Revised, October 8, 1998

Poverty Rates Fall, but Remain High for a Period
With Such Low Unemployment

Gains Strongest for Blacks and Hispanics

Click here for the PDF
of this report.

Strong economic growth and low unemployment reduced poverty and raised incomes in 1997, with especially strong gains occurring among minorities. A number of years of growth have returned the poverty rate and median household income to the levels at which they stood in 1989, the last year before the recession of the early 1990s. The Census data show that 13.3 percent of Americans lived in poverty in 1997, down from 13.7 percent in 1996.

The poverty rate remained high, however, for a year in which the unemployment rate averaged 4.9 percent, its lowest level in 24 years. The poverty rate in 1997 was at about the same level as it was in 1987 through 1989, years in which unemployment averaged between 5.3 percent and 6.2 percent. The 1997 poverty rate was substantially above the poverty rates for every year of the 1970s, even though the unemployment rate was close to or above six percent for more than half of the years of the 1970s. More than one of every eight people in the United States continue to live in poverty.

Depth of Poverty Grows

On average, poor families became poorer in 1997. The average amount by which families that are poor fall below the poverty line increased $200, the Census figures show, from $6,395 in 1996 to $6,602 in 1997. (These figures are both expressed in 1997 dollars.)

This increase in the depth of poverty for the average poor family appears to be related to a weakening of safety net programs in 1997; the decline in the number of families receiving assistance was much greater than the decline in the number of families that are poor. The proportion of poor families with children that receive basic cash assistance that can lessen the severity of their poverty has decreased. The proportion of such families receiving food stamp assistance also has fallen, although food stamp receipt is not reflected in the Census Bureau's standard measures of the incidence and depth of poverty.

The lack of greater progress in reducing poverty also reflects a broader, long-term trend — income growth over the past quarter century has been unevenly spread. Average household income in 1997 exceeded average income in 1989 only for the top two-fifths of the population. For the rest of the population, average income was at or below the 1989 level, after adjusting for inflation. Income gains have been largest for the five percent of households with the highest incomes.

Income Inequality Tied for All-time High;
Income Gains Largest for High-Income Households

The Census data show that while the increase in income inequity between 1996 and 1997 was not statistically significant, the increase in inequity over the course of the current business cycle — the period from 1989 to 1997 — has been sizeable. Every income group except those at the top of the income scale received a significantly smaller share of national pre-tax income in 1997 than in 1989. The highest-income groups received larger shares.

In 1997, the 20 percent of households with the largest incomes received nearly half — 49.4 percent — of the national income, while the other 80 percent of the population divided the other half of the national income. Although the 49.4 percent of the national income going to the top fifth of households did not represent a statistically significant change from the 49 percent figure for 1996, it tied for the largest share of national income the top fifth has received in any year on record. It also significantly surpassed the share of national income the top fifth of households received in 1989. Each one percent of national income equals $51 billion.

The top five percent of households received 21.7 percent of the national income in 1997, also tied for the highest level on record. Meanwhile, the shares of income that each of the bottom four fifths of the population received in 1997 were at or below their shares for all other years on record. The share going to the middle three-fifths of the population combined was at a record low.

9-24-98pov-f1.jpg (12365 bytes)In addition, the incomes of households at the top of the income spectrum grew more between 1996 and 1997 than the incomes of households at the middle or the bottom.

These data are for households. Household income data are the data the Census Bureau emphasizes in its reports and briefings. Household income data comprise the most comprehensive data on income trends among the U.S. population because these data include virtually everyone in the United States, whether they are, for example, individuals living alone, families with children, or other types of households. The Census Bureau also releases information on family incomes. The term "families" refers to related people who reside together; it does not include individuals who live alone or with other individuals to whom they are not related.

If only families are examined, income growth from 1996 to 1997 is more broad-based than if all households are examined. The bottom four fifths of families experienced average income gains between 2.6 percent and 3.5 percent in 1997, after adjusting for inflation. The average income of the top fifth rose 4.5 percent.

