After PRWORA: Barriers to employment, work, and well-being among current and former welfare recipients
from Poverty Research Insights, Fall 2004
By Kristin S. Seefeldt, Gerald R. Ford School of Public Policy, The University of Michigan.
Download printable version (pdf)
Overview
Shortly after welfare reform was implemented, numerous studies were launched with two goals: (1) to understand better how the personal and family characteristics of welfare recipients affected their ability to move from welfare to work and (2) to follow them over time and determine how they fared under a new set of rules.
In this issue of Poverty Research Insights, we highlight the findings from a number of these studies and discuss the policy implications raised in various reports.
During the welfare reform debates of the 1990s, and even after passage of the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996, many researchers and policy makers expressed concern about the strong work requirement provisions and the time limits on cash assistance. To receive benefits under the Temporary Assistance to Needy Families (TANF) program created by PRWORA, recipients must work or participate in work-related activities, and receipt of those benefits is limited to 60 months during a recipient’s lifetime, or less at state option.
Studies of welfare recipients conducted before welfare reform indicated that many lacked prior work experience, had low levels of education, and thus would have a difficult time succeeding in the labor market. However, others argued that the welfare system itself had trapped recipients — most of whom were single mothers — on the rolls for long periods of time through rules that typically deducted a dollar of benefits for every dollar earned from a job.
Given the costs associated with working (e.g., taxes, child care and transportation) and the likelihood that the jobs available to most welfare recipients paid relatively low wages, the incentive to remain on welfare was high for many single mothers.
Studies of current and former welfare recipients
Passage of welfare reform resulted in a proliferation of studies on welfare recipients. Major research organizations initiated their own studies or were funded by foundations and states to study particular welfare reforms in particular locations. The Office of the Assistant Secretary for Planning and Evaluation (ASPE) within the U.S. Department of Health and Human Services commissioned more than 30 studies of so-called “welfare leavers,” i.e., families who had transitioned off of welfare, applicants who were formally or informally diverted from welfare, and current welfare recipients.
A few random assignment studies, most begun prior to welfare reform, also provide information on the impacts of particular welfare policies, such as time limits, work requirements, and work incentives, on the employment and other relevant outcomes.
This article will draw from many of these research projects but will focus particular attention on the results from three studies: two are panel studies of welfare recipients over time and one is a national study. For each study, the primary purpose is to examine the circumstances of welfare recipients in-depth and track well-being by surveying individual welfare clients.
The Women’s Employment Study (WES) is a five-wave panel study of an initial sample of 753 welfare recipients residing in a single urban Michigan county. In-person interviews were conducted in 1997, 1998, 1999, 2001, and 2003, with response rates of between 86 and 92 percent at each wave. WES is nationally recognized for the depth of information it gathers, including measures on the extent of numerous mental and physical health problems and experiences of domestic violence as well as detailed information on employment experiences.
Where possible, WES uses previously validated measures that allow researchers to make comparisons with national estimates. WES has followed a single cohort over a longer period of time than any of the other post- welfare reform studies. This study is conducted by the Michigan Program on Poverty and Social Welfare Policy, at the Gerald R. Ford School of Public Policy, University of Michigan.
Like WES, Work First New Jersey (WFNJ) is a panel study, involving around 2,000 welfare recipients; the sample was drawn from the entire state’s caseload, as opposed to that of a single county. The study, conducted by Mathematica Policy Research, measures clients’ economic progress, employment patterns, material hardship, health status, living situations, and other outcomes over the five- to six-year period after they entered the TANF program.
WFNJ respondents, all of whom participated in New Jersey’s welfare reform program during 1997 and 1998, have been interviewed five times (annually, 1999-2003), with response rates of 80 percent or more in each survey round. The study also follows a later cohort of welfare recipients, who came on the rolls during 2000 and 2001 (and were surveyed in 2002).
Finally, the National Survey of American Families (NSAF), conducted by the Urban Institute, gathered data on a nationally representative sample of adults and children through cross-sectional telephone surveys conducted in 1997, 1999, and 2002. The
scope of NSAF is broad, with the study aiming to provide a comprehensive overview of the well-being of U.S. families. It contains an over-sample of low-income households and asks questions on topics ranging from employment and earnings, health status, health insurance coverage, and use of health services, to knowledge and use of social services and attitudes about work and welfare.It is important to note that these studies do not use the same measures in their surveys; thus caution must be exercised when making comparisons across studies. At least some, and perhaps a great deal of the differences between the studies’ findings may also be due to the demographic composition of the welfare caseload across states.
