CHAPTER 5
FACTORS AFFECTING POVERTY AND WELFARE USE

Poverty is the largest single factor that drives people to apply for governmental assistance, commonly called "welfare." Many researchers agree that the major factors that create poverty are family size, family background, low educational achievement, unemployment, underemployment (for example, part-time workers who want to work full-time), low wages, and the prevailing economic conditions in the labor market.

Since 1974 children have been the poorest group in America, a status previously held by the elderly. The major reason for the decline in poverty rates among the elderly was social insurance benefits, nearly all of which come from Social Security. In 2002, 16.7 percent of all children in the United States lived in families with incomes below the poverty level. (See Figure 3.2 in Chapter 3.) Changes in household and family composition, particularly the increase in the number of single-parent families, have contributed to high poverty rates, especially the high rate of child poverty. According to the U.S. Census Bureau's Poverty in the United States: 2002 (Washington, DC, 2003), "Families with a female householder and no husband present made up half of all families in poverty."

ASSISTANCE FROM THE GOVERNMENT

Women and children are most affected by poverty. The two largest cash assistance programs for families with children have been Temporary Assistance for Needy Families (TANF), which replaced Aid to Families with Dependent Children (AFDC) under the 1996 welfare reform law, and the Earned Income Tax Credit (EITC). States currently provide cash assistance from their TANF block grants to families that meet the work requirements (mandatory after two years on assistance) and who have not exceeded the five-year limit for assistance. The EITC is available to needy working households, and most of it goes to families with children.

Other cash and noncash programs intended to help women, children, and the elderly include Supplemental Security Income (SSI), the Low Income Home Energy Assistance Program (LIHEAP), Medicaid, the Food Stamp Program, the National School Lunch Program, the Child Care Food Program, the School Breakfast Program, the Summer Food Service Program, and Supplemental Food for Women, Infants, and Children (WIC).

Many poor families do not qualify for welfare benefits because of their income or resources. Most of these families are more likely to receive income from social insurance programs such as Social Security (benefits for children whose parents are dead, retired, or disabled) and unemployment compensation. Even though Social Security benefits are not means-tested (they are not based on financial need), they do help reduce the child poverty rate.

Poor Families More Likely to Need Government Assistance

In its 2003 report to Congress, Indicators of Welfare Dependency, the U.S. Department of Health and Human Services (HHS) examined dependence on welfare. The HHS found that, as one would expect, poor households relied far more heavily on government assistance to survive than did nonpoor households. In 2000 families with incomes below the poverty line received half (50 percent) their income from wages or salaries, compared to 87 percent of nonpoor families. Poor families received one-third (30 percent) of their income from means-tested public assistance programs, including AFDC/TANF, SSI, and food stamps, compared to less than 1 percent of nonpoor families. (See Figure 5.1.) Indicators of Welfare Dependency reported that means-tested benefits accounted for 34.5 percent of the income of poor families with children age five or under in 2000.

FAMILY STRUCTURE OF WELFARE RECIPIENTS

Single-Parent Families

An increasing number of children are being raised by one parent, usually the mother. The proportion of single-parent families grew rapidly between 1970 and 1990, while the proportion dropped for families headed by married couples. Since then the structure of American households and families has remained relatively stable. According to the U.S. Census Bureau in America's Families and Living Arrangements (Washington, DC, 2001), in 2000, 76.8 percent of all family households were families headed by married couples, down from 82.5 percent in 1980 and 87 percent in 1970. Meanwhile, the proportion of single-parent families headed by males rose from 2.4 percent in 1970 to 2.9 percent in 1980 and 5.6 percent in 2000. (See Table 5.1.)

According to the Census Bureau, the proportion of single-parent families headed by females grew from 11 percent in 1970 to 15 percent in 1980 and to 17.6 percent in 2000. (See Table 5.1.) Single-parent families, especially single-parent families headed by women, usually earn much less than families headed by married couples.

