Social work advocacy in the post-TANF environment: lessons from early TANF research studies.

Anderson, Steven G. ; Gryzlak, Brian M.

Social work is distinguished by its history of services provision and advocacy on behalf of people who are poor. The NASW Code of Ethics asserts that the profession's "social change efforts are focused primarily on issues of poverty, unemployment, [and] discrimination" (p. 5), and calls on social workers to "advocate for changes in policy and legislation to improve social conditions in order to meet basic human needs" (NASW, 2000, p. 27). However, the profession also has been criticized for deviating from this historical mission (Specht & Courtney, 1994), and it has been argued that this drift has contributed to waning support for public assistance programs (Stoesz, 1999).

The passage of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA; P.L. 104-193) and subsequent implementation of state Temporary Assistance for Needy Families (TANF) programs present new challenges for social workers advocating on behalf of low-income people. PRWORA eliminated poor families' federal entitlement to public assistance and imposed strict employment and training requirements. When coupled with the devolution of public welfare decision making to the states, a need arose for advocacy coalitions at the state level (Schneider & Netting, 1999).

Because these welfare policy changes received considerable bipartisan support, strong political incentives exist to label TANF programs as successful. Thus, the approximately 50 percent declines in welfare caseloads since TANF passage have fueled speculation that welfare reform is working well (Loprest, 1999). Yet, caseload reductions do not necessarily imply program success, and a focus on such declines may divert attention from the well-being of people leaving welfare.

We assume that social workers should play a major role in monitoring TANF programs, conducting TANF research, and using knowledge gained from program assessments to advocate for low-income people. Schneider and Netting (1999) have delineated changes in advocacy processes needed to enhance social workers' effectiveness in decentralized social services environments such as TANF. However, the profession has not adequately articulated the substantive TANF-related policies on which social work advocates should target their efforts, and TANF research findings generally have not informed advocacy discussions.

This article synthesizes early research findings on selected outcomes for people leaving TANF and uses these findings to develop related advocacy strategies. We review initial TANF research findings on employment and wage experiences, the stability of TANF exits and related recidivism rates, and the use of support services by those who leave. We recommend social work advocacy strategies related to these issues at the policy, administrative, and direct practice levels. We also discuss areas in which further research would foster better understanding of TANF program results and advocacy needs.

Developing substantive TANF advocacy strategies is particularly timely, because of political circumstances and increasing knowledge emerging about TANF program results. TANF research findings from many state studies are now available, which allow for initial interpretations of TANF program experiences. Simultaneously, the accumulation of unspent TANF funds in most states, and TANF reauthorization discussions create realistic windows of opportunity for advocates.

Program Review Approach

One of the greatest difficulties in examining TANF program effects is that the devolution of program responsibility has resulted in a plethora of programs with different services and rules. This program diversity extends beyond the state level, because many states allow program variations at the local level. Although many states and localities have released reports on TANF program experiences, national and cross-state findings are limited. Only one study has presented national findings on TANF recipients (Loprest, 1999), and this study mixes pre- and post-TANF experiences. Several researchers have reported cross-state findings based on studies from selected states (Acs & Loprest, 2001; Brauner & Loprest, 1999; Parrott, 1998; Tweedie, Reichert, & O'Connor, 1999; U.S. General Accounting Office, 1999b). Single-state TANF leaver studies are much more common, because most states track TANF leavers through surveys, administrative data analyses, or both. However, these studies vary greatly in methodological techniqu es and rigor, as well as in the questions asked.

Our approach involved two strategies for analyzing early state findings. We first reviewed cross-state studies to determine common program findings. Because these studies typically included few large states or else did not include substantive areas of interest, we reviewed studies from the 20 states with the largest 1999 TANF caseloads. Only studies that met the following methodological criteria were included in our analysis. First, the studies must have used either surveys or administrative data based on random sampling techniques. Second, survey findings had to be based on telephone or in-person interviews. Third, because the welfare policy environment changed so dramatically with the implementation of TANF, only studies that examined welfare leavers after TANF program implementation were included. Fourth, state studies that were limited to special populations of leavers, such as recent applicants or those who were sanctioned or had exceeded time limits, also were excluded. Fifth, because people who remain on welfare often differ from people who leave, we only included studies that provided separate analyses of leavers. Finally, we did not review studies that excluded people who recycled to TANF, because these studies underrepresent leavers with unsuccessful outcomes.

Studies were accessed by searching welfare monitoring projects and state departmental Web sites and through telephone calls to state TANF agencies. Seventeen of the 20 states released studies on TANF recipients or leavers between 1997 and 2000. However, only 12 studies met all of the criteria and were included in our analysis. Of these, eight provided statewide data. Of the remaining four states, Virginia included all counties that had fully implemented TANF work programs (roughly two-thirds of the state caseload), New Jersey included six counties, Ohio was conducted in Cuyohoga County (Cleveland), and Louisiana was conducted in New Orleans. Table 1 provides basic information on the studies reviewed. For simplicity of presentation, we refer to all of these studies as "state" studies unless otherwise specified.

The three issues selected for analysis are important in assessing program outcomes for TANF leavers, and each provides realistic opportunities for social work advocacy. Other TANF issues deserve similar attention, such as the characteristics and special problems of those who remain on TANF, quality of employment and training programs and caseworker assistance, application of sanction policies, and effect of time limits. In addition, research results on noneconomic effects of welfare reform, such as effects on psychological well-being and relationships with children, require research scrutiny. However, the focus on the three issues presented here allows presentation of sufficient detail to develop useful advocacy strategies around a critical subset of issues.

Two additional differences in the studies reviewed are noteworthy. First, comparability between states sometimes is limited according to whether findings are based on administrative records or survey data. Although less influenced by recall problems than surveys, administrative data analyses often are limited in the depth of content provided. Thus, some states use each of these approaches in a complementary manner. A second technical limitation results from variations in interviewing times of one to 24 months, although most studies conducted interviews five to 12 months after TANF exit. Such data collection timing issues can substantially affect results, because both program and other effects are sensitive to the passage of time. Consequently, we restricted some analyses to studies with similar interviewing periods if observation points outside this range appeared to affect findings.

Findings

In the following sections, we present program themes that emerged from our analyses. Even after applying the indicated selection criteria, the studies reviewed differed both in question content and technical procedures. Therefore, the findings are intended to convey only approximate ranges about various program outcomes, as opposed to precise population estimates. For less commonly covered program themes, specific state examples are presented to illustrate issues requiring additional attention.

