When social program responsibilities trickle down: impacts of devolution on local human services provision.

The Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA) shifted responsibility for public assistance from the federal government to the states. This study examined early impacts of this devolution and related program reductions on local service authorities in Illinois. Based on surveys from 101 large townships responsible for administering General Assistance, medical assistance, and emergency needs programs, we found that 60 percent of these localities experienced increased service demands. These demands not only placed pressure on limited local programming funds, but also transformed local service populations in subtle and unintended ways. Reports of bureaucratic mistreatment and confusion also were common as states implemented PRWORA changes. Local responses to increased service demands were variable, with many localities increasing expenditures but expressing reservations about longer term funding given local tax limits. Follow-up surveys with 40 township officials two years later found that a declining economy and impending Temporary Assistance for Needy Families (TANF) five-year time limits were intensifying township program concerns. The implications of these findings for the development and monitoring of state and local public assistance systems are discussed.

**********

The 1996 Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) not only substantially reduced benefit entitlements for disadvantaged citizens, but also provided states with unprecedented latitude to design their own public welfare systems. Proponents of this intergovernmental transfer of responsibilities argued that states would be more sensitive to local conditions, and that state program experimentation could result in innovative program development. Critics countered that service devolution stripped the poor of minimal income protections, and that states might engage in a "race to the bottom" to cut benefits (Ellwood, 1996; Greenburg, 1996).

The impact on existing local service systems as federal entitlements were eliminated received minimal attention in these service devolution debates. Yet, local governments traditionally have been the service provider of last resort for poor persons through General Assistance programs (Lav, Lazere, Greenstein, & Gold, 1993), which raises the specter that disentitled persons may increasingly turn to local governmental programs for assistance. General Assistance coverage has been cut back in recent years and varies widely across states, but 40 states still provided programs in 1998 (Gallagher, Uccello, Pierce, & Reidy, 1999). A related issue concerns the impact of welfare reforms on local voluntary providers as the public sector decreases program commitments.

Mayors in large cities recognized these issues, and argued that social service devolution could result in unfunded mandates for local governments (Goshko, 1995). This prospect was particularly daunting given that welfare recipients often reside in large urban areas facing ongoing fiscal crises (Fuchs, 1998; Kahn & Kamerman, 1998). Nonetheless, these concerns were subsumed as governors argued for the shifting of program control from the federal to the state level (Weir, 1998).

Local social services issues similarly have received limited empirical research attention since PRWORA changes were implemented. Yet, the capability and willingness of local governments and providers to meet service demands will be critical to the well-being of poor persons, particularly as they encounter TANF time limits and other PRWORA service restrictions. Issues related to the appropriate roles of local governments in providing services for the indigent also will become more pronounced in this changing public welfare environment.

Local governmental officials thus can provide important and overlooked perspectives on welfare reform impacts. Based on a survey of township officials who administer General Assistance and related social services in Illinois, this study examines how early welfare reform implementation affected local service delivery systems in one state. In addition to analyzing whether PRWORA and other program restrictions increased caseload pressures among different client groups, we explore local officials' views on the specific policy changes that may have affected local programs. We also demonstrate how discretionary administrative actions associated with social service devolution may have subtle but significant program effects. The implications of these findings for low-income persons, local social service programs, and state-local intergovernmental relations then are discussed.

Potential PRWORA Impacts on Local Service Delivery Systems

General Assistance programs often serve clientele similar to those affected by PRWORA changes (Karger & Stoesz, 1998). Generally operated by counties, townships, or other local governmental units, these programs not only provide limited cash assistance to persons without other means, but also sometimes include medical, emergency, and other services (Gallagher, Uccello, Pierce, & Reidy, 1999). Consequently, benefit reductions mandated by PRWORA and other 1990's federal and state program changes may affect General Assistance and related local programs in several ways. Four groups of service recipients are especially likely to be affected: food stamp recipients, immigrants, SSI recipients, and TANF recipients.

PRWORA reduced food stamp benefits in several ways. The per person per meal costs were cut from 80 to 66 cents, which will result in food stamp expenditure reductions of almost 20 percent by 2002 (Karger & Stoesz, 1998). In addition, food stamp program benefits were limited to three months every three years for unemployed able-bodied single adults between the ages of 18 to 50, with three month extensions for those laid off from their jobs. These single adults traditionally have been the primary General Assistance recipient group (Karger & Stoesz, 1998).

