December 20, 2004

by Arloc Sherman

PDF of this analysis
Press Release:  HTM | PDF

View Related Reports

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.

Two recent, little noticed government reports spotlight the strained living conditions of millions of U.S. families.  One report shows that the number of Americans threatened by hunger inched upward for four straight years from 1999 to 2003, with one in eight people falling in this category in 2003.  The other report examines a long list of hardships that are tied to poverty status.  Our own analysis, combining measures used in that report, finds that 8 million poor and near-poor children live in households that have experienced recent hunger, severe crowding, or problems paying bills so serious that phone or utility service was shut off.

The first report, from the U.S. Department of Agriculture, found that the number of American households ever experiencing hunger during the year rose by nearly one-quarter from 1999 to 2003.  Further, the proportion that was either hungry or threatened by hunger — a group the government terms “food insecure” — has edged higher for four straight years.  Nearly 13 million households containing 36 million Americans were food insecure in 2003.

The second report, from the U.S. Department of Health and Human Services, took an unusual, more comprehensive look at hardships affecting households with children.  Besides hunger and food insecurity, HHS examined difficulties paying bills and meeting other basic needs, as well as housing safety problems (such as broken windows and exposed wires) and severe crowding.  (See box.)  A new Center analysis, using the same raw survey data and hardship measures used in the HHS report, finds that:

These cross-cutting hardship findings are based on government survey data gathered in 1998.  Although the data are thus six years old, other data suggest that the situation for low-income families today is likely to be similar to that in 1998.  The number of poor Americans was higher in 2003 than in 1998.  Moreover, data from the Urban Institute show that rates of food hardship and housing hardship among low-income parents were virtually unchanged from 1997 through at least 2002.

Under both sets of findings, from USDA and from the Center analysis of HHS measures, the levels of hardship shown are almost certainly far lower than they would otherwise be because of federal housing, nutrition, and other basic assistance programs.


Further USDA Findings on Hunger and Food Insecurity

Hunger and problems affording adequate food affect millions of American households each year and, after declining in the late 1990s, have grown somewhat more common during recent years.

More than 3.9 million households, containing 9.6 million Americans, experienced “food insecurity with hunger” at least once during the year in 2003, according to the USDA.[2]   The number of Americans in households experiencing hunger rose from 7.8 million in 1999 to 9.6 million in 2003, an increase of 24 percent.

The government considers hunger a severe form of the broader problem of “food insecurity.” More than 36 million Americans lived in households that were “food insecure,” with or without hunger, in 2003; this means that at “some time during the year, these households were uncertain of having, or unable to acquire, enough food for all their members because they had insufficient money or other resources.”

Hunger is usually a recurring problem.  “On average, households that are food insecure with hunger at some time during the year experience this condition in 8 or 9 months during the year,” notes USDA.  Nearly three out of every four households that were hungry during the past 12 months were also hungry during the previous 30 days.

The percentage of households experiencing hunger, like the overall rate of food insecurity and the poverty rate, declined slightly in the late 1990s but increased during the subsequent period of labor market weakness.  The hunger rate declined from 3.7 percent in 1998 to 3.0 percent in 1999, but increased in small steps to 3.5 percent by 2003.  Among households below the poverty line, 13 percent sometimes experienced hunger in 2003, compared with 14 percent in 1998.


Further Findings Using the Cross-Cutting HHS Hardship Measures

The HHS report brings together data on an unusually wide range of hardships, from hunger to crowded housing.[3]  HHS examined 14 distinct types of hardships related to “basic needs and food security” and “housing safety and overcrowding.”  All 14 hardships were closely related to poverty status, occurring two to eight times more often among poor families as among non-poor families with incomes above twice the federal poverty line, data in the report showed.

The Center’s own computer analysis of these data extends the HHS findings in two ways: first, by looking cumulatively across the measures to reveal how many families experience combinations of different hardships, and second, by looking at a measure of poverty that is based on annual rather than monthly income.[4]

The Center found that one in three poor households with children (34 percent) experienced at least one of the following hardships:  hunger in the last four months; lived in severely crowded housing (with 1.5 or more people per room); or had their phone or utilities shut off in the last 12 months.  These households contained approximately 4 million dependent children under 18.  (See Figure 1.)  (As explained below, the definition of crowding used by the HHS study is more extreme than the standard federal definition and understates the prevalence of crowding and hardship overall.)

