income and towards younger generations that depend more on wages and salaries. As a result, the flat tax would reinforce the intergenerational transfers now occurring through Social Security and Medicare.

/9/ Gravelle, 1994, Table 2.2, p. 44; and Congressional Budget Office, 1987, p. 67.

/10/ In the absence of special transition rules extending additional tax preferences to existing capital, the market value of the existing capital stock would fall, because NEW capital investment would be untaxed and therefore more valuable relative to old capital. The dimensions of this "capitalization effect" could be very large. For example, the financial assets owned by households exceed $7 trillion, and most of those assets are owned by high income households. A 5 percent reduction in the value of these assets would produce a one-time wealth loss of $350 billion (Gale, et al., 1996).

/11/ Ibid.

/12/ Some advocates of a national sales tax have proposed creating a new income transfer for low income people as a proxy for an exemption on initial income.

/13/ Gate, et al., 1996. This estimate assumes that low income families receive no new transfer benefits to offset their additional tax burdens.

/14/ The generational impact of both a national sales tax and the saving-exempt plan would be the opposite of the flat tax. By focusing the tax burden on the act of consumption, and exempting the act of saving and investment, they would shift more of the tax burden on to the elderly who consume most of their income.

/15/ Karoly, 1994.

/16/ For a thorough review of the findings discussed here, see Triest, 1993.

/17/ Browning and Johnson, 1984.

/18/ Ballard, 1988.

/19/ Triest, 1993.

/20/ In the 1980s, a study appeared that seemed to confound these conclusions: It found that after the tax rate cuts of 1981, men increased their working hours by 6 percent relative to prior trends from 1967 to 1980. But closer analysis revealed that the greatest increases in hours worked were among lower income men whose total tax rates had remained constant or actually increased through much of this period. These men worked more not because their tax rates drove them to, but because they had to, whatever their tax rate. Women also worked harder in the 1980s, and once again the greatest increases in hours were found among lower income women whose total federal marginal tax rate did NOT fall -- although unlike men, higher income women also worked longer hours. See Bosworth and Burtless, 1992.

/21/ The value of these deductions is phased out at a rate of 3 percent of the amount of adjusted gross income (AGI) exceeding the threshold level of $114,700 for joint returns, or $57,350 for married persons filing separately (1995 tax year). Deductions for medical expenses, investment interest, and casualty, theft, or wagering losses are not covered under this phaseout provision.

/22/ The personal exemptions are reduced at very high income levels, a rate of 2 percent of their value for every $2,500 (or fraction of $2,500) that adjusted gross income exceeds $172,050 (1995 tax year, joint returns). Thus, the personal exemptions for a married couple are phased out entirely when their adjusted gross income exceeds $232,050.

/23/ In addition, major deductions other than fringe benefits are mostly phased out at high income levels.

/24/ Slemrod, 1994a.

/25/ Gravelle, 1994, Table B.1, p. 294.

/26/ Payroll taxes impose a higher marginal tax rate on low and middle income workers (15.3 percent) than on highly paid people (2.9 percent).

/27/ However, the premiums for Medicare Part B, covering in- hospital costs, increase with income.

/28/ Social Security benefit levels increase every year, even adjusted for inflation, because the benefit formula adjusts the wage base that determines benefits not only for inflation but also for economy-wide real wage gains over the beneficiary's working lifetime. An average two-earner couple who retired in 1960 could expect to receive benefits worth $102,000 (1993 dollars) over the course of their retirement. The same couple retiring in 1980 would receive checks totaling $208,400; if they retired in 1995, they could expect to collect $226,600; by 2010, the couple would expect to receive $261,700. (Steuerle and Bakija, 1994, Table 5.3, p. 107).

/29/ Ibid.

/30/ These calculations adjust the tax payments paid to Social Security for inflation and for a 2 percent average annual real rate of return, equal to the long-term real rate of return for government securities.

/31/ Steuerle and Bakija, 1994, Table 5.3, p. 107. For example, a low wage two-earner couple retiring in 1995 should receive net retirement benefits totaling $62,600 -- less than the $78,600, net, that they would receive if they had earned average wages but more than the $37,100 net they would collect if they had earned high wages.

/32/ Ibid.

/33/ Ibid.

