For other versions of this document, see http://wikileaks.org/wiki/CRS-RL31597 ------------------------------------------------------------------------------ ¡ ¢ ¢ Prepared for Members and Committees of Congress ¡ ¢ ¢ The downturn in the stock market in 2002 prompted renewed interest in tax relief for income from corporate dividends, partly to stimulate the economy. It is also sometimes argued that corporate shareholders are unfairly taxed twice on their corporate earnings, once by the corporate tax and again at the individual level, and that these shareholders are disproportionately elderly. Critics of dividend tax relief cite concerns about large revenue costs and concerns that tax benefits go to well-off taxpayers. Temporary dividend relief (through 2008) was enacted in 2003 (P.L. 108-27); P.L. 109-455 extended the benefits an additional two years. Dividend tax reductions have less stimulative effects on the economy than government spending or tax cuts for lower and moderate income individuals, because it is not as likely to directly increase spending, which is the most effective way to stimulate the economy. Dividend tax reductions that increase demand for corporate stock, while possibly boosting market values and spending out of wealth, would also decrease consumption spending or shift funds from bonds that could raise interest rates, both limiting any expansionary effects. Economists have traditionally criticized the current tax system, which imposes both a corporate and individual income tax, for creating distortions in the allocation of the capital stock that can be large relative to revenue raised. Economic analysis, however, disputes the unfairness argument because behavioral responses cause the true burden of the corporate tax to be spread to other income (based on most economic analysis, to all capital income). Most analysis also suggests that the reduction in the additional tax paid on corporate equity investments would tend to benefit higher income individuals, an issue of importance for some. A barrier to reducing the double tax is the significant revenue loss. Rough order-of-magnitude calculations suggest that dividend deductions by firms could result in annual revenue losses well in excess of $100 billion per year; dividend exclusions at the individual level could cost $25 billion per year. The difference reflects the small share of dividends subject to individual tax (about a third): most of the revenue cost of a dividend deduction for firms would be associated with non-taxable entities (pensions, individual retirement accounts, life insurance companies, and foreigners). Costs could be reduced with a partial exclusion or capped exclusion. Corporate tax integration has been studied many times. Efficiency gains should be considered in the light of revenue needs and distributional effects. The revenue cost could be reduced to a few billion dollars by a capped exclusion (such as the $400 exclusion in previous law). Such a proposal would not have any of the economic efficiency effects, however--rather, it would merely be a windfall for relatively well-off taxpayers. Exclusions of small amounts of passive earnings are sometimes proposed based on eliminating the need to file certain tax schedules, but such a small exclusion would only reduce filing requirements significantly if it included interest as well as dividends. This report will be updated to reflect legislative developments. ¡ ¢ History of Dividend Taxation .......................................................................................................... 2 Macroeconomic Impacts ................................................................................................................. 3 Issues of Equity and Distribution .................................................................................................... 4 Dividend Taxation, Economic Efficiency, and Saving .................................................................... 8 Administrative and Other Issues ..................................................................................................... 9 Methods of Integration and Dividend Relief and Their Costs....................................................... 10 Full Integration........................................................................................................................ 10 Dividend Relief ....................................................................................................................... 12 Assessment .................................................................................................................................... 13 Table 1. Dividends by Income Class, 2005 ..................................................................................... 6 Table 2. Distribution of Income: Capital, Labor and Total.............................................................. 7 Author Contact Information .......................................................................................................... 14 ¡ ¢ T he downturn in the stock market in 2002 prompted renewed interest in tax relief for income from corporate dividends. President Bush had expressed interest in providing tax reductions for income from corporate dividends as part of a plan that would provide economic stimulus by increasing stock market values during an August economic summit. In 2003, the President proposed a plan for eliminating individual taxes on dividends (and taxes on capital gains arising from retained earnings) as the centerpiece of his $674 billion (over ten years) stimulus plan: the dividend proposal accounted for $364 billion of the total. In early May 2003, the House adopted H.R. 2 which provided for a 15% maximum tax rate on both dividend and capital gains income; this provision eventually became part of the bill, enacted as P.L. 108-27. The lower rates were scheduled to expire after 2008. P.L. 109-222 enacted in 2006 extended the provisions another two years. The original dividend and capital gains relief costs $50.8 billion from 2006-2015, and P.L. 109-222 adds an additional cost of $50.89 billion through the budget horizon (through 2016). As with other temporary provisions, legislation will be required to make these changes permanent. The treatment of dividends and capital gains is also related to an issues that has been given some consideration by the administration and is also contained in a tax reform bill introduced by Chairman Rangel of the Ways and Means Committee, namely to lower the statutory corporate tax rate. With lower taxes on dividends and capital gains, it is much easier for a lower corporate tax rate to produce a tax