I.  Introduction

The Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996 revolutionized welfare policy. Ending Aid to Families with Dependent Children (AFDC) — the 60-year-old federal cash welfare program for poor families — the act granted unprecedented authority and responsibility for public assistance policies and programs to the states, established a new form of aid known as Temporary Assistance for Needy Families (TANF), which provides funds to the states via block grants, and placed a five-year lifetime limit on federally assisted cash benefits for most families. States were permitted to exempt from the federal time limit no more than 20 percent of their average monthly caseloads and also faced increasingly stringent requirements to place more welfare recipients into jobs or work preparation activities. In the aftermath of PRWORA, states have further "devolved" much of the responsibility for welfare to local welfare agencies and other entities.

Congress enacted PRWORA and President Clinton signed it out of the profound conviction that the existing welfare system was a failure, but without much prior research to suggest what the likely effects of the new law’s provisions might be. PRWORA’s proponents expected the changes to spur policy innovation, lead more families to become self-supporting, and encourage marriage while discouraging out-of-wedlock births. The law’s critics predicted dire effects on poor families and the neighborhoods in which they live — more poverty, hardship, homelessness, domestic violence, and crime.

The fundamental premise underlying the Project on Devolution and Urban Change (Urban Change for short) is that the effects of PRWORA — whether positive, negative, or mixed — will be felt with special intensity by residents of the nation’s big cities, where long-term welfare recipients and other poor people are increasingly concentrated and employment opportunities are often limited. The Urban Change project is a five-year, multicomponent study of PRWORA’s implementation and of its effects on poor families with children, the communities in which they live, and the institutions that assist them. The study is taking place in four of the nation’s largest urban counties — Cuyahoga, Ohio (which includes Cleveland); Los Angeles, California; Miami-Dade, Florida; and Philadelphia, Pennsylvania. These counties (also referred to as the study’s sites) were selected to represent a mix of older Northern industrial cities and younger Sunbelt cities, with different local economies, welfare grant levels, and ethnic mixes. All four counties, however, account for a disproportionate share of the TANF recipients in their respective states. Table 1 describes the study’s principal features, while Table 2 summarizes key characteristics of the study sites.

The study is being undertaken by the Manpower Demonstration Research Corporation (MDRC), a nonprofit, nonpartisan organization that develops and evaluates interventions designed to improve the self-sufficiency and well-being of economically disadvantaged populations, in cooperation with researchers from other institutions at or near the study sites. The project is supported by a consortium of 11 foundations, which are listed at the front of the report.

This report, the first from the project, centers on case studies of the four sites. These case studies contain information from two of the project’s components: the implementation study, which explores welfare agency policies and practices, and the ethnographic study, which centers on in-depth interviews with welfare families living in poor neighborhoods at the sites. Although the data for this report were collected approximately 10 to 20 months after the passage of PRWORA and the story has continued to unfold since that time, many of the issues and dilemmas identified in the early round of research are ones with which the sites are still grappling.

II.  Findings in Brief

• The study found that within a relatively short time — and typically with little prior experience in putting major pre-PRWORA reforms in place — the welfare agencies in these counties made significant strides in communicating a new welfare message, changing over to a work-first approach, mandating participation in welfare-to-work services, and designing new institutional structures.

Before the passage of PRWORA, many states had experimented with welfare policies and programs, but they generally tested these initiatives in relatively small jurisdictions. Cities rarely had the resources to require all recipients to participate in their welfare-to-work programs, which were geared mostly toward building human capital gradually through education and job skills training, not toward early entry into the labor force as emphasized by PRWORA. (Thus, of the Urban Change sites, only Los Angeles County had operated a mandatory, work-first-oriented welfare-to-work program before the 1996 federal legislation.)

In the face of these challenges, the Urban Change sites advanced on several fronts. Considerable effort went to ensuring that recipients heard the new message that welfare is only temporary and that they needed to find a job. All the sites adopted a work-first approach, usually requiring recipients to engage in job search before they could enter other work preparation activities. (To date, however, other work-first activities — work experience and community service jobs — have not been widely implemented, although Philadelphia is gearing up for a large community service employment program.) The sites began requiring all recipients except those specifically exempted by law to take part in welfare-to-work activities. And Cuyahoga and Miami-Dade Counties developed new institutional structures for carrying out their welfare reform initiatives.

• Much remained to be done, however, in terms of changing the culture of welfare agencies, sharpening and clarifying welfare’s new messages, developing and carrying out plans for dealing with especially disadvantaged recipients, enhancing job placement efforts, ensuring ongoing benefits for former recipients who have made the transition to low-wage employment, and improving recipients’ ability to keep jobs and to move up to better ones.