New Working Poor: Employment Experiences of TANF Leavers

Employment most commonly was offered as the principal reason for leaving TANF in open-ended survey responses about exit reasons. All of the state studies examined employment rates at various time points, most often when respondents were interviewed during the first year after leaving TANF. The majority of respondents were employed when interviewed, with employment rates ranging from 55 percent to 65 percent for the studies that provided statewide data. The only study that reported findings well outside this range was New Orleans, Louisiana, where only 34 percent of respondents were employed. Overall, the state employment ranges closely paralleled National Survey of America's Families (NSAF) employment rates reported by Loprest (1999), as well as other cross-state analyses (Acs & Loprest, 2001; Brauner & Loprest, 1999; Tweedie et al., 1999). Although these employment rates often are presented in positive terms, it is noteworthy that between 35 percent to 45 percent of the leavers in most study states were not working when interviewed.

Average wages alone were inadequate to raise working leavers above the poverty level in all the states that presented wage rate information except New York. Because of differing methodologies, the studies reviewed included both hourly and quarterly earnings rates. Four states reported mean hourly pay rates, and these ranged from $7.10 to $7.70 an hour among surveyed leavers (AZ, IL, OH, VA). Three states (AZ, IL, NY) used administrative wage files to estimate mean earnings in the quarter after exit and found levels ranging from $2,142 to $3,868. Assuming fulltime year-round work (35 hours for 50 weeks), a wage rate of $7.80 an hour or $3,420 per quarter was required to lift a family of three above the 1998 poverty level. Again, these data closely parallel findings from the previous cross-state studies.

TANF proponents have argued that the work experience gained by leavers after exit will result in human capital formation and subsequent wage growth (Anderson, Halter, Julnes, & Schuldt, 2000). The Arizona, Illinois, and New York studies used administrative wage files to examine earnings growth over the first year after TANF exit and found aggregate mean earnings growth of 9 percent to 15 percent. The Acs and Loprest (2001) study reported similar earnings growth in additional state studies. These data do not allow a determination of the extent to which earnings increase reflected additional hours worked versus wage increases, but research by Cancian, Haveman, Kaplan, Meyer, and Wolfe (1999) suggested that increases in hours worked are particularly important.

Leavers worked an average of 35 to 37 hours per week in the four states that reported mean hours worked (AZ, IL, VA, WA), suggesting that working leavers generally found full-time work. Occupational data also were reported in only four studies (IL, MA, VA, WA), and occupational reporting categories differed somewhat. Services, clerical and administrative support, and sales were the most frequently reported categories.

Double Revolving Door: In and Out of Jobs and On and Off TANF

Pre-TANF research demonstrated that welfare leavers often moved in and out of jobs, and that these job transitions sometimes were accompanied by returns to welfare (Edin & Lein, 1997; Harris, 1996). The state studies similarly found considerable job instability. Four states asked respondents if they had worked at any time between leaving TANF and being interviewed, with the interviews occurring six to 16 months after exit (IL, MA, VA, WA). Between 85 percent and 87 percent reported ever having worked since leaving TANF in these states, which was 20 to 30 percentage points higher than the work levels at the time of interviews. This common pattern of work at some point in time, coupled with employment inconsistency, is illustrated by the Cuyohoga County Ohio study. Although 67 percent of leavers were working when interviewed six months after exit, and 87 percent had worked at some point since leaving TANF, only 48 percent had worked consistently.

Similarly, Illinois and New York used administrative wage files to analyze employment stability; only 40 percent of New York leavers and 39 percent of Illinois single-parent leavers worked in all four quarters the first year after exit. In comparison, 66 percent of the leavers in New York and 70 percent of Illinois single-parent leavers worked at some point during this first year. Even those who worked in all four quarters may not have worked consistently, as administrative wage file comparisons indicate only whether a leaver worked at any time during a quarter.

Given such job instability, it is not surprising that TANF leavers frequently recycled back to TANF. Although most states reported on recycling, only four provided cumulative recidivism rates for the first year after leaving (AZ, IL, MA, NY). These rates ranged from 21 percent to 35 percent, and are similar to the 29 percent rate reported in the NSAF study (Loprest, 1999). Most of this recycling occurred in the first three to six months after exit. Because each of these states applied criteria that a person must have left for one to two months to be considered a leaver, these data do not capture very short-term recycling incidents.

Six states asked those who recycled back to TANF to indicate why returns had been necessary (AZ, CA, IL, OH, VI, WA). Employment- or earnings-related reasons were offered most commonly in each of these states, with recyclers typically indicating they had lost or been unable to find jobs. The Illinois study found that TANF recyclers were more likely to be poorly educated, African American, never married, lacking work experience, and living in the metropolitan Chicago area than leavers who remained off TANF. These findings suggest that a sizable subset of leavers experience barriers that complicate their efforts to remain off TANF.

The role of noncompliance with program requirements and related sanctions also is important when considering TANF recidivism. Six state studies indicated that noncompliance issues ranked only behind employment as a reason for TANF case closure (AZ, IL, LA, NJ, VA, WA), but information on sanctioned cases generally was not provided. However, many states have conducted separate studies on sanctioned cases, which have been well summarized by the U.S. General Accounting Office (GAO) (2000). The GAO study estimates that about 23,000 cases did not receive benefits in an average month in 1999 because of full-family sanctions. Furthermore, the state studies reviewed by GAO found that people who left TANF as a result of sanctions were, on average, less educated and less likely to become employed than people who left TANF for other reasons. The most common reported reason for sanctions was failure to comply with employment and training requirements, and nearly one-third of all people sanctioned eventually had full ben efits restored.

Social Contract and Government Assistance: Support Services Use by Leavers

Efforts to improve supports for those who leave welfare preceded TANF, with expansions of the earned income tax credit (EITC), child care subsidies, and medical coverage. Along with food stamps, these benefits form a package of supports intended to assist leavers in making the transition from welfare. The state reports generally do not provide consistent information across this entire range of services. However, findings on specific services suggest both common use patterns and variations between state programs.