In addition to being affected by general food stamp and TANF reductions, immigrants were targeted for specific cuts under PRWORA. Most legal immigrants were prohibited from receiving SSI and food stamps until they became citizens. In 1997, SSI benefits were restored for most legal immigrants who were in the country when PRWORA was passed, and 1998 changes restored food stamp benefits for immigrant children, older immigrants, and disabled immigrants who entered the country before August 1996. Nonetheless, most immigrants admitted to the United States after PRWORA passage remain ineligible for both food stamps and SSI.

Those awaiting determinations of SSI eligibility historically have been a substantial component of the General Assistance caseload (DiNitto, 2000). Such persons not only receive General Assistance while awaiting SSI decisions, but also may remain on General Assistance and utilize other local services if SSI is denied. While not directly impacted by PRWORA, other changes in the SSI program could increase local service demands. For example, 1996 federal legislative changes have eliminated SSI eligibility for persons with alcohol or other substance abuse problems (Zelenske & Yates, 1996). Further, Illinois eliminated a state program in 1995 that had provided cash and medical assistance for persons awaiting SSI eligibility determinations (State of Illinois, 1995).

TANF program implementation may affect local social service programs in several ways. General Assistance programs traditionally have provided limited services to families with children, but such families may turn to local service programs after leaving TANF because of time limits or sanctions (Karger & Stoesz, 1998). In addition, many states are requiring applicants to engage in job search programs before becoming eligible for TANF (Nathan & Gais, 1999), which may result in the increased use of local emergency needs programs during these waiting periods.

The economic circumstances experienced by persons who leave TANF also may affect local services demand. Studies of leavers have found that most persons who leave TANF find jobs (Loprest, 1999; Acs & Loprest, 2001). However, earnings generally are below the poverty level, and the jobs often are unstable (Acs & Loprest, 2001; Anderson & Gryzlak, in press; Anderson, Halter, Julnes, & Schuldt, 2000). Given these circumstances, TANF leavers may turn to local services to supplement marginal incomes or to obtain transitional assistance after losing jobs.

The program restructuring accompanying service devolution also may have more general service impacts. Based on PRWORA implementation studies in twenty states, Nathan and Gais (1999) have argued that the major devolution of services has been to local entities, because state human service bureaucracies have granted discretion to local organizations to develop new service arrangements. This has created wide variations in local programming within states. Other implementation studies have found major deviations from intended policies as local providers struggle with new rules, expectations, and staffing needs (Brodkin, 1997; Iverson, 2000). Recipients may be confused as they negotiate these decentralized service networks, and may turn to local governments to test service possibilities.

The Illinois Program Context

In Illinois, all townships are required to establish General Assistance programs, which include both cash assistance and limited medical services (Illinois Compiled Statutes, 1998). General Assistance recipients must participate in workfare, job search, and job training programs to receive assistance unless exempted. Townships also have the option of delivering emergency assistance to indigent persons, typically by providing goods or service vouchers to those in need.

Although a small percentage of townships with high poverty levels and low property values receive state funding assistance (Johnson & Walzer, 1996), most township General Assistance programs are funded almost exclusively through local property taxes. Townships establish their own program rules and regulations within broad parameters established in the law. Substantial differences in eligibility standards, payment levels, service options, and administrative service structures result in this decentralized system. This is typical of the wide variations in General Assistance programs found in other states (Gallagher, et al., 1999)

Methodology

To determine whether PRWORA changes were affecting Illinois township programs, the authors developed a mail questionnaire in conjunction with the Township Officials of Illinois (TOI), a voluntary association comprised of Illinois township officials.

In designing the questionnaire, the authors examined the PRWORA legislation, background information on the Illinois General Assistance program, and related research studies to determine program relationships between PRWORA and General Assistance programs. We also met with TOI members to assure that questioning reflected issues of importance to township officials. The resulting questionnaire included both closed and open-ended questions detailing what types of programs were available in each township, how PRWORA changes may have affected local programs and target populations, and how townships had responded to any resulting changes in service demands.

Only the largest 200 townships were included in the study, because most of the over 1,400 Illinois townships have small populations and little social services programming. The City of Chicago was excluded from the sample, because the Illinois Department of Human Services (IDHS) administers the General Assistance program in Chicago without local governmental involvement. However, all of the townships in the densely settled metropolitan area surrounding Chicago were included, so the sample contained both the outlying Chicago metropolitan area and all other large urban areas in the state. The sample townships comprise about 81 percent of the total Illinois population outside Chicago.

The questionnaires were distributed to the selected townships through two methods. First, one of the authors attended regional meetings of the township officials, made a brief presentation on the purpose and importance of the study, and personally delivered the questionnaire to township officials in attendance. For the remaining townships, the questionnaire was sent to the township supervisor with a letter from TOI introducing the study and soliciting cooperation. Project staff then conducted extensive telephone follow-ups both to encourage questionnaire completion and to answer technical questions. Township social services staff members generally completed the questionnaires.