shelter for wealthy individuals by lowering the combined rate of the tax.1 It is sometimes argued that corporate shareholders are unfairly taxed twice on their corporate earnings. First, corporate profits are taxed at the corporate level under the corporate income tax and then, when corporate profits are distributed as dividends to shareholders, they are taxed under the individual income tax. Concern has also been expressed about the effects of the current tax treatment of dividend income on the elderly. An analysis of the taxation of dividend income requires an understanding of the relationship between the individual and corporate income taxes. The corporate income tax actually pre-dates the individual income tax and has always been imposed as a separate tax. As a result, income from corporate equity investments is subject to higher tax rates than income from other investments. Economists have frequently criticized the current tax system, which imposes both a corporate and individual income tax, for creating inefficiencies in the allocation of the capital stock. Economists, however, generally reject the unfairness argument because economic models suggest that behavioral responses cause the true burden of the corporate tax to be spread to other income (based on most analysis, to all capital income). Studies of how to integrate the individual and corporate income taxes have occurred with regular frequency over the years. However, the corporate income tax raises significant amounts of revenue and adds progressivity to the federal tax system. Hence, proposals for full or partial (dividend relief) integration of the corporate and individual income taxes have significant revenue and distributional consequences. This report begins with a brief history of the tax treatment of dividend income. Next it discusses the effectiveness of dividend relief for stimulating the economy. The report then turns to the more 1 For further discussion see CRS Report RL34229, Corporate Tax Reform: Issues for Congress, by Jane G. Gravelle and Thomas L. Hungerford. See CRS Report RL34249, The Tax Reduction and Reform Act of 2007: An Overview, by Jane G. Gravelle, for a review of Chairman Rangel's bill. ¡ ¢ traditional issues surrounding corporate tax integration and dividend relief: equity and distributional issues, economic efficiency, administrative issues, and the revenue costs for alternative relief proposals. ¢ ¡ The modern corporate income tax came into being in 1909 as part of the Payne-Aldrich Tariff Act. The tax was set at 1% of net income over $5,000. Net income included dividends paid to stockholders. The modern individual income tax came into being in 1913 as Section II of the Underwood- Simmons Tariff Act. It imposed a 1% tax on the net income of citizens and residents of the United States (the 1% levy was referred to as the normal tax). In addition, the 1913 Act levied an additional tax or surtax on an individual's net income in excess of $20,000. The surtax rates ranged from 1% to 6%. For purposes of the normal tax, net income did not include dividends paid from the net earnings of corporations subject to the corporate tax. However, dividend income was included in an individual's net income for purposes of computing the surtax. The Revenue Act of 1936 (sometimes referred to as the Undistributed Profits Tax Act) made significant changes in the tax treatment of dividend income. First, corporate dividends paid to individuals were subject to both the normal individual income tax and the surtax. Second, the act imposed a new surtax on the undistributed net income of corporations. The new corporate surtax consisted of five graduated rates ranging from 7% to 27%. The 1936 Act was designed to prevent what was considered a "leakage" in the individual income tax system. The tax rate on corporate income was lower than the individual surtax rates. Corporations could reduce the net tax on corporate source income by retaining earnings rather than paying them out to stockholders as dividends (where they would be subject to the surtax). This was considered a tax avoidance scheme for upper income taxpayers and the 1936 Act was imposed as a means of forcing corporations to pay out their earnings as dividends where they could be taxed to the individual stockholders. The 1936 Act was vigorously opposed by corporate interests as a detriment to investment and growth. The Revenue Act of 1938 essentially repealed the tax on undistributed corporate profits. It significantly reduced the surtax rate structure on undistributed profits and applied the surtax to corporations with net incomes over $25,000. Moreover, the surtax was applicable only for calendar years 1938 and 1939, after which it expired. From 1939 until 1954, there was no special corporate or individual income tax treatment of dividend income. The 1954 Act which recodified the tax (the Internal Revenue Code of 1954) introduced a new dividend exclusion for individuals. For married couples, each spouse could exclude the first $50 of dividend income received with respect to stock owned by that spouse. Hence married couples could have a maximum exclusion of $100 if both spouses received at least $50 of dividend income. Single individuals were allowed to exclude up to $50 of dividend income from taxation. In addition, taxpayers were granted a tax credit equal to 4% of the dividends they received in excess of the exclusion. ¡ ¢ The Revenue Act of 1964 increased the dividend exclusion for married couples to $100 for each spouse (maximum of $200 per joint return). It also increased the dividend exclusion to $100 for single individuals. For tax year 1964, it reduced the dividend tax credit to 2% of dividends received in excess of the exclusion. For tax years after 1964, the act repealed the dividend tax credit. In 1980, the Crude Oil Windfall Profits Tax Act increased the maximum dividend exclusion for joint returns from $200 to $400 and allowed the exclusion regardless of which spouse earned the income. For single individuals, the dividend exclusion was increased from $100 to $200. The 1980 Act also expanded the exclusion to cover interest income. These changes were to be effective only for tax years 1981 and 1982. After 1982, the exclusion was to revert to its previous law levels and coverage ($100 exclusion of dividend income for each individual with respect to stock owned by that individual). The Economic Recovery Tax Act of 1981 repealed the interest and dividend exclusion for tax years beginning after December 31, 1981. However, the 1981 Act reinstated the previous law exclusion of up to $100 of dividend income from taxation. For joint returns, a $200 dividend