April 13, 2005


Treasury Department Release Creates Misleading Impression
About Taxes that High-Income Taxpayers Pay
By Joel Friedman, Isaac Shapiro, and Robert Greenstein

PDF of full report

View Related Reports

If you cannot access the files through the links, right-click on the underlined text, click "Save Link As," download to your directory, and open the document in Adobe Acrobat Reader.

The Treasury Department recently updated a fact sheet on who pays the federal income tax.  The fact sheet makes two main points:  that “the individual income tax is highly progressive — a small group of higher income taxpayers pay most of the individual income tax each year,” and that the burden these taxpayers bear has increased as a result of the tax cuts enacted under the Bush Administration.[1]

The Treasury release, along with other statements by Administration officials, are designed to counter arguments that the tax cuts enacted since 2001 are tilted to those at the top of the income scale.  The fact sheet, however, is misleading.  By focusing only on the federal income tax and leaving out all other federal taxes, it creates misimpressions both about the degree to which the tax code is progressive and about who is benefiting the most from the recent tax cuts.


How Progressive Is the Nation’s Tax Code?

The Treasury fact sheet focuses only on the individual income tax, one of the most progressive taxes.  Not surprisingly, it shows that high-income households pay a large share of this tax.  But the degree to which high-income households shoulder the tax burden is lessened once payroll taxes are taken into account.  Although the progressive individual income tax is the single largest source of federal revenue, the payroll tax — which is regressive — is a very close second, accounting for between 35 percent and 40 percent of all federal taxes in recent years.[3]  The payroll tax places a greater burden on low- and middle-income families than the income tax because it is levied at a flat rate on the first dollar of wages and does not apply to wages above $90,000 in 2005.  According to CBO, more than three-quarters of workers paid more in payroll than in income taxes in 2000, the latest year for which these data are available.[4]

As a result, the CBO data show that the top one percent of households paid 21.1 percent of all federal taxes in 2002, significantly below the 33.0 percent of individual income taxes that CBO estimates this group paid.[5]

The CBO data also indicate that in 2002, the top one percent of households received 13.4 percent of the before-tax income in the nation, eight percentage points less than the share of total federal taxes they paid.  This indicates that the federal tax system overall is progressive, but not to the steep degree the Treasury fact sheet implies.

Furthermore, taxes also are paid to state and local governments, which tend to rely more heavily on regressive sales taxes and excise taxes.  When state and local taxes are included, the gap between the percentage of the national income that high-income individuals receive and the percentage of the taxes they pay narrows further, a point that the noted tax expert Joseph Pechman demonstrated nearly two decades ago.

Estimates by the Citizens for Tax Justice confirm this conclusion.  They show that in 2004, when more of the recently enacted tax cuts were in effect than in 2002, the overall share of federal, state, and local taxes that high-income households paid was only a little bit larger — less than two percentage points larger — than their share of the national income.  Similarly, the share of taxes that middle- and lower-income households pay was just a little more than one percentage point lower than their share of the national income, according to these CTJ estimates.[6]  It is only because the progressive federal tax system offsets the regressive effects of state and local taxes that the U.S. tax system as a whole is even mildly progressive.


Who Gains The Most From The Tax Cuts?

According to the Treasury Department fact sheet, “The President’s tax cuts have shifted a larger share of the individual income taxes paid to higher income taxpayers.  In 2005, when most of the tax cut provisions are fully in effect (e.g., lower tax rates, the $1,000 child credit, marriage penalty relief), the projected tax share for lower-income taxpayers will fall, while the tax share for higher-income taxpayers will rise.”  These statements appear intended to create the impression that the tax cuts were distributed in a highly equitable manner, giving households that do not have high incomes more than their fair share.

Other Administration statements make this argument even more directly.  In its latest Economic Report of the President, the Council of Economic Advisors wrote that “the tax relief passed during the President’s first term increased the overall progressivity of the Federal tax system.”[7]  A White House spokesperson boldly claimed that “any way you slice it, the president’s tax cut made the system more progressive.”[8]  But these assertions that the Administration’s tax cuts have made the tax code more progressive “any way you slice it” do not hold up under scrutiny.

