CAPITAL GAINS TAXATION: DISTRIBUTIONAL EFFECTS

CRS Report for Congress
Capital Gains Taxes: Distributional Effects
September 24, 1999
Jane G. Gravelle
Senior Specialist in Economic Policy
Government and Finance Division


Congressional Research Service ˜ The Library of Congress

ABSTRACT
Several different measures of the distribution of the capital gains tax are presented. These
measures examine the absolute and relative distribution across income classes, the effects on
the distribution of taxes, and the proportion of the population affected by the tax. These
measures are presented for 1999 and indicate that capital gains taxes are concentrated among
high income individuals. This report will not be updated unless new data become available.



Capital Gains Taxes: Distributional Effects
Summary
Many types of data have been presented to illustrate who pays capital gains taxes
(and who might benefit from a reduction in these taxes). These different approaches
include absolute measures of distribution (such as how the tax is distributed relative
to the distribution of the population and the average tax paid), relative measures of
distribution (whether after-tax incomes would become more or less equal without the
tax), measures of the distribution of tax liability, and measures of who pays the tax.
These measures are presented for 1999 and indicate that capital gains taxes are
concentrated among high income individuals. Those with earnings over $200,000,
who constitute the top 1.8 percent of income, account for 78.6 percent of capital
gains taxes. While the average capital gains tax paid is $476, the average for the
highest income class is $20,536 and the average for the bottom half is less than $10.
Capital gains taxes contribute to a progressive tax system: while capital gains taxes
average 1.3 percent of disposable income, they account for 5.7 percent in the highest
income bracket and less than one tenth of one percent for the bottom 70 percent of
the population.
Some of this concentration in higher income classes occurs because of the
concentration of taxes at higher income levels. However, capital gains taxes are also
concentrated relative to other taxes. The capital gains tax is 4.5 percent of total
federal income, payroll and excise taxes; however, it is 14 percent of total taxes in the
highest income bracket, and less than one half of one percent for the bottom 70
percent of the population. For the income tax alone, capital gains taxes are 8.1
percent of total income taxes, but 16.2 percent of income taxes in the top income
class. In the bottom 70 percent of the distribution, the capital gains tax is less than
one percent of income taxes.
About a quarter of taxpayers who pay a capital gains tax are in the bottom 70
percent of the distribution, while 11 percent of capital gains taxpayers are in the top
1.8 percent of the population. About 12 percent of all taxpayers pay a capital gains
tax; in the highest income class, 75 percent pay a capital gains tax.
This report will not be updated unless new data become available.



Contents
Measures of Absolute Burden......................................1
Measures of Relative Burden and Progressivity.........................2
Effects on the Distribution of Tax Liability.............................3
Who Pays the Capital Gains Tax?...................................6
Limitations of Distributional Measures................................7
Appendix: Sources of Data........................................8
List of Tables
Table 1: Distribution of Capital Gains Taxes Across the Population and the Average
Tax, 1999 Income Levels......................................2
Table 2: Effects of the Capital Gains Tax on the Distribution of After-tax Income,
1999 Income Levels..........................................3
Table 3: Distribution of the Capital Gains Tax Compared to the Distribution of
Federal Income Payroll and Excise Taxes, 1999 Income Levels .........4
Table 4: Effects on the Distribution of Federal Individual Income Taxes, 1999 Income
Levels .................................................... 5
Table 5: Who Pays Capital Gains Taxes? 1999 Income Levels .............6
Table 6: Data on Number of Taxpayers, Percentage Taxable and Income, 1999.8
Table 7: Data on Taxes Paid and Capital Gains Income, 1999..............9