The increase in the number of single-parent families was most dramatic among African-Americans and Hispanics and less so among whites. From 1970 to 2000, the proportion of white families headed by married couples declined from 89 percent to 81 percent. During the same period, the proportion of white single-parent families headed by males rose from 2.2 to 5.1 percent, and the proportion of white single-parent families headed by females grew from 9 to 13.9 percent. (See Table 5.1.)

America's Families and Living Arrangements also reported that among Hispanics, the proportion of families headed by married couples dropped from 81 percent in 1970 to 67.9 percent in 2000. The proportion of single-parent families headed by males rose from 4 to 8.7 percent, and the percentage of single-parent families headed by females increased from 15 to 23.4 percent. (See Table 5.1.)

The largest increase in the proportion of single-parent families occurred among African-Americans. While the

proportion of families headed by married couples fell from 68 percent in 1970 to less than half (47.8 percent) in 2000, the proportion of single-parent families headed by men grew from 4 to 8.1 percent, and the percentage headed by women increased from 28 percent to 44 percent of African-American families. (See Table 5.1.)

Families with Children

Single-parent families make up a large proportion of families with children under eighteen years of age. The Census Bureau reported that in 2000, 27 percent of families with their own children were single-parent families. Two-parent families made up 73 percent of all family groups with their own children, down from 87.1 percent in 1970 and 78.5 percent in 1980. Meanwhile, the proportion of families with their own children headed by men rose from 1.3 percent in 1970 and 2.1 percent in 1980 to 5.2 percent in 2000. Similarly, the proportion of families headed by women with their own children rose from 11.5 percent in 1970 to 21.9 percent in 2000. (See Table 5.1.)

BY RACE. Table 5.2 shows the living arrangements of children in March 2002. African-American children are far more likely to live with a single parent than are white or Hispanic children. According to the U.S. Census Bureau study Children's Living Arrangements and Characteristics: March 2002 (Washington, DC, 2003), 53 percent of African-American children lived at that time with one parent; 48 percent of those children lived with their mothers only. Thirty percent of Hispanic children lived

with one parent, while 70 percent lived in two-parent families. Twenty percent of white children lived with one parent, while 80 percent lived with two parents.

The same study reported that in 2002 a higher percentage of African-American children (9 percent) than whites (4 percent) and Hispanics (6 percent) lived with neither parent. In part, this is because African-American children are more likely to live with grandparents without the presence of either parent.

DIVORCE

The divorce rate in the United States has risen markedly in every decade since the 1950s, although the increase slowed somewhat in the 1990s. In 1985 almost one-quarter of those living in the United States who had ever been married had also been divorced at one time or another. According to U.S. Census Bureau data, by 2002 the number of people who had been divorced more than quadrupled from 1970, from 4.3 million to 20.9

million. This translated to 9.4 percent of the total U.S. adult population, compared to 2.8 percent in 1970. Further, there were more single women then men because men are more likely to remarry after divorce. Table 5.3 shows the numbers of married, never married, separated/divorced, and widowed men and women each decade from 1950 to 2002.

Divorce is a major factor in why women receive welfare, as divorced women are seldom as well off financially as they were when they were married. The fact that divorced women are more likely than men to receive custody of their children adds an additional financial burden. In 2002, 84.4 percent of custodial parents were female. (See Table 5.4.)

NEVER-MARRIED ADULTS

The number of people age fifteen or older who had never married rose from 25 percent in 1970 to 32 percent in 2002. The proportion of those who have never married has increased as young adults delay the age at which they marry. (See Table 5.3.)

Single-parent women are more likely never to have been married (31.2 percent in 2002) than single-parent men (20.3 percent in 2002). In 2000, 47.4 percent of African-American mothers, 16 percent of Hispanic mothers, and 34 percent of non-Hispanic white mothers with children under age eighteen had never been married. In addition to a growing trend away from marriage, these percentages could also be explained by marriages following the birth of the first child.

