CDA Working Paper: The Economic and Fiscal Effects of Ending the Federal Double Taxation of Dividends

Report Taxes

CDA Working Paper: The Economic and Fiscal Effects of Ending the Federal Double Taxation of Dividends

February 3, 2003 29 min read Download Report

Contributors: Norbert Michel, Ralph Rector and Alfredo Goyburu

On January 7, 2003, President George W. Bush unveiled a multi-faceted proposal to improve the nation's economic growth. One of the most important features of his plan calls for abolition of the current federal double taxa­tion of corporate dividends paid to individual shareholders. Economic analysts at the Center for Data Analysis (CDA) at The Heritage Foun­dation found, in a study of a dividend reform proposal similar to President Bush's, that end­ing the double taxation of dividends would improve the nation's economic growth, employment level, and other economic indica­tors over the next 10 years.

For example, CDA estimates indicate that the employment level would average 285,000 additional jobs from 2003 to 2012. In addi­tion, CDA analysis has found that ending this double taxation would reduce federal revenue by $64 billion over ten years, or 79 percent less than an estimate that does not account for the effects of greater economic activity follow­ing the proposal's implementation. The CDA's $64 billion estimate is slightly more than one-fifth of the $364 billion cost estimated by the United States Department of the Treasury for President Bush's proposal.[2] The CDA and Treasury analyses consider slightly different proposals, but this cost difference is largely due to the more realistic estimation method used by the CDA.

The Treasury Department employs an erro­neous "static" approach to estimate the reve­nue effect of tax law changes, while the CDA uses dynamic simulation, a method that accounts for the impact that federal tax policy may exert on economic growth.[3] Figure 1 shows that the estimation method chosen can make a large difference in the projected reve­nue loss. The figure compares the CDA's own static and dynamic projections of the federal revenue change resulting from a particular plan to end the double taxation of dividends.

This double taxation[4] has two stages. The first stage occurs when the federal government taxes shareholders on corporate income through corporate taxes. The second occurs after the corporation has distributed part of the post-tax profits to the shareholders in the form of dividends. In this second stage, the federal government taxes shareholders on their dividend income through the personal income tax.

Economists have long argued that the double taxation of dividends reduces the after-tax return on capital in the nation's economy and thus discourages investment-in other words, purchases of new business equip­ment and machin­ery.[5] This reduced investment in turn weakens economic growth. Conse­quently, eliminating the double taxation would spur invest­ment and improve the economy's long-term growth. Recognizing these economic benefits, several nations, including Australia, France, Italy, Canada, Germany, Japan, and the United Kingdom, have abolished or reduced their double taxation of corporate divi­dends.[6]

One recent legislative proposal to abolish this double taxation in the United States was spon­sored by Representative Christopher Cox (R- CA).[7] The Heritage Foundation's CDA used this proposal to illustrate the economic and federal fis­cal effects of ending the double taxation of divi­dends.[8] To estimate these effects, Heritage analysts employed the DRI-WEFA U.S. Macroeconomic Model and the Center's own Individual Income Tax Model. Assuming the reform becomes law in September 2003, the investigation found that:[9]

  • GDP Increases
    During the period from 2003 through 2012, the Cox proposal would increase the nation's gross domestic prod­uct (GDP) by an inflation-adjuste[10] $32 billion per year on average, com­pared to what it would otherwise have been. GDP would be at least $22 billion higher in 2004 and no less than $45 bil­lion higher in 2012 if the pro­posal were to be implemented. (See Figure 2.)
  • Employment grows
    The pro­visions in the Cox bill would enable the econ­omy to support 325,000 more jobs by 2012. (See Figure 3.) With these addi­tional jobs in the economy, the unemployment rate would be 0.2 percent lower throughout the period2005- 2012 than cur­rent projections indicate.
  • Investment strengthens
    Over the 10-year period from 2003 through 2012, the proposal would result in an aggregate increase of at least $253 billion (adjusted for inflation) in non-residential investment. Because of this higher level of investment, the nation's non-residential capital stock would be $175 billion higher in 2012. (See Figure 4.)
  • Disposable income picks up
    Under the Cox legislation, disposable personal income would average an inflation-adjusted $56 billion higher from 2003 through 2012. (See Figure 5.) This higher level would raise annual dis­posable personal income by $192 per person on average during the period. For a family of four, this increase would correspond to $768 more in disposable income on average each year.
  • Personal savings increases
    The proposal would increase personal savings by an infla­tion-adjusted average of $18 billion per year from 2003 through 2012.
  • Higher economic growth reduces the "cost" to the Treasury by over 70 percent
    The CDA's own static estimates suggest the pro­posal would reduce federal revenue by about $300 billion from 2003 through 2012.How­ever, the CDA's more realistic dynamic esti­mates show that the proposal would reduce federal revenue during the period by a total of $64 billion.(See Figure 1.) During the last five years, the proposal would be nearly revenue neutral, since the improved economic growth caused by the legislation would, in turn, increase tax collections. (See Table 3). For rea­sons discussed below, these estimates do not take into account the way in which the pro­posal's effect on capital gains tax collections would change federal tax revenue.