exclusion was allowed without regard to which spouse actually received the dividend income. This reinstated dividend exclusion became effective in tax year 1982. The dividend exclusion for individuals was ultimately repealed by the Tax Reform Act of 1986, which lowered tax rates and broadened the tax base. The issue of dividend tax relief resurfaced in the 105th Congress when several bills were introduced to lower the tax on dividend income. The downturn in the stock market during 2002 prompted President Bush to include dividend tax reductions in his FY2004 stimulus package. In early May 2003, the House adopted H.R. 2 which would provide for a 15% maximum tax rate on both dividend and capital gains income; this proposal because part of the final bill but was adopted as a temporary provision through 2008. The Tax Increase Prevention and Reconciliation Act of 2005, adopted in 2006, extended the 2003 reductions through 2010. One objective of dividend relief was to stimulate the economy in the short term. Normally a tax benefit favoring individuals with high permanent incomes (such as a capital gains tax cut, or, as suggested subsequently, dividend relief) is a relatively ineffective way to stimulate the economy because these individuals tend to have a higher propensity to save, and it is spending, not saving, that stimulates the economy. The most effective economic stimulus is one that most closely translates dollar for dollar into spending.2 Direct government spending on goods and services would tend to rank as the most effective, followed by transfers and tax cuts for lower income individuals (who have a higher propensity to consume). 2 For a discussion of the effectiveness of alternative stimulus proposals, see CRS Report RS21136, Government Spending or Tax Reduction: Which Might Add More Stimulus to the Economy?, by Marc Labonte; CRS Report RS21126, Tax Cuts and Economic Stimulus: How Effective Are the Alternatives?, by Jane G. Gravelle; CRS Report RL31134, Using Business Tax Cuts to Stimulate the Economy, by Jane G. Gravelle; and CRS Report RS21014, Economic and Revenue Effects of Permanent and Temporary Capital Gains Tax Cuts, by Jane G. Gravelle. ¡ ¢ By these standards, dividend relief tends to rank relatively low as an effective stimulus. Whether relief is provided directly to corporations or to individuals, the initial recipients tend to have low short run propensities to spend. Increased cash flow to corporations would not be expected in theory to increase corporate spending, especially in a downturn, and empirical evidence, while showing a small positive relationship between investment spending and cash flow, has significant limitations. It is therefore reasonable to expect that most of a cut at the corporate level would be used to pay down debt or paid out as dividends rather than spent immediately. If relief is provided directly to individuals, the effect will also be limited because dividends are concentrated among higher income individuals who tend to save more. One argument that might be made for choosing dividend tax relief as a stimulus tool is that it would increase the value of the stock market and thus investor confidence (as well as spending through a wealth effect). Such a link is weaker, more uncertain, and perhaps more delayed, than a direct stimulus to the economy via spending increases or cuts in taxes aimed at lower income individuals. Indeed, it is possible that dividend relief could introduce some contractionary elements through portfolio shifts. Increased demand for stocks may raise stock prices, but the translation of this effect into investment stimulus will likely be felt with a substantial delay. If the increased demand for corporate stock comes at the expense of investment in noncorporate assets, however, then the negative effects on investment could occur more quickly than positive ones depending on how fast a contraction in supply leads to a smaller investment. If investors move funds directly out of non-corporate business investments, there will be an immediate contraction in investment spending for those areas where most spending is in new assets. If funds are moved from debt financed assets, the amount of debt available will fall and interest rates will rise, which could have a more immediate effect on investment than a rise in the stock market (and also offset the stock market increase). ¢ Because of the existence of both a corporate and individual income tax, income from corporate equity investments is subject to higher tax rates than other investments. For example, $1 of net corporate income would be subjected to the corporate tax rate of 35%, generating $0.35 of tax and leaving $0.65 of after-tax income. If the $0.65 of after tax income were paid out as a dividend to a shareholder in the 25% marginal individual income tax bracket, then an additional $0.1625 of tax would be owed (25% of $0.65). Hence, the tax rate on $1 of net corporate income paid out as a dividend would be 51.25% ($0.35 corporate plus $0.1625 individual). Retained earnings, which increases the value of the firm's stock, may eventually be taxed as capital gains income to the shareholder when the shareholder disposes of the stock. For retained corporate earnings, however, the problem of high tax rates caused by taxing corporate source income twice is minimized because individual capital gains tax rates are lower than ordinary income tax rates, taxes on capital gains income are deferred until the asset is disposed of, and the value of the asset may benefit from the step-up in basis rules on the death of the taxpayer. Two distinct equity issues have emerged during the discussion of dividend relief proposals. The first is the argument that holders of corporate stock are subject to an "unfair" double tax. This issue is an issue of horizontal equity: are holders of corporate stock treated unfairly compared to holders of corporate bonds or other investments? The second is the question of how the tax is ¡ ¢ distributed across incomes in the economy, and that depends on whether the burden of the tax is more likely to be borne by higher income individuals. These equity issues hinge on the behavioral responses leading to the shifting and ultimate incidence of the corporate tax. Analyses of the corporate income tax generally suggest that the extra tax imposed on corporate equity falls on owners of corporate stock in the short run but is spread to other incomes (either other capital income or labor income) in the long run. Because returns to corporate equity are taxed at higher rates than the returns to other assets, capital should migrate out of the corporate sector (inducing higher rates