Revised May 31, 2001

Pathbreaking CBO Study Shows Dramatic Increases in Income Disparities in 1980s and 1990s:
An Analysis of the CBO Data

by Isaac Shapiro, Robert Greenstein, and Wendell Primus

PDF version of this report
HTML of press release

PDF of press release

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.

A pathbreaking study of income and tax trends since 1979 that the Congressional Budget Office issued last week(1) shows that dramatic increases in income disparities — especially between the wealthiest one percent of Americans and the rest of the society — occurred in both the 1980s and the 1990s. The study also shows that the percentage of income that Americans pay in federal taxes declined for every income group between 1979 and 1997 (the period the study covers) and, by one key measure, fell the most for those with the highest incomes. The CBO study provides the most comprehensive and reliable information that any researcher or research institution has developed on these matters and will now become the definitive data on these issues for the period examined.

Table 1

Average After-Tax
Income Gains, 1979-97

Top 1%


Middle fifth


Bottom fifth


The CBO study shows that from 1979 to 1997:

The CBO study lays to rest the oft-stated, but now clearly mistaken, notion that although income disparities grew rapidly in the 1980s, they stopped growing in the 1990s. To the contrary, the study shows that income disparities continued to grow sharply over the past decade.

Average After-Tax Income
Gains, 1989-97

Top 1%


Middle fifth


Bottom fifth


In addition to these data on income trends, the CBO study provides the most comprehensive data available on federal tax burdens over the past two decades. The study includes data from 1979 through 2001 on the percentage of income that each income group pays in federal taxes. The CBO study highlights one striking finding in this area — the percentage of income that Americans pay in federal taxes declined between 1979 and 2001 among every income group. Furthermore, the study reports that "Households in the top 1 percent of the distribution had the largest percentage-point fall in effective tax rates" (i.e., in the percentage of income paid in federal taxes). Thus, even before enactment of tax legislation that will provide very large tax cuts to the top one percent of households, that group already has received larger federal tax reductions over the past 22 years than any other income group.

The tax legislation President Bush is about to sign will accelerate the trends of growing income disparities between the wealthiest individuals and other Americans. The average percentage gain in after-tax income that will result from the legislation will be about three times greater for those in the top one percent of the income distribution than for those in the middle of the income spectrum, and more than seven times greater than the average percentage gain among the bottom fifth of the population. After-tax income disparities, already larger now than at any other time in more than two decades, will widen further as a result of the tax legislation.

This analysis examines the CBO study and the data it contains. The analysis begins with a brief discussion of why the data in the study eclipse data previously available on these matters.

Figure 1

The New CBO Data

In its report, CBO combines Census data with Internal Revenue Service data from income tax returns. In developing this data set, CBO consulted many of the nation's foremost experts on the measurement of incomes.

The data in this study are superior to other data on income trends. Census data are often cited on these matters, but experts agree that the CBO data are substantially stronger.

In short, CBO draws upon both Census data and IRS data. Each of those data sets is better than the other for certain types of data on certain income groups. CBO has used the strongest elements of each of these data sets to build a data set that is superior to either the Census or IRS data by themselves.(3)


The Long-term Trend

The CBO data cover nearly two decades. These data show that the income gains among those at the top of the income spectrum greatly outpaced the gains among all other income groups throughout the 1979-1997 period. Between 1979 and 1997, average income stagnated among the bottom part of the population, increased modestly for the middle of the population, and soared among those at the top of the income spectrum.

Table 3

Shares of National
After-Tax Income




Top 1%



Bottom 40%



As a result, the gaps between the top and the middle of the income distribution are much larger now than they were at the end of the 1970s, as are the gaps between the top and the bottom of the income distribution. After-tax income disparities were larger in 1997 than in any other year measured by CBO — and dramatically larger than they were two decades ago. This also can be seen in the data the CBO study presents on changes in the proportion of the national income that each income group receives.

