Washington, DC 20002
Iris J. Lav
Board of Directors
David de Ferranti, Chair
The World Bank
John R. Kramer, Vice Chair
Tulane Law School
Henry J. Aaron
University of Texas
Barbara B. Blum
Marian Wright Edelman
Children’s Defense Fund
James O. Gibson
Center for the Study of Social Policy
Beatrix Hamburg, M.D.
Cornell Medical College
Hill and Knowlton
Richard P. Nathan
Nelson A Rockefeller
Institute of Government
Johns Hopkins University
Chairman, The Price Company (Retired)
Robert D. Reischauer
Juan Sepulveda, Jr.
The Common Experience/
William Julius Wilson
POVERTY INCREASES AND
MEDIAN INCOME DECLINES
FOR SECOND CONSECUTIVE YEAR
Ranks of the Poor Increase by 3 Million Since 2000
Census data released today show that poverty
increased and median household income fell in 2002 for the second
consecutive year. The number of poor people increased by 1.7 million to
34.6 million; the poverty rate rose from 11.7 percent to 12.1 percent;
and median household income fell by $500, or 1.1 percent, to $42,409.
There were 3 million more poor people in 2002 than in 2000, the last year
before unemployment began to rise.
In addition, those who
were poor became poorer on average. The poverty rate — the percentage of
people with incomes below the poverty line — was lower in 2002 than in
most other years of the 1980s and 1990s, although higher than in most
years of the 1970s. But the basic measure of the depth, or severity, of
poverty — the average amount by which the incomes of those who are poor
fall below the poverty line — was greater in 2002 than in any other
year on record, with these data going back to 1979.
The rise in poverty and decline in median
income primarily reflect the increase in unemployment in 2002. The
unemployment rate averaged 5.8 percent in 2002, up markedly from 4.7
percent in 2001 and 4.0 percent in 2000.
“Misplaced priorities by Congress and the
President are making the increase in poverty larger than it needs to be,”
noted Center executive director Robert Greenstein. “The temporary
federal program Congress set up to help the long-term unemployed is
significantly weaker than the comparable program established in the
recession of the early 1990s. That’s an important reason why the
number of workers who ran out of those federal unemployment benefits
without finding work was twice as big in 2002 as at a similar point in
the last downturn.
“Also, Congress and the President chose to
exclude low-income working families from the increased child tax credit
benefits that went to better-off families this summer,” Greenstein
added. “Yet this year’s tax legislation will give people earning $1
million or more an average tax cut of $93,000.”
Greenstein also noted that federal fiscal
relief to states has not been adequate and that the majority of states,
facing budget shortfalls, have cut child care assistance for low-income
working families, making it more difficult for those families to remain
Further Increases in Poverty and Declines in
Income Likely in 2003
So far in 2003, unemployment has been higher
than it was in 2002. In addition, the number of long-term unemployed —
those out of work more than half a year — has increased dramatically,
from 650,000 a month in 2000, to 810,000 in 2001, 1.55 million in 2002,
and an average of 1.89 million so far in 2003. Adding to the
deterioration of circumstances facing poor households, average hourly
wages for low-paid workers also have fallen in 2003, and a number of
states that faced budget deficits have instituted budget cuts in basic
assistance programs in 2003. These developments suggest that poverty may
increase and incomes decline for a third straight year in 2003.
Increases in Poverty Rates
Among the findings from the data the Census
Bureau released today are the following:
in poverty rates and in the number of people in poverty were widespread
across various population groups, but in a number of cases, the increases
in poverty rates were not large enough to pass statistical significance
tests. For some groups, the increase in the poverty rate was not
statistically significant, but the increase in the number of
people who are poor was significant.
released today show that poverty rose among the four basic measures of
child poverty, with the increase in three measures being large enough to
be statistically significant. The increase in one measure — the
poverty rate for all children under 18 — was not large enough to be
The number of such children who were
poor rose 400,000, an increase that was statistically significant.
The increase in the poverty rate for all children under 18
— from 16.3 percent to 16.7 percent — fell just short of being large
enough to pass the statistical significance test.
In addition, the increase in the poverty rate
among related children under 18 — another basic measure of child
poverty — was significant. This rate increased from 15.8 percent to 16.3
in poverty were largest among blacks. The black poverty rate rose from
22.7 percent in 2001 to 24 percent in 2002, and the number of blacks who
were poor increased by 500,000 or 700,000, depending on which
definitional category of blacks is used.
of poor Hispanics increased by 600,000, a statistically significant
increase. The Hispanic poverty rate increased from 21.4 percent in 2001
to 21.8 percent in 2002, but this increase was not statistically
rate for married families also increased significantly, although it
remains far below the poverty rate for female-headed families.
two-thirds of all poor families with children included a worker in 2002.
These individuals typically work a significant amount. In 2001 (data for
2002 are not yet available), workers in poor families with children
worked an average of 44 weeks per year, and an average of 41 hours during
the weeks in which they were employed. (The average exceeded 40 hours a
week because some poor families have more than one worker.)
poverty rates among alternative measures of poverty were mixed. Under a
set of alternative measures that take into account the value of various
non-cash benefits (such as food stamps), the Earned Income
Credit, and taxes, the poverty rate increased. Under a second set
of new alternative measures, the increases in poverty generally were not
statistically significant. This set of alternative measures,
however, did show higher levels of poverty than the standard measure.