of return before tax in the corporate sector as it does so) and into the non-corporate sector (inducing lower rates of return before tax in the non-corporate sector). This process should continue until after tax returns are equated in both sectors on a risk-adjusted basis. This adjustment process means that the equity issue is not one of unfairness to holders of corporate stock because these individuals pay taxes twice. Even though dividend recipients are legally obligated to pay more taxes, they are partially compensated by the higher pre-tax returns that arise from the shifting process and this partial compensation puts them on an equal footing with investors in other assets. Thus the "unfairness" problem has already been addressed by market forces. Rather the equity issue is one of vertical equity: whether the tax on capital income contributes to the progressivity of the tax system and how desirable that progressivity is. For that purpose, the crucial issue is where the tax has been shifted. An extensive literature studying the effects of the corporate tax under many different circumstances concludes that the burden of the extra tax falls on all capital and that the tax on corporate equity income can be viewed, for purposes of distributional analysis, as a tax on all capital income.3 There are several caveats to the conclusion that the corporate tax is a tax on all capital, but none of them support the argument that dividend relief is needed in the interest of fair treatment to dividend recipients. Indeed, if anything, some arguments suggest that dividend relief would produce a windfall gain to recipients. The conclusion that the corporate tax eventually falls on capital income (as opposed to labor income) rests on the assumption of a fixed capital stock. A decline in the capital stock as a result of the corporate tax would cause part of the burden to fall on labor.4 The capital stock could vary for two reasons: savings could change or capital could exit the country. The first effect is not certain in direction but is probably small.5 The effect of a corporate tax on capital outflows and 3 The landmark study is Arnold C. Harberger, "The Incidence of the Corporate Income Tax," Journal of Political Economy, Vol. 70, June 1962. For a review of further developments, see Jane G. Gravelle, "The Corporate Income Tax: Economic and Policy Issues," National Tax Journal, Vol. 48, June 1995. 4 Discussions of the corporate tax sometimes talk of passing the tax forward in prices to consumers. But the tax actually falls, in an aggregate sense, on capital or labor income. The relative price of corporate-produced goods could go up, but the relative price of noncorporate goods would fall, with no aggregate effect on real prices. (Nominal prices in the economy only rise with monetary accommodation and real effects would not occur in any case.) Consumers who prefer corporate goods might be burdened while those who prefer noncorporate goods would benefit, but there is no aggregate effect on consumers per se, other than in the underlying effects on labor and capital income. 5 Opposing income and substitution effects make the effect on savings of altering capital income taxes uncertain. Direct empirical studies of savings elasticities have found mixed results, but generally report small elasticities that can be positive or negative. See Eric Engen, Jane Gravelle, and Kent Smetters, "Dynamic Tax Models: Why They Do the Things They Do?" National Tax Journal, Vol. 50, September 1997. ¡ ¢ incidence depends on the mobility of capital and the substitutability of products, the degree to which the tax is a territorial versus a residence-based tax, and the degree to which the imposition and magnitude of a U.S. corporate tax causes other countries to adopt similar taxes.6 But even in the case where a territorial tax is assumed and no response from other countries occurs, empirical measures of capital and product mobility suggest that the burden still largely falls on capital income.7 Moreover, even with estimated equity capital outflows, the corporate income tax encourages an inflow of debt capital and the effects on debt-financed capital, which is more mobile, could offset, or more than offset, the effects on equity-financed capital. If firms have market power and are able to earn a return higher than normal, some fraction of the tax is likely to fall on owners of corporate stock indefinitely. This tax should appear as a one time decline in prices (when the corporate tax is imposed). That is, the burden of the corporate tax is not falling on current owners but is already capitalized in prices. Hence, a cut in taxes would result in a windfall gain. One theory of corporate tax incidence argues that dividend taxes are capitalized in asset values, and, thus, the burden of the normal tax on dividends is also a one-time windfall tax, on owners of corporate stock, with no current consequences. This theory, called the "new view" relies, however, on the inability of firms to repurchase their own shares, which is not restrained in the U.S. tax system (but was in the British tax system when this theory was first introduced by a British economist). Table 1 shows data on the distribution of dividends and Table 2 shows the distribution of capital income in general. Both show significant concentration among higher income individuals. Over 40% of dividends are received by the top 2% of returns (incomes in excess of $200,000) with almost 80% of dividends received by the top quarter (incomes in excess of $50,000). Note that ownership of stocks may be more common in the middle classes through indirect ownership via pensions and retirement savings accounts. The return on these investments is not, however, effectively subject to tax. 5002 ,ssalC emocnI yb sdnediviD .1 elbaT detsujdA fo tnecreP fo tnecreP fo noitcarF egarevA emocnI ssorG ni snruteR ni sdnediviD egarevA htiw snruteR rof sdnediviD )sdnasuoht$( ssalC hcaE ssalC hcaE sdnediviD sdnediviD htiw snruteR sdnediviD enon 3.1 2.1 28$ 03 067,2$ 5 rednu 5.8 5.0 94 11 654 6 A residence-based tax applies to all capital owned by U.S. residents regardless of where invested, while a territorial tax applies depending on the location of investment. The current corporate income tax is a mixture of the two. 7 See Jane G. Gravelle and Kent Smetters, "Does the Open Economy Assumption Really Mean That Labor Bears the Burden of a Capital Income Tax?," Advances in Economic Analysis & Policy: Vol. 6: No. 1, Article 3, 2006, available at http://www.bepress.com/bejeap/advances/vol6/iss1/art3/. Note that there have been some recent cross-country empirical studies that find the burden falling on