The income story remains largely the same, however, whether families or households are examined — either way, those at the top of the income scale fared better than others. For example, the average income of the top five percent of families grew 5.7 percent, or $12,680 last year, far more than the income growth in either percentage or dollar terms for any other group of families. The family income data also indicate that income inequality among families remains at an all-time high and that between 1989 and 1997, only upper-income families experienced substantial income gains.(1)

Poverty Rates Decline, Incomes Rise for Blacks and Hispanics

The brightest parts of the new Census report are the strong gains registered by Blacks and Hispanics. Poverty dropped for both groups in 1997, with the Black poverty rate falling from 28.4 percent in 1996 to 26.5 percent in 1997, an all-time low. The Hispanic poverty rate also dropped substantially, declining from 29.4 percent in 1996 to 27.1 percent last year.

Median household income also climbed for both groups, rising 4.3 percent, or a little over $1,000, for Blacks and 4.5 percent, or $1,150 for Hispanics. These strong gains suggest that minorities — the groups with the highest unemployment rates — benefit disproportionately when the overall unemployment rate drops below five percent and employers reach increasingly into low-income and minority populations for new hires. These data also suggest that keeping unemployment at its current low levels is especially important to making continued progress in raising incomes and reducing poverty among Blacks and Hispanics.

Gains were especially strong in 1997 for Black and Hispanic female-headed families. Median income for Black female-headed families climbed a robust eight percent in 1997, after adjusting for inflation, while for Hispanic female-headed families, median income surged 10.3 percent. In addition, the poverty rate for black female-headed families fell from 43.7 percent in 1996 to 39.8 percent in 1997. (The poverty rate for Hispanic female-headed families also appeared to decline, but the change was not statistically significant. The poverty rate for non-Hispanic white female-headed households appeared to rise, but here, too, the change was not significant.)

Looking at changes over the course of the current business cycle — that is, since 1989 — the new Census data show that Blacks have experienced larger percentage decreases in poverty and increases in income than either Hispanics or non-Hispanic whites. The median income of Black households climbed 6.2 percent between 1989 and 1997, after adjusting for inflation. By contrast, for non-Hispanic whites, median household income was essentially unchanged (it rose one percent, but the change was not statistically significant), while median income for Hispanics fell 5.5 percent over this period.

Similarly, while the Hispanic poverty rate and poverty rate for non-Hispanic whites were both at about the same level in 1997 as they had been in 1989, the overall Black poverty rate dropped from 30.8 percent in 1989 to the record low of 26.5 percent in 1997. Poverty rates for Black children, Black female-headed families and Black married-couple families also have declined substantially since 1989 and all set or remained at record lows in 1997.

Even so, poverty rates for Blacks as well as for Hispanics continue to be much higher than poverty rates for non-Hispanic whites. Some 8.6 percent of non-Hispanic whites were poor in 1997, a rate less than one-third that for Blacks and Hispanics. Similarly, median household income in 1997 was more than 50 percent higher for non-Hispanic whites than for both Blacks and Hispanics.

Poverty at All-time Low for Elderly, but not for Working-age Adults

In 1997, the poverty rate stood at or below its 1989 level for the youngest and oldest age groups, but not for working-age adults. Children continued to have the highest poverty rate of any age group at 19.9 percent, although this poverty rate has returned to its 1989 level.

The Census Bureau reported that children constituted 40 percent of all people who were poor in 1997 although they made up only 26 percent of the population. The Census data show an extremely high poverty rate for one group of children — those under age six who live in female-headed families with no spouse present. Their poverty rate was 59 percent last year.

Among working age adults 18 to 64, the 1997 poverty rate of 10.9 percent remained significantly above the 1989 poverty rate of 10.4 percent. This likely reflects the long-term downward trend in wages for lower-paying jobs.

For the elderly, the 1997 poverty rate of 10.5 percent ties for the lowest rate on record. Almost all elderly individuals receive Social Security benefits. Since Social Security benefits are tied to average wages earned during a worker's career and each new group of retirees has earned higher average wages than previous groups of retirees, Social Security benefits have been steadily rising. This pushes down elderly poverty rates. The poverty rate for elderly people of 10.5 percent in 1997 compares to a 24.5 percent rate in 1970 and a rate of approximately 35 percent in 1960. Census data show that in 1997, the elderly poverty rate would be close to 50 percent in the absence of Social Security.

