Taxing Capital Income / Introduction

Whether to tax wealth or income from wealth is a perennial issue. In 1942, President Roosevelt considered adding a large, new consumption tax exempting savings as a way to finance World War II, but ultimately settled on making the income tax much more robust. Blueprints for Basic Tax Reform, a 1976 Treasury Department study, examined both comprehensive income and consumption tax options, without picking a best option. Sweeping income tax reform followed ten years later in the Tax Reform Act of 1986. Rather than abandoning taxation of capital income, the 1986 legislation reinforced the taxation of income for all sources in a lower-rate, broader-based, structure. In 2005, the President’s Advisory Panel on Federal Tax Reform offered a choice of income and consumption tax options, but the income tax option included new tax breaks on capital income, and the consumption tax option featured a supplemental tax on dividends and capital gains to maintain progressivity. If history is any guide, ambivalence on this issue is the norm even among disciples of tax reform.

There are a variety of consumption tax proposals; each aims to exempt most capital income and eliminate the inherent tax penalty on saving built into the income tax. Since wealthy people own most capital, these options would appear to redistribute tax burdens from the rich to everyone else. Many analysts believe that taxing only consumed income would promote economic efficiency, and some believe that the extent of redistribution could be limited. Are the claimed gains in economic efficiency from encouraging saving enough to justify the alleged loss in fairness from reducing tax burdens on those most able to pay?

This book aims to examine not just these abstract arguments but also the practical issues that shifting from an income tax to a consumption tax would raise. Theoretical analysis typically compares the instantaneous replacement of comprehensive progressive income taxes with equally comprehensive, but usually less progressive, consumption taxes. The models often suggest that the largest efficiency gains would result primarily from imposing new taxes on retirees and other past savers, who are in no position to alter the investments they already made or their labor supply.

Such models are misleading in several ways. First, neither the United States nor any other nation has a truly comprehensive income tax. Capital income is taxed at widely varying rates, and some is not taxed at all. Some analysts and advocates question whether we actually do tax capital income under our income tax and, if so, by how much.¹ Joel Slemrod and several discussants examine that question in part 1.

Whether or not we tax capital income, a more fundamental question is whether we should. To answer this question, one must consider taxes as they are or can be applied, not as imagined abstractions. If capital can only be taxed imperfectly under an income tax and if exempting capital income from tax opens new opportunities for tax avoidance under a consumption tax, is it clear which system is fairest, or simplest, or even most efficient? Eric Toder, Kim Rueben, and George Zodrow consider different perspectives on those questions in part 2.

Two practical difficulties involve problems in taxing capital income when it can be moved freely across borders and when complex financial transactions can be used to conceal it. If capital income is exempt, how can government prevent taxpayers from masking earnings from labor as exempt capital income? Is it possible to exempt capital income without exempting a significant share of labor income as well? In part 3, Julie Roin looks at the implications of global enterprises for the sustainability of the income tax, Ed Kleinbard develops an innovative new system that attempts to solve the principal problems in the current porous income tax, and Joseph Bankman and Michael Schler try to determine which tax regime is likely to be more successful or more shelter-proof in light of institutional constraints. That these authors cannot agree, even among themselves, underscores the challenge.

This volume will not settle the debate over whether to tax income or consumption, but will illuminate many practical issues that policymakers and the public should consider in evaluating proposals that call for converting income to consumption taxes. Although the contributors have rendered difficult topics comprehensible, their analysis is grounded in reality, which makes it complex. Given the stakes, we believe that working through these analyses is a worthwhile investment.

What is Capital Income?

Measuring or even defining the income from capital—essentially returns from assets or saving—may seem straightforward, but it is not. Some capital income is easy to recognize: interest on bonds, dividends paid on stock, profit from the sale of a house, royalties paid to the owner of a patent, and rent received for leasing an apartment. But measuring capital income is more difficult if one adjusts for inflation or depreciation. Measuring the returns from a business, which may have branches all over the world, is vastly more complex.