The New Jersey sample is over half African American, more than a quarter Hispanic, and 19 percent White. WES respondents are fairly equally split between African Americans and Whites, with the vast majority living in urban areas. NSAF respondents were 22 percent Hispanic, 42 percent White, Non-Hispanic, and 37 percent Nonwhite, Non-Hispanic, but living in many different states and in a mix of urban and rural settings. NSAF respondents were also receiving welfare when interviewed, whereas some WES and WFNJ participants had left the rolls at the time they were surveyed.
Barriers to employment
A basic assumption underlying PRWORA was that welfare recipients would be capable of finding jobs and transitioning from welfare to employment. Most states adopted a “Work First” strategy that emphasized job search and job readiness skills. Little time was devoted to training recipients or to assessing needs that they might have.
Prior research documented that low levels of human capital, such as a lack of a high school degree, lack of specific job skills, and limited work experience were “barriers” to employment, and welfare recipients were more likely than other women to face these types of barriers. When recipients are compared to non-recipients with the same schooling and family characteristics, recipients leave jobs at higher rates and earn lower wages than do non-recipients (Hershey and Pavetti, 1997). This suggests that previously unmeasured barriers, including mental health problems, psychiatric disorders, substance abuse or dependence, and physical health or disability also may be important.
The studies listed above attempted to document the prevalence of these barriers among the welfare population. Table 1 highlights some of the findings.
All studies find high levels of barriers among their samples of current and former recipients. Between 30 and 45 percent lack a high school diploma or GED. At least 15 percent and perhaps as many as half lack significant work experience prior to welfare reform. The prevalence of physical health problems ranges from 19 percent in WES to 36 percent in the New Jersey study. And mental health, domestic violence, and substance abuse problems are all experienced by significant minorities of the samples.
In part, the varied findings between studies are likely due to measurement and definitional differences. For example, WES defines “low work experience” as having worked in less than 20 percent of one’s adult years prior to the start of the panel study. In the New Jersey study, the measure is not having worked at all in the two years prior to TANF entry, and in NSAF, it is not working in the past three years.
Definitions related to mental health provide another example of differences between the three studies. While the WES measure of health problems is more restrictive than those used in the other two studies, for its mental health and substance abuse measures WES uses diagnostic screening criteria for psychiatric disorders based upon the DSM-IV manual, which is used by clinical professionals. NSAF relies on respondent self-report of having experienced mental health problems, while the New Jersey figure reflects the proportion who had actually been in treatment for a mental health problem.1
The extent to which drug and alcohol problems are significant for welfare recipients is a question that has received a great deal of attention from researchers and policy makers. Some advocates speculate that the majority of recipients suffer from substance abuse problems (for further discussion, see, the National Poverty Center’s Policy Brief #2, Substance Abuse and Welfare Reform).
Table 1: Prevalence of potential barriers to employment among welfare recipients
In the first wave of the WES study, only five percent of respondents met the diagnostic screening criteria for alcohol and/or drug dependence, i.e., significant functional impairment resulting from use of the substance. In New Jersey, five percent of respondents said they had sought treatment for a drug or alcohol problem in the past year. At any given survey wave in WES, though, about 20 percent of respondents report that they have used illicit substances — primarily marijuana — in the previous 12 months.
Comparisons with national data yield another perspective on the barriers to employment facing welfare recipients. Relative to a national sample of women of the same ages, WES respondents were much less likely to have graduated from high school and much more likely to have experienced transportation problems, to have met diagnostic screening criteria for health and mental health problems, and to report poor child health and severe physical abuse.
For example, just under a third lacked a high school diploma or GED, compared to about 13 percent of women nationally; more than a quarter met the diagnostic screening criteria for major depression, again compared to about 13 percent of women in a national sample; and 15 percent experienced domestic violence, as opposed to roughly three percent of all women (Danziger et al, 2000).
The three studies described above all focus primarily on welfare recipients living in urban areas. A study of welfare recipients in Nebraska finds similar or even higher levels of barriers among rural versus urban dwellers. Rural clients were somewhat more likely to report having substance abuse problems (22 percent versus 14 percent for urban clients). They were also more likely to have experienced domestic violence in their lifetimes (50 percent as opposed to 40 percent for urban clients). Although not significantly different, 37 percent of rural clients had a major depressive disorder, compared to 30 percent of clients living in urban areas (Meckstroth, Ponza, and Derr, 2002).