The surveying was conducted during the first six months of 1999, and so reflects early local experiences related to PRWORA implementation. A total of 101 townships returned surveys, representing a response rate of 51 percent. The mean population size for responding townships was 43,000. In addition to basic General Assistance cash and medical assistance provided by all townships, respondents were asked if they included emergency assistance or other services in their programs. Seventy percent of respondents provided emergency assistance, while 55 percent provided other services such as food pantries, help with heating assistance applications, and shelter assistance.

The following presentation relies primarily on qualitative analysis of responses to the open-ended survey questions. The authors coded and organized all open-ended question responses according to thematic content (Strauss & Corbin, 1990). Quotes then were selected from these thematic categories to illustrate the most commonly cited issues. To maintain the integrity of respondent comments, these quotes are presented verbatim.

To explore whether these initial local perceptions may have changed as program experience accumulated, we conducted a more limited follow-up survey with 40 township officials from the sample in late 2001. These surveys again questioned officials about TANF impacts on local programming. In addition, the follow-up questions focused on whether impending TANF time limits and the economic downturn were creating new problems.

Findings Based on Initial Survey

This section presents results from the initial surveys. We first describe the township officials' perceptions of welfare reform impacts on local caseloads, as well as whether the characteristics of persons who apply for services are changing. Then, we explore how administrative practices and confusion over new program rules may affect local service demands. Finally, respondent views are presented on how local governments are responding to these changes, as well as on the possibilities and limitations of such responses as welfare reform initiatives are implemented more fully.

New Demands from a Changing Client Population

Officials first were asked if federal and state program changes had led to increases in local service demands. Sixty percent of respondents indicated that services demands had increased. Of those who indicated that service demands had increased, nearly 92 percent cited changes in SSI eligibility requirements as a contributing factor, which was closely followed by 85 percent citing TANF changes and 72 percent indicating Food Stamp requirements (Table 1).

Respondents generally viewed SSI eligibility requirements as having the greatest impacts on local caseloads. This is not surprising, given General Assistance's predominance in serving single persons with disabilities (Halter, 1996). A series of SSI-related federal and state program changes and administrative actions were seen as causing these increased caseload pressures. The state elimination of cash and medical assistance for persons awaiting processing of their SSI applications was mentioned most often. Some local officials viewed this as a direct cost shift from the state to local governments. As one respondent stated: "Previously, the state would give them a medical card and money while they were applying for SSI--now the townships must take care of these cases." Another added: "We saw a 166% increase in 1997-98 over the previous fiscal year in township General Assistance pending SSI recipients... It is evident the increases by this township in dollars and recipients is due to SSI and emergency assistance cases which used to be carried by the state programs."

Several respondents also complained of increasing time frames for processing federal SSI applications. These processing delays probably resulted from case review backlogs related to changing federal eligibility requirements (Zelenski & Yates, 1996), as well as attempts by new client groups to obtain SSI given the more restricted TANF program environment. Because such "pending" SSI claimants received General Assistance, respondents indicated that SSI processing backlogs caused longer lengths of time on General Assistance and related increases in township medical expenditures.

While respondents most often mentioned caseload or expenditure impacts of federal and state changes, they also referred to subtler effects. For example, several officials discussed how the SSI and related medical changes had transformed the character of their local General Assistance program:

   The elimination of interim assistance at the state level to individuals
   pending SSI has had the highest impact on our General Assistance caseload.
   The inability of these clients to cooperate with our community work program
   has shifted the focus of General Assistance from a welfare to work program
   to a disability payment system. The elimination of SSI recipients with
   diagnoses related to substance and alcohol addiction from the federal rolls
   has had the same effect.

   The revised eligibility [requirements] for SSI and Medicaid have changed
   our caseload from "temporarily unemployed" to unemployed and needing
   medical assistance, denied SSI, denied Medicaid, as well as temporarily
   unemployed.

The aggressive state implementation of TANF and Food Stamp employment and training requirements also affected local caseloads. Most directly, persons cut off from TANF or Food Stamps for non-compliance with education and training requirements subsequently applied for General Assistance. In other cases, the prospect of tougher requirements discouraged persons from applying for state assistance. For example, one respondent stated, "The more recent changes that IDHS TANF applicants must seek employment for 30 days prior to any assistance has increased the recent load".

