• Email
  • Single Page
  • Print

What Happened to Welfare?

Three families are clearly inadequate to assess the impact of welfare reform, and DeParle also reviews more systematic evidence from other sources, including surveys and various experiments that treated groups of families differently. Three questions seem particularly important in assessing the legislation’s impact: whether single mothers’ material standard of living rose or fell, how pushing mothers into the labor force affected their children, and how the new system affected the number of women raising children on their own.


DeParle stopped by one evening in 1999 to visit Angie, who was propped up in bed beside her on-and-off boyfriend, Marcus:

Great heaps of stuff spilled everywhere: report cards, pay stubs, unopened bills, CDs, an iron, mounds of dirty clothes. A milk crate held a battered TV, and Marcus sprawled on the bedcovers…. Angie was in an expansive mood. All in all, she figured, it hadn’t been a bad year. She had worked three jobs and bought a car. She had earned $18,500, a personal record. She had started a retirement plan. She had been shot at by Marcus and ripped off by Opal. She had lost her lights twice and run low on food more times than she could count. Someone had just egged her car….

So how had the new law changed her life? Had ending welfare worked? While I had posed versions of the question before, they never seemed to grab her, and I was starting to understand why. On welfare, Angie was a low-income single mother, raising her children in a dangerous neighborhood in a household roiled by chaos. She couldn’t pay the bills. She drank lots of beer. And her kids needed a father. Off welfare, she was a low-income single mother, raising her children in a dangerous neighborhood in a household roiled by chaos. She couldn’t pay the bills. She drank lots of beer. And her kids needed a father. “We’re surviving!” is all Angie said. “‘Cause that’s what we have to do.”

Were her kids proud that she works? It was a question that often arose when I talked about Angie with middle-class friends, most of whom took it as an article of faith that the answer was yes. Angie paused. “I don’t think the kids think about that,” she said. “They’d like it if I’d just sit around with them all day.” She raised her voice to mimic a squeal: “‘Why you always at work?’ Shoot! Why you think I gotta work? Ain’t none a you got a job!” It was possible, of course, that the kids felt prouder of her than she knew and that the power of the example she set would become clearer with time. I asked her if she thought her struggles to grind out a low-wage living would encourage the kids to stay in school. “Do I think they’re going to finish high school? Hell, no!” Angie said.

DeParle’s calculations suggest that Angie was a little better off economically after leaving welfare. Her after-tax income during her last three years on AFDC had averaged $21,500 a year ($8,400 from AFDC, $4,800 from food stamps, and $8,300 from wages and tax credits). When she left welfare, she lost her AFDC benefits but kept most of her food stamps, earned more, and got a bigger tax credit for working. Her average annual income after taxes rose $3,400, but she also had more expenses. Her children continued to get Medicaid, but she became ineligible. Anything she spent on medical care had to come out of her extra $3,400. Because she had to work more, she also had to pay more for transportation.

Angie earned significantly more than most mothers who left welfare. But because she had four children and Wisconsin’s AFDC benefits were unusually high, the check she lost when she left the rolls was far more than most AFDC mothers had been getting. What she ended up with—a small cash gain but some additional expenses—seems to have been typical of others who went off welfare across the country.

The mothers who left the welfare rolls are only half the story, however. States also reduced the rolls by “diverting” would-be applicants into job searches and employment programs, and these tactics discouraged others from applying for welfare. We cannot identify these families, so there is no way to determine whether they are now better or worse off than they would have been under AFDC. But we can make inferences by examining trends in poverty and material hardship for all unmarried mothers.

Figure 1 shows changes in the overall unemployment rate and the official poverty rate for unmarried women who headed their own households and had children under eighteen. (For a family of three in 2004, the poverty line was just over $15,000. It is adjusted every year for inflation.) In 1989, when the Reagan-era boom petered out, 43 percent of these families were officially “poor.” In 2000, when the Clinton boom came to an end, only 33 percent were officially poor. At that time skeptics often attributed declining poverty among unmarried mothers to the unusually tight labor market and argued that we would not be able to assess the true cost of welfare reform until the next recession. That recession began in 2001. Unemployment peaked at 6 percent in 2003, which was about the same as the rate in 1994. The poverty rate for single mothers was only 36 percent in 2003, compared to 44 percent in 1994. It was still 36 percent in 2004, when unemployment was about what it had been in 1996.

The official poverty rate has serious flaws. It omits food stamps, free medical care, housing subsidies, and taxes. So it is important to see whether direct measures of material hardship tell the same story. The Agriculture Department’s Food Security Survey (FSS) is a good place to begin. It asks mothers whether money problems forced them to skip meals or cut the size of their meals at any point during the previous twelve months. Between April 1995 and April 2001, as the welfare rolls were being cut in half, the fraction of single mothers who said they had to limit what they ate fell from 17.7 to 12.5 percent.5 By 2003, with unemployment slightly above its 1995 level, the percentage of mothers who reported cutting back what they ate had risen, but only to 13.8 percent.