CHILD SUPPORT

In 2002, 21.5 million children under the age of twenty-one lived with a custodial parent. Most of these children lived with their mothers. Of the nearly fourteen million custodial parents in 2002, about 11.3 million were women, according to the U.S. Census Bureau. (See Table 5.4.)

Child support is becoming an increasingly important source of income for women, with the growing number of families headed by women, coupled with the time limits now in place for receiving cash assistance. About 63 percent of custodial mothers and 38.6 percent of custodial fathers were awarded child support in 2002. (See Table 5.4.) In that same year, more than 7.9 million custodial parents were entitled to receive child support. The other 5.5 million custodial parents did not receive financial support from an ex-partner. Of those custodial parents receiving child support, about 7.3 million had legal agreements established by a court or other government institution. Another six million had some sort of nonlegal agreement. (See Figure 5.2.)

According to the U.S. Census Bureau, in Custodial Mothers and Fathers and Their Child Support: 2001 (Washington, DC, 2003), 73.9 percent of custodial parents who were due child support in 2001 actually received partial or full payment. While the overall proportion of custodial parents who received partial or full payments dropped from 75.8 percent in 1994 to 73.9 percent in 2001, the proportion receiving full payment increased from 36.9 percent in 1994 to 44.8 percent in 2001. (See Table 5.4.)

Those Living in Poverty

Child support is often not enough to keep custodial mothers and their children out of poverty. According to the U.S. Census Bureau study, in 2001 about 23.4 percent (1.5 million) of custodial parents had family incomes considered below the poverty level. Twenty-five percent (2.8 million) of custodial mothers lived below the poverty threshold. Only 14.7 percent (307,000) of custodial fathers were below the poverty threshold during the same period. Almost all (93.6 percent) custodial parents below the poverty line in 1999 were mothers. (See Figure 5.3 and Table 5.5.)

The U.S. Census Bureau reported that approximately 36 percent of custodial parents who were due child support and were poor did not receive it in 1999. In addition, a similar percentage (30.9 percent) of custodial parents who were poor were not awarded child support in 1999. Of parents awarded child support, 64 percent of poor custodial parents actually received payments in 1999, compared to 74 percent of all custodial parents. (See Table 5.5.)

Child Support Received

In 1999 custodial mothers received an average amount of $3,844 in child support for the year. The average amount for custodial fathers was $3,175. According to the Census Bureau, the mean total income of custodial mothers who actually received child support was 36 percent smaller than that of custodial fathers who received child support ($24,983 versus $39,047). (See Table 5.5.)

One of the reasons women tend to have lower incomes than men is that fewer women participate in the workforce permanently and full-time. Custodial Mothers and Fathers and Their Child Support: 2001 reported that although about 82 percent of custodial mothers who received child support worked in 2001, 52.3 percent of them worked full-time year-round, whereas 71.7 percent of custodial fathers held full-time, full-year jobs. Despite the disparity between the sexes, the percentage of custodial mothers working full-time, year-round jobs increased from 40.9 percent in 1993 to 52.3 percent in 2001.

Socioeconomic Factors

Other socioeconomic factors that prevent custodial mothers from lifting themselves out of poverty were highlighted by the U.S. Census Bureau in 2003. They included:

• Of poor women owed child-support payments in 1999, 63.2 percent actually received payments. (See Table 5.5.) For all custodial mothers supposed to receive payments, the receipt rate for payments due was nearly 75 percent.
• Some 67.3 percent of white women received child-support payments that were due to them in 2002, compared to 19.3 percent of African-American women and 11.9 percent of Hispanic women. (See Table 5.4.)
• The percentage of never-married women who received payments in 2002 was much lower (21.8 percent) than that of ever-married women (78.2 percent).
• Women with at least a high school diploma were more likely to receive the child support due them (90.2 percent) than women with less education (9.8 percent).

