A primary objective of the 1996 welfare reform law, the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA), is to raise the percentage of welfare recipients who work or are preparing for work. Welfare agencies, which have long struggled to achieve this goal, were stymied in the past by weak labor markets, large caseloads, insufficient resources, and organizational resistance to change. In this report, the Manpower Demonstration Research Corporation (MDRC) takes an in-depth look at four urban counties — Cuyahoga (Cleveland), Los Angeles, Miami-Dade, and Philadelphia — to see whether and how circumstances have changed since welfare reform was passed. The report addresses five sets of questions:

1. How have the counties responded to the welfare-to-work provisions in PRWORA? What work-related services, mandates, and incentives have they put in place?
2. How have participation rates and expenditures on welfare-to-work programs changed since PRWORA? Do the counties now emphasize different kinds of activities than they did several years ago?
3. What is the role of case management under PRWORA? How are case managers working with clients to move them into employment?
4. What is the role of the workforce development system in moving welfare recipients into employment? Specifically, how are workforce development agencies using funds allocated through the U.S. Department of Labor grant program for services targeted at the hard-to-employ?
5. How might the counties’ experiences inform federal and state welfare policies? What are the implications for PRWORA, which comes before Congress for reauthorization in 2002?

The overall conclusion presented here — based on field visits and analyses of program participation and expenditure data supplied by the study sites — is that the four counties have indeed become more employment-focused since welfare reform. This is evidenced by the programs they have implemented and the percentage of welfare recipients who are working or engaged in job preparation activities. The change appears to be attributable to PRWORA in combination with other policies and conditions that have made it easier for welfare agencies to move welfare recipients into employment.

This report is one of a series from the Project on Devolution and Urban Change, a study that MDRC is conducting with the aim of understanding how PRWORA is being implemented and what effects it is having on welfare recipients, low-income neighborhoods, and organizations that serve the poor in big cities. The underlying premise is that the changes brought about by the law — whether positive, negative, or mixed — will be felt most acutely in urban areas, where welfare recipients and poverty are most heavily concentrated.

I.  Welfare-to-Work Provisions Under PRWORA

PRWORA rewrote the rules for how cash assistance and welfare-to-work programs are operated in the United States. Before 1996, poor families who met federal and state eligibility requirements were guaranteed cash assistance through Aid to Families with Dependent Children (AFDC). PRWORA replaced AFDC with Temporary Assistance for Needy Families (TANF), thereby ending the entitlement to welfare. Most poor families may now receive federally funded cash assistance for a maximum of five years (or less, at states’ discretion). PRWORA also changed the federal mechanism for funding state welfare programs. Formerly, states received federal reimbursement for 50 percent to 75 percent of their actual expenditures on AFDC, along with a capped matching grant to provide welfare-to-work services. States now receive a fixed annual allocation, or block grant, based on their rate of spending before 1996 on AFDC and on welfare-to-work and related programs. The block grant does not go up if a state’s welfare caseload increases, but neither does it go down if its caseload falls.

Traditionally, block grants have allowed states greater discretion in the use of federal funds. PRWORA indeed offers flexibility in the design and operation of cash assistance programs, but it is prescriptive with regard to welfare-to-work policies. It requires states to engage 50 percent of TANF recipients in employment or job preparation activities by 2002. At the same time, it rewards states for lowering their welfare caseloads by reducing the 50 percent participation requirement by one percentage point for each percentage point drop in the caseload relative to that in 1995/1996. The law identifies 12 activities that states may count toward the participation rate. In contrast to regulations in effect before 1996 — which were widely interpreted as encouraging “human-capital”-building activities like education and training — PRWORA emphasizes paid or unpaid work and activities designed to move welfare recipients into employment quickly. The law states that TANF recipients should participate in activities at least 30 hours per week, including at least 20 hours in work or job search. The law also includes a “work trigger” that requires recipients to be working after receiving cash assistance for 24 months, though it gives states flexibility in determining how work is defined. States are expected to impose financial sanctions on adults who do not work or participate in assigned activities.

To supplement the funds available to states through the TANF block grant, the 1997 Balanced Budget Act created the Welfare-To-Work (WTW) grant program (the abbreviation WTW is used here to distinguish this program from others funded through the TANF block grant). The WTW grant program, which is administered through the U.S. Department of Labor and state and local workforce development agencies, is intended to provide work experience and job preparation services to long-term and hard-to-employ TANF recipients and to noncustodial parents of children receiving TANF. The grants were originally intended to be used over a three-year period, though many states have received extensions.