The Center also found considerable hardship among those living a little above the poverty line:  1.6 million near-poor households with children — with annual income between one and two times the poverty line — experienced hunger, severe overcrowding, or phone or utility shutoffs.  These households contained close to 4 million children and made up 19 percent of near-poor households with children.

Hardship Measures Examined by HHS

“Basic Needs and Food Security” Hardships
(ever experienced in the last 12 months, unless noted)

1.   Did not pay rent or mortgage.
1b. Evicted for failure to pay rent or mortgage.
   Did not pay oil, gas, or electricity bill.
2b. Lost gas, oil, or electricity for failure to pay.
   Telephone disconnected for failure to pay.
   Needed to see doctor or go to hospital but did not.
   Needed to see dentist but did not.
6.   Food insecure (in the last 4 months)
6b. Food insecure with hunger (in the last 4 months)

“Housing Safety and Overcrowding” Hardships
(ongoing at the time of the survey)

  1. Problem with pests
  2. Leaking roof or ceiling
  3. Broken windows
  4. Exposed wires
  5. Nonworking plumbing
  6. Holes in wall or ceiling
  7. Holes in floor
  8. Overcrowding (more than 1.5 persons per room)

HHS also examined the presence of durable goods (a refrigerator and a gas or electric stove).  At least 98 percent of poor families had these items.  Durable goods are not included the Center’s analysis.

Items 1b, 2b and 6b above are not counted separately in the Center’s count of multiple hardships.  See footnote 5

In total, 3.2 million poor and near-poor households, with nearly 8 million children, experienced hunger, severe crowding, or phone or utility shutoffs, the Center found.  And 3.6 million poor and near-poor households experienced three or more of the 14 HHS hardships.

The relatively high rates of hardship among near-poor families suggest that having income slightly above the federal poverty line is often not enough to protect families from hardship.  The federal poverty line (which was $18,810 a year for a family of four in 2003) may be a low standard for low-income working families, who comprise the majority of poor and near-poor families and who may face high child care costs and other work expenses. 

The Center also found that experiencing multiple hardships was common among families in poverty.[5] More than half of families below the poverty line, containing nearly 6 million children, had two or more of 14 distinct hardships examined by HHS.  (See Figure 2.)[6]

Although some higher-income families, with incomes above twice the poverty line, report hardships, the hardships reported by families in poverty are notable in three ways.  First, poor families are much more likely to experience any hardship than the non-poor.  Second, and relatedly, they are much more likely to experience multiple hardships.  Most poor families who experience any hardship experience a total of three or more hardships.  Third, poor families’ hardships are more severe:  poor families that report experiencing a hardship in a particular area are more likely than non-poor families to report experiencing a severe hardship in that area.  For example, among families that report missing a utility payment, poor families are more likely to have their utilities shut off as a result.


HHS Data Likely Understate Hardship

The Center analysis uses the same data source used by HHS: data gathered in the summer of 1998 by the Census Bureau's Survey of Income and Program Participation (SIPP), which tracked a representative sample of 40,000 American households for 4 years starting in 1996.

The data used by HHS and the Center analysis are conservative in key respects.  For example, the definition of crowding used by HHS (more than 1.5 persons per room) differs from the standard government definition and thus understates the extent of crowding in particular and hardship in general.  If crowding is defined as U.S. Department of Housing and Urban Development defines it (more than 1 person per room) the figures above would have shown that substantially more poor families — 46 percent rather than the 34 percent shown above —  experienced hunger, crowding, or phone or utility shutoffs.

It is also noteworthy that the HHS estimates of food insecurity differ from other government food insecurity figures in a number of ways, including the time frame in which the hardship occurs and the battery of questions that was asked.  For example, the HHS study relies on the SIPP survey, which only counts food insecurity experienced during the four months preceding the survey.  By contrast, most USDA reports (including the food insecurity report cited earlier) rely on data from a different government survey and focus chiefly on households who experience food security at any time during the last 12 months.  These USDA figures show levels of annual food insecurity well above the estimates here[AS1] .[AS2]   Because food insecurity is the most common type of hardship reported among poor families in the HHS report, these differences have a substantial effect on overall estimates of hardship.