labor and by amounts far in excess of the tax, but there are serious problems with these studies. See CRS Report RL34229, Corporate Tax Reform: Issues for Congress, by Jane G. Gravelle and Thomas L. Hungerford. might explain why the bottom 20% receive 3% of dividends but less than 1% of capital income. incomes are really higher income individuals with large discrepancies between economic and tax basis incomes, which concentrated towards higher income individuals. It is always possible that some returns with low adjusted gross The differences in these distributions may reflect the income measure rather than indicating that dividends are less8 .9991 rebmetpeS ,repaP ATO ,sisylanA xaT fo eciffO tnemtrapeD yrusaerT .S.U ",ygolodohteM lanoitubirtsiD yrusaerT .S.U" ,ninorC ennA-eiluJ :ecruoS 8.41 9.01 6.13 %1 poT 4.92 9.42 8.15 %5 poT 5.04 0.73 5.16 %01 poT 7.65 6.55 2.37 elitniuQ poT 3.12 6.32 1.51 elitniuQ ht4 6.21 7.21 2.9 elitniuQ dr3 2.7 3.6 2.4 elitniuQ dn2 7.2 7.1 7.0 elitniuQ mottoB emocnI latoT emocnI robaL emocnI latipaC erahS fo egatnecreP fo egatnecreP fo egatnecreP noitalupoP latoT dna robaL ,latipaC :emocnI fo noitubirtsiD .2 elbaT run or long run measures, dividend taxes are borne by higher income individuals. individuals and about 73% of capital income is received by the top 20%.8 Thus, under either short Table 2 shows that about 30% of capital income in general is received by the top 1% of .5002 ,snruteR xaT emocnI laudividnI ,emocnI fo scitsitatS ecivreS euneveR lanretnI no desab snoitaluclac SRC :ecruoS 305,851 78 662,831 2.53 2.0 +0001 595,42 18 849,91 8.8 4.0 0001-005 677,9 96 927,6 5.51 .02 005-002 946,3 74 117,1 6.51 0.8 002-001 742,2 23 917 3.6 8.7 001-57 589,1 32 264 1.7 6.31 57-05 506,1 71 572 5.2 907 05-04 334,1 31 291 2.2 4.01 04-03 615,1 11 061 2.1 5.6 03-52 811,1 01 901 9.0 3 .7 52-02 913,1 01 331 2.1 3.8 02-51 401,1 01 701 0.1 7.8 51-01 287 9 76 7.0 0.9 01-5 sdnediviD sdnediviD ssalC hcaE ssalC hcaE )sdnasuoht$( htiw snruteR htiw snruteR sdnediviD ni sdnediviD ni snruteR emocnI ssorG rof sdnediviD fo noitcarF egarevA fo tnecreP fo tnecreP detsujdA egarevA ¢ ¡ ¡ ¢ It is also sometimes argued that the "double tax" that results from including dividend income in both the corporate and individual income tax systems tends to disproportionately affect elderly taxpayers. There is some empirical evidence to suggest that the elderly tend to hold investments in less risky corporate assets and that these assets tend to pay higher dividends than other investments. However, as pointed out earlier in this report, economic analysis rejects the unfairness argument in general because behavioral responses cause the true burden of the corporate tax burden to be spread to all capital income. The empirical evidence also shows that in 2000, only 21% of those individuals aged 65 years or older actually received dividend income. Interest income, which was received by over 58% of this age group, was a much more significant source of capital income for individuals aged 65 years or older. Hence, the concern that including dividend income in the individual income tax disproportionately affects the elderly does not appear to be supported by either economic theory or the empirical evidence. ¡ ¢ Taxes on dividends (and on capital gains that reflect already taxed retained earnings) result in a heavier taxation of corporate equity capital than is the case for debt-financed corporate investment or for noncorporate investment. All equity investments in business assets are taxed more heavily than investments in owner-occupied housing, but the tax is greater in the case of corporate equity. These differentials in tax burdens create a distortion that favors noncorporate investment over corporate investment. Taking into account treatment of debt (which is deducted by the firm and taxed to the lender), corporate investment is taxed at about twice the rate of noncorporate investment. The system also favors debt finance over equity, and retained earnings over dividends. Moreover, because part of the personal tax is collected as a capital gains tax, the system contributes to a lock-in effect for assets that discourages the sale of stocks. It is on these grounds that many economists criticize a separate corporate tax. Additional taxes also could have effects on savings responses, although the empirical evidence suggests that taxes on capital income have a small effect that can be either positive or negative. (Theoretically an increase in tax burdens can decrease or increase savings because of offsetting income and substitution effects.) It is also possible that dividend relief proposals would reduce savings because of the increased incentive to pay out dividends and the possibility that such actions would lead to less savings. But it is not the savings effect per se, but the distortion in choice that is costly in an efficiency sense.9 The loss in efficiency in the economy due to the corporate tax has been estimated at various levels depending on the model used, but may be quite large compared to the revenue the tax produces. For example, the Treasury integration study reported full integration to result in an efficiency gain of between 0.13 and 0.73% of consumption depending on the model and whether revenues were replaced.10 This share amounts to $12 billion to $67 billion in 2000 (before the 9 Measuring the efficiency is complicated by issues of how revenue is to be replaced if the tax is repealed. Assuming a fixed level of revenue, the efficiency effects depend on how the distortions arising from capital compare to other distortions that may be larger under different tax systems. 10 See U.S. Department of the Treasury, Integration of the Individual and Corporate Tax Systems, Taxing Business Income Once (1992). See also Jane G. Gravelle, The Economic Effects of Taxing Capital Income, Cambridge, MIT Press, 1994. ¡ ¢ recession began), at a time when corporate tax revenues were close to $200 billion.11 However, more limited forms of relief would result in smaller gains. For example, a dividend exclusion at the individual level (which would eliminate the individual tax) would produce about $10 billion to $48 billion efficiency gain; a dividend credit designed to eliminate the corporate level tax and retain the individual level would result in slightly larger efficiency gains falling between full integration and dividend exclusion. A dividend deduction at the firm level would probably produce gains of similar magnitude. Only about a third of dividends actually show up on individual income tax returns as taxable dividends. (The amount potentially subject to tax would be somewhat higher if dividends paid to