Capital income consists of three types of returns:

• The first element is compensation the owner receives for deferring use of the funds invested in an asset. The longer the deferral, the larger the total payment. For example, total interest paid over the life of a 10 year bond typically exceeds total interest paid over the life of a six-month note.
• The second component is compensation to the owner for bearing risk. Those undertaking risky projects have to be paid more on average to give up control of their funds than do those undertaking safe investments. The higher the risk, the higher must be the expected (or average) return.
• The third component is an extra return that can arise on unique assets, or from market power, or luck. That extra return is not necessary to compensate the owner for deferring use of the funds or for bearing risk. Extra returns may be called windfall profits, infra-marginal returns, super normal returns, or rents. Super normal returns may derive from many sources, including unexpected changes in demand for land or natural resources, monopoly power, or even legislative favor.

Why Taxation of Capital Income Is Important

Whether and how much to tax capital income is important for at least five reasons. First, for a given amount of income tax revenue, taxing capital less means taxing labor more. Capital income taxes therefore critically influence the distribution of tax burdens.

Whether this trade-off holds over the long run is a matter of considerable debate. Some believe that capital income taxation may indirectly make workers worse off over time by reducing their wages. According to this position, capital taxes lower saving, thereby raising the cost of capital and depressing investment, which slows the growth of worker productivity and wages. Chapters 2 and 3, by Toder and Rueben and by Zodrow, and the commentaries by Auerbach and Weisbach, illustrate these disagreements.

Second, taxation of capital income determines the relative treatment of savers and consumers. If all income is taxed, whether saved or consumed, the consumer pays tax once on income that finances the consumption, but the saver pays both the tax on this initial income and, later, an additional tax on the capital income earned on the savings. Taxing that return means that deferred consumption is taxed more heavily than immediate consumption.² This second round of tax may distort the timing of consumption and favors those who consume now over those who wait.

Third, the consumption tax debate is intimately intertwined with the debate over progressive taxation. By definition, capital income flows disproportionately to owners of capital—that is, those holding wealth. Capital income is much less evenly distributed than labor income. Capital income taxes, therefore, fall disproportionately on the wealthy. For given tax rates on labor income or consumption, how heavily capital is taxed partly determines the relative taxation of the rich and the poor. Some suggest that taxing heavy consumers (mostly rich people) at much higher rates than light consumers (mostly those with lower incomes) can achieve progressivity. To mimic the distribution of tax burdens under the income tax while exempting saving could require extremely high consumption tax rates, often over 100 percent of the value of the consumption, since high-income families spend only a fraction of their incomes.

Fourth, measuring capital income is notoriously hard and getting harder as globalization and the complexity of capital transactions increases. In fact, much of the cost of complying with and administering the income tax goes into identifying and computing capital income. How the rules for taxing capital income are designed interacts with these developments to determine the cost of compliance and administration.

Many factors contribute to this complexity:

• By its very nature, capital income involves combinations of transactions that occur over two or more periods. Thus, investors must often keep detailed records.
• Increasingly, these transactions involve currencies of different nations, the relative values of which fluctuate.
• Capital income is computed net of the decline in value of capital goods from use and obsolescence, but this loss is generally estimated (for example, from depreciation schedules) and may deviate significantly from the actual loss of value for any given asset.
• Capital income and expense are not adjusted for inflation for tax purposes. When inflation is 4 percent, a bond-holder who receives a 6 percent interest rate must pay tax on the full 6 percent, but the real income is only 2 percent of the value of the bond—the 6 percent of interest income less the 4 percent drop in the real value of the bond. Similarly, bond issuers are entitled to deduct all of the interest but do not have to pay tax on the gain in wealth resulting from a drop in the real value of the debt they owe. This situation creates opportunities for tax avoidance (see section on “tax arbitrage”). (Indexing the measurement of capital income and expense for inflation, however, would create significant new complexities.)
• Capital assets increase or decrease in value constantly, but gains or losses are typically taxed, if they are taxed at all, only when an asset is sold.³ That is, taxes are usually based on realizations rather than accruals of income. Calculating these gains and losses usually entails records spanning many years. The deferral of tax until an asset is sold is valuable to investors, both because it is effectively an interest-free loan from the government and also because it creates tax avoidance opportunities (especially when deductions may be taken before the corresponding income is taxed).
• Increasingly, capital transactions involve many parties to whom various interests in gains and losses are legally assigned. Some investors pay no U.S. tax on much of their income. This group includes insurance companies (which pay no tax on income earned on some reserves held on behalf of policyholders), nonprofits (such as foundations, hospitals, and universities), and foreign investors. Through these so-called “tax indifferent parties,” investment planners not only avoid taxes but turn the tax system into a subsidy machine by assigning deductions to investors with tax liabilities while assigning income to tax-exempt entities or to taxpayers who face low rates.
• Transactions may be designed to reduce tax liability by exploiting differences in tax rules applying to different assets and legal entities. If transactions are without “economic substance,” they are illegal in principle, although the application of this common law doctrine varies from jurisdiction to jurisdiction and from case to case. Determining when there is enough economic substance to legitimate a transaction is an arcane exercise. Because the IRS audits only a fraction of such transactions, some taxpayers now engage in an “audit lottery,” undertaking transactions that may be, or are known to be, illegal in the expectation that auditors will never catch them.