II.  The Implementation Environment

During the period covered by the Urban Change study — from 1997 through early 2001 — the conditions for operating welfare-to-work programs were quite favorable, both nationally and in the study sites. A growing economy and declining unemployment made it easier for clients to find jobs, and the federal Earned Income Credit made low-wage jobs more attractive by supplementing people’s earnings according to their income level and household size. In addition, all the states represented in the Urban Change study raised the ceiling on the amount of earnings that welfare recipients can keep before experiencing a reduction in their welfare check (a policy known as the “earned income disregard”), making it easier for families to combine work and welfare and increase their monthly income.

The improvement in the national economy during the late 1990s and early 2000s was accompanied by rapid declines in welfare caseloads. In the study sites, the declines from 1996 to 2000 ranged from 36 percent in Los Angeles to more than 50 percent in Cuyahoga and Miami. Owing to PRWORA’s block-grant funding structure, the drop in caseloads left the states and counties with a substantial surplus that could be used for a variety of welfare-related purposes, including welfare-to-work programs. Because the participation rate that states are required to meet is lowered when welfare caseloads decline, there was also less pressure on states to enroll large numbers of welfare recipients in work activities. Indeed, among the states represented in the Urban Change study, the caseload declines were so steep that the participation rates the states had to achieve dropped to less than 10 percent in 1999.

III. Welfare-to-Work Policy Choices in the Urban Change Sites

• All the Urban Change counties place a high priority on providing welfare-to-work services to TANF recipients. A “work-first” emphasis predominates.

The present research reveals that the welfare-to-work programs in the Urban Change sites have received significant attention from welfare administrators and staff. Moving away from the education and training focus that predominated before 1996, the sites have increasingly emphasized job search, short-term vocational training (six months or less), and unsubsidized employment. Cuyahoga, Miami, and Philadelphia made this programming shift in response to PRWORA; Los Angeles was already moving in this direction when PRWORA was passed.

• Participation in basic or postsecondary education is generally limited to clients who combine work and school or who take steps to enroll in such activities on their own.

In all the counties, clients are encouraged to pursue education goals part time while they work, and those who do so can receive help in paying their education- and training-related expenses. In Los Angeles, case managers generally approve education and training programs in which clients enrolled before being called in to the county’s welfare-to-work program, provided that the education or training appears to lead to employment. Philadelphia gives clients considerable room to pursue their personal education and training goals before the work trigger kicks in after 24 months.

• In all four counties, participation requirements apply to a larger proportion of the welfare caseload than in the past. The sites have adopted strikingly different sanctioning policies to address noncompliance.

PRWORA significantly expanded the proportion of welfare recipients who are subject to work participation requirements. When calculating participation rates — that is, the number of participants divided by the number of people required to participate — states may exclude from the denominator only single parents with children under 12 months old. Accordingly, California and Pennsylvania require participation of parents whose youngest child is age 1 or older. In Cuyahoga County and Florida, the youngest child may be no older than 3 months and 6 months, respectively, for the parent to be exempt.

The sites have adopted markedly different policies for dealing with clients who do not comply with welfare-to-work requirements. Both Miami and Cuyahoga cut off the entire family from TANF assistance when the adult is noncompliant. In Los Angeles, noncompliance results in removal of the adult from the TANF case while children continue to receive benefits. Philadelphia’s policy is in between: During the first 24 months on cash assistance, only adults may be sanctioned; after 24 months, the entire family’s grant may be terminated.

• With respect to time limits, two of the counties have implemented work-trigger rules. The other two counties emphasize interim time-limit policies, in which families are expected to leave welfare before exhausting their lifetime eligibility.

PRWORA requires welfare recipients to be involved in work after 24 months on cash assistance, but the details are left to states and localities. Philadelphia adopted a broad definition of work, one that encompasses job search and work preparation activities as well as paid and unpaid work. In Los Angeles, the work trigger means that clients must at least participate in an unsubsidized job or an unpaid work activity; however, the “countdown” toward the work trigger does not begin until clients finish job search without finding work and complete a vocational assessment.

Neither Cuyahoga nor Miami has implemented work-trigger policies. These counties instead emphasize interim time-limit policies that require families to go off TANF before exhausting their lifetime eligibility. In Cuyahoga, interim time limits go into effect after 36 months; in Miami, they go into effect after 24 or 36 months, depending on case characteristics.

• Expanded earned income disregard policies have emerged as an important element of the counties’ welfare-to-work strategies.