trusts are included.) The growth of pension and IRA funds over the past few years and the increasing investment of those funds in the stock market may have diminished the degree of double taxation (since these assets are subject to zero tax rates). These non taxable forms of investment may be subject to different distortionary effects: non-corporate equity investment may not be a feasible alternative for these institutional investors, there is no incentive for these investors to prefer retentions to dividends, and no lock-in effect. The total tax burden is smaller, as well, which reduces the savings distortion. At the same time, the distortions between equity and debt may actually be larger, since they are not moderated by the more favorable treatment of capital gains income (relative to interest as well as dividends) under the individual income tax. (These nontaxable forms were not treated as affecting the margin in the above estimates.) There is some agreement that many forms of corporate tax integration or dividend relief would complicate tax administration. Other forms would simplify it. Dividend exclusions would be relatively easy and would simplify tax filing, but certain forms of dividend relief that are designed to eliminate the corporate level tax (rather than the individual level tax) could add to tax administration problems. For example, shareholder credits would require a gross-up and credit of the dividend taxes paid at the corporate level, a more complicated revision. An administrative argument for integrating the corporate income tax is the fact that many types of firms can now take advantage of many of the benefits of incorporation without paying the tax. These types of firms include Subchapter S corporations, LLC's, and limited partnerships.12 These forms of business organization complicate tax collections, and their expansion may be of some administrative concern. Nevertheless, they remain a very small part of business activity and other measures could be taken to limit their growth if needed. An important potential complication of certain types of revisions is how to treat preferences. For example, firms receive a variety of tax benefits, which reduce their effective tax rate below their statutory rate. Should firms get a deduction for the dividend at the statutory tax rate, while having profits taxed at a lower effective tax rate? Many people think they should not, but dealing with preferences (by allocating them between taxed retained earnings and nontaxable dividend 11 The net collection from the additional corporate tax would be somewhat smaller because additional income from the lower tax is subject to individual tax. 12 See CRS Report RL31538, Passthrough Organizations Not Taxed as Corporations, by Jack H. Taylor, for a description of the various forms of pass-through organizations. ¡ ¢ deductions) is a potential complication of dividend relief. President Bush's proposal would provide for such an allocation for preferences. Proposals that require individual tax relief for retained earnings by adding to basis, a feature of the President's proposal, would complicate tax compliance because annual basis adjustments would have to be tracked. There are several methods of integrating the corporate and individual income tax. They can be divided into full integration, which covers both dividends and retained earnings, and dividend relief alone (partial integration). There are three basic approaches to full integration: taxation on a partnership basis, a credit system that uses the corporation as a withholding device, and elimination of taxes at the individual level. A variation of this latter proposal would increase basis by the amount of retained earnings which would in the very long run tend to eliminate (on average) the tax on capital gains. A credit system can be designed to allow or not allow refundable credits of corporate taxes to tax exempt shareholders. There are also three corresponding dividend relief approaches: deductions by the firm, a credit system using the corporation as a withholding device, and exclusion of dividends at the individual level. These approaches can vary dramatically in their revenue costs, and thus we report revenue losses for several of the alternatives. Precise revenue effects are beyond the scope of this analysis and only general magnitudes will be discussed. The purest approach to full integration would be taxation on a partnership basis, so that each shareholder pays a tax on his or her pro-rata share of corporate income. This approach has the effect of eliminating the corporate level tax and taxing individuals on their shares of corporate income. These methods become complicated when stocks change hands many times over a year. Shareholder allocation would lose the entire corporate tax of about $200 billion (at 2000 income levels). However, individual income would increase by the amount of net corporate income and this increase would produce additional individual tax revenues which would offset the cost of lost corporate receipts. Moreover, retained earnings will be taxed currently in the hands of shareholders at ordinary rates (with the basis of stocks increased accordingly). However, according to data in the National Income and Products Accounts, only about 34% of dividends paid by corporations subject to the U.S. tax appear in taxable dividends of individuals because of the large shares held in pensions, individual retirement accounts and by foreigners.13 13 In 1999, according to National Income and Products Accounts data, corporations paid $379 billion in dividends ($503.8 billion less $124.5 in intercorporate dividends) while the IRS reported $129 billion of dividends in adjusted gross income. According to discussions with the Bureau of Economic Analysis (BEA), about $157 billion of dividends were associated with Subchapter S firms not subject to the corporate tax, leaving a total of $222 billion. However, the IRS requires distributions from mutual funds to be reported as dividends regardless of whether their underlying assets pay interest or dividends. About $54 billion of reported dividends are estimated to be interest, leaving a net total of $75 (continued...) ¡ ¢ With an individual tax rate of 30% and 34% of income subject to tax, the effective rate is 10% (0.30 X 0.34) and the cost would be reduced by $20 billion (0.10 X $200 billion). There is also a gain from the taxing of previous retained earnings (approximately $95 billion) at the higher ordinary rates which is about $6 billion).14 Thus the net cost of this form of integration would be about $174 billion. The credit approach to integration has the corporation serve as