Medical Coverage. As part of the 1996 welfare reforms, [section]1931 was added to the Social Security Act to "de-link" Medicaid from cash assistance. As a result, TANF leavers continue to be eligible for Medicaid if they meet the state's income and resource limits that were in effect for the Aid to Families with Dependent Children program in 1996. The only exception to this is that states have the option to terminate Medicaid coverage for nonpregnant adults if the TANF case closure was a result of noncooperation with TANF work and training requirements (Schott & Mann, 1998).

In addition, TANF leavers are eligible for up to 12 months of Transitional Medicaid Assistance if they left TANF because of increased work hours or income from employment, and some states have extended such coverage for up to three years (Families USA, 1999). Additional eligibility is provided to children both through low-income Medicaid extensions and through new state Children's Health Insurance Programs (CHIP). For example, Medicaid coverage for all children below the poverty level is being phased in by 2002, and coverage for pregnant women and children under age 6 with incomes below 133 percent of the poverty level also is required (Schott & Mann, 1998).

Collectively, these policies should allow Medicaid or CHIP coverage for nearly all children and most adults in the initial periods after leaving TANF. In addition, children's coverage is not time limited as long as family incomes hover near the poverty level, and even at higher income levels in many states. However, falling Medicaid rolls in the period after TANF was implemented raise questions concerning whether states have properly implemented Medicaid policies and whether leavers understand the services to which they are entitled (Ellwood, 1999; U.S. General Accounting Office, 1999a).

Only five states provided survey data on health insurance coverage for both leavers and their children (AZ, FL, IL, VA, WA). Of these, insurance coverage rates for leavers ranged from 49 percent to 74 percent. Coverage for children was 7 percent to 21 percent higher in these states, ranging from 70 percent to 87 percent. Despite these state variations, the overall coverage patterns parallel Loprest's (1999) NSAF findings, which indicated that 59 percent of adult leavers and 75 percent of their children had medical coverage.

Medicaid was the most common type of insurance coverage. Use rates varied substantially according to whether states reported use at a point in time versus at any time since leaving. Of the four states that provided data on Medicaid use by leavers at the time of interviews (IL, LA, OH, WA), coverage varied from 40 percent to 59 percent. Lack of knowledge about eligibility rules was cited in all states that asked about reasons for nonuse (AZ, FL, IL, VA, WA). For example, only 44 percent of Florida leavers thought they might be eligible for Medicaid after leaving TANF, and in Georgia 36 percent of leavers thought that Medicaid ended with TANF.

Employer-based health coverage was very limited for working leavers, with the percentage of employed leavers covered through employers ranging from only 20 percent to 28 percent in four state studies (AZ, NJ, VA, WA). These low rates apparently resulted from two factors. First, leavers often were not offered medical coverage through their jobs. Second, many who were offered declined the coverage, partially because of cost issues. In Washington, for example, 36 percent of working leavers were offered medical coverage through their jobs, but only about half of them enrolled in the plans.

Food Stamps. Because food stamp income eligibility limits are much higher than those used for TANF, most families continue to be eligible for food stamp benefits when they leave TANF. Food stamps are not time limited, so leavers may continue to supplement earnings with food stamps indefinitely if they meet income requirements. Nonetheless, national food stamp caseloads have fallen rapidly since PRWORA (Zedlewski & Brauner, 1999).

Six states reported on food stamp use at the time of the interviews (IL, LA, MA, OH, VA, WA). Food stamp use rates were lower than Medicaid use rates in all of these studies except Louisiana, and variations in use rates between states also were greater. Food stamp use ranged from a low of 27 percent in Massachusetts to 54 percent in Virginia. As with Medicaid, lack of knowledge or confusion about eligibility commonly was offered as a reason for not receiving food stamps. Lack of use was not necessarily reflective of lack of need, because more than one-fourth of leavers in most state studies reported having food shortages (AZ, FL, GA, IL, LA, NJ, OH, WA).

Child Care. Unlike Medicaid, states are not required to provide child care subsidies for families who leave TANF. When PRWORA was passed, several federal child care funding streams were combined into the Child Care and Development Fund (CCDF) block grant. States also have the option of transferring up to 30 percent of their TANF block grants to provide child care. These changes allow states to substantially increase child care funding (Long & Clark, 1997). However, given the TANF emphasis on work, the demand for child care services also has increased. The U.S. Department of Health and Human Services (2000) has estimated that only 12 percent of those eligible received federally supported child care in 1999.

Five of the state studies provided data on child care subsidy use at the time of interviews (IL, LA, OH, VA, WA). These rates ranged from 3 percent in the New Orleans, Louisiana, study to 22 percent in the Cuyohoga County, Ohio, study. Relatives or other nonlicensed providers were used more often than licensed providers in each of the five states that asked how children were cared for (AZ, IL, MA, VA, WA). Washington, Illinois, and Virginia reported that leavers estimated average monthly out-of-pocket child care costs of $141 to $211. The Washington study illustrates the importance of public subsidies in reducing child care costs; although those with state subsidies paid 4 percent of their income for child care, those without subsidies paid 14 percent.

Earned Income Tax Credit. This refundable tax credit is the most important governmental income support for the working poor. The EITC has been expanded three times since the early 1980s and now incurs more federal government expenditures than TANF. Working families with two or more children can receive an EITC of up to $3,888 (Sweeney et al., 2000). The EITC is received through the federal income tax system, but one does not have to owe federal taxes to apply. Only four states asked leavers about their knowledge and use of the EITC (AZ, IL, MA, NJ). Response patterns were similar in these states, with a substantial number of leavers not knowing about the credit and even fewer receiving it. For example, of the Massachusetts leavers who were working, only 61 percent knew about the EITC, and only 45 percent said they had claimed it.

Developing Advocacy Strategies

The preceding findings are limited both by the number of state studies and differing study methodologies, which suggest caution in generalizing findings to other TANF program environments. Nonetheless, the similarity of many key findings, both between states and with previous studies, provides confidence that these data can usefully point to common advocacy directions. Variations among states also are to be expected given the administrative discretion granted to states and demonstrate the importance of state and local program monitoring. As additional state studies become available, the data presented here can serve as bases for program comparisons.

The findings indicate the need both for broad TANF-related policy revisions and for improvements in implementing existing policies. Social workers operating at the policy, administrative, and direct services levels all can play useful roles in advocating for such TANF improvements. Claims of TANF program success to date have been based largely on the unprecedented caseload reductions that have occurred. Therefore, in developing an overarching advocacy orientation, social workers must reject this narrow standard and instead argue for the use of economic and social well-being measures in evaluating program outcomes.