Less direct local program effects resulted from the creation of an expanded working poor population that labored for low wages in unstable jobs. Several respondents noted that IDHS quickly removed recipients from the TANF and Food Stamp rolls when they found work, but that these jobs often did not last. Respondents contended that such persons often turned to General Assistance and emergency assistance programs for help as they transitioned to new employment, because they faced delays in re-applying for state assistance or else were too discouraged to apply. As one respondent summarized: "I see the biggest gap is the time delay in receiving benefits when people are in between jobs."

Administrative Practices and Policy Confusion

Over one-quarter of respondents indicated that changing state administrative practices, as opposed to formal policy changes, had increased local caseload pressures (Table 1). Two sets of administrative issues were prominent. First, several respondents suggested that the IDHS offices discouraged potential recipients from applying for state benefits, or else made the process of applying unnecessarily difficult. The alleged disincentives included increasing waiting times for eligibility determination appointments and application processing, requesting excessive information from clients, and closing cases due to single missed appointments or questionable failures to meet verification requirements. Several respondents also spoke of rude bureaucratic treatment as conveying a signal that service applications were unwelcome. Collectively, these practices were seen as leading some persons to turn instead to local programs for help.

   From the feedback we've been getting from clients, it seems that IDHS is
   discouraging people from applying. It seems that they aren't being told
   what help might be available, but only what demands will be placed on them,
   without discussing much about how IDHS can help them overcome obstacles.

   It appears that the delivery of services is the change that is having the
   biggest impact. The preaching of the party line that `There's no free lunch
   anymore--you'll have to get out and work like the rest of us' is fine to an
   extent, but it seems that clients are sometimes treated rudely, and that
   obstacles are being placed in their way at times, rather than being helped
   to cope with obstacles to self-sufficiency.

   The state sits on some cases, treats people rudely, and loses applications,
   and some people give up and come to the township food pantry or emergency
   assistance program.

Other respondents indicated that program confusion and complexity discouraged applications for state services. This confusion was seen partially as inherent in the major systems changes that were occurring, but also as resulting from inadequate IDHS caseworker training on new procedures, requirements, and benefits. As one respondent said in questioning the knowledge of caseworkers: "Clients are being told erroneous information regarding initial eligibility or continued eligibility". Another added that "Mass confusion between caseworkers and new policies creates misunderstandings among applicants and recipients-correct information is hard to come by".

Variations In and Limits to Local Responsiveness

Given that local governments have limited available funding and are restricted in raising revenues by tax limits or caps, there is no assurance that those townships experiencing increased demands will provide additional services. Other options include revising program rules to reduce service eligibility, or developing waiting lists for service receipt. Consequently, we questioned respondents about how their townships had responded to increased demands.

Table 2 shows that 45 percent of respondents indicated that increased service demands had resulted in greater General Assistance expenditures. In 16 percent of responding townships, and 33 percent of those that had experienced General Assistance expenditure increases, townships either had raised local taxes or borrowed funds as a result. Other townships were able to meet new expenditure pressures by using fiscal reserves, carry forward funding, or other slack resources. However, this was considered a viable option only in the short run. As one said, "The increased demand for all services is making our cash on hand disappear at a much faster pace, meaning eventually more taxes."

Only 5 percent of respondents said that their townships had reduced services. However, 16 percent cited longer service application processing times, and changing administrative practices also appeared to restrict service access in some jurisdictions. One respondent neatly summarized how the problems of increased demand, limited funding, and resulting staffing constraints interacted: "Staff was reduced due to budgetary constraints caused by increased applicants. Due to these reductions in staff, processing time is longer". Processing delays were exacerbated by the fact that, due to the new federal and state program restrictions, many ineligible persons presented themselves at township offices because they lacked other options. As one respondent pointed out: "We have seen more intake time regarding these people. High percentages are not eligible, but it [still] takes time for the intake". Administrative reactions to increased demands also resulted in service reductions that, while real from the client perspective, would be unnoticed if only formal policy was scrutinized. "They [client applications] have been handled with stricter guidelines, and [we are] enforcing all requirements".

Several respondents noted that the strong economy had militated against short-term local program impacts, or that local pressures might increase when TANF five-year time limits later were implemented. As a result, concerns were expressed about the longer-term effects, and how service deficiencies may be created given local financial limitations. As one respondent said when discussing the possibility of local tax increases to meet rising service demands: "We find this not feasible due to the poor resources in our community and the lack of jobs". Another added, "Our ability to increase taxes is limited due to tax caps and the fact that we have been close to the rate limit for some time".