Employment, welfare receipt, and economic well-being after welfare reform
Although the prevalence of employment barriers is quite high, many welfare recipients have moved off of the rolls and into employment. Soon after welfare reform was implemented, many recipients had found jobs. In August, 1998, two years after PRWORA’s passage, just under 70 percent of WES respondents were employed. A smaller proportion, 43 percent, were employed in New Jersey at the time of the first survey in 1999. In the NSAF, among those who left welfare between 1997 and 1999 and had not returned to the rolls, 64 percent were working (Loprest, 2002).
By 2002, more than half, 54 percent, of WFNJ respondents were employed (Wood, Rangarajan, and Deke, 2004.) Among WES respondents in 2002, monthly employment rates averaged about 68.5 percent, about the same as 2001 and a slight decline from 2000 (calculations from Danziger and Johnson, 2004).2 Figure 1 shows employment rates for the WES sample over the entire seven-year study period.
Figure 1: Monthly Employment Rates, Current and Former Welfare Recipients. Mean Months Worked = 68% of 79 Months.
Sources: Danziger and Johnson, 2004.
Not all employment is stable, however. In the New Jersey study, three out of every four employed clients experienced a job loss at some point, with “quits” being the most common reason for losing a job (Wood, Rangarajan, and Deke, 2003). Voluntary quits accounted for about one-fifth of all job exits in WES, although that figure dropped to one-tenth of exits during the recession (Danziger and Johnson, 2004). A number of different ASPE-funded, state-level studies of welfare leavers find that on average, about 60 percent of leavers are employed at any given time (Acs and Loprest, 2004). Some are between jobs, and some return to welfare.
However, as employment increased, welfare receipt decreased, even among former recipients who lose employment. All women in the WES sample received TANF in February, 1997, the month the sample was drawn. Four years later, only 23 percent were still on the rolls. Likewise, in New Jersey, 21 percent of the original sample, all of whom were Work First New Jersey participants between July 1997 and December, 1998, received welfare at the time of the fourth survey in 2002.
Employment has not necessarily translated into large improvements in economic well-being for all of those who have left welfare. ASPE commissioned a synthesis of its state leaver studies and found that about a quarter or more leavers experience food hardships at some point after leaving TANF, including problems having enough money for food or having food last for the month. Similar percentages experience trouble paying rent or utilities (Acs and Loprest, 2004). What is not clear, though, is whether hardships increased due to leaving welfare; some studies show that hardships decrease or remain the same after exit (Acs and Loprest, 2004).
Related to the issue of health insurance loss is the debate over whether or not it “pays” to leave welfare for work. While in many states welfare benefits are not overly generous (in 2000 the median monthly TANF benefit for a family of three with no other income was approximately $390), going to work might mean the loss of some or all of those benefits, including loss of Food Stamps and Medicaid.
Additionally, recipients might incur work-related expenses such as child care and transportation. Danziger et. al (2002), using WES data, found that those who left welfare for work had much higher monthly net household incomes than those not working and on welfare ($17,265 vs $8,244). However, analysis of data from the Three-City Study, which follows low-income families (including welfare recipients) in Boston, Chicago, and San Antonio, does not find similar returns to leaving welfare for work (Moffitt and Winder, 2003).3
While welfare rolls have dropped to historically low levels, some recipients still remain on assistance, putting them at risk for reaching TANF’s 60 month time limit. WES respondents who have accumulated the greatest number of months (at least 40 months during a 60 month period) on TANF are more likely to have human capital problems as well as persistent personal and family challenges, such as children with health problems, domestic violence, and drug use (Seefeldt and Orzol, 2004).
What has happened since the recession?
Despite the record-high levels of employment among welfare recipients and single mothers more generally, there are indications that the economic downturn of the early 2000s is affecting welfare recipients and welfare leavers. According to data from the National Survey of America’s Families, the proportion of recent welfare leavers who were working and off of TANF was lower in 2002 than it was in 1999— 42.2 percent compared to 49.9 percent (Loprest, 2003a).