a withholding agent so that individual shareholders would receive a credit for taxes already paid by the corporation. An important issue in this approach is whether the credit would be refundable to tax-exempt investors (including foreign investors and tax exempt holders, such as pension funds). If the credit is refundable, the cost would be the same as the shareholder allocation discussed above. However, if the credit is restricted to taxable shareholders, the cost would be much smaller. Only $68 billion of corporate taxes (34% of the $200 billion) would be associated with taxable shareholders and would be lost. However, the full offsets above would still occur ($20 billion of additional taxes on the increase in individual income (0.3 X $68 billion) and the additional offset of around $6 billion (retained earnings taxed at a higher ordinary rates), for a net total of $42 billion.15 A third approach is to eliminate all individual level taxes, which would include taxes on dividends and an adjustment for capital gains taxes that are collected on corporate stock. This approach would retain a tax at the corporate level but not at the individual level. This cost would be highly sensitive to the effects of current stock market values, but would be less expensive if already accrued gains are not included or if the capital gains adjustment were to adjust the basis of stock for retained earnings shares. In the latter case, which is the method used in the President's proposal, the cost of excluding dividends would be about $23 billion (0.3 X $75) and the cost of step-up would be about $3 billion (0.1X $94.5 X0.34). This method prevents a preference for dividends over retained earnings, but would probably be quite complicated because it would require taxpayers to keep track of a series of basis adjustments for each type of stock. In its 1992 corporate tax integration report, the Treasury also discussed imposing a tax only at the business level, called the comprehensive business income tax (CBIT) that would apply to interest (...continued) billion. Thus, out of total dividends paid by U.S. corporations only 34% showed up on U.S. individual tax returns. (See Table 8.19, National Income and Product Accounts; the total includes dividends in personal income and dividends paid to foreign holders, but not intercorporate dividends. Also see Thae S. Park, "Comparison of BEA Estimates of Personal Income and IRS Estimates of Adjusted Gross Income," Survey of Current Business, November 2000.) Note that dividends do increase eventual pension benefits, but the tax treatment of pensions (deduction of investment and earnings and tax on benefits) is equivalent to eliminating the tax on earnings, in this case, dividends. 14 According to data in table 1.19 of the National Income and Product Accounts, domestic corporate business in 1999 retained $94.5 billion in earnings. If the capital gains rate is approximately halved because of deferral and step-up in basis at death, it is about 10%, and the difference between those rates, (20%) multiplied by 34% (to reflect taxable share) and by $95 billion provides an offset of about $6 billion. 15 The purpose of the credit imputation system is to accomplish the same effect as the shareholder treatment: to eliminate the corporate level tax and retain the individual tax. However, a tax continues to be collected at the corporate level and shareholders get a credit. The mechanism is to impute a grossed up corporate income to the individual and include that amount in income on which an individual tax would be paid while also taking a credit for the corporate tax. For example, suppose earnings are $100, the corporate rate is 35%, and the individual rate is 30%. The after corporate tax earnings are $65 and it is assumed they are all paid out. However, one does not begin with the $65; rather this $65 is grossed up by dividing it by one minus the tax rate or (1-0.35), to yield the original $100 (i.e. $65/(1-.35) equals $100). The individual gets a credit for the corporate tax of 0.35 times grossed up earnings (or $35), and pays an individual tax of 0.30 on grossed up earnings (or $30). The excess of the credit over the tax is $5 ($35 minus $30) which, added to the dividend of $65 provides total earnings of $70. ¡ ¢ and profits. This proposal would allow no deduction for interest at the firm level (for either corporations or unincorporated businesses). Individuals would pay no tax on interest or dividends and capital gains on corporate stock would either be excluded or the basis adjusted to reflect retained earnings. This approach solves some of the revenue problems associated with corporate tax integration, but concerns some economists about the possibility of discouraging investment from abroad because of the lower after tax rates that would be expected to result. This system would also be applied to non-corporate businesses (perhaps with an exemption for small business). Note that the costs estimated in this section should be adjusted to reflect current income levels, and thus would be higher than those provided. Many difficulties are associated with full integration, in part involving the need to track and adjust the basis for capital gains arising from retained earnings. When other countries have integrated their taxes, they have generally done it only with respect to dividends, a form of partial integration. Dividend relief, however, can create incentives to distribute profits and does not have efficiency gains as large as its corresponding full integration method. We calculate revenue costs in this case for 1999, the latest year with data on taxable dividends. Three methods of dividend relief correspond to the three full integration methods. Rather than shareholder allocation of all corporate income, a deduction for dividends can be allowed at the firm level. A credit imputation system confined to dividends can be allowed instead of a system reflecting all corporate earnings. And relief at the individual level can be provided through a dividend exclusion. An apparently simple approach (but one that actually has significant problems with implementation) is a deduction for dividends. In 1999, corporate dividends (excluding Subchapter S earnings) were $222 billion. However, the corporation is likely to pass on as additional dividends to shareholders some or all (even more than 100%) of the tax savings from the dividend deduction. Making the presumption that the tax is distributed as a dividend, and using a 35% tax rate, we could estimate that dividends paid represented $342 billion before tax ($222/(1-0.35)). The direct cost of the dividend deduction would be $120 billion (0.35 