The FSS also asked mothers whether there was a time when their children were not getting enough to eat. The proportion who said this was the case fell from 10.6 percent in 1995 to 7.8 percent in 2001 and had only risen to 8.0 percent in 2003.6 Both of these measures are consistent with the official poverty rate in suggesting that even when unemployment was 6 percent, single mothers did better after welfare reform than before.

The proportion of unmarried mothers living in someone else’s home is another indicator of financial stress. Some mothers live in someone else’s home by choice, but most get their own place when their income rises. Twenty-two percent of unmarried mothers lived in someone else’s home in both 1989 and 2000.7 When welfare reform passed, its critics also predicted a surge in the number of families living in public shelters. The United States does not collect national data on trends in homelessness, but in Milwaukee, which cut about ten thousand families from its welfare rolls, DeParle reports that the number of families in shelters on an average night rose by only forty-one. On December 12, 2004, Boston counted 1,157 homeless children in the city, down from 1,274 a decade earlier.8 Nationally, the proportion of children not living with either their mother or their father was the same in 2004 as in 1994.

These measures of material well-being can be summarized in two ways. Defenders of welfare reform stress the fact that the proportion of single mothers who cannot afford to rent their own housing is no higher today than in 1996, and that the proportion who report not having enough to eat has fallen significantly. Critics of welfare reform stress the fact that at least a million single-parent families still have trouble putting food on the table and that even larger numbers cannot afford an apartment of their own. Welfare reform is clearly a success relative to the dismal situation that prevailed in the United States before 1996. But the country is still a long way from achieving the goals that more compassionate societies set for themselves.

Welfare reform has been particularly tough on single mothers who can neither work regularly nor qualify for public assistance. Some of these mothers have sick children or elderly parents who require constant attention. Some are addicted to drugs or alcohol. Some are mentally ill, physically disabled, chronically sick, or retarded. Most states allow some of these mothers to stay on public assistance, but no state seems to cover all of them. For mothers who are unable to work regularly, welfare reform has often been a terrible experience.


Until 1996 opponents of work requirements had argued that most single mothers wanted to work but were either unable to find jobs or unable to afford child care. These arguments were sometimes correct, but many welfare mothers were in fact working “off the books” without informing the welfare department. Once states began requiring welfare recipients to work, the number of people on welfare fell and “official” employment rose.

Although the overall unemployment rate was about the same in 2004 as in 1995, the fraction of single mothers with paid jobs rose from 62 to 70 percent.9 The reasons for this increase are obvious: welfare was much harder to obtain, and the economic benefits of work had risen. We do not yet have hourly wage data for 2004, but for women near the bottom of the economic hierarchy inflation-adjusted hourly wages rose 14 percent between 1994 and 2003.10 Increases in the Earned Income Tax Credit made the gains for single mothers even larger. In addition, more generous child care subsidies lowered working mothers’ expenses.

Still, not everyone who left the welfare rolls went to work. Last March, 8 percent of single mothers told the Census Bureau that they had neither received welfare nor worked during 2004, and that no one else in their household had worked either. In March 1995 this figure was less than 5 percent.

Most of these mothers were poor, but it is hard to know exactly how poor. Taken at face value, the incomes they reported were often so low that they should have been living in the streets during 2004. Perhaps some were. But some of them had income that they chose not to report, and others lived with boyfriends whose presence they did not mention. Some got by without any income during the previous year because they had been living with their parents. Because of such possibilities, Census Bureau estimates of extreme poverty must be treated cautiously. As I have already indicated, more direct measures of material hardship suggest that single mothers are a little better off today than they were before welfare reform.

  1. 5

    Scott Winship and Christopher Jencks, “How Did the Social Policy Changes of the 1990s Affect Material Hardship Among Single Mothers? Evidence from the CPS Food Security Supplement,” Kennedy School of Government, Working Paper RWP04-027, June 2004 (ksgnotes1.harvard.edu/Research/wpaper.nsf/rwp/RWP04027/$File/rwp04_027_Winship_Jencks.pdf). The comparisons between 1995 and 2001 use the common screen.

  2. 6

    Tabulations by Scott Winship.

  3. 7

    I am indebted to Joseph Swingle of Wellesley College for all the estimates derived from the Census Bureau’s Annual Demographic Survey.

  4. 8

    Friends of Boston’s Homeless, “2004 City of Boston Homeless Census: Homeless Families Increase Dramatically,” available at www.fobh.org/census.htm, accessed August 26, 2004. The title of this report refers to an increase between 2003 and 2004.

  5. 9

    Employment rates are monthly averages taken from US Bureau of Labor Statistics, “Employment Characteristics of Families,” 1996 and 2004, available at bls.gov/schedule/archives/all_ nr.htm#FAMEE.

  6. 10

    Economic Policy Institute, The State of Working America, 2004/2005, Washington, D.C., 2005, Table 2.8. The estimated gain is for the 10th percentile of the female distribution.

  • Email
  • Single Page
  • Print