Monthly employment rates of WES recipients, all of whom were receiving welfare in early 1997, have also slowed. While, as noted above, more than three-quarters of the sample were working in November, 1999, the height of the economic boom, employment rates subsequently declined. In August 2003, roughly 64 percent of the sample were employed (Danziger and Johnson, 2004).
Neither data from NSAF nor from WES indicate that there have been major changes in job quality for welfare leavers. Median hourly wages for welfare leavers in 2002 were not significantly different from those who left welfare in 1999 (Loprest, 2003a). Also unchanged is the proportion of WES respondents (28 percent) who are in jobs providing an annual income that exceeds the federal poverty line and allows them to maintain health insurance. However, six and a half years after welfare reform was implemented, about half of working respondents were not in those types of jobs: despite having accumulated work experience (on average 54 months), almost two-fifths were not employed at all (Danziger and Johnson, 2004).
Also of concern is the still small but growing proportion of welfare leavers who appear to be “disconnected” from work and welfare. In New Jersey, about one in four survey respondents who were on TANF in 1997 and 1998 were, in any given month subsequent to that, neither working nor receiving welfare (Wood and Rangarajan, 2003). Among this group, just under a quarter were living with a working spouse or partner, about another quarter were receiving benefits from either the Supplemental Security Insurance (SSI) or Unemployment Insurance (UI) programs, and others were in between jobs; about two-fifths of this group lack other sources of support.
Analyses of the NSAF find that about one in seven recent welfare leavers are not working, not living with a working spouse or partner, and not receiving welfare or disability (Loprest, 2003b). Compared to working welfare leavers, these families were more likely to cut back on food intake (48 versus 30 percent) or to have had food run out during the year (63 versus 43 percent).
Additionally, the proportion of leavers disconnected from work and welfare increased from 9.8 percent in 1999 to 13.8 percent in 2001, but was slightly lower in 2002 because fewer families left welfare after 1999.
Results from the Women’s Employment Study show a similar increase in the percentage of women who are both not working and not receiving welfare benefits in a given month – from 4.4 percent in 1997 to 15.3 percent in 2001. WES respondents who have been disconnected for more than a quarter of the months between 1997 and 2001 have a signficantly greater number of barriers to employment — including human capital, health, and mental health problems — and are more disadvantaged on a variety of economic measures. (Turner, Danziger, and Seefeldt, 2004).
Policy Implications
Greater numbers of single mothers, including welfare recipients, entered the labor force after welfare reform. And on average, those who moved from welfare to work have higher incomes than those who remain on welfare.
However, as Danziger and Wang (2004) note, many welfare leavers still remain poor despite working. The researchers propose consideration of policies that would raise the net income of workers, such as increased child care subsidies, expanded access to health insurance, raises in the minimum wage, or implementation of state-level earned income tax credits.
Wood and Rangarajan (2003) suggest that employment retention and advancement services or linkages to community agencies might be helpful for the least stable families leaving welfare, for example, those who were sanctioned or employed leavers with low levels of education. For those still on welfare or entering the rolls, assessment, referrals, and the use of a wide array of services has been suggested as a way to help address employment barriers (Danziger and Seefeldt, 2002).
Helping disconnected families, Loprest (2003b) notes, may be particularly challenging. Welfare offices may not follow up with these families, because they have left the rolls; and some families may not wish to return or may have personal and other family challenges (e.g., physical and mental health problems) that impede their ability to navigate the system. Loprest suggests that despite the challenges, attempts should be made to reconnect these families with TANF benefits and also work to assess and address any employment barriers.
ENDNOTES
(1) In the later cohort of WFNJ clients studied (those who were on TANF between 2000 and 2001), a higher proportion, 16 percent, reported a diagnosed mental health problem (Wood, Rangarajan, and Deke, 2004.)
(2) The higher employment rates in WES compared to WFNJ may be due to earlier welfare exits since the WES sample were on welfare in early 1997, whereas only some WFNJ sample members were on welfare in 1997.
(3) For more detail on this debate, see the forthcoming issue of the Journal of Policy Analysis and Management, Volume 24, Number 2, 2005.
CITATIONS
Acs, Gregory and Pamela Loprest. 2004. Leaving Welfare: Employment and Well-Being of Families that Left Welfare in the Post-Entitlement Era. Kalamazoo, MI: W.E. Upjohn Institute for Employment Research.