The sites generally adopted a first-things-first approach, preparing at the outset for services recipients would need first and deferring planning for services that recipients might encounter later in their welfare tenure. In late 1997, when most of the field research for the implementation study was conducted, it appeared that many welfare agency officials were just beginning to think about the upcoming challenges. Los Angeles County was something of an exception in this regard; having already made the transition to a mandatory work-first model, welfare administrators there were in a somewhat better position to think about next steps than their counterparts at the other sites.

Clearly, many challenges lay ahead. For one thing, many details of the welfare changes remained unclear to both line staff members and recipients. In particular, staff did not communicate clearly to recipients the tradeoff between supplementing low earnings with welfare and conserving months of eligibility for a time when the need might be greater, or the fact that recipients might remain eligible for Food Stamps, Medicaid, and child care if they left welfare for employment. Other issues demanding administrators’ attention included: dealing with multiproblem recipients, strengthening the welfare agency’s labor market connections, and helping recipients move not just off welfare but out of poverty.

• To date, the sites have not seen a fraying of their "social safety nets."

The real test of the ability of the new policies to protect the nation’s poorest citizens, and poor children in particular, will come with the imposition of time limits — or the next economic downturn. The sites’ post-PRWORA sanctioning practices have generally not been much more punitive than their pre-PRWORA counterparts, although recipients in Miami-Dade County have been sanctioned more frequently. Moreover, while welfare diversion efforts were just getting under way at the time of the research visits, their purpose was to provide emergency cash aid and services to families that might otherwise go on welfare, not to create roadblocks to the application process.

• The participants in the ethnographic study were in favor of many of the welfare reform provisions but expressed anxiety about their consequences.

Respondents believed that work requirements would prod people to take the necessary steps to self-sufficiency and would help root out those who were undeserving of welfare assistance. At the same time, however, these recipients were concerned about PRWORA’s decreased emphasis on education, about their ability to supervise their children adequately while working, and about what would happen when they reached their time limits. Many were also fearful that they would not be able to find jobs at wages sufficient to meet their monthly expenses.

The next section of this Executive Summary discusses some of the policy options open to the states and the choices made by the Urban Change sites. The final section elaborates on the key findings summarized above.

III.   Devolution of Authority and Policy Choices of the Urban Change States

A principal goal of PRWORA is to move welfare recipients into jobs. Toward that end, the act imposes stringent requirements that increasing percentages of TANF recipients either work or participate in activities geared toward their getting jobs quickly, rather than in education and job skills training. PRWORA further seeks to affect various aspects of recipients’ reproductive, marital, and parental behavior, explicitly stating that among the goals of TANF are to "prevent and reduce the incidence of out-of-wedlock pregnancies" and to "encourage the formation and maintenance of two-parent families." (Accordingly, it mandates penalties for recipients who do not comply with child support enforcement efforts.) PRWORA also terminated many federal benefits for immigrants and allows states to end TANF assistance to legal immigrants who arrived in the United States before August 1996, when the legislation was signed. It places a five-year ban on TANF and most other federally funded, means-tested benefits for many families arriving after August 1996, unless they meet narrow exemption criteria; states may use their own moneys to provide cash assistance to these recent immigrants. (Federal legislation passed in 1998 reinstated Food Stamps eligibility for children under 18 who were lawfully admitted before August 1996.)

Under AFDC, states had received open-ended federal funding for welfare expenditures, but it came with many regulations attached. States could deviate from these regulations only by obtaining waivers to do so. Under PRWORA, in contrast, each state receives a federal block grant in a fixed amount that depends on the state’s pre-TANF allocation; that allocation, in turn, was tied to the size of the state’s AFDC caseload and the level of spending per recipient. Because the block grant amounts are related to past spending levels, and national welfare caseloads declined by 30 percent between January 1993 and September 1997, most states have experienced a substantial windfall (that is, they have gotten much more funding under the block grants than they would have received under the old AFDC system). PRWORA’s provisions further require states to maintain their own spending at 80 percent of their former spending level (75 percent if the state meets specific work participation requirements); these are the maintenance-of-effort funds.

Together, the federal block grants and the states’ maintenance-of-effort funds add up to unprecedented amounts of money for states to spend on behalf of poor families. And they have unprecedented freedom to decide how to spend it. Thus, states can run programs that are so different from "welfare as we have known it" as to be unrecognizable, or they can use their former AFDC and welfare-to-work programs as a framework on which to build incrementally. States can place whatever time limits they choose on the receipt of aid, decide which families should receive extensions to or exemptions from the time limits, or use their own moneys to pay for families that have been on the rolls more than five years, sidestepping the federal time limit altogether. They can add extra steps — and roadblocks — to the application process and restrict eligibility for benefits, and they can offer services and payments that will enable families to avoid going on welfare. They can increase or reduce welfare grants, determine how much of recipients’ earnings to disregard (that is, not count) in calculating benefit amounts, and even replace welfare benefits with wages paid in return for work or with noncash assistance. States can change the mix and amount of welfare-to-work and other services for current recipients and add new services for former recipients and other families who constitute the "working poor." And they can impose "sanctions" (that is, financial penalties) of any amount, including the entire grant, on recipients who lack "good cause" reasons for noncooperation.

The Urban Change states have responded to this flexibility in different ways. Some of their key policy choices are summarized in Table 3. Of particular note are the following:

1. Three states have adopted both lifetime limits and shorter, intermediate time limits on welfare receipt. Both California and Pennsylvania, for example, have five-year lifetime limits and also "work-trigger" time limits that require recipients to work after 24 months in order to continue receiving cash assistance.
2. A critical feature differentiating California’s plan from those of the other states is that when the household reaches either the work-trigger or the lifetime limit, only the adult’s share of the grant is eliminated.
3. All four states have considerably liberalized their earned income disregards, excluding at least half the earnings of recipients who have found employment when calculating the welfare benefits for which they may be eligible.
4. The states have instituted different provisions aimed at regulating recipients’ personal behavior. California and Florida, for example, seek to curb out-of-wedlock childbearing through "family caps," which limit or bar increases in the assistance grant for single mothers who give birth to children conceived while they are receiving welfare.
5. All the state plans include the provision of subsidized child care, along with federally mandated transitional Medicaid, for former recipients who exit the welfare rolls for jobs that are low-paying or do not offer health insurance. In addition, the plans in Los Angeles and Miami-Dade Counties provide for education and training services for recipients who become employed, to allow those placed in low-level jobs to upgrade their skills and positions.

IV.  Key Findings to Date

The four Urban Change sites have adopted different policies and have confronted distinctive issues in implementing their welfare reform initiatives. It is possible, however, to identify a number of cross-cutting themes, challenges, and concerns.

The Policy Process and Its Results

• Three of the Urban Change states have passed down to the localities some of the decision-making authority granted to them under PRWORA.

Pennsylvania is the exception: All Pennsylvania Department of Public Welfare district offices adhere to uniform statewide policies. Although Florida’s welfare system is also state-administered, localities are required to create specially constituted agencies, known as WAGES (Work and Gain Economic Self-Sufficiency) Coalitions, which have significant authority to develop and implement their own welfare reform programs within state guidelines. In both Ohio and California, welfare-to-work programs are operated by the counties, and counties receive their allocations from the state in the form of block grants, an arrangement that gives the localities further decision-making power.

• The Urban Change sites have not, so far, used their policy-making authority to impose a draconian regime on their poorest citizens.

So far, the worst fears of PRWORA’s critics have not materialized in these sites. The Urban Change states instituted their welfare reform policies during a period of considerable prosperity, and they have not slashed benefits and services to create a "race to the bottom." They have not drastically curtailed eligibility; indeed, by liberalizing their earned income disregards, they have maintained continuing eligibility for working people whose grants would have been terminated under the previous rules. They have also preserved TANF benefits for most immigrants.

The arrival of time limits constitutes a major test of the effects of welfare reform on disadvantaged families. Another test may come if many low-wage employees are thrown out of work when the next recession strikes. At that point, states’ welfare rolls would likely expand again, and the costs of benefits and services would grow, placing new pressures on state treasuries depleted by lower tax revenues.

New Players Inside and Outside the Welfare Agencies

• Welfare reform has brought new organizations and actors to the fore.

In several sites, there have been important changes in the key players. The local WAGES Coalitions in Florida have already been cited. Miami-Dade County also provides the best example among the Urban Change sites of the privatization of welfare services. New performance-based contracts let by the local WAGES Coalition posed difficulties for smaller nonprofit service providers, who could not maintain operations under the new terms. Lockheed Martin IMS, a for-profit company, was able to handle the financial responsibilities, and was awarded the largest provider contract; it then subcontracted with approximately 30 community-based organizations, which supply most of the actual services.

In Cleveland, the Cuyahoga County Board of County Commissioners saw PRWORA changes as an unusual opportunity to redesign the county’s human service delivery system from the ground up. The commissioners turned to a private consulting firm, McKinsey & Company, for help in planning the reorganization of the existing welfare department, and one of the McKinsey consultants was appointed to head a new agency created to serve the county’s TANF caseload. Welfare reform in Los Angeles has seen the emergence of agencies providing domestic violence counseling, substance abuse treatment, and mental health services as potentially important players in the policy-making process.

• Creating new administrative and organizational mechanisms has proved time-consuming and has sometimes diverted energy from the provision of services.

In Miami-Dade County, the time needed for the development of the WAGES Coalition and the establishment of collaboration with other key agencies delayed the implementation of many services. The full range of services and supports envisioned in the local plan became available to recipients 15 months after the program officially began — and just nine months before clients would begin to reach the first time limit. While Cuyahoga County’s reorganization of its human services department may well improve services to recipients in the longer term, it has also meant that top administrators have had to concern themselves not merely with service provision but also with the logistics of agency transformation.