The CBO data go back to 1979. It is likely that if these data were available for years before 1979, they would show the disparities in after-tax income are wider now than for a much longer period. Census data on before-tax income show that income disparities were wider in the latter half of the 1990s than at any point in the post-World War II era; it is possible the CBO methodology would find the same result for after-tax income if the relevant data were available.

Table 4

Average After-Tax Gains, 1989 to 1997
(adjusted for inflation, in 1997 dollars)


Top 1%

Next 4%

Next 15%

Fourth 20%

Middle 20%

Second 20%

Lowest 20%

Percentage gains








Dollar gains








Source: Congressional Budget Office

Income Disparities Grew Rapidly in the 1990s

It is conventional wisdom that income disparities grew sharply in the 1980s. Some observers have argued, however, that income disparities have stabilized since then rather than continuing to grow. This argument is based on Census data that appear to show that income disparities have remained relatively constant since 1993.

These Census data are seriously deficient, however, because they fail to capture much of the income growth that has occurred at the top of the income spectrum in recent years. As noted, the Census data do not include capital gains income and omit a significant portion of other income that households with the highest incomes receive. The Census Bureau regards its data on those who have very high incomes as being sufficiently unreliable that it does not publish — or even attempt to break out — information on the top one percent of households.

The more complete CBO data provide a more accurate picture. The pace of the growth in income disparities may have slowed somewhat in the 1990s, but income gains continued to be remarkably uneven. From 1989 to 1997, the average after-tax income of the bottom fifth of the population remained unchanged, while income rose by an average of less than six percent for the middle fifth of the population. But the average after-tax income of the top one percent continued to shoot up, rising 36 percent — close to $180,000 per household — over this period. (See Table 4.) With the release of the CBO study, there is no longer any question that income disparities continued to widen in the 1990s.

Income disparities grew especially quickly from 1995 to 1997

Furthermore, income disparities grew more sharply between 1995 and 1997 — the last two years the CBO study covers — than in any other two-year period since 1979. This suggests the possibility that the growth in disparities in after-tax income may have accelerated in the latter half of 1990s.

Disparities appear to have grown further since 1997

Census data for 1998 and 1999 indicate that all income groups have gained somewhat since 1997. The CBO study concludes, however, that at least through 1999, those gains continued to be concentrated at the top. CBO reports: "Although CBO's database on taxes and income ends in 1997, information from tax returns suggests that the recent rapid rise in the share of income going to the top of the distribution and in the share of individual income taxes those households pay has continued in 1998 and 1999."

IRS data are currently available through 1998. Those data show that the average after-tax income of the top one percent of the population grew 13 percent between 1997 and 1998, and that this was more than three times the average percentage income gain for the bottom 90 percent of the population. (These IRS data reflect income after federal income taxes are subtracted but do not net out other taxes.)


The Effects of Changes in Tax Burdens

CBO notes in its study that the share of federal income taxes that is paid by those at the top of the income distribution has increased over the past two decades. CBO finds, however, that this is primarily the result of the increased concentration of income among the very affluent and is not the result of an increase in tax burdens on high-income households. Since high-income households receive a much larger share of the nation's income than they did two decades ago, they pay more of the nation's taxes. As CBO puts it, "increasing inequality of income led to similar shifts in the distribution of tax liabilities."

Indeed, the CBO study demonstrates that the effective tax rates paid by those at the top of the income spectrum — that is, the percentage of income that they pay in federal taxes — have not increased. To the contrary, the effective federal tax rate these households pay declined between 1979 and 1997, as did the effective tax rates that all other income groups pay.

Moreover, the top one percent of taxpayers experienced a larger percentage-point drop in their effective tax rate over this period than did any other group. In 1979, the top one percent of households paid an average of 37.3 percent of income in federal taxes. In 1997, they paid 33.3 percent, a decline of four percentage points. (The bottom fifth of households experienced the next-largest percentage point drop in its effective federal tax rate. Since the average tax rate these households pay was lower to begin with, this group received the largest proportionate reduction in tax burdens. Even with this decline in federal taxes, however, the average after-tax income of the bottom fifth of households fell $100 over the 1979 - 1997 period.)

