Personal, family, and multiple barriers of long-term welfare recipients.

Taylor, Mary Jane ; Barusch, Amanda Smith

The passage of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (P.L. 104-193) initiated a series of events that resulted in millions of welfare recipients being permanently denied public cash assistance (Duncan, Harris, & Boisjoly, 1997). At the time of the bill's passage, it was asserted that former welfare recipients would be able to find work, maintain employment, and provide financial support for their families. Studies of welfare from a macro perspective, including the views that emphasize labor market and human capital variables (for example, Lerman & Ratcliffe, 2000; Maryland Department of Human Resources, 1998; Rockefeller Institute of Government, 1999), may support the notion that welfare recipients can be trained to work and absorbed into a labor market in which their skills are needed. This approach grossly underestimates the effect of personal and familial barriers for some welfare recipients. Specifically, long-term welfare recipients exhibit complex issues that limit their ability to find and maintain employment.

The study discussed in this article explored personal and familial barriers to employment for long-term recipients and the presence of multiple barriers. The study was unique in that the sample included only long-term welfare recipients. Data were collected through interviews with a representative statewide sample of urban and rural long-term recipients, and administrative data were obtained to supplement interviews.

It is important for social workers to understand the complexity of employment barriers for welfare recipients for a number of reasons. First, there is a widespread assumption that welfare time limits have been "successful"--that former recipients are finding and maintaining employment (U.S. Department of Health and Human Services, 1999). This is true only in part. For some former recipients, employment is part-time, short-term, tenuous, and less financially rewarding than receipt of public assistance. Second, collateral agencies such as mental health agencies, child welfare agencies, health clinics, child mental health facilities, schools, and food and clothing banks are likely affected by former welfare recipients whose financial needs are not being met in the labor market (Taylor & Barusch, 2000). Agencies in which social workers are employed are likely to see an increase in the need for services as the support systems for former welfare recipients disintegrate over time. Third, women and children who are poor are a historically vulnerable group served by our profession, and it is this group that is being targeted by current welfare reform policies.

Experience with TANF

A large percentage of welfare recipients have had difficulty achieving economic independence within the five-year time limit set by the federal government for the Temporary Assistance to Needy Families (TANF) program. Testifying before the U.S. House of Representatives Committee on Ways and Means, Pavetti (1996) reported that 70 percent of the public assistance caseload had received welfare for more than 24 months and 48 percent had been dependent on public assistance for more than 60 months. A wide range of studies examined factors that influence an individuals duration on welfare. These factors include education and labor market experiences, personal and family characteristics, mental health, substance abuse, and domestic violence.

Personal and family characteristics clearly affect the likelihood that a family will remain on public cash assistance. Factors associated with length of time on welfare include age, gender, number of children, marital status, child support, education, and work history (Blank, 1989; National Governor's Association, 1998; Pavetti, 1997; Peterson, 1995; Sherman, Amey, Duffield, Ebb, & Weinstein, 1998; Spalter-Roth, Burr, Hartmann, & Shaw, 1995; Taylor, 1999). Other factors related to decreased likelihood of working include presence of disabilities; job availability; primary responsibility for toddlers or infants; and family supports, for example, child support or earnings from other sources (Pavetti, Olson, Pindus, & Pernas, 1996). Olson and Pavetti (1996) identified eight obstacles that may affect a recipient's ability to transition from welfare to work, including physical disabilities or health limitations, mental health problems, health of behavioral problems of children, substance abuse, domestic violence, involvement with the child welfare system, housing instability, low basic skills, and learning disabilities.

Education and labor market experiences influence recipients' ability to leave welfare for work. Those most likely to be affected by time limits appear to have had little education or work experience (Harris, 1991; Meyer & Cancian, 1996; Parrott, 1998). Welfare recipients with little work experience tend to stay on welfare longer than recipients with more work experience and also tend to work in low-wage jobs (Bane & Ellwood, 1994).