Capital Gains Taxes: Distributional Effects
Many types of data have been presented to illustrate who pays capital gains taxes
(and who might benefit from a reduction in these taxes). These different approaches
include absolute measures of distribution (such as how the tax is distributed relative
to the distribution of the population and what the average tax is), relative measures
of distribution (whether after-tax incomes would become more or less equal without
the tax), measures of the distribution of tax liability, and measures of who pays the
tax.
All of these measures are discussed below, and all are based on 1999 income
levels and data provided by the Joint Committee on Taxation. Data supporting the
calculations are presented at the end of this discussion. 1
Measures of Absolute Burden
Table 1 provides a measure of absolute burden, which shows that the burden is
highly concentrated in the higher income classes. Those with incomes over $200,000
constitute 1.8 percent of the population, but account for 78.6 percent of the tax. 2
Those with incomes over $100,000 constitute about 8 percent of the population but
pay over 90 percent of the tax. By contrast, the bottom third of the population, and
even the bottom half, pay virtually none of the tax, although that is in part because a
larger fractions of individuals in lower income groups do not have income tax liability.3
The middle third of taxpayers pay about 1 percent of the tax. The average tax is
$476; in the lower income and middle income classes tax cuts are under $10, while
the tax cut in the highest income bracket is about $20,000.


The analysis does not take into account the already legislated lower tax rates for1
property held for five years that are not in effect yet. If it did, the overall magnitude of the tax
would be smaller. It also does not take account of corporate capital gains taxes.
The income measure used is an expanded income definition which is adjusted gross2
income plus tax-exempt interest, employer contributions for health and life insurance,
employer share of the FICA tax, worker’s compensation, nontaxable social security benefits,
insurance value of Medicare benefits, alternative minimum tax preference items and excluded
income of individuals living abroad. This definition is narrower than that used by Treasury;
however, the important point to focus on in understanding the meaning of the income classes
is the share of the population in each class.
Note that this distribution is slightly different from the distribution of capital gains3
income, because it is corrected for the fact that some taxpayers have no tax liability and also
for the smaller tax rates for those in the 15 percent bracket.

Table 1: Distribution of Capital Gains Taxes Across the Population and
the Average Tax, 1999 Income Levels
Income ClassPercent of TotalPercent ofAverage Tax Per
TaxpayersCapital GainsReturn
Tax
Under $10,000 16.1 0.0 Less than $1
10,000 to 20,000 19.0 0.0 Less than $1
20,000 to 30,000 14.6 0.2 $7
30,000 to 40,000 11.5 0.5 21
40,000 to 50,000 9.4 0.7 33
50,000 to 75,000 14.3 2.8 94
75,000 to 100,000 7.2 4.4 295
100,000 to 200,000 6.1 12.7 993
200,000 and over 1.8 78.6 20,536
Total 100.0 100.0 476
Source: Congressional Research Service calculations based on Joint Committee on Taxation
data.
Measures of Relative Burden and Progressivity
Some argue that the analysis above is not very meaningful because income is
more concentrated at the higher income levels, and any income tax is concentrated
among higher income individuals. An alternative way to assess distribution is to look
at the effect on the relative distribution of income and to ask whether the tax makes
incomes more equal or less equal. One can also use this approach to determine
whether the provision makes the tax system less progressive or more progressive. The
way to measure this type of effect is to compare the distribution of the tax with the
distribution of after-tax income. If both distributions are the same, then the tax
benefit is distributionally neutral. If higher income individuals pay a larger share of
the tax relative to their after-tax income, the tax makes incomes less equal and
increases the progressivity of the tax system.
Table 2 presents data on these measures of distribution, which indicate that the
capital gains taxes reduce income inequality and increase progressivity. For example,
those with incomes over $200,000 receive 18 percent of after-tax income, but pay 78
percent of the capital gains tax. Similarly, those with incomes over $100,000 who
receive a third of the after-tax income, pay 90 percent of the tax. These differences
are also reflected in the capital gains tax as a percent of after-tax income. At lower