The variations in program outcomes among states demonstrate the importance of understanding state political contexts, and of learning more about successful program models from other states. Furthermore, because devolution of services has been accompanied by increased contracting of services to both nonprofit and profit agencies, advocacy efforts must not be limited to public decision makers and agencies. In particular, employment and training services and many other support services often are provided by non-government agencies, so advocates must clearly consider the needs, capabilities, and limitations of these non-government services providers.

In the following sections, we use findings from the research studies reviewed to outline policy and implementation changes intended to foster the well-being of TANF leavers. The recommended approaches are not exhaustive but rather respond to the issues examined in this article. We likewise recognize that attention to some approaches will preclude attention to others. However, given the decentralized TANF program environment, we believe that providing a range of options allows advocates to select strategies that are politically feasible in and responsive to varying program contexts. Because the state studies leave many important questions about leaver experiences unanswered, issues requiring further research attention also are presented.

Advocacy Related to TANF Exits and Wage Rate

The state findings show fairly consistent employment rates for TANF leavers, generally in the 55 percent to 65 percent range. Average wage rates in the early months after exit also are similar among the states and are insufficient to raise most family incomes above the poverty level. This suggests that, at least among these early leavers, TANF has been much more effective in reducing welfare caseloads than in reducing poverty.

These low wages underscore the importance of continued development of income support strategies for TANF leavers and other low-income workers. When advocating for welfare reforms in the early 1990s, President Clinton and his supporters argued that people who worked full-time should not remain in poverty (Danziger, 1999). When coupled with a growing perception that welfare reform has increased work effort, this argument should remain potent.

Perhaps the most effective income enhancement strategy for low-wage workers is to further expand the EITC. The EITC already provides up to a $2 per hour wage subsidy for workers receiving the maximum credit. Limited expansion would ensure that even full-time minimum wage workers would receive incomes above the poverty level without the need to obtain food stamps (Wilson, 1996). Eleven states have developed state earned income tax credits to complement the federal EITC (Sweeney et al., 2000), and such policies represent another viable strategy for increasing income for TANF leavers and low-income workers more generally. These state earned income tax credits can be based on a simple percentage of the federal EITC; the credit can reward work effort and provide relief from regressive state sales and property taxes.

A related strategy, which focuses on the administrative and direct levels, is to ensure that eligible workers receive the EITC. As a services component to improving the economic prospects of leavers, TANF agency administrators should ensure that caseworkers routinely provide information on the EITC both to working TANF recipients and to TANF leavers. Such information needs to be provided in clear, simple language, and should stress that the credit is available even if taxes are not owed. Presenting examples of EITC effects for specific earnings rates, such as full-time work or part-time work at a common leaver wage level, may demonstrate more clearly the substantial benefits available. Similar efforts by TANF employment and training contractors and other agencies that serve people who are poor would be useful. The tax preparation services offered by nonprofit social services agencies in some communities also hold promise for improving EITC use.

Another approach favored by some advocates is to increase the minimum wage. Even with expansions in the 1990s, the minimum wage remains below 1970s levels (Mishel, Bernstein, & Schmitt, 2001). Although average leaver wages exceeded the minimum wage in the state studies reviewed, TANF recipients and leavers who enter minimum wage jobs would directly benefit from such increases. Furthermore, increasing the minimum wage may provide upward pressure on wages, which could help TANF leavers earning more than the minimum wage.

Living wage strategies move beyond minimum wage policies by passing local ordinances requiring businesses that benefit from government assistance to pay their workers at least minimally established wages. The wage levels are determined as part of the local living wage campaigns, but generally are sufficient to raise earnings at least to the poverty level. These policies often stimulate political opposition, and offer less short-term potential for working poor people than either EITC or minimum wage strategies. Nonetheless, living wages policies have been passed in over 50 communities, including several large cities (ACORN Living Wage Resource Center, 2000). Although the number of low-wage workers affected by living wage ordinances is small, these measures help promote public debate about socially agreed-on minimum levels of compensation for work. Social work advocates, therefore, may find it useful to engage in community living wage campaigns as part of a longer-term strategy to elevate worker compensation i ssues more firmly onto public decision-making agendas.

State studies to date usually have focused narrowly on average employment levels. Consequently, the diversity of employment outcomes experienced by leavers, and the causal factors affecting work effort, are important areas for further research. For example, Moffitt and Roff (2000) found worse employment outcomes for leavers who have been sanctioned; have longer periods of welfare dependency; and have poor health, low education, or younger children. Danziger, Kalil, and Anderson (2000) have reported that about one-third of a sample of current and former Michigan TANF recipients met diagnostic criteria for depression, posttraumatic stress, or generalized anxiety disorder, and that these and other employment barriers are associated with lower work levels.

Given that 35 percent to 45 percent of the leavers in most of the state studies reviewed were not working when they left TANF, and that time limits will force more ill-prepared people from TANF, these research findings reinforce the need for greater attention to hard-to-serve TANF recipients and leavers. More research is needed on the specific employment barriers that nonworking leavers encounter, the survival strategies they use in the absence of regular employment, and the manner in which they differ from employed leavers on exit reasons, work experience and preparation, and other factors. Research on the sanctioning procedures used by states, the appropriateness and consistency of sanctioning enforcement by caseworkers, and the effect that different sanctioning policies have on the incidence of nonworking leavers also would be useful.

Advocacy Related to Employment Instability and Recidivism

Research to date indicates that between 21 percent and 35 percent of TANF leavers will recycle back to the program within one year. This most commonly occurs because leavers cannot find employment, lose their employment, or receive inadequate compensation to provide for their families. Although the improvement of wage supplements and support services discussed in other sections should alleviate earnings problems, measures designed to improve job quality, transitional income support, and assistance after job loss are needed.

Most states have reinforced the tendency of leavers to labor in low-wage and unstable jobs by focusing on "work first" employment and training policies, which serve to push welfare recipients into the first available job. This model is relatively easy to administer and has allowed states to meet federal work participation requirements. However, such an orientation typically has come at the expense of developing longer-term educational and training opportunities (Greenburg & Laracy, 2000; Van Lare & Griener, 2000). This is ironic, given that research has shown that wage returns for education have increased in recent years (Wilson, 1996).

Consequently, it is important that state TANF offices develop a mix of employment and training options responsive to the needs of individual recipients. This requires, at the policy and administrative levels, the establishment of educational opportunities beyond the high school level, as well as the development of vocational opportunities tailored to local employment markets. For example, Illinois, Kentucky, Maine, and several other state programs allow extended college attendance as an employment and training activity, and many public--private vocational training partnerships are emerging (Newman, 1999; Sweeney et al., 2000).

At the direct services level, sound case management models can help recipients select appropriate work and training options and allow monitoring of progress in these activities. For example, in Union County, New Jersey, a case management team, including a social worker, a job coach, a job developer, and other specialized staff, develops a services plan with recipients and then follows up with employers after job placement (U.S. Department of Labor, 1999). In Rhode Island, caseworkers encourage job retention by maintaining contacts with clients after they enter employment (Sweeney et al., 2000). Herr, Wagner, and Halpern (1996) also have developed intensive postemployment services models to provide tangible support and encouragement as people with multiple barriers begin working.

Although improving the quality of job opportunities for TANF leavers is important, many still will move into low-paying and unstable jobs. Therefore, improving transitional income supports for leavers is critical, as is the need to offer timely support when job situations fail. One approach for improving initial income stability is to allow working TANF recipients to continue receiving some TANF benefits until they reach reasonable earnings levels. Since the passage of PRWORA, most states have improved their earned income disregard policies, which increase the portion of earnings that TANF recipients can keep. This allows recipients to combine earnings and TANF benefits at higher income levels before totally losing TANF benefits. Nonetheless, families in most states become ineligible for TANF at income levels well below the poverty level, so further expansions of these earnings disregard policies would be useful.

Another approach is to provide transitional income support or asset-building capabilities to people when they first leave TANF. Income support could be provided through stipends for workers with low earnings, with the intent to assist with work-related expenses (Center on Budget and Policy Priorities, 1998). Recipients who earne(l enough to leave TANF also could be allowed to keep TANF benefits for a transitional period (Anderson, Halter, & Gryzlak, 2000). Such income supports could be financed from TANF block grant funds or with state "maintenance of effort" funding required by PRWORA. Texas has established a program that allows stipends to pay for activities intended to promote job stability and mobility (Sweeney et al., 2000).

Some analysts also have suggested establishing individual development accounts (IDAs) (Sherraden et al., 2000). These savings accounts match individual contributions with government or other funds, with the savings then used by individuals for purposes such as education, training, or initiating a business. This strategy recognizes that many people have difficulty earning their way out of poverty without increasing their human capital or other assets. TANF funds may be used to support IDAs, and nine states are using TANF funding for this purpose. Limited federal funding also is being provided through the federal Assets for Independence Act (P.L. 105-285), and several states also have allocated state funds for IDAs (IL, IN, IA, MN, NC, PN).

Finally, as long as recipients leave TANF for low-wage jobs, high job turnover is to be expected. Consequently, support systems improvements are needed for those who lose jobs. One strategy is to expand the unemployment compensation system, which currently serves only one in six low-wage workers (Center on Budget and Policy Priorities, 1998). Such shortcomings result because TANF leavers often do not have sufficient earnings to qualify. Revising the law to allow benefits after shorter employment spells, or else establishing separate unemployment assistance with TANF funds, could alleviate this problem.

Job stability and recidivism among TANF leavers also point to the importance of expediting re-entry to TANF, food stamps, and other supports after jobs are lost. Given that most recyclers return to TANF within six months after exit, they rarely can rely on savings for support as they await reapplication decisions. Therefore, policies and administrative practices that provide emergency benefits while applications are being processed, or else offer priority processing for leavers who have made good faith work efforts, should be encouraged.

Advocacy Related to Support Services and Case Management

Among the support services examined, two policy directions appear particularly important. First, the lack of health care coverage continues to hinder welfare-to-work transitions. The federal government and the states have made progress with Medicaid extensions and children's health initiatives. However, Medicaid time and income limitations for adult leavers in most states still are united are inadequate, given that most low-wage workers do not receive health insurance through their jobs. This may encourage low-wage workers to decrease work to receive medial benefits. Extending Medicaid or other health insurance coverage to all low-income adults and children without time limits is needed to reinforce work efforts. This coverage should be available to people with income up to 200 percent of the poverty level. New Jersey, Rhode Island, and Wisconsin already have been authorized to the use excess CHIP funds to provide adults coverage.

The inadequacy of child care assistance is a second major support services issue. Although states have improved child care coverage for TANF leavers, quality services at reasonable costs still are lacking. Consequently, it is not surprising that leavers often cite child care problems as employment barriers in the state leaver studies. Child care issues include costs, lack of available providers, and difficulties in meshing work and child care schedules. Child care services are even less likely to be available through employers than health insurance, which argues for the development of non-time-limited child care subsidies for low-income workers. Illinois has established such a non-time-limited program, which provides subsidies to families making 50 percent or less of the state median income on a sliding-fee scale. Several other states, including Florida, Missouri, Ohio, and Pennsylvania, have expanded training, startup initiatives, or higher fees for care during nondaytime hours.

At the administrative and direct practice levels, a primary concern should be ensuring that TANF leavers and other low-income people have easy access to existing support services. The state studies show troubling patterns of underutilization and lack of knowledge about support services intended to assist welfare-to-work transitions. Yet, high utilization rates in some states suggest that sound implementation strategies can substantially increase services use. Administrative advocacy should focus on developing program models, such as one-stop service centers, that allow recipients to access a variety of support services. Administrators likewise can improve access to support services by developing simple and coordinated application procedures, and straightforward materials for informing clients about benefits. Adequate training and skill development for caseworkers also is needed.

The establishment of outreach programs is another useful administrative strategy, particularly in areas with hard-to-serve clients. For example, in California, Fresno County food stamp workers regularly visit schools, health centers, shopping places, and community centers to distribute information and take applications (Chen, McLaughlin, & Manalo-LeClair, 2000). Los Angeles County has used food banks and pantries to distribute information about food stamps.

Direct services workers, both in public and contracted social services agencies, are the other critical actors in ensuring access to support services. Yet, TANF program development follows a long period of public agency caseworker deprofessionalization, and workers have faced information overload problems as they implement new employment and services policies. There is little evidence to indicate that states have upgraded the skills of caseworkers in this changing services environment, which suggests that caseworkers may be poorly equipped to perform more sophisticated case management functions.