Follow-up Survey Issues

The follow-up surveys were conducted two to two and one-half years after the initial surveys. They were designed to determine if the basic policy and program themes described in the previous section were changing, or if new issues were emerging. Like other states, Illinois was beginning to experience the effects of an economic recession at this time, as evidenced by budget cuts at the state level and increasing unemployment rates. In addition, TANF five-year time limits were to be implemented for the first time within six to nine months of the interviews.

Analysis of the 40 responses indicates that service demands had continued to increase. The declining economy was most often cited as the reason for more recent caseload increases. In addition, nearly all respondents anticipated increasing demand in the next year if the economy did not improve. Such delayed effects are common in General Assistance programs, as unemployed workers exhaust unemployment insurance or other benefits before turning to local programs for help. One respondent also spoke of a bumping process that may affect workers with poor skills who were pressured to enter the work force because of welfare reform:

   We are seeing an increase in people who worked for the temporary agencies.
   It is kind of a bumping down process. As family workers have gotten laid
   off, they have gone to temporary agencies for work. This has led to an
   increase in demand for temporary jobs, which has forced the less skilled
   temporary workers to turn to General Assistance. Our caseloads have nearly
   doubled in the last year or so.

Economic downturns always have impacted General Assistance programs. However, the fact that this more recent recession was the first to occur in the context of constrained state public welfare eligibility clearly had township officials worried about the interactive effects of a poor economy when coupled with welfare reform. As one respondent said: "The situation is some combination of a downturn in the economy and welfare reform... If the economy is bad and people can't find jobs, where do they get the money to get by in this welfare reform environment?" Several respondents similarly indicated that economic conditions will shape the impact that time limits have on their townships. As one official noted: "There is a relationship between what is happening with the economy and those who will be leaving TANF. If the economy stays stable there will not be a problem."

While concerns about the economy were most common, some respondents also noted that township pressures resulting from welfare reform had increased since the initial period of TANF implementation. Time delays in local IDHS offices in implementing new TANF requirements were seen as contributing to these later demands.

   It took a year or two before things were really enforced. Now it is picking
   up a lot.

   It took a while for both the TANF and the Food Stamp programs to strictly
   enforce their work requirements. However, when they did, many people failed
   to comply and turned to the township for help. We would send them back to
   IDHS if they were out of compliance, but it still took us time to process
   their applications.

One official also suggested that IDHS administrative practices resulted in many unsuccessful TANF exits, with these persons then turning to the townships for help: "IDHS caseworkers are pushed to make quotas, whether they have a real opportunity for the TANF recipient or not. They have admitted this to us... the TANF people then come to the townships as a last resort after they have been kicked off."

The other dominant theme in the follow-up interviews concerned how welfare reform was continuing to affect the caseload mix, the services sought, and the nature of inquiries at township General Assistance offices. The many respondent complaints about the increased staff time required to explain service availability to persons arriving at the township offices were most notable in this respect. We had hypothesized that our initial findings of confusion about available services might abate as clients adjusted to new program rules and as the state and local social service agencies shared program knowledge and refined their intergovernmental working relationships. This had not occurred by the time that follow-up interviews were conducted.

   I have also found that there are more and more people coming here asking
   questions, wanting to know about where to turn.

   We have found there are many more people who are confused about where they
   can find help. Also, we have found that there are some private agencies who
   do not have a clear understanding about what is going on.

   There is much more traffic--more coming to us. People come here confused.
   Now they may not be eligible, but they come here because we are the last
   organization they seek out.

Such increases in confused clients, as well as associated information demands, led a few respondents to think more systematically about incorporating information and referral activities into the office's basic mission. As one respondent stated: "We have become more of a resource center. More people have questions and it seems there are many who don't know where to turn". Another saw that such a role may also require townships to follow through more aggressively with other agencies on behalf of clients: "We might become more inclined to become information providers... we also may have to become more advocacy oriented."

Given the confused and last resort nature of many client interactions noted by township officials, it is not surprising that several respondents also spoke of increasing pressures on township emergency services. "We are seeing more use of township food pantries and school supplies than ever before--the emergency needs programs are starting to be really pressed," said one official. These increasing emergency needs could result in both program cutbacks and shifts in program emphases.

   I have had to cut back on our Emergency Assistance program, because the
   funding was going to run out.

   We are hard-pressed to do what is needed ... there is more emergency
   assistance being given than General Assistance.

   The concern that we have here is that when time limits really occur we may
   have to re-think what we are doing. Normally, we are helping singles. Do we
   change and provide service to families first and do so more on an emergency
   basis?

Both the caseload pressures and the programmatic responses appeared to vary substantially across responding townships. Nonetheless, the combined effects of welfare reform, service devolution, and the declining economy were frustrating many officials and leading them to question the direction of welfare reform.