and moderate income levels, the capital gains tax is less than a tenth of a percent of
disposable income; at the highest income level it is almost 6 percent of income.
Overall, capital gains taxes are slightly over one percent of disposable income.
Table 2: Effects of the Capital Gains Tax on the Distribution of After-
tax Income, 1999 Income Levels
IncomePercentagePercentageCapital Gains Tax
ClassDistribution ofDistribution ofas a Percent of
Total DisposableCapital GainsDisposable
IncomeTax Income
Under $10,000 1.7 0.0 0.00
10,000 to 20,000 7.0 0.0 0.01
20,000 to 30,000 8.3 0.2 0.03
30,000 to 40,000 8.8 0.5 0.07
40,000 to 50,000 9.2 0.7 0.09
50,000 to 75,000 18.5 2.8 0.20
75,000 to 100,000 12.7 4.4 0.45
100,000 to 200,000 15.9 12.7 1.03
200,000 and over 17.8 78.6 5.67
Total 100.0 13.0 1.29
Source: Congressional Research Service calculations based on Joint Committee on Taxation
data
Effects on the Distribution of Tax Liability
Income taxes (other than those in the form of a refundable credit) inevitably do
not affect lower income individuals because they pay no income taxes. Measures of
distribution of a tax cut are often presented in the form of a percentage change in tax
liability, perhaps in part for this reason. However, changes in the distribution of tax
liability do not really tell us anything about inequality of incomes. For example, a
proportional reduction in tax liability in a progressive tax system makes after-tax
income shares less equal and, of course, benefits higher income individuals more in
an absolute sense. Thus it cannot be said to be a distributionally neutral measure.
Distributional indexes of tax burdens have, however, been used for certain measures



of progressivity (although other, and more common, measures are based on the
distribution of pre- and post-tax income). 4
Note that percentage changes can be very deceiving as measures of tax benefit.
For example a taxpayer with tax liability of $10 who received a reduction in tax of $5
would have a 50 percent reduction in tax liability, which is a very large percentage.
Yet it is not a very meaningful tax cut. Indeed, the percentage change in liability is
undefined, mathematically, when taxes are zero and percentage changes approach
infinity when taxes approach zero.
In any case, the following two tables show the distribution of the tax compared
to the distribution of tax liability. Table 3 refers to federal income, payroll and excise
Table 3: Distribution of the Capital Gains Tax Compared to the
Distribution of Federal Income Payroll and Excise Taxes, 1999 Income
Levels
IncomePercentagePercentageCapital Gains
ClassDistribution ofDistribution of theTaxes as a
Federal Income,Capital Gains Tax Percent of
Payroll and ExciseIncome and
TaxesPayroll Taxes
Under $10,000 0.4 0.0 0.1
10,000 to 20,000 2.1 0.0 0.1
20,000 to 30,000 5.3 0.2 0.2
30,000 to 40,000 6.8 0.5 0.3
40,000 to 50,000 7.9 0.7 0.4
50,000 to 75,000 17.9 2.8 0.7
75,000 to 100,000 14.3 4.4 1.4
100,000 to 200,000 19.9 12.7 2.9
200,000 and over 25.4 78.6 14.0
Total 100.0 100.0 4.5
Source: Congressional Research Service calculations based on Joint Committee on Taxation
data


For a discussion of progressivity indices, see Donald W. Kiefer, “Distributional Tax4
Progressivity Indices.” National Tax Journal, Vol. 37, December 1984, pp. 497-514.