Government Assistance in Obtaining Child Support

In 1975 Congress established the Child Support Enforcement program to ensure that children had financial support from both parents. Though improvements in paternity establishment and child-support collections have followed, much more remains to be done. In 1999 less than half (45 percent) of custodial parents due child support received full payments. Twenty-nine percent received partial payments, and one-quarter (26 percent) received none at all. (See Table 5.5.) Provisions in the Personal Responsibility and Work Reconciliation Act of 1996 (PL 104-193, or PRWORA) strengthened and improved child-support collection and enforcement activities.

Under PRWORA, states must have child-support assignment (requirement that families receiving assistance must assign child-support rights to the state) and good-faith cooperation requirements for TANF participants. The cooperation provision requires the custodial parent to provide the name and other identifying information about the absent parent. However, states have the flexibility to develop their own "good cause" exceptions to the cooperation requirements, such as a serious threat of domestic violence and/or sexual abuse. In addition, they may decide the penalty for noncooperation, although PRWORA mandates a minimum penalty of 25 percent of the monthly cash assistance.

The welfare-reform law (PRWORA) also required states to have child-support automation systems in place by October 1, 2000. Each state must have in effect a computerized child-support enforcement system to account for funds, to record data, to facilitate the collection and disbursement of support payments, and to keep records that improve the state's ability to locate missing parents and/or their assets. Enforcement procedures include denying or revoking driver's licenses, withholding wages, and seizing income tax refunds or unemployment compensation benefits. Data must be kept confidential for several reasons, including the possibility of family violence.

Child-Support Assurance

A family that receives assistance in the state's TANF program must assign any child-support rights to the state in order to reimburse some of the cost of that assistance. If child support is then collected on the case, the state is required to give a predetermined portion to the federal government. As an alternative to this provision, a state may choose to use its maintenance-of-effort funds (the 80 percent of its historic welfare expenditure level a state is required to spend on welfare) to run a Child Support Assurance program (CSA). Any eligible family with a support order could then choose to participate in the CSA program rather than TANF.

In a CSA program, children with child-support orders are guaranteed to receive a minimum amount of support each month. The child-support agency will collect the ordered support and, if necessary, supplement it up to the assured level. If the noncustodial parent does not pay the support, the CSA agency pays the family the guaranteed support and steps up

its enforcement efforts for nonpayment. In return, the custodial parent would have to use the state's child-support system, which keeps records of support collections and disbursements. Collecting back payments would allow the state to recover or partially recover the cost of the guaranteed payment. In addition, states would determine at what point assistance should be phased out as a family's income increases.

With a Child Support Assurance program, the state saves money. When the state is using federal money from the TANF program, it must turn over a portion (maximum 50 percent) of the child support it gets from the TANF recipient to the federal government. In a state-funded CSA program, the recipient keeps the child-support payment, and the state pays the difference between the TANF assistance amount and the assured child support. The state saves money in this case because no money goes to the federal government. In addition, the CSA program recipient continues getting the assured support payment each month following employment, encouraging custodial parents to enter and stay in the paid labor force.

Government Enforcement

About 4.9 million custodial parents made 10.5 million contacts with child-support enforcement offices (IVD office), the department of social services, or other government agencies for assistance related to child support in 2001. Twenty-seven percent of the contacts were to collect unpaid child support, 22.9 to establish legal agreements, 17 percent to receive Temporary Assistance for Needy Families or Medicaid, and 13.9 percent to locate their child's other parent.

UNEMPLOYMENT COMPENSATION

In order to qualify for unemployment compensation benefits, an unemployed person usually must have worked recently for a covered employer for some period of time and for a certain amount of pay. In 2000 about 125 million individuals were covered by unemployment compensation—97 percent of all wage and salary workers and 89 percent of the civilian labor force. Most of those not covered were the self-employed, agricultural or domestic workers, certain alien farm workers, and railroad workers (who have their own unemployment program).