Finally, capital income is often hard for administrators to spot. This problem is particularly serious when taxpayers receive income from abroad. Some foreign nations—so-called “tax havens”—encourage investors to hide assets there by refusing to cooperate with U.S. tax collectors. Some taxpayers go to great effort to hide capital income, forcing administrators to spend time and money to try to find out where capital income goes.

Despite this complexity, it is not clear that exempting capital income from tax would simplify compliance or administration. Under current law, capital gains are taxed at lower rates than ordinary income, and a whole industry is devoted to contriving schemes to transform highly taxed labor income into lightly taxed capital gains. The introduction of low tax rates on dividends has created similar incentives to transform labor income into dividends. If capital income is entirely tax-free, the payoff from such tax shelters would grow, although the disallowance of deductions for interest expense would complicate such schemes. Joseph Bankman and Michael Schler (chapter 6) reach opposite conclusions on whether replacing the income tax with a consumption tax would simplify tax compliance and administration.

CURRENT TAXATION OF CAPITAL INCOME

The classical definition of income is the amount that can be spent without depleting net worth. In the case of labor earnings, the concept seems straightforward. Workers can consume wages and salaries without changing their net worth. Wages and salaries are clearly taxable under an income tax. Computing capital income is inherently more complex because it accrues over time. In practice, different rules apply to different forms of capital income and to different taxpayers.

A Panoply of Rules

Table I.1 illustrates the variety of tax rules governing asset transactions and asset income. The diversity of rules means that disputes are bound to arise over the classification of particular transactions. Should a payment to a lender by a borrower be treated as interest (taxable to the recipient and deductible to the payer) or repayment of principal (neither taxable nor deductible)? If people sell stock, are they doing so as investors, in which case long-term gains are taxed at lower than normal rates and losses are deductible only up to limits, or as stock traders, in which case gains are fully taxable and losses are fully deductible against ordinary income as they accrue (whether or not the asset is sold)?

TABLE I.1. CURRENT LAW TAX TREATMENT OF CAPITAL INCOME AND EXPENSE FOR SELECTED ITEMS

Table I.1 does not begin to convey the complexity associated with taxation of capital income. Kleinbard, for instance, demonstrates how the tax treatment of complex financial instruments can result in what he calls a “tax pastry,” with some income being recognized currently, some deferred, some favored, and some penalized, depending on the instrument’s formal characteristics.

International Transactions

International transactions pose particular challenges that are growing in importance. Globalization facilitates both legal and illegal attribution of income to foreign rather than U.S. sources. As Julie Roin demonstrates (chapter 5), cross-border trade in finished and unfinished goods and financial instruments has increased sharply. It is not unusual, for example, for a company to produce, in several countries, components that are assembled into finished goods in another country for sale in still other countries. This company could have financed construction of each factory with loans or stock sales in still other countries. The United States taxes all income, wherever earned, of corporations chartered in the United States. But it imposes taxes only when foreign subsidiaries of U.S. companies transfer income to their U.S. chartered parents. Before the funds are transferred to U.S. parent corporations, other nations have typically taxed the income earned by these foreign subsidiaries. To prevent multiple taxes on the same income and high cumulative tax rates, the United States allows U.S. chartered companies a credit for foreign taxes paid. The credit is intended not to exceed the tax liability that would have resulted had the income been earned in the United States.