taxes (thus excluding the corporate income tax and the estate and gift tax). Table 4
refers to the individual income tax. They also show the percentage reduction in tax
liability. Note that in both cases, the capital gains tax cut is more concentrated among
higher income individuals than is existing tax liability. For example, in table 3, the
highest income class that pays 79 percent of the capital gains tax pays 25 percent of
total taxes (excluding the corporate tax) and pays 39 percent of the income tax. The
data on capital gains taxes as a percentage of income, payroll and excise tax liability
also show that capital gains taxes are proportionally greater for high income
taxpayers. Capital gains are 4.5 percent of taxes overall, but 14 percent of the taxes
paid by those in the highest income class. For all other taxpayers, capital gains taxes
as a share of total taxes are less than the average. Similarly, in table 4, in the case of
the income tax, the capital gains tax accounts for 8.1 percent of overall income taxes,
the 16.2 percent of taxes in the highest income group. For all other income classes,
capital gains taxes as a percent of the income tax are less than the average.
Table 4: Effects on the Distribution of Federal Individual Income Taxes,
1999 Income Levels
IncomePercent of FederalPercent ofCapital Gains
ClassIndividual Income Capital GainsTaxes as a
TaxesTaxesPercent of
Income Taxes
Under $10,000 0.0 0.0 0.0
10,000 to 20,000 0.0 0.0 0.0
20,000 to 30,000 2.4 0.2 0.7
30,000 to 40,000 4.2 0.5 1.0
40,000 to 50,000 5.7 0.7 1.0
50,000 to 75,000 14.3 2.8 1.6
75,000 to 100,000 12.7 4.4 2.9
100,000 to 200,000 21.4 12.7 4.8
200,000 and over 39.3 78.6 16.2
Total 100.0 100.0 8.1
Source: Congressional Research Service calculations based on Joint Committee on Taxation
data



Who Pays the Capital Gains Tax?
Another measure that is sometimes used to assess the distributional effect of a
tax is to examine where the bulk of the population that pays any of the tax falls. For
almost any tax, such a measure will tend to report that more low and middle income
taxpayers pay a tax than high income individuals simply because there are many more
people who fall in the lower and middle income classes. Moreover, without measures
of magnitude, such measures are not very informative in determining burdens (or
benefits of reducing taxes). Nevertheless, Table 5 reports this type of measure. It
does demonstrate one of the reasons that capital gains tax cuts benefit high income
individuals using virtually any distributional measures. Only 12 percent of individuals
pay capital gains taxes, largely because many individuals do not have any income from
capital gains (and, in the lower income brackets, because of lack of tax liability as
well). However, in the highest income bracket, 75 percent of taxpayers pay capital
gains taxes, while in the next highest bracket, 48 percent pay.
Table 5: Who Pays Capital Gains Taxes? 1999 Income Levels
IncomePercent ofCumulativeCumulative
ClassIndividuals inShares Below anShares Above an
Each Class withIncome LevelIncome Level
a Capital Gains
Tax
Under $10,000 0.1 0.1 100.0
10,000 to 20,000 0.9 1.5 99.9
20,000 to 30,000 3.5 5.7 98.5
30,000 to 40,000 8.2 13.4 94.3
40,000 to 50,000 12.6 23.2 86.6
50,000 to 75,000 20.5 47.4 76.8
75,000 to 100,000 29.6 64.9 52.7
100,000 to 200,000 47.5 88.8 35.1
200,000 and over 74.8 100.0 11.2
Total 12.1
Source: Congressional Research Service calculations based on Joint Committee on Taxation
Data



The third column of table 5 shows the shares of individuals with a capital gains
tax by cumulating from the bottom. For example, 23 percent of those who pay
capital gains taxes have incomes under $50,000 and 47 percent of taxpayers who pay
capital gains taxes have incomes under $75,000. Column 4 cumulates in the other
direction. For example, it indicates that 76.7 percent of individuals who pay capital
gains taxes have incomes above $50,000. These numbers suggest that the capital
gains taxes are atypical in their concentration among higher income individuals.
Limitations of Distributional Measures
While these distributional measures provide a picture of where the burden of the
capital gains tax falls, there are certain limitations to these measures. One potential
problem is the measure of income. This measure of income is based on data taken
from tax returns, and it is not the exact equivalent of economic income, since there
are certain imputed and accrued income items, including unrealized capital gains, that
are not included. At the same time, the tax measure differs from cash income as well.
These limitations of the data can be addressed in part, however, by focusing on the
population shares represented by each income class.
A problem specific to examining the capital gains tax is the fact that capital gains
can be realized in large and uneven amounts. For example, a taxpayer may have sold
a large asset that yields a large amount of gain. Thus, some of the taxpayer’s who5
have large gains may be classified as having higher income simply because of the gain.
A recent study of capital gains during the 1980s suggests that, while there is some
tendency of annual snapshots to overstate the gain at high income levels, it is typically
not a serious enough distortion in most years to alter the basic pattern of distribution.6


