[an error occurred while processing this directive]

Income Inequality 

A BRIEF LOOK AT POSTWAR U.S. INCOME INEQUALITY 

by Daniel H. Weinberg 


Are the rich getting richer and the poor getting poorer? 
Historical Census Bureau statistics on income can shed some light
on that debate.  Although the Bureau has been measuring incomes
for a half-century and a large number of factors have been
identified as contributing to changes in inequality, the root
causes are still not entirely understood.

The Census Bureau has been studying the distribution of income
since the late 1940's.  The first income inequality statistics
were published for families and came from the annual demographic
supplement to the Current Population Survey (CPS).  The most
commonly used measure of income inequality, the Gini index (also
known as the index of income concentration)/1/, indicated a
decline in family income inequality of 7.4 percent from 1947 to
1968.  Since 1968, there has been an increase in income
inequality, reaching its 1947 level in 1982 and increasing
further since then.   The increase was 16.1 percent from 1968 to
1992 and 22.4 percent from 1968 to 1994 (see Figure 1)./2/

Living conditions of Americans have changed considerably since
the late 1940's.  In particular, a smaller fraction of all
persons live in families (two or more persons living together
related by blood or marriage).  Therefore, starting in 1967, the
Census Bureau began reporting on the income distribution of
households in addition to families.  By coincidence, 1968 was the
year in which measured postwar income was at its most equal for
families.  The Gini index for households indicates that there has
been growing income inequality over the past quarter-century. 
Inequality grew slowly in the 1970's and rapidly during the early
1980's.  From about 1987 through 1992, the growth in measured
inequality seemed to taper off, reaching 11.9 percent above its
1968 level.  This was followed by a large apparent jump in 1993,
partly due to a change in survey methodology./3/  The Gini index
for households in 1994 was 17.5 percent above its 1968 level.

Income inequality measures such as the Gini index or shares of
aggregate income are particularly sensitive to changes in data
collection measures.  A change that may only affect a relatively
small number of cases (especially those in the upper end of the
income distribution) can affect these measures, while having
virtually no effect on median income.  We are unable to determine
what fraction of the measured increase in income inequality
between 1992 and 1993 was due to changes in survey administration
between those two years, though our analysis suggests there was
nonetheless a real increase in inequality between 1992 and
1993./4/

Figure 2 illustrates the increasing share of aggregate household
money income received by the highest income quintile (households
with incomes above $62,841 in 1994)/5/ -- 49.1 percent in 1994
and 46.9 percent in 1992, up from 42.8 percent in 1968 -- and the
declining share for households in the middle 60 percent and those
in the bottom quintile (incomes below $13,426)./6/  During that
same period, the share received by households in the top 5
percent of the income distribution went from 16.6 percent in 1968
to 18.6 percent in 1992 and 21.2 percent in 1994.

Yet another way to look at the change in inequality involves the
income at selected positions in the income distribution.  As
Figure 3 shows, in 1994 dollars the household at the 95th
percentile in 1994 had $109,821 in income, 8.2 times that of the
household at the 20th percentile, which was $13,426 (the
comparable 1992 ratio was 7.9)/7/.  In contrast, in 1968, the
household at the 95th percentile had but 6.0 times the income of
the household at the 20th percentile.

A parallel way to look at this change examines the average (mean)
household income in each quintile (see Figure 4).  The average
income of households in the top quintile grew from $73,754 in
1968 to $96,240 in 1992 and $105,945 in 1994.  In percentage
terms, this growth was 30 percent from 1968 to 1992 and 44
percent from 1968 to 1994.   During the 1968-1994 period, the
average income in the bottom quintile grew by only 8 percent,
from $7,202 to $7,762 (7 percent from 1968 to 1992)/8/. 
Consequently, the ratio of the average income of the top 20
percent of households to the average income of the bottom 20
percent went from 10.2 in 1968 to 12.5 in 1992 and 13.6 in 1994. 

Yet one more way to look at the income distribution corrects for
family size changes over the period, by examining the change in
the ratio of family income to its poverty threshold.  Poverty
thresholds vary by family size and composition reflecting
consumption efficiencies achieved through economies of scale
(i.e., families of two or more persons can share certain goods
such as housing)./9/  A ratio of 1.00 thus indicates that the
family has an income equal to the poverty threshold for its size
and composition.  The average ratio in the bottom quintile in
1968 was 1.04, while the average in the top quintile was 6.13. 
By 1994, these ratios were 0.92 and 9.22, respectively, (and 0.89
and 8.39 in 1992), also indicating a widening income gap (see
Figure 5).  The ratio for the middle quintile also rose, from
2.80 in 1968 to 3.26 in both 1992 and 1994.

In sum, when money income is examined, each of these indicators
shows increasing income inequality over the 1968-1994 period. 
But, are there other perspectives that change this story?