Perhaps the major challenge in taxing international transactions is determining where the income was actually earned. This determination is important because tax rates vary widely from country to country and among types of assets. Taxpayers have an incentive to exploit these differences by allocating income to low-tax jurisdictions and deductions to high-tax jurisdictions. When members of a commonly owned corporate group trade among themselves, transactions are not at “arm’s length” and the companies have considerable flexibility in setting the prices paid for trade in such items as intellectual property, professional services, and commodities. The resulting controversies over “transfer pricing” employ a small army of lawyers, accountants, and economists to help companies defend their decisions and to help tax authorities challenge them. These controversies and the associated costs have little redeeming social value and generate considerable cost. Roin believes that a perfect system of measuring capital income in an international context is infeasible when some tax havens have an incentive to help international scofflaws. However, increased enforcement, conforming rules and information-sharing agreements among countries, and improved measurement systems would help. Many of these issues associated with taxing companies based on geographical location of activity do not go away with consumption taxes.

Tax Arbitrage

Nonuniform taxation of different assets, liabilities, and taxpayers creates the opportunity for tax arbitrage. Tax arbitrage denotes the way in which taxpayers are able not simply to save taxes by investing in tax-preferred assets, but actually to leverage or multiply up tax breaks through various combinations that usually involve borrowing and lending (or analogous transactions) at the same time (Steuerle 1985). Sometimes the tax rate will end up being negative.

The costs of tax arbitrage transactions can offset a significant share of the tax savings they produce. Especially in the case of complex transactions, these costs take the form of fees to attorneys and brokers and other transactions costs.

Multiple Layers of Taxation

While tax arbitrage can reduce or eliminate tax burdens (or even turn them into subsidies), multiple layers of taxation can result in higher-than-intended tax rates. Ordinary corporate profits, for example, may be taxed in full at the corporate level and then taxed again at the individual level when dividends are paid or shares are sold. Although the low tax rate on dividends and capital gains can make certain investments attractive tax shelters when the underlying asset is not taxed at the business level, the additional individual-level tax can constitute double taxation if income is fully taxed at the business level. Layers can also be added through the multinational and multistate systems that Roin examines (chapter 5), as well as in the complex financial instruments of the Kleinbard analysis.

PROPOSALS TO REDUCE OR ELIMINATE CAPITAL INCOME TAX

Given these difficulties, it is not surprising that proposals for reforming the taxation of capital income come in many forms, including complete replacement of the income tax with a consumption tax, removal of various taxes on capital income, and changes in how the income tax treats capital income.

Consumption tax proposals come in many flavors. Many states and localities have adopted general and specific sales taxes. All developed countries other than the United States have adopted value-added taxes (VAT). The VAT is assessed on each business approximately on the difference between its sales revenues and the cost of goods and services purchased from other companies. When investment outlays are fully deductible in the year the investment is made, value-added taxes are equivalent to consumption taxes. If only depreciation is deductible, value-added taxes fall on income. In most countries, sales taxes and consumption-type value-added taxes supplement income taxes. Some U.S. groups have suggested that a large retail sales tax should replace personal and corporation income taxes (and perhaps other taxes as well).

Because consumption is defined as income less change in net worth, consumption can be taxed by setting the tax base equal to income less net saving. This approach requires either some form of wealth accounting or measures of all transactions into and out of designated accounts (gross saving or deposits less gross “dissaving” or withdrawals) to approximate the change in net worth.

Value added is the sum of labor compensation and business cost flow. VATs tax both components at the business level. “Flat taxes” are VATs that tax labor compensation at the household level and cash flow at the business level. For instance, if a company buys inputs for $100, pays wages of $700, and sells output for $1,100, its total value added is $1,000 ($1,100 in sales minus $100 of inputs). Under the VAT, the $1,000 is taxed at the company level. Under a flat tax, the $1,000 is divided into two parts. The $700 labor component is taxed at the household level, and the remaining $300 is taxed at the company level. In almost all VATs and flat taxes, the company tax is deferred if earnings are retained and invested, as investments produce immediate deductions. In that case, profits effectively are not taxed until consumed.

One important difference distinguishes the two taxes. Under the flat tax, the household tax may be applied to earnings above an exemption—a zero rate for some below the exempt amount—which introduces an element of progressivity not present in the VAT. The so-called “X tax” not only includes an exemption, but also applies progressive rates to labor income above the exemption. Both the flat tax and the X tax represent coherent reforms that could replace the income tax system.

Some so-called consumption tax plans propose selective or piecemeal modifications to the current system. In many cases, the principal effect of such changes would be to open new tax avoidance opportunities. One such proposal would permit unlimited deposits in individual retirement accounts or other accounts receiving similar preferential tax treatment while continuing to allow deductions for interest payments. Under this scenario, people would be able to borrow, deduct interest costs, and deposit funds in tax-free accounts, saving nothing on balance but generating tax savings. Legislators could pass complex rules to try to limit such schemes, but the financial reward is so great from such arbitrage transactions that the rules would be unlikely to keep up with the tax avoidance schemes.

WHAT DO WE KNOW ABOUT TAXING CAPITAL INCOME?

Although analysts disagree intensely about many aspects of capital income taxation, they concur substantially on some important conclusions. To begin with, taxes on capital, like all taxes, distort behavior. Most analysts agree that taxing capital income sometimes treats savers unfairly relative to consumers who have the same initial ability to pay tax but do not save. Accounting for capital, tax planning, and sheltering income undeniably contribute to tax complexity. Other agreed-upon sources of complexity, distortion, and unfairness include taxing capital income on a realization basis, improperly measuring capital income in the presence of inflation, and determining the location of income in companies with international transactions and operations in multiple jurisdictions.

Do We Tax Capital Income?

Analysts do not agree how much capital income is actually taxed. If effective capital income tax rates are as low as some suggest, the transition to some form of consumption tax might not be especially jarring (and might offer even larger efficiency gains than implied in simple theoretical models). Unfortunately, how much capital income is taxed is not clear. Is all capital income taxed, including the return to risk taking and windfall or monopoly profits, or just the return to waiting, which forms a small part of total capital income? Is the relevant rate in measuring the burden of capital income taxation the average tax rate or the marginal tax rate on new investments?

Joel Slemrod tries to answer these vexing questions in chapter 1. He presents three measures of the capital income tax burden—capital taxes as a share of capital income, the hypothetical tax rate on a new project, and an empirical measure of the marginal effect of taxation on returns to capital. The different measures suggest that the effective marginal tax rate on returns to saving is between 14 and 24 percent. Jane Gravelle questions some of Slemrod’s assumptions in her comments on chapter 1. She estimates that the actual burden is somewhat higher than he does. She also stresses the importance of differences in tax rates on various sorts of capital.

Should We Tax Capital Income?

A key issue in assessing tax reform proposals is the size of current tax distortions to consumption, saving, and investment and the potential efficiency gains from switching to a consumption tax base. Economists agree that many potential gains in efficiency implied from conversion to a consumption tax arise from taxing “old capital”—that is, from taxing the consumption financed out of previously taxed income. Such taxation, however, raises issues of fairness and administrative feasibility. Apart from fairness, it is unclear that such a windfall tax on old capital could ever be imposed without distorting behavior. Taxpayers would have to believe that the confiscation of old wealth will not be repeated. Without such faith, they may shape economic decisions to shield themselves from the risk of anticipated tax increases, for example, by avoiding saving or deferring investment while awaiting the next set of rule changes.

Many opponents of consumption taxes are concerned more about an erosion of tax progressivity than about the merits of one tax base over another. Many consumption tax proposals, along with companion proposals to eliminate the estate tax, would reduce progressivity. Whether this outcome is necessary is also subject to disagreement. In chapter 2, Zodrow argues that progressivity could be maintained under a consumption tax. To do so would require curtailing almost all tax preferences that benefit the wealthy and minimizing transition relief for old capital. If these steps are not taken, Zodrow acknowledges that maintaining progressivity might require extremely high (and possibly unacceptable) maximum tax rates as well as possibly retaining an estate tax. Thus, the desirability of tax reform may hinge on the political feasibility and effectiveness of assessing such taxes on the wealthy.