The poverty rate also was low, in comparison to other years, for female-headed families with children.(2) The poverty rate for these families was as low or lower in 1997 as in any other year in the past 40, except for 1979. Nevertheless, the poverty rate for this group remains quite high, at 41 percent, which is nearly six times the poverty rate for married-couple families with children. (The Census data also show that the proportion of families with children that are female-headed has stopped rising; it has remained essentially unchanged for the past six years. This percentage remains above its levels in previous decades, however, which is one of the reasons that the overall poverty rate continues to be higher than it was in the 1970s.)

One group for which the poverty rate trends are less favorable consists of individuals of all ages who live alone or with others to whom they are not related. Some 20.8 percent of these people, termed "unrelated individuals" by the Census Bureau, were poor in 1997, a rate higher than the 19.3 percent poverty rate for this group in 1989. Low-income single individuals are likely to have been affected adversely by the erosion in wages paid for low-skilled work. In addition, over the past decade, assistance programs have shrunk for low-income single individuals who are not elderly or disabled (or have disabilities not covered under the Supplemental Security Income program). Many states have eliminated state cash assistance (often referred to as general assistance) for these individuals, while changes in federal law that took effect in 1997 ended SSI benefits and Social Security disability benefits for most disabled individuals whose disability stems from a current alcohol or drug addiction.

Decline in Number of People Receiving Assistance Outstrips Decline in Poverty

The new Census data show that in recent years, the declines in the number of people receiving basic assistance targeted on needy families have far outstripped the decline in the number of people who are poor. This suggests that sizable numbers of people who are poor are receiving less aid from means-tested assistance programs than in the past. The Census data also indicate that this decrease in assistance is one of the reasons that poverty rates have not declined more in the past few years amidst a stunning performance by the economy, and these data help explain why the average poor family has become poorer.

From 1995 to 1997, the number of poor people in female-headed families with children fell 4.3 percent. But the number of people receiving welfare assistance for poor families with children dropped 22.6 percent, more than five times as large a percentage. (Most of the families that receive welfare assistance are female-headed families with children.) Just between 1996 and 1997, the number of people receiving welfare assistance fell by nine times as large a percentage as the number of poor people in female-headed families with children; the number of poor people in such families fell 1.7 percent in 1997, while the number receiving welfare assistance in an average month fell 15.7 percent.

Figure 3Similar developments mark the food stamp program. Between 1995 and 1996, the number of poor people remained statistically unchanged, but the number of people receiving food stamps in an average month fell 1.2 million, or nearly five percent. This trend accelerated in 1997. From 1996 to 1997, the number of poor people fell 955,000, or 2.6 percent, while the number receiving food stamps plunged an additional 3.1 million, or more than 12 percent. During the two-year period from 1995 to 1997, the decline in the number of people receiving food stamps — 4.4 million — was five times greater than the decline in the number of people living in poverty.

The food stamp figures are especially noteworthy because the income limit for food stamps is slightly above the poverty line; as a result, families moving from public assistance to low-wage work that leaves them in poverty do not lose eligibility for food stamps. These data indicate that the reductions in the number of households receiving food stamps have exceeded reductions in need and that the proportion of poor people receiving basic food assistance to help them secure an adequate diet has declined.

Center director Robert Greenstein noted that anecdotal evidence suggests that many families no longer receiving cash assistance are not receiving food stamps either. "Practices in some states to dissuade families from receiving welfare aid may be having the unintended effect of discouraging many working poor families from receiving food stamps even though they remain eligible for this assistance," Greenstein said.

The effects of these declines in public assistance and food stamp receipt show up in new, unpublished Census data on poverty rates under alternative measures of poverty that count non-cash benefits such as food stamps as income. These Census data show that the diminished coverage of means-tested programs such as food stamps and cash welfare assistance resulted in 700,000 fewer children being lifted out of poverty by means-tested programs in 1997 than in 1995.(3)

In addition, the data show that the percentage of poor children receiving cash welfare assistance, as well as the percentage receiving food stamps, has declined in the past two years. For example, in 1995, some 61.5 percent of poor children received cash assistance, while in 1997, some 53.3 percent did.(4)

At the same time, the Census Bureau's alternative measures of poverty show strong positive effects from the expansions of the Earned Income Tax Credit enacted in 1990 and 1993. Due to the EITC expansions, federal tax policy lifted 400,000 more children out of poverty in 1997 than in 1995 and 1.8 million more children out of poverty last year than in 1989.