Administration Assertions Ignore More Relevant Data

Data from the Congressional Budget Office and the Urban Institute-Brookings Institution Tax Policy Center indicate that the tax cuts enacted since 2001 benefit high-income households the most and are exacerbating the concentration of after-tax income at the top of the income scale.  (The Center described the CBO data in an analysis issued in September 2004.[9])  New data from the Tax Policy Center show that in 2005:

Table 1:  Impact of Enacted Tax Cuts, 2005

Income Class

Average tax cut

Percentage growth in after-tax income

Middle 20 percent



Top one percent



Over $1 million



Source:  Urban-Brookings Tax Policy Center

As the preceding discussion indicates, there is a range of statistics available to assess the impact of tax cuts on different income groups.  Economists generally believe that examining how tax cuts alter the “after-tax” income of households is the most appropriate measure of the distributional impact of tax cuts.  After-tax income represents the best measure of the income that a household has available to spend or save; it also takes into account the relative tax burden of each household.  (See the box on page 6 for a discussion of the impact of the tax cuts on different income groups when measures to pay for the tax cuts eventually are taken into account.)

By contrast, the Administration tends to focus on the share of taxes paid by different income groups.  As noted, the Administration’s preferred statistic tells us that the higher-income group paid a slightly larger share of the much smaller amount of federal income taxes that were collected after the tax cuts.  But this statistic is not particularly meaningful for assessing the distribution of the tax cuts.

The flaw in the Administration’s favored statistic can be seen by taking the Administration’s case to its logical extreme.  Assume, for instance, that all taxes were eliminated except that high-income households had to pay taxes equal to $1 a year.  The high-income group would be paying 100 percent of all taxes.  Under the Administration’s logic, the tax code would be much more progressive as a result of this change, because those at the top would be paying 100 percent of the taxes.  Yet this statistic would have little meaning because the revenue base would have essentially been eliminated.

While not to this extreme, the revenue base is in fact eroding, in substantial part due to the tax cuts.  Federal revenues have fallen well below their historical average.  In 2005, CBO estimates that federal revenues will equal 16.8 percent of GDP — a level that is lower than the average levels for the decades of the 1960s, 1970s, 1980s, and 1990s.  (This low level of revenues is the primary cause of the large deficit in 2005.)

Administration Focus on 2005 Leaves Out Tax Cuts

By focusing on 2005, the Treasury fact sheet leaves out tax cuts that are not yet in full effect and that benefit only the wealthiest households.  Thus, the Treasury figures are not representative of the impact of the full range of the tax cuts that were enacted in 2001 and 2003 and that the Administration is seeking to make permanent.

Assessing the Tax Cuts Along with Options to Pay for Them

The estimates on the distributional impact of the tax cuts discussed in this report reflect only the “benefit” side of the equation, ignoring the impact of the steps that ultimately will have to be taken to cover the costs of the tax cuts.  Yet, an assessment of the effects of the tax cuts, particularly over the long term, is incomplete if it does not take into account the estimated impact of measures that ultimately will have to be taken to pay for the tax cuts.

This is especially true of efforts to assess the impact of proposals to make the tax cuts permanent, because those impacts will unfold over a period in which the nation’s underlying fiscal position (even without the tax cuts) will be sharply out of balance due to the retirement of the baby-boom generation.  As a result, the cost of the tax cuts eventually will have to be offset.

To date, the tax cuts have been financed by higher deficits and increased borrowing.  Borrowing postpones, but does not eliminate, the need to pay for the tax cuts.  Paying for the tax cuts ultimately will require reductions in federal programs, increases in federal taxes, or some combination of the two.