Although the unemployment compensation system covers 97 percent of all wage and salary workers, only 44 percent of unemployed workers received unemployment benefits in 2002 according to the Ways and Means Committee of the U.S. House of Representatives (BackgroundMaterial and Data on the Programs within the Jurisdiction of the Committee on Ways and Means, Washington, DC, 2004). This compares with a peak of 81 percent of the unemployed receiving benefits in April 1975 and a low of 26 percent in June 1968 and October 1987.

Unemployment compensation varies widely by state. Figure 5.4 shows the percentages of unemployed receiving benefits in each state in 2002. The states with the highest rates of those receiving unemployment compensation were Connecticut and Massachusetts; South Dakota and Arizona had the lowest rates.

In 2003, 2.8 percent of the workforce covered by unemployment compensation received benefits, at a time when the total civilian unemployment rate was about 5.8 percent. The average weekly unemployment compensation benefit was $260. The amount of average weekly compensation (in 2003 dollars) rose about 17 percent between 1999 and 2003. (See Table 5.6.)

While the maximum a state may offer is thirty-nine weeks of coverage (except for special programs), all states provide up to twenty-six weeks of benefits, except Massachusetts and Washington, which offer thirty weeks. Benefits vary dramatically from state to state. In 2002 the average weekly benefits in Massachusetts ($360), New Jersey ($331), Washington ($329), Minnesota ($318), and Colorado ($313) were significantly higher than those offered by Puerto Rico ($107), Alabama ($167), Mississippi ($168), Arizona ($176), and Louisiana ($197). (See Table 5.7.)

Unemployment Compensation Important in Avoiding Poverty

Two studies examined the extent to which unemployment insurance serves as a safety net for low-income workers. The first, conducted by the General Accounting Office (GAO), Unemployment Insurance: Role as Safety Net for Low-Wage Workers Is Limited (Washington, DC, 2000), examined the rates at which low-wage and higher-wage workers received unemployment compensation between 1992 and 1995. The GAO found that while low-wage workers are twice as likely as higher-wage workers to be unemployed, they are only half as likely to receive unemployment

insurance benefits. They attribute this in part to the sectors in the economy in which low-wage workers tend to be employed, such as the service and retail sectors, which often do not provide unemployment insurance.

However, a later study conducted by Ann Rangarajan, Walter Corson, and Robert G. Wood of Mathematica Policy Research did find evidence that the unemployment insurance system was protecting low-wage workers following the enactment of welfare reform in 1996. In Is the Unemployment Insurance System a Safety Net for Welfare Recipients Who Exit Welfare for Work? (Washington, DC, 2001), they examined a group of former welfare recipients who exited welfare between July 1997 and June 1998. They found that between 50 and 60 percent of persons leaving welfare for work are eligible for unemployment insurance, compared to 20 to 35 percent found in earlier studies. Nonetheless, almost 40 percent of those who leave welfare for work are ineligible for benefits. Some of those who leave the welfare rolls are ineligible for benefits because they quit their jobs. This study was conducted during a strong economic period, and more research is needed to determine whether unemployment insurance provides a safety net to low-wage workers and those leaving welfare during periods of slower economic growth and recessions.

Minorities Hardest Hit by Unemployment

The unemployment rate of African-American and Hispanic workers is higher than that of whites. According to the Bureau of Labor Statistics, in June 2004 the unemployment rate for white workers was 5.0 percent, compared to 10.1 percent for African-Americans and 6.7 percent for Hispanics. Women who head families have a higher unemployment rate than either married men or married women with spouses present in the household. The unemployment rate for women maintaining families in December 2003 was over twice that of married men (2.5 times greater) or married women (2.15 times greater). (See Table 5.8.)

THE FEDERAL MINIMUM WAGE—A RELATIVE DECLINE

The federal minimum wage dates back to the passage of the Fair Labor Standards Act (PL 75-718) in 1938, which established basic national standards for minimum wages, overtime pay, and the employment of child workers. (The minimum wage is a "cash wage" only and does not include any fringe benefits. Consequently, the total compensation for minimum-wage workers is even lower than the total compensation for higher-paid workers, who

generally receive some kind of benefits in addition to wages. Most minimum-wage workers do not receive any benefits.) The provisions of the act have been extended to cover many other areas of employment since 1938.