/34/ From this perspective, Social Security is not becoming more progressive with the passage of time: Payroll taxes will continue to burden low earners more than high earners, and high earners will continue to receive larger gross benefits once retired than low earners.

/35/ The remaining excise tax revenues come from special levies on airline tickets, non-gasoline fuels, firearm sales and firearm dealers, heavy trucks and trailers, large tires, automobiles selling for more than $32,000 or with low gasoline-mileage ratings, superfund polluters, coal, vaccines, fishing equipment, and even bows and arrows.

/36/ Congressional Budget Office, 1990.

/37/ Ibid.

/38/ In 1993, there were 2,097,430 deaths in the United States of people over age 24, and only 27,506 taxable estates, or 1.3 percent. (U.S. Department of Health and Human Services, 1993; Internal Revenue Service, 1995a, Table 1, p. 104).

/39/ Internal Revenue Service, 1995a, Table 2, p. 109.

/40/ Some of these differences in tax burdens reflect differences in incomes. Within each income quintile, the average elderly family or household has less income than the average non- elderly family -- although the average elderly household also has fewer members to support. However, within each income quintile, the average family with children has MORE income than the average childless family, principally because families with children are more likely to include two adult wage earners. This suggests that the tax advantages of having children are greater than indicated by the comparisons of effective tax rates by income quintiles; but at the same time, a family with children generally has more people to support than childless families with the same income.

/41/ Gravelle, 1994, Table 2.3, p. 47.

/42/ Current workers building private pensions, however, may receive preferential tax treatment for the income they are saving for retirement, comparable to the preferential tax treatment of capital income that benefit many elderly persons.

/43/ Consider two families in the 15 percent tax bracket, each with a total income of $35,000 a year, one childless and the other raising two children: All else being equal, the income taxes paid by the second family would be $750 less than those paid by the first ($2,500 exemption per child = $5,000 x 0.15 = $750).

/44/ At the same time, the additional costs of progressivity for the richest 1 percent of families -- 9.3 percent of incomes that average $486,000 a year -- are greater than the total income of the average American family.

/45/ In 1975, federal revenues claimed 18.5 percent of the nation's GDP; by 1995, federal taxes took in 19.2 percent of GDP. Nor has federal spending as a share of the economy risen over this time: In 1975, federal expenditures equaled 22 percent of GDP; by 1995, federal spending accounted for 21.9 percent of GDP (Executive Office of the President, 1995, Analytical Perspectives, Table 1.2, pp. 15- 16).

/46/ For a review of these studies, see Sheffrin, 1994.

/47/ Hite and Roberts, 1991.

/48/ This may suggest that the respondents might favor sharply progressive income tax rates, if they knew that other forms of federal tax are generally less progressive. But this inference also should be qualified, because research shows that most people UNDER- estimate their own marginal tax rates and tax liabilities. Therefore, those favoring higher rates might incorrectly assume that they would not be subject to them.

/49/ Gale, et al., 1996.

/50/ Put another way, since wages and salaries comprise barely 50 percent of the income produced by the top 1 percent and more than 80 percent of the income of everyone else, and the flat tax on the personal side covers only wages, salaries, and pension benefits, it will inevitably shift tax burdens onto lower income working families. Flat tax advocates argue that their business tax reforms would reduce capital income distributions, lessening the disparity. Since all capital income would be tax exempt at the personal level, this argument does not affect the proposal's overall distributional implications unless the flat tax raises the effective tax burden on business income.

/51/ For discussion, see Auerbach and Kotlikoff, 1987.

/52/ Groups that favor a national sales tax argue that the rate can be as low as 16 percent, but that rate assumes that the entire consumption base will be taxed. As the political process exempts various types of goods, however -- whether to lessen the burden on the poor or exempt certain services that are difficult to tax under a sales tax -- the rate required for revenue-neutrality inevitably rises. For example, excluding housing, medical services, life insurance, education expenses, food consumed at home, and utilities would shrink the consumption tax base by almost 50 percent, which means the tax rate would have to be almost twice as high.

/53/ One recent study that tried to measure increases in people's net worth (as a percentage of their cash incomes), by income quintile found that families in the bottom two income quintiles have average NEGATIVE net worth saving rates of -4.4 percent and -3.3 percent. But in the top quintile, the estimated rate rose to 6.6 percent (and the sample used by researchers did not include the top 1 percent of the income distribution; Sabelhaus, 1993).