How the Double Taxation of Dividends works

The double taxation of dividends[11] is one of the clearest examples of the way the nation's current tax law reduces the return on capital and, there­fore, the incentive to invest. The following exam­ple illustrates the effect of this double taxation.

Consider $100 in pre-tax profit earned by a cor­poration in the flat 35 percent bracket. Suppose that, after paying the $35 in federal corporate taxes, the firm distributed the remaining $65 to a shareholder. Suppose, further, that this individual was in the 27 percent personal income tax bracket. This shareholder would pay $17.55 in personal income taxes on these dividends. This second round of taxation would leave only $47.45 of the original $100 in corporate profits. In other words, for every $100 in pre-tax profits, the federal gov­ernment would absorb approximately $52.55 in taxes.

In contrast, consider the taxes the shareholder might have paid if that person could have received the dividend before the firm paid corporate taxes. In this case, the corporation would have paid the shareholder all $100 in the form of a dividend. The shareholder would then have paid $27 in per­sonal income taxes on the dividends, leaving that investor with $73 out of the $100 in pre-tax cor­porate profit. As this example shows, the double taxation of corporate dividends reduced the share­holder's return on capital from $73 to $47.45-a reduction of 35 percent (or $25.55). In the aggre­gate, this lower return on capital means that there is less investment than there would otherwise have been.


Heritage economists use dynamic simulation to project the economic and fiscal effects of proposals for tax changes. This method contrasts with the static approach used by the U.S. Department of the Treasury and the Congressional Joint Commit­tee on Taxation (JCT) , which assumes that federal tax policy does not affect economic growth.

In determining the fiscal effects of tax change proposals, the static approach does take into account some of the ways taxpayers alter their tax reporting and filing in response to changes in tax law. For example, the static approach takes into account that taxpayers could increase their item­ized deductions or shift compensation from tax­able to tax-exempt (or tax-deferred) forms in response to certain changes in the tax laws. How­ever, the static approach does not take into account the way investors and workers alter their consumption, investment, saving, and work effort in response to changes in tax policy. This is a major shortcoming of the static approach because economic theory suggests that tax policy changes bring about such alterations.[12]

Such changes in taxpayers' behavior could affect important macroeconomic variables, including employment, personal income, and GDP. Thus, changes in tax law often exert an impact on the nation's economy. The static approach necessarily ignores these impacts, leading to systematic inac­curacies in the estimates of the fiscal effects of tax policy changes.

In contrast, The Heritage Foundation uses dynamic simulation in evaluating the fiscal and economic effects of tax policy proposals. Dynamic simulation takes into account the impact that tax policy legislation can exert on taxpayers' economic decisions, such as consumption, investment, sav­ing, and work effort. Dynamic simulation, there­fore, can reflect changes in macroeconomic variables that new tax policies can cause.

For example, if a tax rate reduction were to strengthen national economic growth and there­fore increase the tax base, a resultant increase in tax collections could partially offset the federal revenue losses caused by the rate reduction. Static analysis would not take such an offset into account and therefore would overestimate the net decline in federal tax collections resulting from the tax rate reduction. Dynamic analysis would include this offset because it would take full account of the economic benefits that the tax rate reduction could cause. It would also capture the ways in which these benefits could strengthen the economy, bol­ster the tax base, and ameliorate the reduction in tax collections.