USDA and Other Recent Data Suggest That These Cross-Cutting Findings Remain Relevant

While the cumulative hardship data used by HHS and the Center were gathered in 1998, and are thus six years old, several facts suggest that conditions of hardship among poor families have not changed very much.

Similarly, the Urban Institute found little difference in the prevalence of food and housing hardships among low-income families with children between 1997 and 2002.  Using their own independent survey data, the Urban Institute concluded that:

“Although the strong economy helped ease food hardship among low-income parents between 1997 and 1999, that gain was erased over the succeeding three years.  Housing hardship among all parents did not diminish between 1997 and 2002, even though overall family income improved somewhat.  Moreover, low-income single parents reported significant increases in housing hardship, particularly between 1999 and 2002.  These findings reflect the difficulties parents have in stretching their incomes to cover the food and housing needs of their families, especially when the economy is weak and housing costs are rising.” [7]

Because food and housing hardships comprise the bulk of the hardships reported in the 1998 hardship data, and because neither type of hardship has shown signs of major change, it seems unlikely that the overall hardship picture has improved much since the data were collected.

While hardship and poverty figures for 2004 are not yet available, there is little reason to expect dramatic improvement when the 2004 figures become available.  Early indicators for 2004 show that the employment rate has merely held steady (at 62.3 percent of all adults in the first 11 months of the year in both 2003 and 2004), while wages at the bottom of the economic ladder have not kept pace with inflation[AS3] .

Moreover, aspects of the safety net have weakened.  For example, in December 2003, Congress allowed the Temporary Emergency Unemployment Compensation program for long-term unemployed workers to expire.  Partly as a result, a record 3 million long-term jobless workers exhausted their regular unemployment benefits during the first ten months of 2004 before they found a job.[8]


Poverty and Hardship Would Be Higher Without Anti-Poverty Programs

Millions of Americans receive assistance meeting their basic needs through a range of public benefit programs.  These include social insurance programs like Social Security and unemployment insurance, income-tested programs like Temporary Assistance for Needy Families, housing and nutrition assistance, and tax credits for working families. 

Altogether, income and non-cash benefits from these programs lifted 28 million people above the poverty line in 2002, reducing by one-half the population that would have otherwise been poor.  The programs also provided vitally needed assistance in helping people meet their food, housing, and other basic needs.

The levels of hardship shown here would almost certainly be far higher were it not for these and other public antipoverty efforts.

End Notes:

[1] The author owes a special thanks to Ying Wang for extensive data preparation.

[2] Mark Nord, Margaret Andrews, and Steven Carlson, Household Food Security in the United States, 2003. U.S. Department of Agriculture: 2004.

[3] Measures of Material Hardship — Final Report. U.S. Department of Health and Human Services: 2004.

[4] In determining household poverty status, the HHS report looks at income during the particular month in which the Census Bureau’s hardship survey was administered.  The 12-month income measure used by the Center corresponds more closely with the official annual income definition used in other federal reports.  Using a 12-month measure also helps avoid a problem noted in the HHS report, namely, that monthly income may not reflect the income of the household at the time the reported hardship actually occurred, which may be up to 12 months prior to the survey.

[5] In counting the number of distinct hardships that households experience, the Center did not count three hardships considered by HHS (eviction, utility shutoffs, and hunger) because they represent extreme instances of other hardships already counted (namely, failure to pay rent, failure to pay utilities, and food insecurity). 

[6] The Center analysis finds that 72 percent of poor households with children faced one or more of the 14 distinct HHS hardships.  This total, however, should be regarded cautiously.  It includes one living condition, infestation by rodents or other pests, that is relatively common at all income levels (although it is most common in lower-income homes).  The total also includes some events that may not always reflect material deprivation, such as failure to obtain needed dental care, which the report notes could simply reflect “a general dislike of dental appointments.” Consequently, a substantial proportion (30 percent) of non-poor families with incomes above twice the poverty line reported one or more of the 14 HHS hardships.

Some higher-income families may also report hardships because they live in parts of the country with high housing costs that strain their budgets, or because they have experienced recent job loss, divorce, or other crises.

[7] Sandi Nelson, “Trends in Parents' Economic Hardship.” Urban Institute, March 18, 2004.

[8] Isaac Shapiro, “Number of Unemployed Who Have Gone Without Federal Benefits Hits Record 3 Million.” Center on Budget and Policy Priorities, October 2004.  Available online at