Substance abuse, family violence, and mental health have been shown to be associated with welfare receipt (Allard, Albelda, Colten, & Cosenza, 1997). Violence and its side effects, including posttraumatic stress disorder (PTSD), may affect a woman's ability to work (Salomon, Bassuk, & Brooks, 1996). Although domestic violence does not directly translate into inability to maintain employment, abuse can interfere with job stability (Tolman & Raphael, 2001). Gerdes (1997) compared the life experiences of women on Aid to Families with Dependent Children with veterans of war. Experiences of both included exposure to violence, physical and sexual abuse, rape, witnessing a homicide or suicide, and criminal victimization. Victims of domestic violence reported symptoms similar to victims of PTSD, including worry, stress, pressure, feelings of being trapped, feelings of powerlessness, and isolation and loneliness.

Method

This article reports a descriptive study of long-term welfare recipients. Expanding on earlier studies of barriers to employment, in-depth, inperson interviews were conducted with a statewide representative sample of 284 long-term recipients of public cash assistance. The legislated lifetime limit on public assistance for respondents was 36 months. Therefore, for the purposes of this study, long-term was defined as 36 or more cumulative months on public cash assistance.

Sample

The sample was drawn randomly from the population of long-term welfare recipients in one southwestern state from 1997 to 1998. It was stratified geographically, and rural recipients were overrepresented to ensure useful data from areas with differing economies and client populations. The recruiting protocol involved an initial letter, followed by postcards, phone calls, and home visits in cases of nonresponse. Each participant in the study received at least two mailed contacts, three phone calls (if they had a telephone), and three home visits. The 284 client interviews represent a 63 percent response rate from the randomly selected sample. Forty-one percent of the sample lived in rural areas, and 59 percent were urban residents.

The 284 long-term recipients interviewed had an average age of 34. The majority (97 percent) were women. Most (74 percent) were white. On average, respondents had completed 12 years of school, with a range from six to 20 years of education. Respondents reported having an average of 2.5 children, with a range from 0 to 10. The mean age of the oldest child in this sample was 11.5 years, and that of the youngest child was 7.2 years.

Respondents reported an average of 2.9 separate episodes of welfare, with a range from one to 15 episodes. They had spent an average of 88 cumulative months (7.3 years) on welfare, ranging from 36 to 336 months (28 years).

Data Collection

Interviews were conducted by trained social workers and graduate students either in the client's home or a location convenient to the client. The interviews lasted between one and three hours. Clients were paid $20 for a completed interview.

The survey instrument included questions on demographics, education, family income, housing, transportation, family characteristics, child care, family background, employment history, welfare history, family health, access to health insurance, self-sufficiency program participation, barriers to work, social support networks, self-esteem and self-efficacy, and domestic violence. In addition, clinically relevant measures of child behavior, individual health, learning disabilities, substance abuse, depression, anxiety disorder, and PTSD were obtained.

Results

The interview data provide an in-depth description of personal characteristics, family characteristics, and multiple barriers to employment among long-term welfare recipients. Personal characteristics included physical and mental health, and education and employment history. In these areas, the following dimensions were examined: general perceived health, learning disabilities, completion of high school or GED, work history, substance abuse, depression, generalized anxiety disorder, and PTSD. Family characteristics included health status, as well as child behavior and domestic violence. Finally, the extent to which respondents experienced multiple barriers to employment was addressed. When feasible, data from this sample were compared with the total state welfare population.

The results of this study indicate sharp contrasts on critical variables between long-term welfare recipients and the general welfare population.

Personal Characteristics

General Perceived Health Status. Respondents were asked to evaluate their general health status as poor, fair, good, very good, or excellent. This measure, known as "self-reported health status," is a widely used indicator (Mjdde-Mossey & Mor-Barak, 1998; White, Tulsky, Dawson, Zolopa, & Moss, 1997). In this sample, about one-quarter (24.6 percent) rated their health as excellent or very good, whereas two-fifths (40.1 percent) rated their general health as fair or poor. More than one-half the sample (53.2 percent) reported the presence of physical health problems, and 34.9 percent of the sample reported that physical health problems prevented them from working. Based on reports from administrative data, this compares with about 12 percent of the total state welfare population that reported physical health problems preventing them from working (Table 1).