The precise mix of program cuts and tax increases that ultimately will be adopted to offset the costs of the tax cuts is not known.  The Tax Policy Center has, however, examined two possible scenarios.  Under one scenario, each household will pay the same dollar amount to offset the cost of the tax cuts (roughly this scenario could occur if the tax cuts were financed largely or entirely through cuts in federal programs).  Under the other scenario, each household pays the same percentage of its income to cover the costs of the tax cuts.  (This is consistent with a financing scenario that combines program cuts and progressive tax increases.)

The Tax Policy Center found that, under either scenario, more than three-quarters of households would not benefit from the tax cuts and would end up losing more than they gain.  Only those with high incomes would be better off.  This result — that the majority of households would be net losers — reflects the fact that the bulk of the tax cuts are going to households with high incomes and that unless the measures ultimately adopted to cover the costs of the tax cuts are aimed at high-income households to the same degree that the tax cuts benefit these households, the net effects will be negative for the majority of households.

In a recent analysis of the Bush tax cuts, Brookings Institution economists and Tax Policy Center co-directors William Gale and Peter Orszag write that their preferred measure to use in assessing the distributional effects of tax changes is the change in after-tax income after accounting for the financing of the tax cuts.  They point out that a distributional analysis of the 2001 and 2003 tax cuts that does not include the impact of financing may yield seemingly contradictory evidence, but that once the financing is included in the analysis, the “the apparent contradictions are removed, and all of the measures show that the tax cuts are regressive.”*

*William Gale and Peter Orszag, “Bush Administration Tax Policy:  Distributional Effects,” Tax Notes, September 27, 2004


Income Gains Have Been Much More Rapid at the Top, Even After the Effects of Taxes Are Taken Into Account

Although households with high incomes pay a larger share of their income in federal taxes than other income groups do, their income after taxes has grown much more rapidly over the past two decades than the after-tax income of other households.  After-tax income is the best indicator of a household’s well-being, because it shows the level of income that a household has to spend, save, or invest.

CBO estimates that the average after-tax income of the top one percent of households grew by a stunning 111 percent — more than doubling — between 1979 and 2002.  In contrast, the incomes of those in the middle fifth of the income spectrum grew by an average of 15 percent over this period, or just 0.6 percent a year.[10]  In dollar terms, the top one percent experienced a $332,800 average increase in after-tax income over this period, while those in the middle saw their incomes grow by an average of $5,700.  (These figures are adjusted to remove the effects of inflation.  The CBO data cover the years from 1979 through 2002.)

Table 2:  Growth in After-Tax Income*




Income Class

Percentage Increase

Increase in Dollars

Percentage Increase

Increase in Dollars

Middle 20 percent





Top one percent





*Adjusted for inflation
Source: CBO

The CBO data also show marked disparities in after-tax income growth since 1990, the first year the Treasury fact sheet covers.  After-tax income grew much faster among high-income households between 1990 and 2002 than among ordinary households, despite the fact that the overall tax rates that the top one percent of households faced were higher in 2002 than in 1990, as a result of deficit-reduction measures enacted in 1990 and 1993.  According to the CBO data:

As a result of the rapid income growth at the top of the income scale, the share of the nation’s income that flows to those with high incomes increased substantially over this period.

According to CBO, the top one percent received 11.4 percent of the nation’s after-tax income in 2002, up sharply from this group’s 7.5 percent share in 1979.[11]

Similarly, the top 20 percent received almost as much income in 2002 as the bottom 80 percent of the population.  That is, the 23 million households with the highest incomes (the top 20 percent) received about as much income as the bottom 88 million households combined.

Given that those at the top of the income spectrum are receiving a larger share of the national income than in the past, it is not surprising that they are paying a larger share of federal taxes.


Tax Burden For Upper-Income Households Has Declined

Finally, the CBO data show that the percentage of income that higher-income households pay in federal taxes was lower in 2002 than in a number of the previous years covered by the CBO study.