In analyzing the economic and fiscal impact of the Cox proposal, CDA analysts made a number of assumptions regarding the alternative minimum tax, capital gains taxation, federal spending, and the date the bill would be enacted. These assump­tions were as follows.

  • Alternative Minimum Tax
    The form of the bill submitted for consideration in the 107th Congress does not clearly state how the divi­dend tax credit should be handled under those parts of the tax code that establish the alterna­tive minimum tax (AMT). Heritage Foundation analysts assumed that taxpayers required to file under the AMT rules would be able to take advantage of the dividend tax credit. If this were not the case, the dividend tax relief for those taxpayers would be negated.
  • Capital Gains Tax
    The Cox proposal would be expected to cause an increase in equity prices. This increase would likely cause inves­tors to adjust their portfolios, perhaps trigger­ing increased capital gains tax liability. Estimating the total increase in capital gains tax collections would require both distribu­tional and basis data that are not readily avail­able to Heritage economists. Therefore, CDA analysts assumed that such collections would remain unchanged relative to the baseline
  • Federal Spending
    Heritage Foundation ana­lysts assumed that Congress would make no government program spending reductions to offset federal revenue cuts expected with the Cox proposal. As a result, any changes in fed­eral spending observed in the simulation are attributable solely to the Cox proposal's effect on the national economy and, in turn, the economy's effect on federal spending.
  • Dividend Increase
    Heritage analysts assumed that ending the double taxation of dividends would increase dividend payouts by 10 per­cent. A portion of this increase would be caused by higher shareholder demand for divi­dends. In response to this higher demand, cor­porations would increase their payouts of dividends out of after-tax profits. The remain­der of this 10 percent increase would be explained by a reduction in the user cost of capital and a corresponding increase in profits. Some of these higher profits would then be returned to shareholders as higher dividends. The combined result of these two effects was assumed to be a 10 percent increase in divi­dends.[13]
  • Date of Enactment
    Heritage economists assumed that the tax reform would become law on September 30, 2003, and apply retroac­tively to dividends received after January 1, 2003. Assuming an earlier date of enactment would have resulted in the proposal's benefits being realized sooner.


Heritage economists used a modified version of the DRI-WEFA U.S. Macroeconomic Model to conduct a dynamic simulation of the effects of Representative Cox's bill.[14] Specifically, Heritage economists developed a baseline by adapting the DRI-WEFA macroeconomic forecast from Sep­tember 2002 to yield the same economic and bud­get projections as those of the Congressional Budget Office (CBO) in August 2002.[15] Thus, the economic baseline employed in this analysis should be comparable to baselines used by the CBO and JCT in analyzing this legislation. The results of the dynamic simulation are displayed in Table 2.

Specifically, the dynamic analysis projects that the Cox proposal would:

  • Increase economic growth
    GDP would increase by an average of at least $32 billion per year (adjusted for inflation) within the period from 2003 through 2012. GDP would be an inflation-adjusted $22 billion higher in 2004 and $45 billion higher in 2012.
    (See Figure 2.)
  • Create more job opportunities
    The proposal would increase the number of jobs by at least 325,000 in 2012. (See Figure 3.) This increase in jobs would correspond to a decline in the unemployment rate of no less than 0.2 percent per year over the next 10 years. (See Figure 3.)
  • Increase investment
    Non-residential invest­ment would average nearly $25 billion per year (adjusted for inflation) higher between 2003 and 2012. By the end of fiscal year 2012, the net capital stock would be at least an infla­tion-adjusted $174 billion higher. (See Figure 4.) The user cost of capital would be about 5.4 percent lower in 2012.
  • Increase disposable personal income
    Dis­posable personal income would increase by an inflation-adjusted average of $56 billion or more per year from 2003 through 2012. For a family of four, this increase in disposable income would correspond to an average of at least $768 per year. (See Figure 5.)
  • Increase personal savings and personal con­sumption
    Personal savings would average an inflation-adjusted $18 billion higher during the 10-year period. Personal consumption expenditures would average an inflation-adjusted $36 billion higher than current pro­jections.
  • Slightly increase consumer prices
    Under the Cox proposal, growth in the consumer price index would average 0.1 percent higher from 2004 through 2008. Over the final four years of the forecast period, increases in the price level would be virtually unchanged in comparison with those of the baseline.
  • Decrease federal tax revenue.
    The Cox divi­dend proposal would reduce total federal tax revenues by a total of $64 billion during its first 10 years. Close to $56 billion of this reduction would take place during the first five years, for an average of $11 billion per year. During the final five years of the simulation period, the tax cut would be virtually revenue neutral, reducing federal revenue by an aver­age of less then $2 billion per year. During this latter five-year period, increases mostly in cor­porate and Social Security tax collections would offset expected declines in personal income taxes. Corporate tax collections would rise because of higher pre-tax corporate prof­its. Payroll taxes would increase because of higher employment levels.[16] (See Table 3.)
  • Increase federal spending
    If Congress were not to reduce federal program spending to off­set the tax revenue reductions caused by this proposal, overall federal spending would rise. Spending would average about $13 billion higher after ending the double taxation of divi­dends. About two-thirds of this increase would result from additional federal interest pay­ments. The rest would be caused by increases in federal expenditures on income-mainte­nance programs for federal and Social Security retirees. These increases in federal income maintenance spending would be caused mainly by higher consumer prices observed during the years from 2004 through 2008. (See Table 4.)