   The federal government and the state government say the locals can do it.
   We can't. This problem is growing, particularly in townships to urban
   areas. We will have to do something ... we need money from somewhere.

   I really think the strategy should be called the trickle-down dump theory.
   They tell us to do it, they dump it on us, but they don't give us the tools
   to do it. Maybe we should become religious organizations and get some
   funds. That way we don't have to ask taxpayers to increase their property
   taxes.

Discussion

The findings from this study provide several insights into the potential impacts of human service devolution on local governments and other local service providers, and also point to areas where additional research would be valuable. Most local governments in our survey were experiencing increased service demands as a result of federal and state program changes. This may have been due partially to temporary adjustments as both poor people and agency staff learned about changing program rules. However, the follow-up interviews suggest more lasting impacts, particularly in times of economic downturn and with the impending implementation of TANF time limits.

Some respondents viewed local caseload increases as a straightforward funding shift resulting from state implementation of new service restrictions. This ironically parallels state complaints about unfunded federal service mandates that stimulated PRWORA devolution debates. The major difference is that the state generally did not legally impose new requirements on localities, as the federal government often did with states in the 1980's and early 1990's. Nonetheless, as the providers of last resort, townships were faced with the choice of increasing expenditures or cutting back services to local citizens in need. Given that expenditure increases are severely constrained by property tax limits and general anti-tax sentiments, this often has left community services staffs in the position of being the ultimate bearers of bad news about more limited resource availability. It also has led many agencies to become more aware of other services as they seek to find whatever help may be available.

Analyses often focus solely on aggregate caseload or expenditure changes that result from new policies, or else on examining program effects only on the groups for whom benefits are targeted. However, our findings illustrate how programs may be transformed in unintended and subtle ways. Federal and state service restrictions set in motion a chain reaction through which disentitled service recipients turned to local units of government that operated related programs. If townships served these new service applicants, program populations sometimes were altered fundamentally. For example, the arrival of new applicants with borderline disabilities re-oriented some township General Assistance programs from work programs for the temporarily unemployed to disability assistance programs. Communities thus may be faced with difficult choices about the service mix they will offer. Devolution proponents would argue that this will encourage increased sensitivity to local service needs. Nonetheless, this local decision-making context of increased service demands coupled with limited resources rarely was emphasized in pre-PRWORA devolution debates.

As was expected, the devolution of services to local governmental units led to differing local service responses. These variations may encourage useful program experimentation, but also may create important equity issues as poor persons are treated differently depending on geographic location (Nathan & Gais, 1999). In addition, continuing devolution may result in a perverse competition in which localities limit services for fear of attracting poor persons from neighboring communities. While respondents generally did not speak of such a "race to the bottom", one reminded us that this issue requires rigorous monitoring as devolving service systems mature.

   We used to have an excellent and very humane Transient Assistance program,
   but we began getting referral business from homeless people hailing from
   other counties. Since the purpose of Transitional Assistance was to assist
   people who'd been stranded, not to invite people into our township to get a
   free night's lodging or whatever, we cancelled the program. Service demands
   plummeted very quickly, and so did our expenditures as a result.

Comments concerning discouragement of service applications, processing delays and excessive paperwork, and rude treatment by caseworkers were troubling. During the debates preceding PRWORA, reform advocates argued that the organizational culture of public assistance offices must change. Public assistance caseworkers were viewed as critical in communicating new expectations about education, training, and work requirements, as well as in informing clients about support services (Bane & Ellwood, 1994; Bloom & Butler, 1995). Yet, PRWORA did nothing to assure that caseworker qualifications and standards meshed with these more substantial case management roles. Previous studies have documented similar inconsistent and unprofessional caseworker and local agency performance during welfare reform implementation (Anderson, 2001; Brodkin, 1997; Iverson, 2000). Additional research on caseworker performance and caseworker-client interactions in TANF and related service bureaucracies is needed to clarify the extent to which such practices are occurring.

The shifting of clients from the federal and state to the local level also suggests an interesting parallel to Halter's (1989) finding that persons eliminated from state General Assistance programs initially turned to relatives for help. However, the willingness or capability of relatives to provide assistance was time-limited, and former General Assistance recipients soon ended up without help or in local shelters and food pantries. In our study, many persons facing new service restrictions at the state level turned to local programs for assistance, and pressures on emergency assistance programs increased. However, tax caps and conflicting pressures from traditional service groups collectively produce tenuous long-term local service prospects for these persons, particularly if demand increases further due to TANF time limits or a declining economy.