Since 1979, the Census Bureau has examined several experimental
measures of income.  These measures add the value of noncash
benefits (such as food stamps and employer contributions to
health insurance) to, and subtract taxes from, the official money
income measure.  The Bureau's research in this area/10/ has shown
that the distribution of income is more equal under a broadened
definition of income that takes account of the effects of taxes
and noncash benefits.  Further, government transfer benefits play
a much more equalizing role on income than do taxes. 
Nonetheless, while the levels of inequality are lower, this
alternative perspective does not change the picture of increasing
income inequality over the 1979-1994 period./11/

Why are these changes in inequality happening?/12/

The long-run increase in income inequality is related to changes
in the Nation's labor market and its household composition.  The
wage distribution has become considerably more unequal with more
highly skilled, trained, and educated workers at the top
experiencing real wage gains and those at the bottom real wage
losses.  One factor is the shift in employment from those goods-
producing industries that have disproportionately provided high-
wage opportunities for low-skilled workers, towards services that
disproportionately employ college graduates, and towards low-wage
sectors such as retail trade.  But within-industry shifts in
labor demand away from less-educated workers are perhaps a more
important explanation of eroding wages than the shift out of
manufacturing.  Also cited as factors putting downward pressure
on the wages of less-educated workers are intensifying global
competition and immigration, the decline of the proportion of
workers belonging to unions, the decline in the real value of the
minimum wage, the increasing need for computer skills, and the
increasing use of temporary workers. 

At the same time, long-run changes in living arrangements have
taken place that tend to exacerbate differences in household
incomes.  For example, divorces and separations, births out of
wedlock, and the increasing age at first marriage have led to a
shift away from married-couple households and toward single-parent
and nonfamily households, which typically have lower incomes.  Also,
the increasing tendency over the period for men with higher-than-
average earnings to marry women with higher-than-average earnings has
contributed to widening the gap between high-income and low-income 
ouseholds.

CONTACTS
Income Inequality -- Edward Welniak (301) 763-3243
Statistical Methods -- Tom Moore (301) 763-4215
Historical tabulations on income and poverty can be found on the
Census Bureau's Internet site, at http://www.census.gov.

ACCURACY OF THE ESTIMATES
All statistics in the report are from the Current Population
Survey and are subject to sampling variability, as well as survey
design flaws, respondent classification errors, and data
processing mistakes. The Census Bureau has taken steps to
minimize errors, and analytical statements have been tested and
meet statistical standards.  However, because of methodological
differences, use caution when comparing these data with data from
other sources. 

NOTES
1. The Gini index ranges from 0.0, when every family (household)
has the same income, to 1.0, when one family (household) has all
the income and is therefore one way to measure how far a given
income distribution is from equality.
2. Part of the increase from 1992 to 1994 is due to changes in
survey methodology; see below.
3. Computer-assisted personal interviewing (CAPI) was introduced
in January 1994 to the Current Population Survey.  As part of the
March 1994 supplement, households were permitted to report up to
$1 million in earnings, up from $300,000, and parallel increases
were made in the reporting limits for selected other income
sources.  Both of these changes affected the data. Analysis of
the 1993 statistics suggests that the increase in the maximum
amounts that could be reported accounts for about 1.8 percentage
points or about one-third of the 1992-1993 increase of 5.2
percentage points.  The contribution of the change to CAPI to the
increase in measured inequality cannot be determined, but may
bring the share of survey methods-related changes in inequality
to over one-half of the 5.2 percentage points.  See Paul
Ryscavage, "A Surge in Growing Income Inequality?", Monthly Labor
Review, August 1995.
4. See U.S. Bureau of the Census, Income, Poverty and Valuation
of Noncash Benefits:  1993, Current Population Reports P60-188,
Washington DC:  U.S. Government Printing Office, February 1995, 
and Ryscavage, op. cit. for a discussion of the 1993 statistics. 
The Gini index of inequality did not change significantly between
1993 and 1994.
5. All dollar amounts are in 1994 dollars and all percentage
increases are corrected for inflation, as measured by the
experimental Consumer Price Index for Urban Consumers.  (The
experimental index uses the official methodology adopted in 1983
by the Bureau of Labor Statistics as applied to the 1968-1982
period; see U.S. Bureau of the Census, op. cit.,  Appendix A.)
6. The respective shares of the middle 60 percent and the bottom
20 percent were 53.0 and 4.2 percent in 1968, down to 49.3 and 3.8 percent in 1992 and
47.3 and 3.6 percent in 1994.
7. Not significantly different from the 1994 ratio.
8. Not significantly different from the 1968-1994 percentage
change.
9. Poverty is defined only for families and unrelated
individuals, not for households.
10. See U.S. Bureau of the Census, op. cit., and U.S. Bureau of
the Census, Income, Poverty, and The Valuation of Noncash
Benefits:  1994, Current Population Reports P60-189, forthcoming.
11. There was no significant difference between the percentage
changes in the Gini index measured using the official income
definition and a comprehensive measure including all income
sources except imputed rent to owner-occupied dwellings.
12. This section is based on Paul Ryscavage and Peter Henle,
"Earnings Inequality Accelerates in the 1980's", Monthly Labor
Review, December 1990; Sheldon Danziger and Peter Gottschalk
(eds.) Uneven Tides:  Rising Inequality in America, New York: 
Russell Sage Foundation, 1993; Lynn A. Karoly and Gary Burtless,
"Demographic Change, Rising Earnings Inequality, and the
Distribution of Personal Well-Being, 1959-89," Demography, v. 32,
no. 3 (August 1995), 379-405; U.S. Council of Economic Advisors,
Economic Report of the President, Washington DC: U.S. Government
Printing Office, February 1992, Chapter 4; and U.S. Council of
Economic Advisors, Economic Report of the President, Washington
DC: U.S. Government Printing Office, February 1995, Chapter 5.

Go to Income Inequality Go to Income Statistics

Created: August 3, 2000 Last Revised: September 21, 2009