The CBO study also found that the average tax rates that all income groups pay declined further from 1997 to 2001. This is a result of the tax cuts enacted in 1997.

At first blush, this story of declines in the effective federal tax rates that every income group pays may seem to be in conflict with the often-repeated fact that federal taxes are at a post-World War II high as a percentage of the U.S. economy. The CBO study explains this seeming discrepancy. As the study notes, the reason is the increased concentration of income at the top of the income spectrum, which has resulted in a larger share of the nation's income being received by those who pay a larger percentage of their income in taxes than the rest of the population does.

These data demonstrate conclusively that claims that American families have been shouldering heavier tax burdens and paying a larger share of their incomes in taxes to the federal government are inaccurate (and deceptive). As the data in the CBO study show, the average tax rates that those at the top pay — along with the average rates that the rest of the population pays — can decline at the same time that federal revenues increase as a share of the economy.

The CBO study also breaks out federal tax burdens by different types of taxes, including individual income taxes, payroll taxes, the corporate income tax, and excise taxes. For the bottom four-fifths of the population, payroll taxes are the largest federal tax. The dominance of payroll taxes has grown over time. CBO reports that in 1979 payroll taxes exceeded income taxes for 56 percent of the population. In 1997, payroll taxes were greater than income taxes for 79 percent of the population.


The Effects of the Tax Bill

It is possible to use the budget surplus to help mitigate these trends of widening income disparities. This could be done by deploying the surplus (whether through tax cuts or program increases) in a manner that raises the after-tax income of low- and middle-income households by a larger percentage than it raises the after-tax income of high-income households. High-income households could still receive a tax cut under such an approach, but not as large a cut as is now in store for them.

The tax bill about to become law would have the opposite effect. It would widen income disparities further.

At a time when the most comprehensive government study of income trends ever conducted demonstrates that disparities in after-tax income grew sharply in both of the last two decades, the tax bill would widen such disparities further and provide much greater benefits to those who have gained the most since the late 1970s than it would provide to other Americans.


Appendix Table 1

Average After-Tax Income by Income Group
(1997 dollars)

Income Category




Percent Change 1979-1997

Dollar Change 1979-1997

Lowest fifth






Second fifth






Middle fifth






Fourth fifth






Highest fifth






81st-95th Percentile






96th-99th Percentile






Top 1 Percent







Appendix Table 2

Average Pre-Tax Income by Income Group
(1997 dollars)

Income Category




Percent Change 1979-1997

Dollar Change 1979-1997

Lowest Fifth






Second Fifth






Middle Fifth






Fourth Fifth






Highest Fifth






81st-95th Percent






96th-99th Percent






Top 1 Percent







Appendix Table 3

Effective Total Federal Tax Rate
(Percentage of Income Paid in Federal Taxes)

Income Category



(before new tax legislation)

Percentage Point Change

Lowest Fifth





Second Fifth





Middle Fifth





Fourth Fifth





Highest Fifth








81st-95th Percentiles





96th-99th Percentiles





Top 1 Percent





Appendix 1. The CBO Study and Income Mobility

The CBO study provides data for each odd-numbered year from 1979 to 1997. For each such year, it places households into different income groups, based on their average income in that year. The study thus shows how the incomes and tax burdens of different income groups compare over time. For example, the study compares the average incomes of the poorest 20 percent of the population in 1979 to the average incomes of the poorest 20 percent of the population in 1997.

Just who is in each income group changes over time. Some families have low incomes for only a few years and move quickly into the middle class. For example, the parents of a young child may be working part time while finishing college. The family's income may be low for a few years, but after both parents graduate from college and obtain better-paying jobs, the family's income may increase substantially.