Many will argue that the intent of TANF is to emphasize client responsibility and not government services provision, which may undercut support for case management services. However, welfare reform debates also focused on increased support for those who worked, and states have expanded many support services. The importance of caseworkers in fulfilling such supportive objectives is indicated by the fact that caseworker support was one of the most valued services by TANF participants in a series of post-employment services demonstration projects (U.S. Department of Labor, 1999).

Therefore, social workers need to advocate for more substantive case management roles in TANF. Critical to these roles is caseworker knowledge about the full range of employment and training opportunities and income and support services options in a service area. Case management development is difficult to sell politically, so demonstration projects testing different case management approaches are a good short-term strategy. Projects that experiment with differing caseload sizes, services options, caseworker qualifications, and casework philosophies are particularly important in this respect. For example, it would be useful to determine if models that empower clients to make choices among education and training options and other supports differentially affect client outcomes.

Opportunities for Advocacy

The current political environment offers useful opportunities for social workers to advocate for progressive changes in TANF-related policies. At the national level, PRWORA must be reauthorized in 2002, which raises the possibility of overarching program changes. Given caseload reductions, some people will argue for cuts in the TANF block grants. However, advocates should instead argue for income and support services improvements, particularly given the public perception that welfare reform has been successful.

Several advocacy groups are developing change agendas around TANF reauthorization (Greenburg & Laracy, 2000). Social workers should review these agendas and consider establishing coalitions around agreed-on issues.

States also are favorable targets for advocacy efforts. The reauthorization serves as a natural time to review state programs, so windows of opportunity for TANF policy changes should be open both preceding and following federal reauthorization. Most states have not expended their full TANF allotments (Lazare, 2000), which has resulted in unspent balances that only can be used for welfare-related services. Advocates can use two complementary approaches in trying to compel states to spend TANF surpluses. First, the publicity accompanying TANF caseload reductions has created the perception that welfare recipients have increased work efforts. When coupled with evidence about the financial and services difficulties still facing working poor people, proposals to improve income and services supports may be better received than in the past. Arguments that such supports are necessary to sustain the welfare exits so far experienced may be a useful corollary to this approach. Second, proposals to take back unspent sta te TANF funds received attention in the fiscal year 2001 federal budget process, and similar proposals will emerge in future budgets and in the reauthorization process. State officials may be amenable to spending TANF surpluses to minimize this possibility if reasonable proposals are made.

As would be expected in a devolved system with limited federal controls, TANF program rules, support services, and services delivery systems vary considerably between states. Therefore, as social workers become involved in state or local advocacy on behalf of TANF recipients and leavers, it is important that they become well-versed in the particular program models and procedures being used in their states. Such local program knowledge, when coupled with information on TANF outcomes such as presented here, can provide the foundation for clearly targeted advocacy efforts in decentralized program contexts.

Table 1

Welfare Leaver Reports


State Study Type of Data

Arizona Westra, K. L., & Routley, Administrative
 J. (2000). Arizona cash
 assistance exit study:
 First quarter 1998 cohort,
 final report, administrative
 and survey data results.
 Phoenix: Arizona Department
 of Economic Security,
 Office of Evaluation.

California California Department of Survey
 Social Services, Data
 Analysis and Publications
 Branch. (2000). CalWORKs
 leavers survey: A statewide
 telephone survey of former
 CalWORKs recipients.
 Sacramento: Author.

Florida Crew, R. E., & Eyerman, J. Survey
 (1999). After WAGES: Results
 of the Florida study.
 Tallahassee: Florida State
 University Press.

Georgia Foster, E. M., & Rickman, D. Survey
 K. (1999). Georgia welfare
 leavers study: Initial results.
 Atlanta: Georgia State
 University, Andrew Young
 School of Policy Studies,
 Applied Research Center.

Illinois Julnes, G., Halter, A., Administrative
 Anderson, S., Frost-Kumpf, Survey
 L., Schuldt, R., Staskon, F.,
 & Ferrara, B. (2000). Illinois
 study of former TANF clients:
 Final report. Springfield:
 University of Illinois,
 Institute for Public Affairs.

Louisiana Mancoske, R. J., Kemp, A. A., Survey
 & Lindhorst, T. (1998).
 Exiting welfare: The experiences
 of families in metro New
 Orleans. New Orleans: Southern
 University, School of Social
 Work, Welfare Reform Research
 Project.

Massachusetts Massachusetts Department of Survey
 Transitional Assistance. (1999).
 How are they doing? A longitudinal
 study of households leaving
 welfare under Massachusetts
 welfare reform. Boston: Author.

New Jersey Liu, A., Gable, M. C., Gold, Survey
 H., & Searcy, C. S. (1999).
 Assessing work first: What
 happens after welfare? Edison:
 Legal Services of New Jersey,
 New Jersey Poverty Research
 Institute.

New York Nelson A. Rockefeller Institute Administrative
 of Government. (1999). After
 welfare: A study of work and
 benefit use after case closing.
 New York: Author.

Ohio Coulton, C., Pasqualone, C., Survey
 Bania, N., Martin, T., Lalich,
 N., Fernando, M., & Li, F.
 (1999). How are they managing?
 A six-month retrospective of
 Cuyahoga County families leaving
 welfare. Cleveland: Case Western
 Reserve University, Mandel School
 of Applied Social Sciences,
 Center on Urban Poverty and
 Social Change.

Virginia Kuhns, C., Gorden, A., Agodini, Survey
 R., & Loeffler, R. (1999). The
 Virginia closed case study:
 Experiences of Virginia families
 one year after leaving TANF.
 Falls Church: Virginia
 Polytechnic Institute and
 State University, Center
 for Public Administration
 and Policy.

Washington Du, J., Fogarty, D., Hopps, Survey
 D., & Hu, J. (2000). A study
 of Washington State TANF
 leavers and TANF recipients:
 Findings from the April-June
 1999 telephone survey. Olympia,
 WA: Department of Social and
 Health Services, Economic
 Services Division, Office of
 Planning and Research.

 Length of Time
State Study after Exit

Arizona Westra, K. L., & Routley, 12 months
 J. (2000). Arizona cash
 assistance exit study:
 First quarter 1998 cohort,
 final report, administrative
 and survey data results.
 Phoenix: Arizona Department
 of Economic Security,
 Office of Evaluation.