The sites’ earned income disregard policies provide a strong work incentive by allowing welfare recipients to keep a portion of their earnings before their cash grants are reduced, thereby boosting family income. For example, in Miami, a family of three without earnings would ordinarily receive a monthly TANF grant of $303. If the head of the case went to work, the family could keep the first $200 earned during the month with no reduction in the cash grant; thereafter, 50 percent of monthly earnings would be disregarded until the family’s income exceeded $806, at which point the family would cease to be eligible for TANF assistance. All the Urban Change sites have adopted similar policies, though the earnings levels beyond which recipients cannot continue to receive benefits differ from site to site.

By encouraging clients to work, earned income disregards help them — and welfare agencies — meet federal welfare-to-work participation requirements. At the same time, earned income disregards may prolong families’ stays on welfare and cause them to use up more months of their welfare eligibility than they would otherwise. Florida is the only state represented in the Urban Change study that allows recipients to “earn back” a month on the time clock for each month during which they work at least 30 hours per week, but Miami officials reported that the policy was difficult to implement because of changing state rules about what type of work counts and shortcomings in their automated system.

• Each of the counties faced significant organizational challenges to revamping its welfare-to-work program. Some challenges were related to the demands of the legislation; others were tied to local objectives or circumstances.

Cuyahoga embarked on a major reorganization of its welfare department and a restructuring of its staff’s responsibilities in order to serve clients more effectively and to prepare them for time limits. These efforts consumed enormous energy during the two or three years after welfare reform was passed. Los Angeles wrestled with the challenge of imposing work requirements on a much larger proportion of the caseload than in the past and the corresponding need to increase services and staffing. Miami’s welfare-to-work program faced rapid turnover at the director’s level, major changes in the contractors responsible for case management functions, and a recent decision by the Florida legislature to restructure the program and turn responsibility for it over to the local workforce development agency. In Philadelphia, state and local officials sometimes clashed over the goals of welfare reform — in particular, the role that education should play in welfare-to-work activities — and differing perceptions of the county’s staffing and budget needs.

IV.  Participation and Expenditure Trends

This examination of the counties’ welfare-to-work participation and expenditure trends relies on data supplied by the states and counties. Readers should be aware that some of the participation estimates were adjusted for duplicate counts. In addition, readers should note that in this analysis participation includes all welfare-to-work activities recorded in state or county systems rather than only those deemed countable under federal regulations. Finally, because of differences in how counties track and report participation and expenditure information, readers are advised to focus on the trends within counties rather than to make comparisons across counties.

• Since welfare reform, all the Urban Change sites increased the percentage of welfare recipients who are working or participating in welfare-to-work program activities (Figure ES.1).

Since 1996, the Urban Change sites have made substantial progress in increasing the percentage of welfare recipients who are employed or participating in welfare-to-work activities. These gains have been achieved amid dramatic reductions in the counties’ TANF caseloads, which substantially lowered the participation rates that states and counties were required to meet.

• Unsubsidized work accounted for a large portion of welfare-to-work participation and was the most common activity in all sites in 1999/2000.

A large proportion of welfare recipients in all the Urban Change sites worked in unsubsidized jobs while receiving TANF. In 1999/2000, unsubsidized work was the most common welfare-to-work activity, beating out traditional program activities like job search, vocational training, and unpaid work experience. The sites’ earned income disregard policies appear to have been largely responsible for this development in that they made it easier for welfare recipients to combine work and welfare. Before welfare reform, welfare recipients with earnings would have lost their eligibility more quickly.

• Participation in basic education activities has declined since passage of welfare reform.

PRWORA shifted the emphasis of welfare-to-work activities away from basic education, and the counties have responded accordingly. Basic education now registers as little more than a blip. In Cuyahoga, for instance, 5 percent of the participants in welfare-to-work programs attended basic education in 1999/2000, compared with 27 percent in 1993/1994. Most clients who were assigned to basic education appear to have combined it with employment.

• Unpaid work experience (“workfare”) was not heavily emphasized.

The Urban Change counties responded to PRWORA’s work emphasis by encouraging welfare recipients to take jobs in the regular labor market rather than by assigning them to community service or other unpaid work experience positions. Cuyahoga and Miami are the only counties that operate sizable unpaid work experience programs, and even there many more TANF recipients work in paid jobs or are assigned to job search than participate in unpaid work.

• The sites’ expenditures on welfare-to-work programs increased dramatically after welfare reform was passed (Figure ES.2).