Philadelphia, in contrast, has benefited from organizational stability in implementing the new policies. The administration has not changed, and there has been no reshuffling of the major players.

A Focus on the Critical First Steps

The Time Limit Message

• Staff in Cuyahoga, Miami-Dade, and Philadelphia Counties have delivered a new message about welfare to recipients: welfare is time-limited.

Welfare staff in these counties sought to communicate similar messages to recipients: "Welfare won’t always be there for you." "Your clock is ticking, whether you believe it or not." "An entry-level job is better than no job." "We are here to help you." (When the Los Angeles research visit took place, the welfare department had not begun to inform recipients systematically about the new time limits, since agency personnel had just completed a draft of the county’s welfare reform plan.)

The specific information about time limits that welfare agencies have chosen to convey is likely to add force to the "welfare is temporary" message in two ways. First, staff have emphasized the time limits that recipients will reach first rather than the lifetime limits — in Philadelphia, for example, staff underscore the fact that recipients must find a job within two years, not that they can receive welfare for a total of five. The recipients interviewed for the ethnographic study understood fairly accurately the length of the time limit that welfare agency officials and line staff had stressed.

Second, line staff said that they rarely, if ever, mentioned the possibility of an extension of or exemption from the time limits. Indeed, some staff members were adamant that they should not discuss these possibilities because doing so might give recipients "false hope" that they would be among the select few to escape the cutoffs.

Work First

• The welfare-to-work programs in Cuyahoga, Miami-Dade, and Philadelphia Counties have all adopted a work-first emphasis.

The pre-PRWORA welfare-to-work programs in these three counties had encouraged recipients to strengthen their educational and vocational skills before seeking jobs. (Los Angeles County had adopted a work-first emphasis in the early 1990s.) Since PRWORA, welfare-to-work activities for most recipients have generally begun with participation in either individual or group job search activities. (The latter, often known as job clubs, are operated by the welfare department, other public agencies, and nonprofit and for-profit organizations.) Education and training remain program options at all four sites, but in general, these activities are reserved for people who have been unsuccessful in finding a job through the initial job search component.

• So far, work first has not meant extensive use of work experience as a pre-time limit program activity.

To date, unpaid work experience has not emerged as a salient welfare-to-work activity across the sites, although it has been used as a follow-up to job search in Cuyahoga and Miami-Dade Counties. In the past, Los Angeles County welfare officials actively resisted developing a large work experience component, reasoning that this would suggest to employers that recipients are unable to hold unsubsidized, private-sector jobs. There are indications that the importance of work experience may increase as recipients who have been unable to find jobs hit work-trigger time limits; holding work experience or community service positions will enable them to retain their welfare benefits. This is particularly true in Philadelphia, where intensive planning for a community service job program is under way.

Increased Participation

• All recipients except those specifically exempted by law are now required to participate in welfare-to-work activities.

Largely because of funding constraints and capacity issues, the pre-TANF welfare-to-work programs in Cuyahoga, Miami-Dade, and Philadelphia Counties were nominally mandatory but effectively voluntary — that is, they mostly enrolled recipients who actively sought out the education, training, and supportive services the programs offered. Los Angeles County’s program also lacked the resources to serve everyone and focused mainly on recipients who had been on welfare three years or longer; for these individuals the program was truly mandatory, and sanctions were imposed for noncompliance.

The cash windfalls resulting from the decline in the states’ TANF caseloads have made new resources available to the Urban Change sites. And the prospect of having large numbers of families reach their time limits with no other source of income has created pressures on the counties to engage their entire caseloads in welfare-to-work activities.

• Bringing so many recipients into their welfare-to-work initiatives has been a major challenge at all sites.

By the time of the implementation research site visits, Cuyahoga and Philadelphia Counties had brought most of their recipients into their welfare-to-work programs and had assigned them to an activity — a massive feat, considering the positions from which the sites had started. In Philadelphia, recipients were phased in within nine months; once called in, they began to participate in job search almost immediately. Philadelphia avoided long waits for assignment by referring recipients to individual job search when group job search openings were unavailable. The site succeeded in coordinating the support services, staff effort, and recipient cooperation needed to reach the agency’s goal of having recipients complete an initial job search activity by the end of 1997, although post-job search services offered to participants were limited. In Cuyahoga County, the process of engaging the full caseload was more hurried: officials felt it was important to familiarize clients with the new rules and engage them in welfare-to-work activities. Thus, the welfare agency called them in en masse over a two-month period to meet with welfare-to-work program staff.

In Miami-Dade County, about half the recipients had been brought into WAGES a year after the beginning of the WAGES program. Senior staff estimated that 15,000 recipients were waiting to begin job club at the time of the implementation research in October 1997; delays resulted in part from the fact that a relatively small number of employment staff members were responsible for leading the week-long employment preparation workshops. A further problem in Miami-Dade was that the state’s computer system was not equipped to identify all people for whom participation would be mandatory.