President Bush has proposed reforming the U.S. tax code to abolish the federal double taxation on corporate dividends. Economists have long argued that this double taxation exerts a harmful effect on the nation's economy because it increases the user cost of capital and therefore reduces investment in the United States. Last fall, Representative Christo­pher Cox introduced legislation that would end this double taxation.

This Heritage Foundation working paper inves­tigates the 10-year economic and fiscal impact of Representative Cox's proposal to abolish this dou­ble taxation. It finds that the proposal would, by the year 2012, improve growth in the nation's GDP, add hundreds of thousands of jobs to the economy, increase investment, strengthen growth in disposable income, and add to the nation's cap­ital stock.

Norbert J. Michel and Alfredo Goyburu are Pol­icy Analysts, and Ralph A. Rector, Ph.D., is a Research Fellow, in the Center for Data Analysis at The Heri­tage Foundation.

Heritage Foundation economists in the Center for Data Analysis (CDA) used a multi-step proce­dure to analyze the budgetary and economic effects of the tax law change proposed by Repre­sentative Cox.

First, CDA economists adapted the September 2002 forecast from the DRI-WEFA U.S. Macro­economic Model to make it congruent with the long-term budget and economic projections pub­lished by the Congressional Budget Office in August 2002.[17] CDA analysts then used this fore­cast as the baseline by which to analyze the effects of the Cox proposal.

CDA economists then used the Center's Individ­ual Income Tax Model to generate a static estimate of the change in federal tax collections resulting from the Cox proposal.[18] This static estimate serves as an essential starting point in analyzing the fiscal impact of proposed changes in tax pol­icy. However, as explained above, to use this esti­mate as an ultimate forecast of the federal revenue lost under the Cox proposal would be to imple­ment an erroneous static approach. The more accurate, dynamic approach would take into account the proposal's macroeconomic effects. These effects include changes in GDP, interest rates, employment levels, price levels, investment, and other quantities. Changes in any of these mac­roeconomic variables could affect tax revenues sig­nificantly.

Next, the Center's analysts introduced these tax collection changes and other implications of the Cox proposal into the adapted DRI-WEFA U.S. Macroeconomic Model. CDA researchers then exe­cuted the simulation and developed a dynamic estimate of the fiscal and macroeconomic effects of the Cox proposal. The researchers noted changes in key macroeconomic and budget variables com­pared with their values in the original adapted ver­sion of the model. Differences in these key variables were attributed to the response of the U.S. economy and federal budget to the tax policy change-that is, the dynamic response. (See Table 2.)

The Simulation[19]

The DRI-WEFA model contains a number of variables that can be altered to simulate policy changes. Using these variables, CDA analysts introduced static-model tax revenue and economic behavior responses attributable to the enactment of Representative Cox's proposal to end the double taxation on corporate dividends. The variables altered include:

      Federal Marginal Tax Rate on Corporate Income
The Cox dividend proposal would significantly reduce the effective federal mar­ginal tax rate on corporate income. CDA ana­lysts altered a variable that controls this quantity in order to reflect this reduction.

      Federal Average Tax Rate on Corporate Income
CDA researchers manipulated a vari­able that controls the federal average corporate tax rate. This variable was changed so that the average rate would remain unchanged com­pared to the baseline value, in spite of the alteration of the federal marginal corporate tax rate.