The first minimum wage instituted in 1938 was $0.25 an hour. Over the years, it gradually increased, reaching $4.25 in 1991. In July 1996 Congress passed legislation that raised the minimum wage to $5.15 in 1997 by means of two 45-cent increases. (See Table 5.9.) In 2004 the minimum wage was still $5.15, although twelve states (including Alaska, California, and Illinois) had minimum wage rates higher than the federal rate, and two states (Kansas and Ohio) had minimum wage rates lower than the federal rate.

The minimum wage remained unchanged from 1981 to 1990. When inflation is taken into account, the minimum wage actually lost about half its value over this period. The increases in 1996 and 1997 still left the real value of the minimum wage well below the 1978 value. (See Figure 5.5.) A person working forty hours a week for fifty

weeks a year at minimum wage ($5.15 per hour) would gross $206 per week, or $10,300 per year, well below the poverty level for a family of three ($15,670 in 2004). For adults, this means that "day laborers" (those without a permanent job who look for a job every day) and those employed in many service jobs for minimum wages are unlikely to earn enough to escape from poverty.

The number of people working at or below the minimum wage dropped sharply from 7.8 million in 1981 to 3.2 million in 1989. The decrease was caused mainly by the sharp decline in the purchasing power of the minimum wage. As the value of the minimum wage dropped, the number of those hired at that minimum level also fell. After the recession of 1990–1991 and the slow recovery in 1992, 4.2 million workers in 1993 earned the minimum wage or less. In 1996 nearly 10 million workers were directly affected by the minimum-wage increase. Often employers use the minimum wage as a standard for low-paying jobs, perhaps paying $1 or $2 above minimum wage for a particular job.

Who Works for Minimum Wage?

Most workers affected by minimum wage increases are adults age twenty and older (72 percent), and many of those who work for the minimum wage (67.1 percent) are employed in the service sector. The majority (63.1 percent) of minimum-wage workers in 2002 were women, while some 60.4 percent of minimum-wage workers were employed in part-time jobs. About 82.1 percent of minimum-wage workers are white, while 12.7 percent are African-American and 14.1 percent are Hispanic. (See Table 5.10.)

While workers must receive at least the minimum wage for most jobs, there are some exceptions in which a person may be paid less than the minimum wage. Full-time students working on a part-time basis in the service and retail industries or at the student's academic institution, certain disabled persons, and workers who are "customarily and regularly" tipped may receive less than the minimum wage.

TRENDS IN TOTAL EARNINGS

In the 1980s and 1990s the median personal income of American men declined substantially, according to the U.S. Department of Commerce. For all males, median earnings (half earned more and half earned less) reached $27,209 in 1973 (in adjusted 2000 dollars) and then began dropping, hitting a low of $23,975 in 1982, a year of deep recession. Males' median earnings then began a slow recovery, reaching $26,825 in 1989. Then, median earnings for men began to fall again, dipping to $24,681 in 1992, lower than the median earnings during the recession of the early 1980s, and about 10 percent below males' earnings in 1973. In 2000 the median earnings of males reached $28,269, an increase of only 3.9 percent since 1973. This trend applied to white, African-American, and Hispanic men.

According to Income in the United States: 2002 (Washington, DC: U.S. Bureau of the Census, 2003), the median income of full-time, year-round male workers in 2002 dollars reached $39,429; for women the figure was $30,203.