Learning Disabilities. The Payne (1998) scale was used to identify clients with potential learning disabilities. This scale consists of a series of nine questions about problems, such as "working with numbers in a column," "filling out forms," "mixing up arithmetic signs," and "difficulty spelling words you know." Nearly one-fourth of the group (22.9 percent) scored in a range that suggests they should be screened for learning disabilities.

Education and Work History. Close to one-third of this sample (32 percent) reported that they had neither a high school diploma nor a GED. This percentage is lower than that for the total state welfare population, of which almost 38 percent had no high school graduation credentials (Table 1). The larger percentage of high-school-educated recipients in the sample may be a reflection of the higher average age of the sample compared with the total state welfare population.

Respondents reported that the longest they had ever worked at one job averaged 18 months. The mean time on one job (18.2 months) was exceeded by the mean duration of unemployment for the same period (27.4 months). Regarding health insurance, only 22.2 percent reported having health insurance in their current job, and only 8.3 percent of those employed reported that they were eligible for participation in a pension plan. Over 30 percent reported that in the past tire years they had never worked more than six months at one job.

Substance Abuse. Respondents were asked questions about their use of alcohol or drugs, such as whether they had ever considered decreasing alcohol or drug consumption, whether a friend or family member had ever suggested they decrease alcohol or drug consumption, and whether alcohol or drug consumption had ever interfered with their job. Results suggest that approximately one in five long-term welfare recipients was currently abusing either drugs or alcohol. Of the total state welfare population, about 10 percent of participants reported a substance abuse problem.

Depression. Two scales were used to measure depression. The Center for Epidemiological Studies Depression Scale (CES-D) is a continuous scale used to measure symptoms of clinical depression (Radloff, 1977). The CES-D provides reliable information on the extent of depressive symptomatology. In addition to the CES-D, a scale using questions from the DSM-III (American Psychiatric Association, 1987) was used. The DSM scale is a dichotomous measure indicating presence or absence of a clinically relevant level of chronic depression. A majority (56.7 percent) of the respondents scored above the CES-D cutoff score, and a large percentage (42.3 percent) scored positively for clinical depression on the DSM-III measure (Table 1). Fewer members of the total state welfare population (15 percent) reported depression.

Generalized Anxiety Disorder. A scale created from the DSM-IV (American Psychiatric Association, 1994) was used to measure presence or absence of generalized anxiety disorder (GAD). GAD is chronic, excessive anxiety about a number of events or activities. Nineteen respondents (6.7 percent) scored positively for presence of GAD (see Table 1).

Posttraumatic Stress Disorder. Questions based on the DSM-IV were used to create a scale measuring PTSD. PTSD is exposure to a traumatic event in which the person witnessed or experienced events that involved actual or threatened death or serious injury. In this study, 15.1 percent of long-term welfare recipients met the DSM-IV criteria for a diagnosis of PTSD.

Family Characteristics

Welfare recipients are often treated as if their employment and welfare trajectories were solely the result of personal characteristics. Clearly, family variables affect individual choices and constraints. Three such measures were considered in this study: family health, child behavior problems, and domestic violence.

Family Health. Respondents often reported that family health problems affected their ability to work. More than one-third (35.6 percent) of the sample reported that at least one of their children had a physical disability or medical condition. Among these, more than one-half (57 percent) of the children were taking medication for the condition, and roughly the same proportion (54 percent) were receiving treatment. Almost one-third (29.5 percent) reported having at least one child with a learning disability. Among these, 17 percent were taking medication and 79 percent were in a special class in school. Just under one-fourth (23.9 percent) reported having at least one child with a mental health condition. Among these, 32 percent were taking medication, and 54 percent were in treatment. Thirty-one percent of the respondents reported that their children had physical, learning, or mental health conditions that required much of their physical or emotional energy. Seventeen percent reported that their spouse or partner had a health problem, disability, or addiction.