The process of expanding the welfare-to-work program to engage the entire Los Angeles County caseload in welfare-to-work activities had yet to occur at the time of the implementation site visit. Officials estimated, however, that enrolling some 150,000 TANF recipients in such activities would require doubling the size of the county’s welfare-to-work program.

• Only Miami-Dade County appears to have consistently sanctioned recipients at a substantially higher rate than before PRWORA.

Florida and Ohio have enacted sanctioning provisions that allow a family’s entire grant to be eliminated from the beginning of its receipt of welfare. In Miami-Dade County, staff commented on the much greater speed and frequency with which sanctions were imposed after the advent of welfare reform. Ethnographic study respondents, too, were more likely than their counterparts in the other sites to say that they had been sanctioned; they frequently complained that they were penalized for missing appointments of which they had not been notified. In Florida, the family’s entire grant, not just the portion attributable to the noncompliant individual, may be cut unless the adult designates someone else to receive payment on behalf of her children, and the Food Stamp allotment may be eliminated as well.

In Cuyahoga County, it appears that for a relatively brief period additional sanctions were imposed in response to the welfare-to-work program call-in. Both staff and clients were also very much aware of a provision of Ohio’s welfare reform legislation that, while technically not a sanction, had the same effect: it made recipients who quit a job without good cause, along with their families, ineligible for welfare for three months. Paradoxically, this provision may have militated against the state’s work-first message by making recipients less willing to take jobs they were not sure they could keep.

Pennsylvania also permits full-family sanctions, but only after the head of the household has been on welfare for 24 months. Staff in Philadelphia indicated that sanctioning, which was infrequent at that site before PRWORA, has not increased since. Conversations with agency staff in Los Angeles suggested that, as in the past, sanctions would be used to enforce participation but would not be emphasized more than they had been (new rules were not yet in effect).

Challenges and Tasks for the Future

Continuing Administrative Needs

• Roles and responsibilities of staff members continued to need clarification.

Various observers have commented on the new role of income maintenance staff under welfare reform: they must deliver the new welfare message to recipients, identify clients’ problems, help clients resolve the problems, and encourage recipients’ work efforts. These new responsibilities, however, are added to the workers’ traditional job tasks of determining applicants’ initial eligibility and grant amounts correctly and adjusting grants as recipients’ circumstances change. At the time of the site visits, not only did income maintenance workers report that painstaking attention to eligibility was as critical as ever, but central office officials and income maintenance office directors also reported almost uniformly that keeping eligibility and grant amount error rates low remained an important priority for them.

• A number of staff members remained confused about the new rules and procedures and expressed a desire for additional training.

This confusion generally concerned detailed implementation issues such as which month a policy was slated to go into effect or how to carry out a certain procedure on the computer. Although agency administrators sometimes maintained that line staff had received an adequate amount of training, the workers themselves often declared otherwise. The confusion of staff members frustrated the ethnographic respondents in the four counties and sometimes led to a counterproductive "wait-and-see" attitude.

• To establish the target dates for time limits, welfare agency staff can use automated data to count the number of months a person has received welfare within their state, but not in other states.

A state’s ability to enforce its own time limits depends on welfare agency staff members’ ready access to information about a household’s prior receipt of welfare in all jurisdictions of the state. Staff in all sites have access to automated data systems containing this information, although the systems are not always user-friendly.

Enforcing the federal five-year limit, however, will require access to a national database of welfare receipt in all states — a database that, at this writing, does not exist. Field research indicates that local welfare office staff do try to ascertain whether a new applicant for assistance who reports that she came from another state was previously receiving welfare in that state. They verify that the applicant’s case in the state she used to live in has been closed before opening a new case. An explicit purpose of this practice is to prevent "double-dipping," or receiving welfare from more than one jurisdiction; it is unclear, however, whether prior receipt in other states is being counted against recipients’ time limits.

Planning the Next Steps

• Three of the Urban Change sites planned initiatives to divert applicants from welfare.

Cuyahoga, Los Angeles, and Miami-Dade Counties all had plans to make one-time payments available to eligible applicants to help them avoid going on welfare — Miami-Dade’s plan provides for emergency services and jobs as well as cash aid — but had not yet begun to implement these plans. Diversion at these sites does not appear to be aimed at placing additional roadblocks in the path of welfare applicants.

• Three of the Urban Change sites had begun to plan for hard-to-serve recipients.

As more readily employable recipients leave the welfare rolls, many of those left behind will be the so-called hard to serve — those facing significant barriers to steady employment. The fact that states can exempt only a limited proportion of the caseload from the federal time limits (and must pay for cases that exceed these time limits using their own funds) makes it important that states give new attention to these clients. Figuring out what services an individual needs in order to acquire and sustain employment is a further challenge, especially because she may face multiple, interrelated difficulties — including, for example, poor basic skills, a limited work history, and depression — that call for a multifaceted intervention.