X $342). As with the case of the partnership treatment, there will be an approximate 10% offset (reflecting the 30% tax rate for the 34% of dividends received by taxable shareholders), for a net cost of $108 billion. This number is likely to be overstated because some firms would not have enough tax liability to absorb the full tax deduction. Some firms pay dividends even though they have no profits, and some firms pay dividends that exceed their taxable income because of tax preferences. Moreover, it would be smaller still if preferences were partially allocated to dividends. That is, an important issue for dividend deduction is whether dividends should be deducted at the full statutory tax rate, even though the firm is taking advantage of tax preferences, or whether preferences should be allocated. The second method of dividend relief, a dividend credit imputation system, is the form most commonly used by other countries. The firm pays a tax and the dividend recipient receives a credit for taxes the firm has paid, based on pre-tax dividends. As in the case of the full integration with a withholding/credit device, there is an issue of how to treat tax-exempt shareholders. If the credit is refundable the costs will be the same as a dividend deduction. If allowed only for taxable ¡ ¢ returns, it would affect only 34% of dividends (34% of $120 billion, or $41 billion) and then have a 30% offset, for a total of about $29 billion. This number, again, could be overstated because of lack of tax liability for some firms and any allocation of preferences. The third and simplest approach is a dividend exclusion, which would eliminate the individual layer of the tax and cost about $23 billion (0.3 X $75 billion of taxable dividends (1999 levels)). The 1992 Treasury study found the exclusion and the credit imputation to be about of similar size (although the individual top rate was a little lower in 1991), so their results are consistent with such an assumption. By any standard, however, these are very large revenue costs. Moreover, they would be larger at current income levels, although it is difficult to project these costs. Personal dividend income, however, grew by 24% from 1999 to 2001 and 9% from 2000 to 2001. Finally, there are proposals to provide a dividend exclusion that is capped at a certain level, such as the $400 exclusion that was provided historically. While this provision would be much less costly, it would provide little or no behavioral response and thus do little to increase investment in corporate equity. The capped exclusion therefore would have little effect on efficiency or the stock market, the main reasons for providing benefits, and would essentially be a windfall benefit for holders of dividends. We can place an upper bound on an estimate of $400 per return by assuming every dividend recipient had $400 of dividends. This estimate produces about $4 billion at the 30% tax rate. This estimate would be lower if singles had half the cap, higher if married couples had twice the cap, and would generally be too high because some individuals would have smaller amounts of dividends, especially given that some dividends are characterized as interest on tax returns. The capped deduction, however, is clearly a different order of magnitude of costs. Another way to reduce the revenue cost is to allow an uncapped, but partial deduction, such as an exclusion for a fraction of dividends received. As noted earlier, these estimates are not very precise. However, they clearly illustrate the very large annual cost of almost any type of full scale corporate tax integration or dividend relief. As one can see by these estimates, limiting the benefits to dividends, and focusing on taxable shareholders reduces the cost of these provisions. However, even in the case with the smallest cost, a dividend exclusion, the cost is still $23 billion at 1999 income levels (which could be considerably larger today). Limiting the benefits to taxable shareholders may accomplish significant efficiency gains for each dollar of revenue loss; the power of such a provision in reducing distortions depends on the extent to which pensions and other plans act as marginal investors. As indicated in the introduction, the traditional arguments for relieving the double taxation imposed by the corporate tax are largely related to economic efficiency, while at least one of the important problems is the potentially large revenue cost. There is also an issue of who bears the burden, with most analysis suggesting that the corporate tax contributes to the progressivity of the overall tax system, an issue of importance for some. Some approaches will reduce administrative and compliance costs, while others will increase these costs. ¡ ¢ These offsetting costs and benefits make the assessment of general corporate tax integration difficult. There are ways of achieving the efficiency gains that were considered by the Treasury Department study while still raising the same amount of revenue and not shifting the tax burden from high to low income individuals. But they would call for a fairly radical change in the current tax structure or an increase in top tax rates, changes that might be very difficult. A partial dividend deduction could be used to scale back the cost but would have more limited efficiency effects. It appears that there are other, more effective, ways of stimulating the economy than a dividend deduction. The analysis suggests that a capped dividend deduction is not likely to achieve goals of fairness, equity and efficiency. As discussed in the section on distribution, the double tax does not lead to an "unfair" burden on corporate stockholders. Any tax on normal return is shifted to capital in general so that, for purposes of measuring tax burdens, the tax can be considered as a general tax on capital income. A capped dividend exclusion would be a windfall for the individuals who receive it and have little or no consequences for marginal investment. Thus it would not have much effect on allocation, efficiency, or the stock market. The capped dividend exclusion might be argued to simplify the tax law, but since recipients of dividends tend to be higher income and have interest bearing assets as well, little simplification will result. More simplification would be achieved by allowing an exclusion that covered interest as well as dividends or limited it to interest income alone. Jane G. Gravelle Senior Specialist in Economic Policy jgravelle@crs.loc.gov, 7-7829 ------------------------------------------------------------------------------ For other versions of this document, see http://wikileaks.org/wiki/CRS-RL31597