Implications

Although PRWORA will undergo scrutiny during 2002 reauthorization discussions, it is unlikely that either devolution or the other major program thrusts of welfare reform will be reversed. Therefore, monitoring and refining decentralized public welfare service systems present important challenges for researchers and human services professionals. As Schneider and Netting (1999) have argued, this requires that human services organizations refocus advocacy efforts at the state, and sometimes local, levels of government. Because state programs are still in a developmental phase and many states have not expended all available TANF funds (National Campaign for Jobs and Income Support, 2000), it may be possible to gain support for TANF program improvements if effective advocacy efforts are launched.

The study findings suggest substantive areas in which such efforts would be useful. The de-skilling of caseworkers has plagued public welfare bureaucracies (Fabricant & Burghardt, 1992), and in a restricted and time-limited program environment the consequences of poor performance have grown. Human services professionals should lobby state and local governments and service bureaucracies for adequate training, qualifications, and caseload sizes that recognize the increased case management responsibilities envisioned under PRWORA. The redefinition and refinement of public welfare case management responsibilities in the new TANF program environment also is needed (Hagen, 1999). The goal should be to assure that the helping components of these new roles are adequately developed, rather than being mere subsidiaries to a case sanctioning emphasis.

Problems of intergovernmental coordination also require careful scrutiny. Workers in various state and local programs often have limited understanding of the services available from other providers. Human services professionals can play important roles in their localities by mapping service networks that cross governmental boundaries, and then working within organizational settings to encourage service referrals and the sharing of information that link clients across otherwise fragmented systems. In addition, continued experimentation is needed with coordinative approaches such as one-stop service centers and more consistent eligibility procedures between programs (Holcomb, Pavetti, Ratcliffe, & Riedinger, 1998).

Limitations of the current study point to additional research needs. First, the study was conducted in only one state program environment. Governmental structures and programs differ considerably between states, so similar research in other states would establish whether the issues elaborated here are widespread. Additional research also is needed on the impacts of devolution on large cities, given their tenuous governmental finances and large concentrations of poor persons. Likewise, it would be useful to systematically examine impacts on shelters, food pantries, and emergency services, as oftentimes these are nonprofit entities not operated by a single governmental authority.

Our follow-up surveys demonstrate the importance of continued monitoring of local program effects over time. Both the declining economy and the impending implementation of time limits had heightened concerns among officials about local devolution impacts. The interaction between these factors in different localities is an especially useful area for study. In addition, later cohorts of TANF leavers may be more disadvantaged than early leavers, and may be more likely to turn to local service providers.

Finally, the characteristics and experiences of persons who leave TANF and other federal and state programs should be systematically compared across localities. In a restricted service environment with little federal oversight, independent evaluations of bureaucratic treatment and functioning are critical. Questions such as whether those who leave TANF are employable, whether they have been provided with options in meeting work and training requirements, and whether they are informed about service supports need to be assessed in diverse local settings. Because of their close interactions with current and former TANF clients in a variety of service settings, human services professionals are uniquely positioned to contribute to the development of information on these questions.

Table 1
Township Officials Perceptions of Changing Local Caseload Pressures
Related to PRWORA

                                                            Percent
                                                           of those
                                                          responding
                                                             that
                                                           federal/
                                                             state
                                                            program
                                                            changes
                                                          had led to
                                                           increased
                                                             local
                                               Percent      service
                                              of sample     demand
                                     Number    (n=101)       (n=61)

Federal/state program changes led      61       60.4%          NA
to increased local service demands

Demand increased because of:

  SSI eligibility requirements         56       55.4%        91.8%

  Food Stamp eligibility               44       43.6%        72.1%
  Requirements

  TANF eligibility requirements        52       51.5%        85.2%

  State administrative changes         26       25.7%        42.6%

  Other                                12       11.9%        19.7%
Table 2
Township Responses to Increasing Service Demand Pressures

                                                           Percent
                                                          of those
                                                         responding
                                                        that general
                                                         assistance
                                             Percent    expenditures
                                            of sample   had increased
Increasing demands led to:         Number    (n=101)       (n=45)

Increased General Assistance         45       44.6%          NA
expenditures

The need to raise local taxes or     16       15.9%         33.3%
borrow funds

The need to reduce services           5        5.0%         6.7%

Longer time periods to process       16       15.8%         26.7%
requests

References

ACS, G., & Loprest, P. (2001). Initial synthesis report of the findings from ASPE's "leavers" grants. Washington, D. C.: U. S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation.

Anderson, S. (2001). Welfare recipient views about caseworker performance: lessons for developing TANF case management practice, Families in Society, 82(2), 165-174.