Danziger, Sandra K., Mary Corcoran, Sheldon Danziger, Colleen Heflin, Ariel Kalil, Judith Levine, Daniel Rosen, Kristin S. Seefeldt, Kristine Siefert, and Richard Tolman. 2000. “Barriers to the Employment of Welfare Recipients.” In R. Cherry and W.M Rodgers, eds, Prosperity For All? NY: Russell Sage Foundation.
Danziger, Sandra K, and Kristin S. Seefeldt. 2002. “Barriers to Employment and the “Hard to serve”: Implications for Services, Sanctions and Time Limits. FOCUS. 22: 1, 76-81.
Danziger, Sheldon and Rucker Johnson. 2004. “From Welfare to the Low-Wage Labor Market,” paper prepared for the Russell Sage Foundation Forum on the Future of Work.
Danziger Sheldon and Hui-chen Wang. 2004. “Does It Pay to Move From Welfare to Work: A Reply to Robert Moffitt and Katie Winder.” Forthcoming in the Journal of Public Policy and Management.
Garrett, Bowena and John Holahan. 2000. “Welfare Leavers, Medicaid Coverage, and Private Health Insurance.” “ The Urban Institute, New Federalism: National Survey of America’s Families, #B-13, http://www.urban.org/url.cfm?ID=309442
Hershey, Alan and LaDonna Pavetti. 1997. “Turning Job Finders into Job Keepers: The Challenge of Sustaining Employment.” The Future of Children, Spring 1997 (pp. 74-86). Los Altos, CA Center for the Future of Children, The David and Lucille Packard Foundation.
Loprest, Pamela. 2002. “Making the Transition from Welfare to Work: Successes but Continuing Concerns,” in Welfare Reform: The Next Act, Alan Weil and Kenneth Finegold, eds. Washington, D.C.: Urban Institute Press.
Loprest, Pamela. 2003a. “Fewer Welfare Leavers Employed in a Weak Economy.” The Urban Institute, Snapshots of America’s Families, No. 5. http://www.urban.org/UploadedPDF/310837_snapshots3_no5.pdf
Loprest, Pamela. 2003b. “Disconnected Welfare Leavers Face Serious Risks.” The Urban Institute, Snapshots of America’s Families, No. 7. http://www.urban.org/UploadedPDF/310837_snapshots3_no5.pdf
Meckstroth, Alicia, Michael Ponza, and Michelle Derr. 2002. “Welfare Clients’ Employment Obstacles in Rural and Urban Areas: Lessons from Nebraska.” Mathematica Policy Research, Issue Brief #4, http://www.mathematica-mpr.com/publications/PDFs/employmentobstacles.pdf
Rangarajan and Wood, 1999. “Work First New Jersey Evaluation. How WFNJ Clients Are Faring Under Welfare Reform: An Early Look.” Mathematica Policy Research,
http://www.mathematica-mpr.com/publications/PDFs/wfnj.pdf
Seefeldt, Kristin S. and Sean Orzol. 2004. “Watching the Clock Tick: Factors Associated with TANF Accumulation.” University of Michigan, National Poverty Center Working Paper #04-09. http://www.npc.umich.edu/publications/workingpaper04/paper9/04-09.pdf
Turner, Lesley, Sheldon Danziger, and Kristin S. Seefeldt. 2004. “Failing the Transition from Welfare to Work: Women Disconnected from Work, Welfare, and Other Sources of Economic Support.” University of Michigan, Program on Poverty and Social Welfare Policy. http://www.fordschool.umich.edu/research/poverty/pdf/disconnected-april04.pdf
Wood, Robert and Anu Rangarajan. 2003. “What’s Happening to TANF Leavers Who Are Not Employed? Trends in Welfare-to-Work.” Mathematica Policy Research, Issue Brief # 6.
http://www.mathematica-mpr.com/publications/PDFs/tanfleave.pdf
Wood, Robert, Anu Rangarajan, and John Deke. 2004. “Early and Later WFNJ Clients: Are Their Experiences Different?” Mathematica Policy Research,
http://www.mathematica-mpr.com/publications/PDFs/earlywfnj.pdf
Zedlewski, Sheila. 1999. “Work-Related Activities and Limitations of Current Welfare Recipients.” The Urban Institute, Assessing the New Federalism Discussion Paper 99-06,
http://www.urban.org/url.cfm?ID=409169.
<< Back to Poverty Research Insights, Fall 2004