The welfare-to-work plans of all sites acknowledge that some welfare recipients face major obstacles to employment and need special services to overcome them. At the time of the field research, Philadelphia had not developed initiatives targeted toward the hard to serve.

Miami-Dade County’s implementation plan for WAGES established a client profiling program, which classifies recipients into categories based on their previous work experience, education level, and length of time on assistance. Clients are considered hard to serve if they have significant barriers to employment, including limited or no work history, substance abuse problems, mental health issues, and a criminal record. WAGES Coalition contracts stipulate that the agencies that provide services have both a longer period and a higher per capita budget to work with these clients than they have for less disadvantaged recipients.

In both Cuyahoga and Los Angeles Counties, special services are planned for domestic violence victims, substance abusers, and those with mental health problems. In both locales, agencies have received contracts to serve recipients who fall into these categories; these recipients may be required to enter treatment either in addition to or as an alternative to regular welfare-to-work activities. Both sites are also planning to institute screening protocols and to train staff on how to identify people with mental health, substance abuse, and domestic violence issues. However, it remains to be seen whether recipients will be willing to identify themselves as suffering from one of these conditions, all of which carry a good deal of social stigma. Thus, the extent to which needy recipients will actually use available treatment resources is uncertain.

• Two of the Urban Change sites had embarked on major job development initiatives.

To help recipients find jobs, welfare agencies must deal with the demand side as well as the supply side of the picture by strengthening their connections with local employers. When the site visits took place, Miami-Dade County was planning an aggressive outreach campaign to major employers and industries, including the airline and hospitality industries. Local welfare offices held job fairs, hotlines were established to inform businesses of the financial incentives associated with hiring welfare recipients, and the local paper provided extensive coverage of local welfare-to-work activities. Los Angeles County was planning to expand the job development activities already begun under its pre-PRWORA welfare-to-work program. The other two sites had not yet developed plans for intensive placement efforts.

In all four counties, many of the employment opportunities are far from the higher-poverty areas where most recipients live. Cuyahoga County welfare officials have secured funding from the state to develop a pilot project linking central-city residents to suburban jobs. The need for similar initiatives is evident in the other sites as well. Developing new public transportation routes, extending the hours when public transportation operates to facilitate night-shift and weekend work, and devising van transport and carpool systems may be necessary to address the geographic mismatch between people and jobs.

• Three of the Urban Change sites planned to provide special job retention and/or skills upgrading services.

Placing a former welfare recipient in a job is unlikely to ensure either that she will be able to keep the job or that she will be able to progress to higher pay. Many recipients who find jobs lose them very quickly, often because of situational problems (such as unstable child care arrangements) or because of interpersonal difficulties at the workplace. After recipients have used up their available time, it will no longer be possible for them to turn to welfare for assistance if they become unemployed. Follow-up counseling provided after recipients go to work may enable them to avert the crises that lead to job loss, or, if that loss is inevitable (or after it has occurred), to move quickly into new positions. Postemployment education and training can help former recipients upgrade their job skills and move out of the low-wage labor market into better-paying, more secure employment.

The Los Angeles County plan for postemployment services is the most comprehensive; it calls for welfare-to-work program case managers to provide counseling centered on job retention and rapid reemployment for recipients who lose their jobs while they remain on assistance and for a year thereafter. Both the Los Angeles and Miami-Dade County plans also include postemployment education and training to enable recipients to upgrade their skills and move beyond entry-level employment. Cuyahoga County will make available moneys intended to help recipients secure and maintain employment when confronted with unexpected emergencies or substantial work-related expenses. How these plans will play out in practice is a subject for future research.

• In general, the sites have not focused on issues related to respondents’ personal lives.

PRWORA emphasizes provisions to promote marriage, reduce out-of-wedlock childbearing, and encourage parental responsibility for their children. These goals have not been a focus of the Urban Change sites, however. Their welfare agencies appear to have translated the "personal responsibility" part of the act’s name as an obligation to work and to assist child support enforcement efforts.

Changing recipients’ conduct in other areas is a much lower priority. Thus, although California and Florida have imposed family caps, front-line workers are not routinely expected to refer recipients to public health clinics for contraceptive assistance. Line staff do not generally inquire about parenting practices or provide recipients with information about parenting programs; indeed, they often lack training about how to approach these sensitive areas.

Remaining Uncertainties

• At the time of the site visits, welfare staff at all levels were concerned about how recipients would fare when time limits fell due, but the sites had not yet enunciated policies concerning what would happen at that point.

Welfare and welfare-to-work officials and line staff members at all sites were worried about what would happen when time limits arrived. They cited the difficulties of placing large numbers of people in jobs and expected a substantial proportion of recipients to reach their time limits without having any means of support. They also believed that the most disadvantaged recipients and their children would fare the worst in the new world of welfare reform.