Child Behavior Problems. Three subscales from the Child Behavior Checklist (CBC) (Achenbach, 1992) were completed for the oldest child in each family. The mean age for these children was 11, with a range from four to 18 years. The group was evenly divided between boys (50.7 percent) and girls (49.3 percent). A total of 47 problem behaviors were examined in relation to aggressive, delinquent, and anxious behavior.

Nearly one-fourth of the respondents reported severe child behavior problems. In fact, 23.2 percent reported CBC scores for their oldest children in the "clinical" range, a point at which professional intervention is strongly advised. These children were evenly divided among boys (50 percent) and girls (50 percent). Data for the total state welfare population indicate about 6 percent of participants reported severe behavior problems with their oldest children (see Table 1).

Domestic Violence. Respondents were asked a series of questions regarding violence involving a spouse or partner. The questions were adapted from the women's employment study by the Survey Research Center, Institute for Social Research (1997) of the University of Michigan. Respondents were asked about their experiences as adults with domestic violence ever as well as in the 12 months before the interview.

The majority of respondents reported being involved as a victim of domestic violence at some time in their adult lives. Seventy-nine percent reported at least one incident of violence involving a spouse or partner. Regarding recent domestic violence, 20 percent of respondents reported at least one incident of violence during the preceding 12 months.

More than one-half of the respondents (62 percent) reported having called the police because of domestic violence. Almost one-half (42 percent) reported having been harassed at work, 36 percent reported having to stay home from work because of domestic violence, and 37 percent had seen a doctor because of domestic violence. Almost one-half the respondents (49 percent) had tried to get a restraining order against a violent partner, and 17 percent had tried to obtain a restraining order within the past 12 months. Almost one-quarter (21 percent) reported that their current partner had committed an act of domestic violence against them.

The extent of "severe" domestic violence was measured by combining six incidents: being hit with a fist; being hit with an object; being beaten; being choked or threatened; having a weapon used against you; or being forced into sexual activity. Almost 75 percent of the respondents had been a victim of severe domestic violence in their adult lives, and 12.3 percent had been a victim of severe domestic violence within the past 12 months (Table 1).

In addition to partner abuse, studied families often experienced high rates of child abuse and neglect. More than one in 10 (11 percent) reported that at least one of their children had been in foster care. A striking 46.3 percent of the sample reported that child protective services had investigated their family for abuse or neglect of their children, compared with fewer than 2 percent of the total state welfare population that reported investigation by child protective services (Table 1).

Multiple Barriers and Employment

This study provides data that indicate long-term welfare respondents also struggle with multiple barriers. Barriers were computed as dichotomous variables for the following:

* mental health

* learning disability

* education

* work history

* physical health problems

* child physical or mental health problems

* domestic violence

* drug abuse

* alcohol abuse

* severe child behavior problems

* child protective services referral.

The most frequently reported number of barriers was three (26.1 percent). Five percent of the sample had six or more barriers. The relationship between number of barriers and employment was found to be statistically significant [[chi square](10) = 18.947, p < .015]. At zero barriers, the percentage of those working at least half-time was almost double those not working. As the number of barriers increased, the percentage of participants working at least half-time decreased (Table 2).

Conclusion

With respect to all measures considered here, long-term welfare recipients experienced high levels of difficulty. Almost one-half of the respondents reported fair or poor health, and more than one-third reported that a physical disability or health problem was severe enough to prevent them from working. When compared with the total state welfare population, long-term recipients were more likely to present mental and physical health problems, to have experienced severe domestic violence, to report drug or alcohol problems, and to have difficulties with their children as evidenced by a behavior problem or child protective services referral. Roughly one-half the respondents reported symptoms of depression. Almost one in four may have had a learning disability and, perhaps as a result, almost one-third had no credential for high school graduation.