      Federal Average Tax Rate on Personal Income
The Cox dividend proposal would abolish the double taxation of corporate divi­dend income by returning a credit that could be claimed against personal income tax liabil­ity. CDA analysts altered this variable to cap­ture the static reduction in federal personal income tax collections resulting from the enactment of Representative Cox's legislation.

      Personal Dividend Income
The Cox divi­dend proposal is expected to boost corporate payments of dividends. This increase would have both short-run and long-run compo­nents. In the short run, existing C-Corpora­tions would increase their dividend payouts as a share of after-tax profits. They would do so in response to shareholder demand. In the long run, the Cox dividend proposal would reduce the user cost of capital. The lower user cost of capital would boost corporate profits, leading to an increase in payouts of corporate divi­dends. CDA analysts recognized this increase through an appropriate change in a model variable that controls personal dividend income.

      Corporate Profits
The Cox dividend pro­posal is expected to increase personal dividend income compared to its level in the baseline forecast. As indicated in the simulation, part of this increased dividend income comes from an increase in firm profitability, as described above. The rest of the dividend increase would represent a shift from corporate retained earn­ings to increased payouts of dividends. CDA economists adjusted a variable that controls after-tax corporate profits to reflect this shift.

The bill sponsored by Representative Cox would eliminate the double taxation of dividends paid by C-Corporations through an imputation credit method similar to that used in several other countries.[20] This method adds an amount equal to the corporate layer of the tax on the distributed dividend to the individual shareholders gross income and then gives the shareholder a tax credit equal to that amount.

The Cox approach has the effect of removing the corporate layer of taxation from dividends by returning it to shareholders at the personal level, via a credit. Although corporations continue to pay income taxes on the dividends they distribute, individuals receiving dividends end up with a lower tax liability to offset the corporate income tax.

This proposal would not change any aspect of taxation at the corporate level. In addition, the shareholders legal obligation to report dividends received as ordinary personal income would remain unchanged. However, in addition to the dividends, shareholders would have to add to their taxable income the amount that each corporation paid in taxes on the profits from which each divi­dend payout came.[21]

By adding the corporate tax payments on each dividend payout to their ordinary personal income, shareholders would be said to be gross­ing up their dividend income. The corporate tax payments on the dividendsthat is, the amount by which the dividend payments would be grossed upwould also become a non-refund­able credit that shareholders could claim against tax liability.

Thus, the gross-up amount would both add to and subtract from each shareholders tax liability. However, the net effect would never be a tax liabil­ity increase. The gross-up would increase the shareholders liability by adding to taxable income. On the other hand, the gross-up would reduce tax liability by serving as a non-refundable credit. The effect of the former can never add more in tax lia­bility than the latter reduces. This is because the gross-up increases the shareholders liability only by the amount of the gross-up multiplied by the shareholders top marginal tax rate, while the shareholders tax liability is reduced by up to the full amount of the gross-up.

Table 1 provides an example illustrating how the Cox proposal works for a hypothetical dual-earning married childless couple in the 27 percent tax bracket[22] during 2003.[23] The couple is assumed to own stock in a company subject to the average corporate tax rate of 35 percent. The cor­porations tax situation is illustrated in the section of the table labeled Corporate Taxpayer (lines 1 to 3). This section shows that corporate tax liabil­ity on pre-tax dividends does not change with the proposal. In both cases, the $100 in pre-tax profits is taxed at the corporate rate of 35 percent, leaving $65 that could be paid to individuals in the form of dividends.

The example illustrated in the table sets aside the effect of state and local corporate taxes and further assumes that all of the $65 is paid to the couple in the form of a dividend. Under both cur­rent law and the Cox proposal, the couple adds the $65 dividend to its other taxable income (line 8). The couples other taxable income, in turn, is calculated the same way under both current law and the proposal (lines 47). The couple starts with $62,000 in wage and salary income and no other type of income (line 4). It then takes its stan­dard deduction of $7,950 (line 5) as well as its personal exemptions totaling $6,100 (line 6). These deductions leave $47,950 in taxable other income (line 7).