After 1996/1997, all the counties invested increasing amounts of their TANF funds into their welfare-to-work programs, despite the fact that the number of people on welfare declined during that period. Cuyahoga’s growth in expenditures was fairly moderate, rising from $14 million in 1996/1997 to $18.3 million in 1999/2000 (an increase of about 30 percent). At the other end of the spectrum, spending in Miami-Dade increased sevenfold, from $8.9 million in 1996/1997 to $63.7 million in 1999/2000. (Note that these figures do not include child care or expenditures made under the WTW grant program.) The counties used the money to hire more staff, increase the number of program slots, develop more intensive services for the hard-to-employ, and expand support services and payments for welfare recipients who are working or participating in job preparation activities.

V. The Role of Case Management

• Case managers played a pivotal role in the implementation of the counties’ welfare-to-work programs, serving as the link between policies and recipients.

In all the sites, case managers played a central role in assigning clients to work activities, monitoring their participation, helping clients access support services, and enforcing program rules. Three of the sites — Los Angeles, Miami, and Philadelphia — chose to keep welfare-to-work case management and income maintenance as two separate, specialized roles. Cuyahoga combined income maintenance and welfare-to-work functions in the newly created role of “self-sufficiency coach” with the goal of enabling staff to address employment issues during every TANF interview and to gain a deeper understanding of each client’s history and circumstances.

• In most counties, case managers had little discretion in assigning clients to welfare-to-work activities.

Because work-first programs generally begin with job search, in most of the sites case managers did not have much flexibility in assigning clients to initial program activities. In Los Angeles and Philadelphia — where the programs had a prescribed sequence — the majority of case managers reported that both initial and subsequent program assignments were based on state or county rules, leaving little room for staff judgment. In Cuyahoga, in contrast, staff indicated having considerable discretion, in keeping with the individualized program approach that administrators in that county wanted to achieve. Early on, some Cuyahoga case managers expressed frustration with the complexity of their role and felt poorly prepared to make decisions, though over time they seemed to grow more confident in performing their duties.

• Across the four counties, case managers reported that enforcement of program rules was a high priority. The use of financial sanctions, however, varied from site to site.

In field interviews and surveys, case managers in all four counties indicated that monitoring and enforcement took between 14 percent and 22 percent of their time. Most case managers indicated that they were prepared to sanction clients who did not comply with participation requirements, though the participation data suggest that Miami staff were far more likely to impose sanctions than were staff in the other sites. In 1999/2000, 61 percent of Miami clients who were subject to the participation requirement were deemed noncompliant and were referred to or receiving sanction in an average month. In Cuyahoga in the same year, the figure was 2 percent.

VI. Implementation of the WTW Grant Program

As already discussed, the U.S. Department of Labor’s WTW grant program was intended to supplement the funds available to states through the TANF block grant. It included formula grants to states (which accounted for the bulk of WTW funding) and competitive grants awarded directly to service providers. This study examines only the formula grants. Because Ohio declined formula funding, the findings pertain chiefly to Los Angeles, Miami, and Philadelphia.

• The use of WTW program funds in the Urban Change sites was hampered by the grants’ timing and restrictive targeting criteria.

Although Congress created the grant only one year after passage of PRWORA, no funds reached local workforce development agencies until late 1998. By that point, state and county welfare agencies had already developed their welfare reform plans and implemented services under TANF. With caseloads falling, many welfare and workforce development officials neither saw a strong need for WTW program funds nor felt pressured to use them. Officials also complained that the targeting criteria contained in the authorizing legislation were administratively cumbersome. (As initially written into law, 70 percent of the funds had to be spent on families who had been on welfare for at least 30 months, were within 12 months of reaching a time limit, or had two or more specified barriers to work). By most accounts, this problem was eliminated by Congress’s broadening the program eligibility rules in July 2000.

• Of the Urban Change counties that received WTW formula grants, Philadelphia spent the largest portion of its allocation and enrolled the most participants. Implementation in the other counties proceeded more slowly.

As of April 2001, Philadelphia had spent more than 80 percent of its WTW grant and enrolled about 17,000 welfare recipients into WTW-funded programs. As of May 2001, Los Angeles had spent about a third of its grant and enrolled nearly 9,000 participants; and Miami spent about 10 percent of its grant and enrolled fewer than 500 participants.

Philadelphia’s program benefited from the strong backing of its mayor. Philadelphia also chose to invest the bulk of its WTW grant in two large programs and to target a broad group of TANF recipients: those who had reached the 24-month work trigger without finding employment. Los Angeles and Miami, in contrast, developed small programs that targeted specialized segments of the population, such as particular groups of non-English speakers and noncustodial parents. Though consistent with the goals of the grant, the latter approach may have contributed to the lower rates of implementation and enrollment in Los Angeles and Miami.