Even if welfare reform policies could accomplish the purported goal of helping heads of households obtain the training, therapy, or education necessary to become self-supporting, long-term welfare recipients would still be at a disadvantage because of family complications. High numbers of respondents' children had serious medical problems, and almost one-quarter of respondents reported children with severe behavior problems. Domestic violence episodes were much more frequent among respondents than among the total state welfare population, and nearly one-half had been investigated by child protective services as an adult parent.

It appears that long-term welfare recipients are qualitatively different from other welfare recipients. They not only have severe mental health, health, and family barriers to self-sufficiency, but also experience multiple barriers. Some of these barriers may improve over time, as children grow older and become more independent and as health problems improve. However, the prognosis for other barriers, including chronic health and mental health conditions, is less hopeful. Many in this segment of the welfare population will not respond to welfare reform by attaining and maintaining employment that pays a living wage.

Findings from this study suggest a number of important implications for social workers. First, at least some of the barriers experienced by long-term recipients should be addressed by welfare agencies when recipients are on public assistance. Efforts toward early identification of barriers (State Policy Documentation Project, 2000) should be replicated and expanded.

Early identification of barriers would allow for agency staff to help recipients prioritize and aggressively address barriers, especially complex issues such as substance abuse and mental health problems, during their time on assistance. Some states have a review period in the final months before recipients reach their time limits (State Policy Documentation Project, 2000). The final review period could be used to apply intensive efforts toward reduction of employment barriers, and to provide resource information recipients may need after leaving assistance.

Many former recipients will enter the ranks of the poor working people. As public sentiment has turned against an entitlement of lifetime assistance, there is an opportunity for social workers to lobby for "safety net" provisions for this population. Gains have been made in expansion of the earned income tax credit and increase in the minimum wage (Danziger, 2000). However, quality child care for working parents, expanded Medicaid or guaranteed health insurance, increased wage supplements, expanded access to food stamps, and supported housing should be available for all working families whose incomes are at or near poverty level.

Finally, findings from this study suggest that some welfare recipients will not be able to support their families through employment. For these families, a network of permanent services should be provided. States could provide ongoing cash assistance for these recipients, possibly with a minimal work of participation requirement. Case management, crisis intervention, and mentoring should be available as well. Social workers must work in their agencies and in their roles as advocates to develop supports for marginal families.

Ultimately the question for state officials may be whether they are willing to either remove children from these families or see them living in homeless shelters and on the streets. Those who find these possibilities abhorrent must seriously consider developing permanent, cost-effective supports for the nation's most vulnerable families.

Table 1
Prevelance of Barriers to Work for Long-Term Welfare Recipients
(36 or more months)

 Long-Term
 Welfare State Welfare
 Recipients Population (a)
 % %
Barrier (N = 284) (N = 7,951)

Mental health
 CES-depression 56.7
 DSM-III depression 42.3 15.0
 Generalized anxiety 6.7
 Posttraumatic stress disorder 15.1
Learning disability (b) 23.0 --
Education (no diploma or GED) 32.0 37.5
Work history (c) 30.4 --
Physical health problems that
 prevent work 53.2
 34.9 11.9
Physical or mental health
 problems-child 42.2 --
Severe domestic violence within the
 past 12 months 12.3

Severe domestic violence ever
 as an adult 73.6 12.0
Drug abuse 19.6 10.2
Alcohol abuse 20.1 --
Severe child behavior problems 23.0 6.1
Child protective services referral 46.3 1.5

(a) Information on Utah state welfare population in July 1998
was drawn from administrative data routinely collected by the
state. In some cases, measures used by the agency for assessment
differed from those used in this study. This was true in the
assessment of mental health problems, as the caseworker simply noted
the presence of the problem rather than conducting a detailed
screening. This was also the case in measurement of domestic
violence and alcohol or drug abuse. In the case of other measures,
such as education, physical health, and child protective services
referral, the study data and the administrative data are more
directly comparable as they are both based on client self-report.
These differences in measurement approach suggest that fine-tuned
comparison of these data might be misleading.