Although some families see their incomes increase over time, studies of income mobility have shown that the majority of low-income families continue to have low incomes for many years. A recent study of earnings mobility by Peter Gottschalk of Boston College and Sheldon Danziger of the University of Michigan found that in the early 1990s, 75 percent of the people whose incomes placed them among the poorest fifth of the population in one year remained in the poorest fifth of the population the following year. While income mobility is greater when a longer period of time is analyzed, Gottschalk and Danziger found that almost half of the young adults (those age 22 to 39) who were in the bottom fifth of the income distribution in 1968 still were in the bottom fifth some 23 years later, in 1991. Moreover, nearly three-quarters of those who were in the bottom fifth in 1968 were in the bottom two-fifths in 1991.(4)

Another question is whether income mobility has increased over time. This is an important issue. Increases in income mobility can offset increases in income inequality. If, however, income mobility has not increased — that is, if it has remained about the same or declined since the 1970s — then the sharp increases in income disparities since the late 1970s reflect true growth in disparities and not merely a reshuffling of the income distribution. The Gottschalk and Danziger study finds that income mobility declined somewhat between the late 1960s and the early 1990s, as the percentage of people changing income category from one year to the next diminished.

The Gottschalk and Danziger study is one of the most recent studies on income mobility. A review of the literature in the field conducted by researchers at the Urban Institute (before release of the Gottschalk-Danziger study) also found that mobility has not increased. The Urban Institute study, by Daniel McMurrer and Isabel Sawhill, concluded: "The evidence on this point is clear: Mobility has not changed significantly over the last 25 years. Indeed, a number of different studies indicate that relative mobility rates in the United States — both short term and long term — have been remarkably stable."(5)

End Notes

1.  The Congressional Budget Office, Historical Effective Tax Rates, 1979-1997, Preliminary Edition, May 2001. <>

2. Two other, somewhat more technical, differences between the CBO data and Census data should be noted. First, CBO places households into income categories after making an adjustment for the number of people in the household. Take two households with incomes of $20,000. Household A has four people in it; Household B has one person in it. Recognizing that it is harder for four people to live on $20,000 than it is for one person, CBO's method considers the household with one person to be better off, and to have higher adjusted household income, than the household with four people. The Census Bureau makes no adjustments for family size, so that a household of four with an income of $20,000 can be placed in a higher income category than a single-person household with income of $19,500.

Second, CBO places an equal number of people into each income fifth. The Census Bureau places equal numbers of households into each income fifth. Since many low-income households consist of elderly individuals, under the Census approach the bottom fifth of the population contains a smaller number of people than each of the other fifths of the population do. As a result, the Census approach understates the amount of income going to the bottom fifth of the population.

3. Over the past year, the Center on Budget and Policy Priorities has released two reports on income distribution based on the IRS data. These reports were written because CBO was still developing its new methodology and had not provided income data for any year after 1995 under its old methodology. The IRS data contained information not otherwise available on income trends among the top one percent of the population through 1998. For the latest of these two reports, see Isaac Shapiro, "The Latest IRS Data on After-tax Income Trends," Center on Budget and Policy Priorities, February 26, 2001.

4. Peter Gottschalk and Sheldon Danziger, "Family Income Mobility - How Much is There, and Has It Changed?" in James A. Auerback, and Richard S. Belous, eds. The Inequality Paradox: Growth of Income Disparity. (Washington, DC: National Policy Association) 1998.

5. Daniel P. McMurrer and Isabel V. Sawhill, "How Much Do Americans Move Up and Down the Economic Ladder," The Urban Institute, November 1996.


Center Staff
Internship Information


To join the Center's e-mail notification list, ask questions,
or send comments, write to
Center on Budget and Policy Priorities
820 First Street, NE, Suite 510
Washington, DC  20002
Ph: (202) 408-1080
Fax: (202) 408-1056