Critical open questions are how officials will respond if large numbers of recipients reach the time limits without employment and how policies will be carried out by line staff; MDRC will examine their resolution in subsequent reports. One possibility is that extensions of work-related and interim-termination time limits will be granted en masse. This is the course Miami-Dade initially plans to pursue; its long-term strategy remains undetermined. Another is that staff will attempt on a case-by-case basis to distinguish the "deserving" from the "undeserving" poor and will grant extensions (or exemptions) to some but not others; if this should be the case, it will be important to examine how staff exercise their discretion.

Still another scenario calls for the large-scale creation of public service jobs. At the time of the site visits, none of the counties was contemplating creating community service jobs on a large scale, although Philadelphia has subsequently moved much further in this direction. The state, the mayor’s office, and a Philadelphia-based foundation have announced an initiative to create 3,000 community service positions in a variety of for-profit and nonprofit organizations. These jobs, which will be six months in length and pay the minimum wage, will be designed specifically for recipients with little or no prior employment experience. Los Angeles County was not planning to address its need for community service employment until a few months before recipients reached their work-trigger time limits in late 1999.

Finally, the late 1990s version of welfare reform may come in for further revision at both the federal and the state levels.

Recipients’ Responses and Expectations

• Recipients have grasped the message that welfare is time-limited.

It should be said, however, that some recipients expressed doubt about whether public officials would "really let children go hungry" if their mothers could not find work by the time they reached their limits.

• Recipients often did not understand important elements of welfare reform: the value of "banking time," the expanded earned income disregards, and transitional benefits.

Staff members told researchers that they advised recipients with other sources of support to consider forgoing welfare assistance, especially in months when they would receive only a small supplemental check, and instead to store up their months of eligibility until they really needed the aid. MDRC researchers were unable to determine how strongly this part of the message was emphasized: most ethnographic study respondents did not appear to grasp the concept of banking time in this way.

As noted previously, all four Urban Change states have substantially bolstered the financial incentives for recipients to work by increasing the amount of money that recipients can earn before their welfare grants are reduced (that is, the earned income disregard). Ethnographic study respondents in Philadelphia did understand that they could keep half of their earnings if they worked part time (although they were uncertain how full-time employment would affect their welfare benefits). The enhanced disregard — in conjunction with job search requirements, awareness of the time limits, and other factors — may have made them more willing to apply for and take part-time, minimum-wage, and outer-ring suburban jobs. (However, these recipients often did not believe they could sustain their jobs once they had reached their lifetime limits and could no longer receive supplemental benefits.) Recipients in the other three sites exhibited less knowledge of the income disregards, although in each of them the new earned income disregard is substantially more generous (if somewhat more complicated to explain) than Philadelphia’s.

Finally, the ethnographic findings indicate that recipients have very limited knowledge of the transitional child care and Medicaid benefits for which they may be eligible when they leave welfare for work. Administrators realize that staff need to explain the benefits more clearly, and that better procedures are needed for ensuring that when a welfare recipient gets a job and her cash case is closed, her household continues to receive the other kinds of assistance to which it is entitled.

• Both individual circumstances and social/economic factors influenced recipients’ views of welfare reform and their own prospects.

Except for Cuyahoga County, the counties in which the Urban Change study is being mounted have had higher unemployment rates than the nation as a whole since the mid-1990s; in 1990, all but one of the neighborhoods in which ethnographic study respondents resided had poverty concentrations of at least 30 percent. As a result, residents of these neighborhoods have often been both physically isolated from jobs and socially isolated from networks of jobholders.

From the fall of 1997 through the spring of 1998, respondents in Cuyahoga County, where unemployment rates were generally at least two percentage points lower than in the other sites, appeared quite optimistic across the board. Cleveland respondents generally assumed that they would find living-wage employment before they reached the time limit. But most did not have a clear sense of how they would go about finding jobs with adequate pay.

Ethnographic study sample members in Los Angeles County (including Mexican-born immigrants, most of whom had lived in the United States for some time) were generally very bleak about their own prospects as well as those of their children. (Unemployment rates in Los Angeles County averaged 6.8 percent in 1997 and ranged between 6.2 and 6.5 percent in the first half of 1998.) Unaware that they would keep all but the adult portion of their welfare benefit upon reaching the time limit, some recipients believed that they and their children might well become destitute.

Unemployment in Miami-Dade County averaged a very high 7.1 percent in 1997, and ranged between 6.5 and 7.3 percent in the first six months of 1998. Respondents interviewed in the winter of 1997 through the spring of 1998 were generally very anxious and depressed about their ability to find work. Recent immigrants were more optimistic in this regard than more settled immigrants or native-born African-Americans.