(b) Based on Payne, N. (1998). Learning disabilities: A report by
the State of Washington Department of Social and Health Services,
Economic Services Administration WorkFirst Division.
Olympia: Washington State Department of Social and Health Services.

(c) Never worked more than six months at one job in past five years.

Table 2
Number of Barriers for Long-Term Welfare Recipients, by Those
Not Currently Working 20+ Hours and Those Currently Working 20+ Hours

 Not Currently Working Currently Working
 20+ Hours per Week 20+ Hours per Week Total
Number (n = 184) (n = 100) (N= 284)
of Barriers % %

0 2.8 4.9 7.7
1 7.0 5.6 12.7
2-4 37.4 19.8 57.2
5-8 17.7 5.0 22.7
Total 64.8 35.2 100.0

SOURCE: Adapted from Danziger, S., Corcoran, M., Danziger, S. H.,
Heflin, C., Kalil, A., Levine, J., Rosen, D., Seefeldt, K.,
Siefert, K., & Tolman, R. (1998, October).

Barriers to the employment of welfare recipients.
Paper pre-sented at the annual meeting of the Association for
Public Policy Analysis and Management, New York.

[X.sup.2] = 18.947.
* p <.015.

References

Achenbach, T. M. (1992). Manual for the Child Behavior Checklist/2-3 and 1992 profile. Burlington: University of Vermont, Department of Psychiatry.

Allard, M. A., Albelda, R., Colten, M. E., & Cosenza, C. (1997). In harm's way? Domestic violence, AFDC receipt, and welfare reform in Massachusetts. Boston: University of Massachusetts, McCormack Institute and Center for Survey Research.

American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed. rev.). Washington, DC: Author.

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author.

Bane, M. J., & Ellwood, D. T. (1994). Understanding welfare dynamics. In M. J. Bane & D. T. Ellwood (Eds.), Welfare realities: From rhetoric to reform (pp. 28-66). Cambridge, MA: Harvard University Press.

Blank, R. (1989). Analyzing the length of welfare spells. Journal of Public Economics, 39, 245-273.

Danziger, S. H. (2000, November). Approaching the limit: Early lessons from welfare reform (Working Paper 195). Retrieved January 21, 2004, from http://www/jcpr.org/wpfiles/ruraldanziger.PDF

Danziger, S. H., Corcoran, M., Danziger, S. H., Heflin, C., Kalil, A., Levine, J., Rosen, D., Seefeldt, K., Siefert, K., & Tolman, R. (1998, October). Barriers to the employment of welfare recipients. Paper presented at the annual meeting of the Association for Public Policy Analysis and Management, New York.

Duncan, G. J., Harris, K. M., & Boisjoly, J. (1997, April). Time limits and welfare reform: New estimates of the number and characteristics of affected families (Working Paper 1). Retrieved January 21, 2004, from http://www.jcpr.org/publications

Gerdes, K. E. (1997). Long-term AFDC mothers and posttraumatic stress syndrome: Is there a connection? Affilia, 12, 359-367.

Harris, K. M. (1991). Teenage mothers and welfare dependency: Working off welfare. Journal of Family Issues, 12, 492-518.

Lerman, R. I., & Ratcliffe, C. (2000). Did metropolitan areas absorb welfare recipients without displacing other workers? Washington, DC: Urban Institute.

Maryland Department of Human Resources. (1998). Highlights from life after welfare: Second interim report. Retrieved April 6, 2001, from www.dhr.state.md.us/life.htm

Meyer, D. R., & Cancian, M. (1996). Life after welfare. Public Welfare, 54, 25-30.

Mjelde-Mossey, L. A., & Mor-Barak, M.E.M. (1998). The conceptual and empirical link between health behaviors, self-reported health, and the use of home health care in later life. Home Health Care Services Quarterly, 17, 71-89.