As described above, the dividend payout the couple receives is added to their other taxable income under both current law and the proposal (line 8). However, under the Cox proposal, the dividend gross-up is also added to the couples taxable income (line 9).[24] Under the proposal, the couple applies the same rate structure to their income as under current rules. Under current law, the couple ends up with a total tax liability of $6,652.50 and an after-tax income of $55,412.50 (lines 15 and 16). Under the Cox proposal, because of the dividend gross-up, the couples tax­able income (line 10) totals $48,050, not $48,015 as under current law. This higher taxable income incurs a pre-credit tax liability of $6,661.95 (line 11). At this point, the filing couple applies the credit (line 12) and is left with a total tax liability of $6,626.95 (line 15)a $25.55 reduction in tax liability.

The Dividend Detail section of Table 1 shows how the Cox proposal reduces the taxes the couple pays on the dividends it received. For example, under current law, the taxpayers individual por­tion of the tax on the dividend is $17.55.[25]

Under the Cox proposal, however, the tax­payers individual portion of the tax on the divi­dend is negative $8 (line 8).[26]Since the personal tax on other income remains unchanged, the tax­payers total tax liability falls by $25.55from $6,652.50 to $6,626.95. Therefore, the Cox pro­posal lowers the effective personal tax on the divi­dend by $25.55 for the couple in this example (line 20).

Under current law, the $100 in pre-tax divi­dend income is reduced by $35 at the corporate level, leaving $65 for the individual, which is fur­ther reduced by $17.55 at the personal level (lines 1719). Adding the $35 tax and the $17.55 tax results in an effective personal tax of $52.55. (Adding lines 18 and 19 results in the total on line 20.) When the $52.55 is subtracted from the orig­inal $100, the individual investor receives an effec­tive dividend of only $47.45. (Subtracting line 20 from line 17 gives the total on line 21.)

Under the Cox proposal, the effective personal tax on the dividend is only $27 ($65 dividend + $35 credit = $100 x 27% = $27). This means that the effective personal dividend is $25.55 higher, for a total of $73 ($47.45 + 25.55). This new effec­tive dividend is exactly what the individual would have received had the original $100 been taxed only at the personal level ($100 x (1-.27) = $73).

While the corporation pays the same tax on the dividend income that it pays under current law, the Cox proposal has the effect of distributing a dividend that was untaxed at the corporate level. The end result is that one layer of taxation on divi­dends is removed, resulting in a higher after-tax rate of return on investment.[27]

[3] <_spanarial27_22_>Forthcoming sections of this paper further discuss the differences between static and dynamic analysis.
[4]Deborah Thomas and Keith Sellers, "Eliminate the Double Tax on Dividends," Journal of Accountancy, November 1994, and Ervin L. Black, Joseph Legoria, and Keith F. Sellers, "Capital Investment Effects of Dividend Imputation," The Jour­nal of the American Taxation Association, Vol. 22, Issue 2 (2000), pp. 40-59.
[7]CDA analysts assumed that the reform would be enacted on September 30, 2003, and applicable retroactively to divi­dends paid after January 1, 2003<_spanarial27_22_>.
[10]All dollar values listed as "inflation-adjusted" are indexed to the general 1996 price level.
[11]Survey of Current Business, November 2002, Table 2, at 1102irs&agi.pdf. 
[15]Congressional Budget Office, "The Budget and Economic Outlook: An Update," August 2002, at showdoc.cfm?index=3735&sequence=0.
[16]To maintain comparability with published CBO long-term projections, projections of changes in federal spending and rev­enue are not adjusted for inflation in this paper.
[17]Congressional Budget Office, "The Budget and Economic Outlook: An Update."
[18]Thomas and Sellers, "Eliminate the Double Tax on Dividends," and Black, Legoria, and Sellers, "Capital Investment Effects of Dividend Imputation."
[21]The table uses CCH projections for the 2003 federal income tax brackets (Schedule Y-1: Married Filing Jointly and Surviving Spouses), deductions, and exemptions. See CCH Incorporated, 2003 Master Tax Guide (Chicago, Ill.: CCH Incorporated, 2002), pp. 25, 102, 309.
[24]See second footnote on Table 1.


Norbert Michel
Norbert Michel

Research Fellow in Financial Regulations

Ralph Rector

Senior Research Fellow

Alfredo Goyburu

Policy Analyst, Transportation and Infrastructure