California California Department of 2-10 months
 Social Services, Data
 Analysis and Publications
 Branch. (2000). CalWORKs
 leavers survey: A statewide
 telephone survey of former
 CalWORKs recipients.
 Sacramento: Author.

Florida Crew, R. E., & Eyerman, J. 1-24 months
 (1999). After WAGES: Results
 of the Florida study.
 Tallahassee: Florida State
 University Press.

Georgia Foster, E. M., & Rickman, D. 6-8 months
 K. (1999). Georgia welfare
 leavers study: Initial results.
 Atlanta: Georgia State
 University, Andrew Young
 School of Policy Studies,
 Applied Research Center.

Illinois Julnes, G., Halter, A., 6 quarters
 Anderson, S., Frost-Kumpf, 7-9 months
 L., Schuldt, R., Staskon, F.,
 & Ferrara, B. (2000). Illinois
 study of former TANF clients:
 Final report. Springfield:
 University of Illinois,
 Institute for Public Affairs.

Louisiana Mancoske, R. J., Kemp, A. A., 1-5 months
 & Lindhorst, T. (1998).
 Exiting welfare: The experiences
 of families in metro New
 Orleans. New Orleans: Southern
 University, School of Social
 Work, Welfare Reform Research
 Project.

Massachusetts Massachusetts Department of 3 months
 Transitional Assistance. (1999).
 How are they doing? A longitudinal
 study of households leaving
 welfare under Massachusetts
 welfare reform. Boston: Author.

New Jersey Liu, A., Gable, M. C., Gold, 5-8 months
 H., & Searcy, C. S. (1999).
 Assessing work first: What
 happens after welfare? Edison:
 Legal Services of New Jersey,
 New Jersey Poverty Research
 Institute.

New York Nelson A. Rockefeller Institute 4 quarters
 of Government. (1999). After
 welfare: A study of work and
 benefit use after case closing.
 New York: Author.

Ohio Coulton, C., Pasqualone, C., 6 months
 Bania, N., Martin, T., Lalich,
 N., Fernando, M., & Li, F.
 (1999). How are they managing?
 A six-month retrospective of
 Cuyahoga County families leaving
 welfare. Cleveland: Case Western
 Reserve University, Mandel School
 of Applied Social Sciences,
 Center on Urban Poverty and
 Social Change.

Virginia Kuhns, C., Gorden, A., Agodini, 10-16 months
 R., & Loeffler, R. (1999). The
 Virginia closed case study:
 Experiences of Virginia families
 one year after leaving TANF.
 Falls Church: Virginia
 Polytechnic Institute and
 State University, Center
 for Public Administration
 and Policy.

Washington Du, J., Fogarty, D., Hopps, 7-9 months
 D., & Hu, J. (2000). A study
 of Washington State TANF
 leavers and TANF recipients:
 Findings from the April-June
 1999 telephone survey. Olympia,
 WA: Department of Social and
 Health Services, Economic
 Services Division, Office of
 Planning and Research.

Original manuscript received January 3, 2001

Final revision received April 19, 2001

Accepted May 8, 2001

References

ACORN Living Wage Resource Center. (2000). Living wage ordinances [Online]. Available: http:11 www.livingwagecampaign.org.

Acs, G., & Loprest, P. (2001). Initial synthesis report of the findings from ASPE's "leavers" grants. Washington, DC: U.S. Department of Health and Human Services.

Anderson, S. G., Halter, A., & Gryzlak, B. G. (2000). Life after TANF: Chicago women talk about their experiences: A report to the Joyce Foundation. Urbana: University of Illinois, School of Social Work.

Anderson, S. G., Halter, A. P., Julnes, G., & Schuldt, R. (2000). Job stability and wage progression patterns among early TANF leavers. Journal of Sociology and Social Welfare, XXVII, 39-58.

Assets for Independence Act, P.L. 105-285, 112 Stat. 2759 (1998) 42 U.S.C.

Brauner, S., & Loprest, P. (1999). Where are they now? VVhat states' studies of welfare leavers tell us. Assessing the new federalism. Washington, DC: Urban Institute.

California Department of Social Services, Data Analysis and Publications Branch. (2000). CalWORKs leavers survey: A statewide telephone survey offomer Cal WORKs recipients. Sacramento: Author.

Cancian, M., Haveman, R., Kaplan, T., Meyer, D., & Wolfe, B. (1999). Work, earnings, and well-being after welfare. In S. H. Danziger (Ed.), Economic conditions and welfare reform (pp. 16 1-186). Kalamazoo, MI: W.E. Upjohn Institute for Employment Research.

Center on Budget and Policy Priorities. (1998). Reinvesting welfare savings: Aiding needy families and strengthening state welfare reform. Washington, DC: Author.

Chen, S., McLaughlin, D., & Manalo-LeClair, G. (2000). Best practices to improve the food stamp program in California. San Francisco: California Food Policy Institute.

Coulton, C., Pasqualone, C., Bania, N., Martin, T., Lalich, N., Fernando, M., & Li, F. (1999). How are they managing? A six-month retrospective of Cuyahoga County families leaving welfare. Cleveland: Case Western Reserve University, Mandel School of Applied Social Sciences, Center on Urban Poverty and Social Change.

Crew, R. E., & Eyerman, J. (1999). After WAGES: Results of the Florida study. Tallahassee: Florida State University Press.

Danziger, S. H. (1999). Introduction. In S. H. Danziger (Ed.), Economic conditions and welfare reform (pp. 1-14). Kalamazoo, MI: W.E. Upjohn Institute for Employment Research.

Danziger, S. H., Kalil, A., & Anderson, N. J. (2000). Human capital, health and mental health of welfare recipients: Co-occurrence and correlates. Journal of Social Issues, 54, 637-656.

Du, J., Fogarty, D., Hopps, D., & Hu, J. (2000). A study of Washington State TANF leavers and TANF recipients: Findings from the April-June 1999 telephone survey. Olympia, WA: Department of Social and Health Services, Economic Services Division, Office of Planning and Research.

Edin, K., & Lein, L. (1997). Making ends meet: How single mothers survive welfare and low-wage work. New York: Russell Sage Foundation.