/54/ The ability of a sales tax to increase savings rests NOT on problematic claims that higher rates of return increase capital formation, but on this shifting of the tax burden towards those who consume, thereby leaving more capital. Auerbach and Kotlikoff (1987) estimate that a pure consumption tax could increase the lifetime incomes of future generations by more than 2 percent, but nine-tenths of these benefits represent transfers of income from older generations to younger and future ones.

/55/ The national sales tax has other serious economic problems. In contrast to a value-added tax, a sales tax cannot distinguish between final retail sales and "final" sales of goods purchased by businesses as intermediate goods. As a result, it would arbitrarily redistribute the tax burden among sectors and firms, and provide a powerful incentive for firms to merge and integrate vertically. A sales tax also could favor foreign-owned firms over American-owned businesses, since foreign owners could repatriate their profits and consume them free of U.S. tax, while U.S. owners would pay the tax when they spent their profits. At the same time, a sales tax would initially favor exporters over importers, since exports would escape the tax while imports were subject to it.

/56/ Caspersen and Metcalf, 1984.

/57/ In theory, a labor-based tax such as the individual side of the flat tax should be roughly equivalent to a consumption tax because all national income is both derived from either labor or saving and either consumed or saved (excepting bequests). However, this equivalence also applies only over very long time periods. While roughly 10 percent of national income is saved and 90 percent is consumed, nearly 25 percent of national income is derived from saving in some form and 75 percent from labor. Therefore, at any point in time, a labor tax will exempt more of the income of the well-to-do than a consumption tax, which in turn at any point in time will exempt more of the income of the well-to-do than an income tax. Since people derive relatively less of their income from labor as their total incomes rise, a labor tax can avoid shifting the tax burden from the affluent to the middle class and poor only by applying even more steeply progressive rates than a consumption tax, and much higher marginal rates than an income tax.

/58/ Auerbach and Kotlikoff, 1987.

REFERENCES

Auerbach, Alan J., and Laurence J. Kotlikoff. 1987. Dynamic Fiscal

Ballard, Charles L. 1988. "The marginal efficiency cost of

Bosworth, Barry and Gary Burtless. 1992. "Effects of Tax Reform on

Browning, Edgar K. and William R. Johnson. 1984. "The trade-off

Caspersen, Erik and Gilbert Metcalf. 1984. "Is a Value-Added Tax

Congressional Budget Office. 1987. The Changing Distribution of

_______. 1994. An Economic Analysis of the Revenue Provisions of

_______. 1990. Federal Taxation of Tobacco, Alcoholic Beverages, and

Denison, Edward F. 1983. "The interruption of productivity growth in

_______. 1985. Trends in American Economic Growth, 1929-1982

Executive Office of the President, Office of Management and Budget.

Gale, William G., Scott Houser, and John Karl Scholz. 1996.

Gravelle, Jane G. 1994. The Economic Effects of Taxing Capital Income

Hall, Robert E. and Alvin Rabushka. 1995. The Flat Tax, Second

Hite, Peggy A. and Michael L. Roberts. 1991. "An experimental

Internal Revenue Service. 1995a. "Estate Tax Returns, 1992-3."

_______. 1995b. Statistics of Income: Individual Income Tax Returns,

Karoly, Lynn A. 1994. "Trends in income inequality: the impact of,

Kasten, Richard, Frank Sammartino and Eric Toder. 1994. "Trends in

Musgrave, Richard A. 1994. "Progressive taxation, equity, and tax

Sabelhaus, John. 1993. "What is the distributional burden of taxing

Sheffrin, Steven M. 1994. "Perceptions of fairness in the crucible of

Slemrod, Joel. 1994a. "On the high-income Laffer curve." In Joel

_______, ed. 1994b. Tax Progressivity and Income Inequality (New

Solow, Robert. 1988. "Growth theory and after." The American Economic

Steuerle, C. Eugene and Jon M. Bakija. 1994. Retooling Social

Triest, Robert K. 1993. "The efficiency cost of increased

U.S. Bureau of the Census. 1995a. Current Population Reports: Series

U.S. Department of Commerce (Bureau of Economic Analysis). 1995.

U.S. Department of Health and Human Services, National Center for