Analysts disagree even on whether and how much the income tax penalizes savers relative to nonsavers. While a case can be made for not taxing normal or safe returns from capital, not taxing exceptional returns, such as those from monopoly, or even returns to risk taking, is harder to justify. In chapter 3, Toder and Rueben argue that only a small portion of total capital income reflects the normal or safe return.

Many consumption taxation advocates hold that whether one pays taxes now or later is not important provided that the amount of tax increases with the rate of return on the underlying asset. For example, if tax rates rise, Alan Auerbach explains that a consumption tax amounts to a tax on capital income because current consumption is treated more favorably than deferred consumption (see comment to chapter 2). But timing of taxes may be important. If one regards taxes as the price of public services and such services provide the legal and social protections essential for the security of capital income, then recipients of such income may properly be held responsible for bearing some of the costs of those arrangements. Deferring the taxes required to pay for today’s benefits well into the future, perhaps to future generations, results in a temporal mismatch between when benefits are received and when they are paid for. Because tax changes occur continuously, it is doubtful whether a fair distribution of payments for benefits received can be achieved with so much deferral.

Simplicity is a frequently claimed virtue of consumption taxes because they would eliminate the need to calculate depreciation and capital gains and to track complex arrangements, such as installment sales where payments and deliveries are made over time. However, businesses would still need to account for capital income to allocate their investments efficiently. Corporations long ago introduced modern capital income accounting to meet their needs to plan operations efficiently and to report to shareholders. If computation of taxable income exactly matched internal business requirements, no appreciable simplification would result from abandoning the income tax. But computation of taxable and business income are not identical. Thus, some accounting simplification might still result from a switch to a tax system that does not impose the requirements of the current income tax. The size of such savings is less clear.

Can We Tax Capital Income?

In many cases, implicit income taxes are likely to be retained even with a conversion to a consumption tax. Government policies require many people and some businesses to measure income for other purposes. For example, in many transfer programs, including cash assistance and subsidies for education, housing, or purchase of food, benefits are reduced as income increases. To reduce benefits without regard for capital income would hardly be sensible, and basing benefits on consumption is neither administratively feasible nor sensible.

Whether or not the income tax is reformed, income accounting will still be required for these other governmental systems. Further, most states and nations use income accounting. Thus, businesses and individuals subject to state or foreign taxation will still have to compute capital income, as Julie Roin notes in chapter 5.

Nor would a switch to consumption taxation eliminate opportunities for tax avoidance and evasion, as Joseph Bankman and Michael Schler explain in chapter 6. They and Julie Roin warn that companies still must allocate their sales and many of their intracompany or related transfers to the right jurisdictions at the right prices. Schler and Bankman point out that a consumption tax of the flat or X type would create new tax planning opportunities. They agree that controlling these opportunities would be difficult, but they disagree on whether complexity and tax cheating opportunities under a consumption tax would worsen or improve.

The fundamental reason why tax reform cannot eliminate complexity, tax avoidance, and evasion is that these problems emerge from politics and the complexity of modern economic activity. For example, whatever the tax system, accounting sophistication and politics are likely to cause legislators to favor small over large businesses. If differences in treatment exist, taxpayers will try to exploit them.

International trade and capital flows create the additional opportunity to hide assets abroad, and a switch to consumption taxation will increase these opportunities. If a filer hides assets abroad, the government loses tax on income from those assets under an income tax. With a flat or X tax, an asset hidden abroad can end up reducing the tax base by the entire value of the asset. For example, hiding $100 of an asset abroad in one year would reduce the income tax base by $5 if the rate of return is 5 percent. Hiding the same asset could reduce the consumption tax base by $100.

Further, a consumption tax would defer massive amounts of tax as filers take deductions for the full value of current investments. Schler worries that deferring so much more tax under a consumption tax than occurs under an income tax, where only depreciation is available upon purchase of assets, will create massive incentives for filers and their accounting and legal representatives to craft new tax shelters that neither legislators nor administrators can fully anticipate. Joseph Bankman views this risk as inadequate grounds for rejecting a consumption tax, given all the ways that capital income can be sheltered under an income tax.