• The counties had difficulty recruiting and retaining the target population.

As originally enacted, the WTW grant contained detailed eligibility criteria that most administrators thought were too restrictive. Even when the criteria were loosened, some programs had trouble recruiting long-term or hard-to-employ participants. Some administrators speculated that the factors that made people hard to employ — personal problems, low skills, poor attitude, and so forth — also made them hard to serve. In addition, the tough sanctioning policies of some welfare agencies may also have removed many hard-to-employ individuals from the rolls before they could be referred to WTW-funded services.

• The WTW grant generally operated independently of other workforce development and welfare programs.

In the three counties that received a WTW grant, program administrators and staff reported that there was little integration between WTW-funded services and other programs run by the workforce development agencies and the welfare agencies, apparently owing to the grant’s timing, reporting requirements, and limited duration.

VII. Implications for Policy

PRWORA’s scheduled reauthorization in 2002 provides an opportunity for federal, state, and local policymakers to assess the accomplishments of recent years and to consider future modifications. The research conducted in the Urban Change sites yields several observations and lessons:

• PRWORA’s goal of getting welfare agencies to adopt a work-first orientation has been realized in the Urban Change sites.

One of PRWORA’s primary objectives was to end needy parents’ dependence on government benefits by encouraging them to prepare for jobs and to work. The Urban Change sites responded by changing their welfare-to-work programs to emphasize quick job entry rather than education and skill-building and by expanding earned income disregards to increase welfare recipients’ financial incentive to work. The sites also extended work requirements to a much larger proportion of the caseload than in the past.

• The work-first strategy adopted by the sites was well suited to the strong economy of the late 1990s and early 2000s. Softer labor markets may call for program modifications.

During the period covered by this study, the ready availability of entry-level jobs made it relatively easy for welfare recipients to find work. The counties took advantage of the strong economy by emphasizing activities like job search and short-term vocational training, and they devoted few TANF resources to developing paid or unpaid employment. (Some WTW grant funds were used for this purpose, particularly in Philadelphia.) Given the recent economic downturn, welfare agencies may need to place greater emphasis on developing such strategies or offering certificate or degree programs that will help welfare recipients compete in the slack labor market, perhaps by combining education and training with work assignments.

• The Urban Change sites used the TANF surplus to expand their welfare-to-work program capacity and to increase participation.

The structure of the TANF block grant — a five-year allocation based on states’ pre-1996 spending levels — generated a substantial surplus during the late 1990s, when welfare caseloads declined. The present study indicates that the Urban Change sites invested a considerable portion of this surplus in expanding their welfare-to-work programs. For example, in 1995/1996, Los Angeles spent $43.6 million to serve 34,000 program enrollees, whereas in 1999/2000, it spent $133.4 million to serve 102,000 enrollees. Though the cost per enrollee (approximately $1,300) changed little, the increased expenditures permitted Los Angeles to reach a much greater proportion of its caseload than in the past. In the other sites, rising program expenditures reflect both an effort to increase the number of people served and a higher cost per enrollee. The sites’ experiences suggest the level of resources that may be necessary to engage a high percentage of the welfare caseload in welfare-to-work activities. High coverage has long been a stated goal, but only recently have states and localities had the funds to achieve it.

• Expansions of earned income disregards have played a major part in the increases in welfare-to-work participation rates.

Earned income disregards may have the dual effect of encouraging welfare recipients to go to work and keeping them on welfare longer. In the short run, this is a win-win situation: Welfare clients who are employed in subsidized jobs gain valuable experience and increase their income, and welfare agencies can count such employment toward the welfare-to-work participation rate required by the federal government. In most states, however, clients who combine work and welfare use up months of assistance that they may need later — for instance, if they lose a job or another source of income. Policymakers might therefore reconsider whether clients who “play by the rules” (that is, work) should be subject to the same lifetime limit on federal cash assistance as are clients who are not working.

• Participation rates of 40 percent or 50 percent may be difficult to attain, even in the most favorable program environments.