The largest asset owned by most taxpayers, their home, was typically not taxed or not5
fully subject to tax (and recent legislation exempts almost all owner-occupied housing from
the capital gains tax).
See Leonard E. Burman, The Labyrinth of Capital Gains Tax Policy: A Guide for the6
Perplexed. Washington, D.C., the Brookings Institution, 1999, pp.100. Taxpayers with
$200,000 or more in income received 57 percent of capital gains in the ten years from 1979-
1988. During the individual years, the share ranged from 59 percent to 91 percent, although
the latter number was very atypical. In five of the years, the share was less than 65 percent;
in seven of the years, it was less than 70 percent, and for all but one years, it was less than 80
percent.

Appendix: Sources of Data
The following two tables provide the basic sources of data used to calculate the
various distributional measures used in this report. Table 6 reports the number of
taxpayers in each class. In order to convert capital gains income to taxes, it is
necessary to adjust by the percentage of returns that have positive tax liability, which
is reported in column (3). It also reports number with capital gains and income in
each class.
Table 6: Data on Number of Taxpayers, Percentage Taxable and
Income, 1999
Income ClassNumber ofPercentage NumberIncome
Taxpayers Taxable With (millions)
(thousands) Capital
Gains
(thousands)
Under $10,000 22,371 7.0 273 $94,036
10,000 to 20,000 26,314 33.8 677 390,202
20,000 to 30,000 20,301 61.0 1,165 502,468
30,000 to 40,000 15,902 79.4 1 639 551,205
40,000 to 50,000 13,082 90.6 1,823 586,599
50,000 to 75,000 19,829 98.4 4,130 1,208,043
75,000 to 100,000 10,042 98.6 2, 982 859,646
100,000 to 200,000 8,461 99.8 4,022 1,105,399
200,000 and above 2,527 99.8 1, 895 1,284,907
Source: Columns (2), (4) and (5) from Joint Committee on Taxation #D-99-49, July 9, 1999;
Column (3) calculated by Congressional Research Service based on data for 1998 in Joint
Committee on Taxation, “Estimates of Federal Tax Expenditures for FY 1999-2003,”
Committee Print, 105 Congress, 2 Session, December 14, 1998. thnd
Table 7 reports the remainder of the data needed to provide these calculations:
the amount of total federal taxes (excluding corporate taxes) paid, the amount of
individual income tax paid, and the distribution of capital gains income.



In addition, a tax rate is provided for each income class based on data indicating
that 71 percent of taxpayers with tax liability pay at the 15 percent rate (Internal
Revenue Service Statistics of Income, Individual Income Tax, 1996). The first five
income classes are assigned an initial tax rate of 10 percent, the next class a weighted
average of 12.9 percent, and the top three classes an initial tax rate of 20 percent.
Table 7: Data on Taxes Paid and Capital Gains Income, 1999
Income ClassTotal Income,IndividualCapital Gains
Payroll andIncome TaxIncome
Excise Taxes(millions)(millions)
(millions)
Under $10,000 5,728 -$8,300 $444
10,000 to 20,000 31,141 - 8,519 1,071
20,000 to 30,000 77,249 18,898 2,244
30,000 to 40,000 98,341 34,291 4,227
40,000 to 50,000 115,336 46,655 4,795
50,000 to 75,000 260,363 116,354 14,638
75,000 to 100,000 207,798 102,779 14,855
100,000 to 200,000 290,271 173,919 42,082
200,000 and above 369,750 319,360 259,958
Source: Joint Committee on Taxation, #D-99-50, July 9, 1999.