Median incomes for all sexes and races have changed by a fifth or less since the early 1970s. Hispanics' earnings increased almost 16 percent, from $28,600 in 1972 to $33,103 in 2002, and non-Hispanic whites' earnings increased by 22 percent, from $38,439 in 1972 to $46,900 in 2002. African-Americans' earnings saw the greatest increase, 47 percent between 1967 and 2002, from $19,734 to more than $29,000. However, the differences between the median earnings of whites, African-Americans, and Hispanics remained fairly constant. African-Americans earned about 42 percent less than non-Hispanic whites in 1972; by 2002 African-Americans' median income was about 38 percent that of whites. In 1973 Hispanics' earnings were 26 percent less than whites', and in 2002, Hispanics' earnings were 29.4 percent of whites' median earnings. (See Table 5.11.)

Female earnings rose 71.4 percent from 1973 to 2000. It should be noted, however, that Income in the United States: 2002 reported the median income of female workers in 2002 ($30,203) was still only 76 percent that of male workers ($39,429).

HAVING A JOB DOES NOT GUARANTEE ESCAPE FROM POVERTY

The working poor are those people who participated in the labor force for at least twenty-seven weeks (either working or looking for work) and who lived in families with incomes below the official poverty level. Over 6.8 million workers in 2001 (4.9 percent of those in the labor force) found that their jobs did not provide enough income to keep them out of poverty. (See Table 5.12.)

Working women had a higher poverty rate (5.5 percent) than working men (4.4 percent). Nearly three-quarters of the working poor were white (72.1 percent), but African-American and Hispanic workers continued to experience poverty while employed at more than twice the rates of whites. African-Americans (9.6 percent) and Hispanics (10.1 percent) with at least twenty-seven weeks in the labor force had a far higher poverty rate than whites (4.3 percent). Younger workers were more likely to be in poverty than older workers. Much of the reason for this is that many younger workers are still in school and work at part-time or entry-level jobs that often do not pay well. (See Table 5.12.)

In general, the lower the educational level, the higher the risk of poverty. According to the U.S. Bureau of Labor Statistics in "The Working Poor in 2001" (Monthly Labor Review, November/December 2003), among workers in the labor force for at least twenty-seven weeks in 2001, those with less than a high school diploma (13.1 percent) had a much higher poverty rate than high school graduates (5.8 percent). Far lower poverty rates were reported for workers with an associate's degree (2.2 percent) or a four-year college degree (1.4 percent). African-American workers, regardless of education levels, had higher poverty rates than white workers. The highest poverty rate (26.3 percent) was for African-American women workers with one to three years of high school.

In 2002 working families headed by married couples without children were least likely to be poor (1.5 percent),

while the presence of children under age eighteen increased the married-couple poverty rate to 5 percent. Single women with families were most likely to be living in poverty (21.3 percent), although the working-poor population included a significant proportion of single men with children (12.7 percent). (See Table 5.13.)

In a family headed by a married couple, a greater likelihood exists that two members of the family are working than exists in a single-parent family. Two-income families are rarely poor. Only 1.2 percent of families headed by married couples with two or more wage earners were poor in 2001. Of the 3.7 million working-poor families, about 47 percent were families maintained by women. Working women who were the sole supporters of their families had the highest poverty rate, 20.8 percent. (See Table 5.13.)

Several factors affect the poverty status of working families: the size of the family, the number of workers in the family, the characteristics of the workers, and various labor market problems. The addition of a child puts a financial strain on the family and increases the chances that a parent might have to stay home to care for the child. While a child in a single-parent family may work, children are usually employed for low pay and at part-time jobs. In addition, the more education a person has, the more his or her job is likely to pay. Single mothers are more likely to have less education than married women.

Finally, the labor market plays a major role in whether a working family lives in poverty. Three major labor market problems contributed to poverty among workers in 2001—unemployment, low earnings, and involuntary part-time employment. Only 0.7 percent of workers who did not suffer from any of these problems were poor in 2001; 21.4 percent of low-paid workers were in poverty. Unemployment accounted for the poverty of

5.4 percent of workers, and involuntary part-time work for 2.2 percent. However, it was the combination of two or more factors that had the most devastating effect on families. Unemployment coupled with low earnings and involuntary part-time employment accounted for 41.3 percent of those in poverty.(See Table 5.14.)