As already noted, the Urban Change counties operated their welfare-to-work programs during the best of economic times, with plentiful jobs, falling caseloads, and large surpluses in funding. Even under such favorable conditions, the data suggest that the counties would have struggled to meet the participation rates set by PRWORA had they not received credit (in the form of reductions in the required participation rate) for reduced welfare caseloads. This study used a broader definition of participation than the federal standard, one that counted activities that are not listed in PRWORA and included all participants, regardless of whether they met the 30-hour weekly requirement. (To a large extent, adopting this more inclusive definition was necessitated by the sites’ data limitations.) Policymakers may want to examine actual state and county participation reports more closely to avoid setting future participation targets that are unrealistically high for a less favorable economic context.

• Disparities in sanctioning policies and practices raise questions about the fairness and effectiveness of formal enforcement.

As noted earlier, the Urban Change sites handled clients who did not comply with welfare-to-work requirements in markedly different ways. Whereas some sites terminated the entire case, others ceased to provide cash assistance for the adult but continued to provide for the children. The likelihood of being penalized for noncompliance also seems to have depended on where a welfare recipient resided. Though the policy in both Miami and Cuyahoga was to cut off the family’s whole grant (rather than only the adult portion), Miami’s high rate of sanctioning suggests that noncompliance almost always resulted in a penalty, while Cuyahoga’s low rate suggests that clients there were given more chances to comply. Given such disparities, federal policymakers may wish to clarify intentions and set guidelines for states and counties to follow.

Although one might expect tougher sanctioning policies to boost participation in welfare-to-work activities, no evidence of such a relationship was observed in the Urban Change sites. In particular, the participation rates in Miami and Cuyahoga were roughly the same, despite their very different sanctioning rates. There may also be a trade-off between strict enforcement and serving the hard-to-employ. Some WTW grant program administrators said that their difficulties in locating eligible clients were due in part to tough sanctioning by welfare agencies. This perception is supported by a report from the Urban Change project in which it was found that welfare recipients who had multiple health problems, experienced physical abuse, were at risk of depression, or had a chronically ill or disabled child were more likely than other recipients to have been sanctioned in the previous year. [1]

• Working with the hard-to-employ may require new program strategies.

In all four sites, program administrators and staff talked about clients who appeared to need more intensive services, such as mental health and substance abuse treatment or longer-term education and training. Although PRWORA does not prevent welfare agencies from making such referrals, the fact that it does not count them toward the participation rate is a deterrent for many program administrators and staff. If self-sufficiency is a goal of welfare reform, policymakers might consider broadening the list of allowable activities to accommodate mental health and substance abuse services and longer-term education and training for clients who need them and who are making progress in their activities.

• Finding ways to promote job retention and advancement is a major concern.

The Urban Change sites currently provide support services, such as transitional child care and medical benefits, to clients who leave TANF for employment. Nevertheless, there is a strong sense among county administrators that former TANF recipients may be vulnerable to setbacks or get stuck in low-paying jobs. Policymakers may want to consider ways to extend transitional benefits to families who are working but remain poor. In addition, policymakers might look for ways to help former recipients gain easier access to job leads, education and training programs, and other services that will help them advance economically.

• Supplementary funding opportunities like the WTW grant should be directed to communities and organizations that have identified needs and solid plans.

The formula grants provided through the WTW program ensured that federal funds were distributed equitably. However, judging from the experience in the Urban Change sites, not all localities were prepared to use the resources. Should such funding opportunities be provided in the future, policymakers may want to make grants contingent on demonstrated interest and need. One option might be to award small planning grants that could be followed up by larger program grants once plans have been thought through and interagency relationships have been established.

• Organizational change is an ongoing process.

When considering the Urban Change counties’ responses to PRWORA, it is important to keep in mind that significant organizational change does not come quickly or easily. Most of the counties took two to three years to plan and begin implementing programs before seeing major results. Four years after the law was passed, the counties were continuing to make modifications and in some cases found that they needed to retrain staff on basic policies and procedures. Any dramatic changes that follow TANF reauthorization would likely require similarly large investments of welfare agency time and resources.

Notes:

[1] Denise F. Polit, Andrew S. London, and John M. Martinez, The Health of Poor Urban Women: Findings from the Project on Devolution and Urban Change (New York: MDRC, 2001)

Ford Foundation John S. and James L. Knight Foundation, Charles Stewart Mott Foundation, The Joyce Foundation, The Pew Charitable Trusts, The Cleveland Foundation, W. K. Kellogg Foundation, The George Gund Foundation, The Robert Wood Johnson Foundation, William Penn Foundation, U.S. Department of Health and Human Services (including interagency funds from the U.S. Department of Agriculture), The James Irvine Foundation, The California Wellness Foundation, The Edna McConnell Clark Foundation,