Ellwood, M. (1999). The Medicaid eligibility maze: Coverage expands, but enrollment problems persist (Assessing the new federalism, Occasional Paper # 30). Washington, DC: Urban Institute.

Families USA. (1999). Losing health insurance: The unintended consequences of welfare reform. Washington, DC: Author.

Foster, E. M., & Rickman, D. K. (1999). Georgia welfare leavers study: Initial results. Atlanta: Georgia State University, Andrew Young School of Policy Studies, Applied Research Center.

Greenburg, M., & Laracy, M. C. (2000). Welfare reform: Next steps offer new opportunities. McLean, VA: Neighborhood Funders Group.

Harris, K. M. (1996). Life after welfare: Women, work, and repeat dependency. American Sociological Review, 61, 407-426.

Herr, T. S., Wagner, S., & Halpern, R. (1996). Making the shoe fit: Creating a work-prep system for a large and diverse welfare population. Chicago: Erikson Institute.

Julnes, G., Halter, A., Anderson, S., Frost-Kumpf, L., Schuldt, R., Staskon, F., & Ferrara, B. (2000). Illinois study of former TANF clients: Final report. Springfield: University of Illinois, Institute for Public Affairs.

Kuhns, C., Gorden, A., Agodini, R., & Loeffler, R (1999). The Virginia closed case study: Experiences of Virginia families one year after leaving TANF. Falls Church: Virginia Polytechnic Institute and State University, Center for Public Administration and Policy.

Lazare, E. (2000). Welfare balances after three years of TANF block grants: Unspent TANF funds at the end of federal fiscal year 1999. Washington, DC: Center on Budget and Policy Priorities.

Liu, A., Gable, M. C., Gold, H., & Searcy, C. S. (1999). Assessing work first: What happens after welfare? Edison: Legal Services of New Jersey, New Jersey Poverty Research Institute.

Long, S. K., & Clark, S. J. (1997). The new child care block grant: State funding choices and their implications. Washington, DC: Urban Institute.

Loprest, P. (1999). Families who left welfare: Who are they and how are they doing? (Discussion Paper No. 99-02). Washington, DC: Urban Institute.

Mancoske, R. J., Kemp, A. A., & Lindhorst, T. (1998). Exiting welfare: The experiences of families in metro New Orleans. New Orleans: Southern University, School of Social Work, Welfare Reform Research Project.

Massachusetts Department of Transitional Assistance. (1999). How are they doing? A longitudinal study of households leaving welfare under Massachusetts welfare reform. Boston: Author.

Mishel, L., Bernstein, J., & Schmitt, J. (2001). The state of working America: 2000/2001. Ithaca, NY: Cornell University Press.

Moffitt, R., & Roff, J. (2000). The diversity of welfare leavers. Welfare, children, and families: A three-city study (Policy Brief 00-2). Baltimore: Johns Hopkins University.

National Association of Social Workers. (2000). NASW code of ethics. Washington, DC: NASW Press.

Nelson A. Rockefeller Institute of Government. (1999). After welfare: A study of work and benefit use after case closing. New York: Author.

Newman, K. S. (1999). No shame in my game: Working poor in the inner city. New York: Alfred A. Knopf.

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.

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

Schneider, R. L., & Netting, F. E. (1999). Influencing social policy in a time of devolution: Upholding social work's great tradition. Social Work, 44, 349-357.

Schott, L., & Mann, C. (1998). Assuring that eligible families receive Medicaid when TANF assistance is denied or terminated. Washington, DC: Center on Budget and Policy Priorities.

Sherraden, M., Johnson, L., Clancy, M., Beverly, S., Schreiner, M., Zhan, M., & Curley, J. (2000). Savings patterns in IDA programs. St. Louis: Washington University, Center for Social Development.

Specht, H., & Courtney, M. (1994). Unfaithful angels: How social work abandoned its mission. New York: Free Press.

Stoesz, D. (1999). Unraveling welfare reform. Society, 36, 53-61.

Sweeney, E., Schott, L., Fremstad, S., Goldberg, H., Guyer, J., Super, D., & Johnson, C. (2000). Windows of opportunity: Strategies to support families receiving welfare and other low-income families in the next stage of welfare reform. Washington, DC: Center on Budget and Policy Priorities.

Tweedie, J., Reichert, D., & O'Connor, M. (1999). Tracking recipients after they leave welfare. Washington, DC: National Conference of State Legislatures.

U.S. Department of Health and Human Services, Administration for Children and Families. (2000). New statistics show only small percentage of eligible families receive child care help. HHS News [Online]. Available: http://www.acf.dhhs.gov/news/ccstudy.htm.

U.S. Department of Labor. (1999, September). Helping new workers keep jobs. Ideas that work (Issue 9) [Online]. Available: www.dol.gov.

U.S. General Accounting Office, Health, Education, and Human Services Division. (1999a). Medicaid enrollment: Amid declines, state efforts to ensure coverage after welfare vary (GAO/HEHS-99-163). Washington, DC: Author.

U.S. General Accounting Office, Health, Education, and Human Services Division. (1999b). Welfare reform: Information on former recipients' status (GAO/HEHS-99-48). Washington, DC: Author.

U.S. General Accounting Office, Health, Education, and Human Services Division. (2000). Welfare reform: State sanction policies and number of families affected (GAO/HEHS-00-44). Washington, DC: Author.

Van Lare, B. L., & Griener, G. (2000). Reauthorization of TANF--An early perspective. Welfare Information Network, Issue Notes (4) [Online]. Available: http://www.welfareinfo.org/.

Westra, K. L., & Routley, J. (2000). Arizona cash assistance exit study: First quarter 1998 cohort, final report, administrative and survey data results. Phoenix: Arizona Department of Economic Security, Office of Evaluation.

Wilson, W. J. (1996). When work disappears: The world of the new urban poor. New York: Alfred A. Knopf.

Zedlewski, S. R., & Brauner, S. (1999). Declines in food stamp and welfare participation: Is there a connection? (Discussion Paper No. 99-13). Washington, DC: Urban Institute.

Steven G. Anderson, PhD, is assistant professor, School of Social Work, University of Illinois at Urbana-Champaign, 1207 West Oregon, Urbana, IL 61801; e-mail: sandersn@uiuc.edu. Brian M. Gryzlak, MSW, is a graduate student, Department of Sociology and Anthropology, Western Illinois University, Macomb.