National Governors' Association. (1998). Preparing youth for the workforce under welfare reform. Retrieved January 21, 2004, from www.nga.org/center/divisions/1,1188,C_ISSUE_BRIEF^D_1859,000.html

Olson, K., & Pavetti, L. (1996). Personal and family challenges to successful transition from welfare to work. Washington, DC: Urban Institute.

Parrott, S. (1998). Welfare recipients who find jobs: What do we know about their employment and earnings? Washington, DC: Center on Budget and Policy Priorities.

Pavetti, L. (1996, May 23). Time on welfare and welfare dependency: Hearings before the Committee on Ways and Means, House of Representatives, 104th Cong., 2nd Sess.

Pavetti, L. (1997). Moving up, moving out or going no-where? A study of the employment patterns of young women. Washington, DC: Urban Institute.

Pavetti, L, Olson, K., Pindus, N., & Pernas, M. (1996). Designing welfare-to-work programs for families facing personal or family challenges: Lessons from the field. Washington, DC: Urban Institute.

Payne, N. (1998). Learning disabilities: A report by the State of Washington Department of Social and Health Services, Economic Services Administration WorkFirst Division. Olympia: State of Washington.

Personal Responsibility and Work Opportunity Reconciliation Act of 1996, P.L. 104-193, 110 Stat. 2105.

Peterson, C. D. (1995). Female-headed families on AFDC: Who leaves welfare quickly and who doesn't. Journal of Economic Issues, 29, 619-629.

Radloff, L. S. (1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1,385-401.

Rockefeller Institute of Government. (1999). After welfare: A study of work and benefit use after case closing. New York: State Office of Temporary and Disability Assistance and State Department of Labor.

Salomon, A., Bassuk, S. S., & Brooks, M. G. (1996). Patterns of welfare use among poor and homeless women. American Journal of Orthopsychiatry, 66, 510-525.

Sherman, A., Amey, C., Duffield, B., Ebb, N., & Weinstein, D. (1998). Welfare to what? Early findings of family hardship and well being. Washington, DC: Children's Defense Fund.

Spalter-Roth, R., Burr, B., Hartmann, H., & Shaw, L. (1995). Welfare that works: The working lives of AFDC recipients. Washington, DC: Institute for Women's Policy Research.

State Policy Documentation Project. (2000, July). Findings in brief: time limits. Retrieved January 21, 2004, from http://www.spdp.org/tanf/timelimits/timelimits

Survey Research Center, Institute for Social Research. (1997). Women's employment study survey instrument. Ann Arbor: University of Michigan.

Taylor, M. J. (1999). Race and regional unemployment as predictors of exit from AFDC. Journal of Social Service Research, 25, 1-18.

Taylor, M. J., & Barusch, A. S. (2000). Multiple impacts of welfare reform in Utah: Experiences of former long-term welfare recipients. Unpublished manuscript, University of Utah, Social Research Institute, Graduate School of Social Work, Salt Lake City.

Tolman, R., & Raphael, J. (2001). A review of the research on welfare and domestic violence. Journal of Social Issues, 56, 655-681.

U.S. Department of Health and Human Services. (1999). President Clinton will announce record numbers of people on welfare ate working as businesses hire from the welfare rolls. Retrieved January 21, 2004, from www.acf.hhs.gov/news/press/1999/w2w0803htm

White, M. C., Tulsky, J. P, Dawson, C., Zolopa, A. R., & Moss, A. R. (1997). Association between time homeless and perceived health status among the homeless in San Francisco. Journal of Community Health, 22, 271-282.

Mary Jane Taylor, PhD, CSW, is associate professor, College of Social Work, and principle investigator, Understanding Families with Multiple Barriers to Self Sufficiency study, University of Utah, 395 South 1500 East, Salt Lake City, UT 84112-0260; e-mail: lmjtaylor@socwk.utah.edu. Amanda Smith Barusch, PhD, is professor, College of Social Work, University of Utah.

Original manuscript received September 29, 1999

Final revision received May 3, 2001

Accepted May 10, 2001