One key point is that these CBO estimates for 2002 do not reflect the full effects of the recent tax cuts that have benefited high-income households handsomely.  The top income tax rate stood at 38.6 percent in 2002, down only slightly from the 39.6 percent rate before the tax cuts.  Today, the top tax rate is significantly lower at 35 percent.  In addition, the recent tax cuts on dividend and capital gains income were not enacted until 2003.  As more of the recent tax cuts have taken effect, the share of income that high-income households pay in federal taxes is lower than it was in 2002.[12]  As just noted, the 2002 level itself was below the level for all years since 1992.



The Treasury Department fact sheet continues a trend by the Administration of using statistics selectively to create questionable impressions about the nature of the nation’s tax burden and the impact of the tax cuts enacted since 2001.  In general, the overall U.S. tax system is only modestly progressive.  The share of federal taxes that high-income households pay has risen in recent decades, but this is in part because these households’ incomes have risen much more rapidly than everyone else’s and they now receive a larger share of the national income.  These trends continue to hold after the effects of taxes are taken into consideration.

Moreover, the percentage of income that high-income households pay in federal taxes has been declining since 1996, was already lower in 2002 than in any year since 1992, and has fallen significantly further since 2002 as more tax cuts have taken effect.

End Notes:

[1] Treasury Department, “Fact Sheet:  Who Pays The Most Individual Income Taxes?” March 2, 2005.

[2] Congressional Budget Office, “Historical Effective Federal Tax Rates, 1979-2002,” March 2005.  Unless otherwise noted, all the CBO data mentioned in this analysis come from this report.

[3] CBO estimates that in 2005, individual income taxes will represent 44 percent of all federal taxes and payroll taxes will represent 38 percent.  After 2005, assuming extension of the tax cuts, the share attributable to individual income taxes is projected to rise somewhat and the share attributable to payroll taxes to fall modestly.

[4] Congressional Budget Office, “Effective Federal Tax Rates, 1979-2000,” August 2003.  The CBO calculation follows the standard practice of attributing to workers the burden of both the employee and employer payroll taxes, with the employer portion being passed on to workers in the form of lower wages.

[5] The CBO estimate (33.0 percent) is slightly lower than the Treasury estimate (33.7 percent).  This small discrepancy exists largely because the two institutions employ somewhat different concepts of income and of the populations their analyses cover.

[6] Citizens for Tax Justice, “Overall Tax Rates Have Flattened Sharply Under Bush,” April 13, 2004.  CTJ estimates that in 2004, the top one percent of households will receive 19.1 percent of the nation’s income and pay 20.8 percent of federal, state, and local taxes.  The CTJ figures reflect concepts of income and of the population covered in its analysis that are somewhat different than those used by CBO and Treasury. 

[7] Council of Economic Advisors, Economic Report of the President, 2005, March 2005.

[8] David Cay Johnston, “The Bill’s Lower Now.  What About 4/15/11?” The New York Times, February 13, 2005

[9] David Kamin and Isaac Shapiro, “Studies Shed New Light on Effects of Administration’s Tax Cuts,” Center on Budget and Policy Priorities, September 13, 2004.

[10] These trends parallel the changes in before-tax incomes.  The CBO data show that the before-tax income of the top one percent of the population grew by 98 percent between 1979 and 2002.  In contrast, the before-tax incomes of those in the middle fifth of the income scale grew by 9 percent over this period.  Among both groups, after-tax incomes grew at a somewhat more rapid rate than before-tax incomes due to the general decline in the rate of taxation.

[11] The share of income going to the top one percent was lower in 2002 than in 1997 through 2001, reflecting in large part the drop in the stock market.  The CBO data now end in 2002.  Since 2002, however, income gains at the top may have outpaced the gains across the rest of the income spectrum.  For instance, the ratio of the pay of the nation’s chief executive officers to the pay of average workers fell from 2000 to 2002, but rose from 2002 to 2003.

[12] CBO estimated in a recent study that between 2002 and 2004 the effective tax rate for the top one percent of the income spectrum would fall by almost 3 percentage points, reflecting the impact of the additional tax cuts taking effect between those two years.  See Congressional Budget Office, “Effective Federal Tax Rates Under Current Law, 2001 to 2014,” August 2004.