In the late summer of 1997, respondents in Philadelphia County, where the unemployment rate averaged 6.8 percent in 1997, saw the city’s lack of jobs as a real problem, and generally predicted catastrophic effects for welfare recipients living in their own neighborhoods and elsewhere in the city. Yet many had high hopes for themselves — especially younger recipients with little work experience. These respondents often pointed to job opportunities in the suburbs as evidence that jobs were indeed available, if they could find a way to get to them.

• Respondents in Cuyahoga, Miami-Dade, and Philadelphia Counties generally reported positive attitudes toward the job search requirements.

Ethnographic respondents often said that they approved of mandatory job search because they believed it would "smoke out" others who were less deserving of assistance than themselves and either force them to work or purge them from the rolls. In this regard, many respondents seemed to hold the same critical views of welfare recipients as does the general public, although their own personal experiences with friends and neighbors on welfare generally did not support such negative opinions. The respondents reported that, with hardly an exception, most people they knew well were "exceptional" rather than "typical," deserving as opposed to undeserving. Conversely, the women often assumed that people they didn’t know so well acted irresponsibly and were neglectful mothers.

Some recipients at these three sites also felt that the requirements would give them the extra push they felt they needed to do what they had wanted to do for a long time: find a job that could sustain their family and give them some hope of upward mobility.

Los Angeles County respondents, in contrast, tended to view the job search requirements negatively because they or their friends had been unsuccessful in finding jobs through the county’s mandatory job search component.

• Many recipients in the ethnographic study were upset by cutbacks in opportunities to combine welfare with long-term schooling.

Some clients eloquently expressed their opinion that what they saw as a "one-size-fits-all" work-first model was shortsighted and would have very bad long-term consequences for recipients’ earning capacity. The majority expressed strong doubts that many of their welfare-reliant neighbors could find work at a living wage without additional education and training.

• Recipients held mixed views about the effects of the reforms on their children, but were generally concerned about how their children would be supervised.

Younger respondents and respondents with more limited employment histories tended to be more positive about the effects of reform on children. These recipients believed that if they worked, they would be able to provide their offspring with both better role models and more material goods, and that their children’s self-esteem would increase correspondingly.

Older recipients, those with more years of prior employment, and more established immigrants were sometimes more cynical. Not only did they doubt whether the jobs would be there, they also questioned whether their families would be better off materially if they worked. Even so, these women still desperately wanted to find living-wage employment rather than to continue to rely on welfare, and they wanted their children to see them working.

Finally, recipients were worried about whether their school-age children would be adequately supervised if they went to work. Few women contemplating full-time work knew of affordable after-school, weekend, or summer programs for children aged six to 12. For older children, some respondents believed that their ability to prevent school truancy and other undesirable behavior (such as delinquency, gang activity, and sexual activity leading to pregnancy) would be impaired by the time constraints associated with full-time work. The quality of child care was also a concern to the women, and some worried that their children would be abused by child care workers.

• Ethnographic study respondents’ actions in the wake of welfare reform varied considerably.

In Miami-Dade and Philadelphia Counties, where time limits were fast approaching, some respondents were participating in job club or individual job search, and some had already done so and had made serious efforts to find employment through these activities. Others were actively looking for work on their own, and still others were trying to enroll in or complete training programs before the time limits were imposed. Finally, some had done nothing to seek employment or training, although most were aware that they would have to begin to do something in the not-too-distant future.

Work requirements were only vaguely understood by Los Angeles ethnographic respondents, who had not yet met with welfare agency staff to learn about the new rules. In Cuyahoga County, where the time limits were still far off, few respondents were actively seeking work or training.

In sum, ethnographic study participants’ conversations with interviewers reflected a mix of high hopes and deep fears — hopes that they could provide their children with better lives through living-wage employment and fears that they would be unable to manage the pressures inherent in combining solo childrearing with low-wage employment. Though, in general, those with more experience in the system were more pessimistic about their own prospects and perceived more potential problems than the relative newcomers to welfare, it was quite common for the same respondent to express both hopes and fears in a single interview. Nearly all respondents wanted to be employed and off welfare, but even the most optimistic among them perceived combining full-time work with single motherhood as daunting. Although some recipients said they would manage to weather the reforms and be better off as a result, a great many predicted catastrophic effects for their neighborhoods and cities.

                                                                                                * * *

These findings from the Urban Change study paint an early portrait of welfare reform as experienced by welfare agency staff members and recipients. Subsequent rounds of implementation and ethnographic research will fill in and expand this picture, while quantitative data from administrative records, client surveys, and other databases will provide more conclusive evidence about how poor people and their communities fare under welfare reform — who benefits and who is left behind.

Charles Stewart Mott Foundation, W. K. Kellogg Foundation, Ford Foundation, The Robert Wood Johnson Foundation, The Pew Charitable Trusts, The George Gund Foundation, The Joyce Foundation, The Cleveland Foundation, The James Irvine Foundation, The California Wellness Foundation, The Edna McConnell Clark Foundation