Anderson, S., & Gryzlak, B. (In press). Social work advocacy in the post-TANF environment: Lessons from early TANF research studies. Social Work.

Anderson, S., Halter, A., Julnes, G., & Schuldt, R. (2000). Job stability and wage progression patterns among early TANF leavers. Journal of Sociology and Social Welfare, XXVII(4), 39-59.

Bane, M. J., & Ellwood, D. T. (1994). Welfare realities: From rhetoric to reform. Cambridge, MA: Harvard University Press.

Bloom, D. & Butler, D. (1995). Implementing time-limited welfare: Early experiences in three states. New York: Manpower Demonstration Research Corporation.

Brodkin, E. Z. (1997). Inside the welfare contract: Discretion and accountability in state welfare administration. Social Service Review, 71, 1-33.

DiNitto, D. (2000). Social welfare: Politics and public policy, fifth edition. Boston, MA: Allyn and Bacon.

Ellwood, D. (1996). Welfare reform as I knew it: When bad things happen to good politics. The American Prospect, (May-June), 22-29.

Fabricant, M., & Burghardt, S. (1992). The welfare state crisis and the transformation of social service work. Armonk, New York: M. E. Sharpe.

Fuchs, E. (1996). The permanent urban fiscal crises. In Vitullo-Martin, J. (Ed.), Breaking away: The future of cities: Essays in memory of Robert F. Wagner, Jr. New York: The Twentieth Century Fund.

Gallagher, L. J., Uccello, C. E., Pierce, A. B., & Reidy, E. B. (1999). State General Assistance programs 1998. Washington, D.C.: The Urban Institute.

Goshko, J. M. (1995). U. S. mayors ask for a "seat at the table". Washington Post, January 26, p. A5.

Greenburg, M. (1996). Racing to the bottom? Recent state welfare initiatives present cause for concern. Washington D. C.: Center for Law and Social Policy.

Hagen, J. L. (1999). Public welfare and human services: New directions under TANF? Families in Society, 80 (1), 78-89.

Halter, A. (1989). Welfare reform: One state's alternative. Journal of Sociology and Social Welfare, XVI (2), 151-162.

Halter, A. (1996). State welfare reform for the employable poor: The facts behind the assumptions. Social Work, 41 (1), 106-110.

Holcomb, P. A., Pavetti, L., Ratcliffe, C., & Riedinger, S. (1998). Building an employment focused welfare system: Work first and other work oriented strategies in five states. Washington, D. C.: The Urban Institute.

Illinois Compiled Statutes, Public Aid, Illinois Public Aid Code. (1998). 305 ILCS 5/6.

Iverson, R. R. (2000). TANF policy implementation: The invisible barrier. Journal of Sociology and Social Welfare, XXVII (2), 139-159.

Karger, H. J., & Stoesz, D. (1998). American social welfare policy: A structural approach. White Plains, NY: Longman.

Kahn, A. J., & Kamerman, S. B. (1998). Big cities in the welfare transition. New York: Columbia University School of Social Work.

Johnson, R. A., & Walzer, N. (1996). Financing township services in the 1990's. Rural Research Report, 7 (6). Macomb IL: Western Illinois University.

Lav, I; Lazere, E.; Greenstein, R.; & Gold, S. (1993). The states and the poor: Budget decisions hurt low income people in 1992. Washington, D.C.: Center on Budget and Policy Priorities.

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

Nathan, R. E, and Gais, T. L. (1999). Implementing the Personal Responsibility Act of 1996: A first look. Albany, NY: The Rockefeller Institute of Government, State University of New York.

National Campaign for Jobs and Income Support. (2000). Poverty amidst plenty: Amount of unspent federal anti-poverty funds grows despite persistent need. Washington, D. C., Author.

Schneider, R. L., & Netting, F. E. (1999). Influencing social policy in a time of devolution: Upholding social work's great tradition. Social Work, 44(4), 349-357.

State of Illinois. (1995). Public Act 89-21.

Strauss, A., & Corbin, J. (1990). Basics of qualitative research: Grounded theory procedures and techniques. Newbury Park, CA: Sage Publications.

Weir, M. (1998). Big cities confront the new federalism. In Kahn, A. J., & Kamerman, S. B., Big cities in the welfare transition. New York: Columbia University School of Social Work.

Zelenske, E., & Yates, T. (1996) Recent legislation eliminates drug addiction and alcoholism as a basis for Social Security and Supplemental Security Income Disability Benefits. Clearinghouse Review, July-August, 401-413.

Steven G. Anderson University of Illinois School of Social Work

Anthony Halter University of Illinois School of Social Work

Brian Gryzlak Department of Sociology Western Illinois University