In chapter 4, Ed Kleinbard acknowledges that financial instruments can be manipulated to avoid tax but believes that this problem can be fixed. He examines two reforms. Under one, most capital income would be taxed at the corporate level but still on a realization basis. Under an alternative that he designed, capital income would be taxed currently on an approximation or “imputed” basis in cases where there are no realizations of income. He also favors a simplified system of accounting for different types of capital costs (such as traditional depreciation). These accounting changes combined with the imputation system would minimize the possibility of tax planning and avoidance. Not everyone agrees, and Kleinbard himself calls for further analysis of his proposed changes.

Conclusion

The debate over taxing capital income is centuries old. Recent research has added theoretical insights and empirical knowledge. Certain tried and true principles of reform remain generally accepted. The best hope for reducing tax-based distortion and simplifying compliance and administration lies in curtailing special exceptions, exclusions, deductions, and credits. This formula applies as much to income as to consumption tax reform. The rate reduction made possible by broadening the tax base further reduces distortions and enhances fairness because the value of remaining tax deductions and exclusions declines when rates are lowered.

We believe that any broad expansion of consumption taxation would use a value-added tax base. All developed countries that have adopted VATs already have income taxes. VAT revenues could finance large personal exemptions under the income tax, freeing many filers from income tax burdens, as Michael Graetz (1997) suggests. Even without direct adoption of a VAT, most well developed proposals for conversion of the income tax base to a consumption tax base have built upon a VAT-like structure. Other attempts, such as personal consumption taxes (based on measuring changes in net worth) and very high retail sales taxes, have floundered over issues of administration and enforcement.

Under any tax system, the wealthy likely will pay most of the tax and far more than they receive in cash or in-kind benefits. Thus, overall fiscal progressivity is almost inevitable, even if the degree and method of achieving progressivity are contentious. This debate is inescapable and desirable.

Estimates of the gains from a switch to a consumption tax all rest on highly abstract models of the economy. The estimates are extremely sensitive to assumptions. Moreover, economists cannot determine with certainty even such key matters as how heavily capital is currently taxed. Under these circumstances, a wholesale switch in the tax system is a large and risky step. Whether the switch will, in the end, raise or lower welfare may depend on politically sensitive matters, such as how much transition relief is provided to owners of capital when the system is changed.

References

Council of Economic Advisors. 2005. Economic Report of the President. Washington, DC: U.S. Government Printing Office.

Graetz, Michael. 1997. The Decline (And Fall?) of the Income Tax. New York: Norton & Norton Co.

Office of Management and Budget. 2006. “Tax Expenditures.” Section 19, Budget of the U.S. Government, Fiscal Year 2007. Analytical Perspectives. Washington, DC: Superintendent of Documents.

Notes

1. President George W. Bush’s Council of Economic Advisers has argued that the switch from an income tax to a consumption tax would be far less disruptive than many assume because so much capital income is already exempt from tax (Economic Report of the President 2005). The Office of Management and Budget made a similar argument in arguing that the baseline tax system for examining “tax expenditures”—that is, tax provisions that deviate from the norm—might just as well be a consumption tax as an income tax (OMB 2006).

2. For example, if the interest rate is 5 percent, then, before taxes, $1.00 today is the equivalent of $1.05 a year from now. If capital income is taxed at, say, a 20 percent rate, 1 cent (20 percent) of the 5 cent return to waiting would be paid in tax. Thus, the tax reduces the return to waiting by 20 percent. If the taxpayer was just indifferent between current consumption and future consumption at the 5 percent tax rate, he or she would not be willing to lend (i.e., would increase current consumption) when saving is subject to tax.

3. Although many gains and losses are recognized for tax purposes only when assets are sold, in the case of some financial market derivatives and some financial asset traders and brokers, accrued gains and losses are computed regularly, even when the underlying assets has not been sold, based on market prices.

Taxing Capital Income, edited by Henry J. Aaron, Leonard E. Burman, and C. Eugene Steuerle, is available from the Urban Institute Press (paper, 6" x 9", 366 pages, ISBN 978-0-87766-737-7, $29.50).