Increases in the Depth of Poverty
In 2002, the average
amount by which people who were poor fell below the poverty line was
greater than in any other year since 1979, the first year for which these
data are available. The average amount by which people who were poor
fell below the poverty line was $2,813 in 2002. (These data use a
measure of income that counts non-cash benefits such as food stamps and
housing subsidies, as well as the Earned Income Tax Credit, as income,
and subtracts income and payroll taxes.)
amount by which people who were poor fell below the poverty line has
increased sharply since 1996, and was 23 percent larger in 2002 than in
1996, after adjusting for inflation.
The number of
people who live in severe poverty — that is, who have incomes below half of
the poverty line — increased by 600,000, to 14.1 million. This increase was
statistically significant. The percentage of people living below half of the
poverty line also increased, but that increase was not statistically
Changes in Median Income
household income fell by $500 between 2001 and 2002, or 1.1 percent. The
drop in median income was concentrated among minorities.
households, median income fell 2.5 or 3 percent (depending on the
category used), a drop of $762 or $913.
Hispanics, median income fell 2.9 percent, or $996.
Asians, median income fell between $1,862 and $2,470, depending on which
definitional category of Asian is used. Among three different
definitions of the Asian category, the decline in median income was
significant. Among a fourth definitional category, the decline in
median income was not significant.
Hardship Among the Poor
In addition to its
normal data on poverty and income, the Census Bureau also released a
report today on a limited set of measures of consumption and
expenditures. This report — which is mostly limited to data from 1998
but also includes some energy consumption data from 2001 — shows that
the living conditions of the poor were notably worse than those of
higher-income groups in 1998. For example, the report shows that poor
families were almost 2.5 times as likely as higher-income families to
have been unable to pay their rent or mortgage at some point in the 12
months before being surveyed and 2.5 times as likely not to have gone
to a doctor when they needed one.
According to Census,
the report is intended to expand the understanding of the nature of
poverty in the United States by complementing the official income-based
measures of poverty. Unfortunately, the report does not include
discussion of other more recent and more important indicators of
material well-being, such as food insecurity and housing and child care
affordability, that would provide a fuller picture of the nature of
poverty in the United States.
data from the Centers for Disease Control and Prevention show that in
2002, some 36.8 percent of poor adults aged 18 – 64 were uninsured, as
opposed to 10.8 percent of adults aged 18 – 64 with incomes above twice
the poverty line. Among children, 14.5 percent of poor children were
uninsured, compared to 5.3 percent of children with incomes above twice
the poverty line.
44.5 percent of poor households with children experienced either food
insecurity or hunger at some point during the year. (Households that
experienced food insecurity were unable to acquire, or were uncertain
of having, enough food to meet the basic needs of all household
members, because they lacked sufficient income and resources.)
study by the Harvard Joint Center for Housing Studies found that in
2001, the number of households in the poorest fifth of the population
(those with incomes below $17,000 in 2001) that paid more than half of
their income for housing rose to
10.6 million. (This
number stood at 10.1 million in 1999.)
Income inequality remained unchanged under
the basic Census measure of inequality and narrowed slightly under some
noting that the Census information on income trends is incomplete and
problematic in one major respect. Limitations in its data prevent the
Census Bureau from accurately capturing income changes at the very top of
the income spectrum. The Census data miss a large share of the income of
very-high-income households; for this reason, the Census Bureau does not
publish data on the incomes of the top one percent. Other data series,
most notably a Congressional
Office series that supplements Census data with IRS data, capture gains
and losses among the top one percent of the population much more
The latest CBO data, which were released a
few weeks ago and cover years from 1979 to 2000, show that the average
after-tax income of the top one percent of the population rose by
$576,000 — or 201 percent — between 1979 and 2000, after adjusting for
inflation, while the average income of the middle fifth of households
rose $5,500, or 15 percent. The average income of the bottom fifth rose
$1,100, or 9 percent, over the 21-year period.
and Policy Priorities released an analysis, “The
New, Definitive CBO Data on Income and Tax Trends,” on
September 23 of the CBO data and other recent data on income trends.
These data indicate that the top one percent of the population received a
larger share of the national income in 2000 than in any year since 1929.
The CBO data do not shed light on changes in income inequality since
in Safety Net Contribute to Poverty
While the increase in poverty primarily
reflects developments in the economy, weaknesses in the safety net —
particularly in the temporary federal unemployment benefits program —
also contributed to it. In 2002, some 2.2 million workers exhausted all
of their unemployment benefits before finding a job and consequently
received neither a paycheck nor an unemployment check for a period of
time. Many of these individuals and their families are likely to have
fallen into poverty. One reason their unemployment benefits ran out
before they were able to find work is that the temporary federal
unemployment benefits program that Congress established last year to aid
long-term unemployed workers and their families during the economic slump
is much more limited and offers significantly fewer weeks of unemployment
assistance than the comparable program that Congress established during
the economic slump of the early 1990s.
addition, data from the U. S. Department of
and Human Services show a steep decline in the proportion of
very-low-income families with children that receive cash welfare
benefits. In the mid-1990s, about eight in every ten families poor
enough to qualify for cash welfare benefits received them. (Families
usually have to be well below the poverty line to qualify for cash aid.)
By 2000, the last year for which HHS has provided these data, only five
of every ten families this poor received these benefits. Given that
welfare caseloads continued to decline as poverty increased, data for
2001 and 2002 will show further declines. Data released by the Census
Bureau today confirm that cash welfare benefits did less to reduce
poverty in 2002 than in any year since at least 1989.
The ranges in poverty rates and the numbers of people in poverty
presented in this analysis reflect the different ways the Census Bureau
defines various racial and ethnic groups.
# # # #
The Center on Budget and Policy Priorities
is a nonprofit, nonpartisan research organization and policy institute that
conducts research and analysis on a range of government policies and
programs. It is supported primarily by foundation grants.