The new Census data are likely to lead to speculation about the effects that changes in the welfare system are having on poverty. Full understanding of those effects must await more time, data, and study. Nevertheless, the new Census data suggest the changes in the system are having different effects on different families. The combination of the welfare policy changes, the increased number of job opportunities created by a growing economy, the rise in the minimum wage and expansions of the Earned Income Tax Credit (which have boosted the compensation for low-paid work), and the extension of Medicaid to more children in low-income working families appear to have spurred significant increases in employment and earnings among female-headed families with children. These increases in employment and earnings appear to have raised incomes and reduced poverty among a substantial number of female-headed families. At the same time, welfare system changes appear to have led to losses of cash and food assistance by a sizeable number of other families, many of whom have not offset these losses of benefits with increases in earnings. Such families have become poorer as a result.

Employment and Minimum Wage Increases Reduce Poverty and Raise Incomes

The new Census data also show that although median weekly earnings of full-time workers rose in 1997, these earnings remain below their level for most of the period since 1970. These long-term trends are particularly unfavorable for male workers. For men working full-time year-round, wages remain $1,500 below 1989 levels, after adjusting for inflation. For women working full-time year-round, wages were $750 higher in 1997 then in 1989.

Median household income is higher than in any year except 1989 despite these less favorable trends in wages because of large increases in the number of people working and the number of hours worked. Labor Department data show that the proportion of adults who worked in 1997 reached the highest level ever recorded. In addition, data from the Economic Policy Institute (EPI) indicate that the amount of time that families work increased significantly between 1989 and 1997, with some of this increase occurring in 1997.

Increases in work hours were particularly important in 1997 in reducing poverty and raising income among low-income families. The EPI data show that among the bottom fifth of families, the average number of hours worked per family rose more than five percent between 1996 and 1997.

One other factor appears to have helped raise wages and family incomes for lower-paid workers in 1997 — the increase in the minimum wage from $4.25 an hour to $5.15 an hour, instituted in two stages in October 1996 and September 1997. A recent EPI study found that the minimum wage increase has contributed to faster growth in the wages of low-wage workers.

End Notes:

1. The Congressional Budget Office uses a different method for measuring incomes for each fifth of the population. The CBO method, which includes an adjustment for household size, is generally regarded by analysts as a better way of measuring changes in income for different income groups than measuring changes for either households or families as the Census Bureau defines those terms. Applying the CBO method to the new Census data shows that in 1997, average income rose by a smaller percentage (1.9 percent) for the bottom fifth of the population than for any other income group and rose by the largest percentage for those at the top of the income scale.

2. The data on female-headed families cited on page 4 of this analysis are for all female-headed families, including those without children. Poverty rates for all female-headed families are lower than poverty rates for female-headed families with children.

3. Among children who were poor before receipt of means-tested benefits, the proportion lifted from poverty by these benefits fell substantially between 1995 and 1997. If means-tested programs had lifted from poverty in 1997 the same proportion of children who otherwise would be poor as these programs lifted from poverty in 1995, some 700,000 fewer children would have been poor last year. (These figures refer to the alternative measures of poverty that count non-cash benefits such as food stamps as income).

4. This percentage might have been expected to decline if poor families with children became less poor, on average, and more of them secured earnings that lifted them above cash assistance income limits in their states while still leaving them below the poverty line. The Census data suggest, however, that this did not occur; to the contrary, poor families with children became poorer, on average, in 1997. Furthermore, the Census data indicate that the decline in receipt of cash assistance by poor families with children was greatest among very poor families with no earnings

The figures cited here on the percentages of poor children receiving cash assistance in 1995 and 1997 reflect average monthly participation by children in state cash assistance programs funded through the Temporary Assistance for Needy Families block grant, as reported by states to the U.S. Department of Health and Human Services, measured as a percentage of the number of children the Census Bureau reports as having been below the poverty line that year.

Get the PDF version of the report!  Click here
for PDF version
of this report

Get the program.  It's fast and FREE!!  
Download the Adobe Acrobat Reader
if you do not already have one.

Additional related reports: