Further Progress, Persistent Constraints:
Findings From a Second Survey of the Welfare-to-Work Grants Program

Chapter I:
Introduction

Contents

1. Context of the Welfare-to-Work Grants Program
2. Changes in the Welfare-to-Work Grants Program
3. Changes in the Evaluation Design
4. The Grantee Survey

A major federally funded initiative has been unfolding over the past two years to move welfare recipients and other low-income Americans into employment. In 1997, the Balanced Budget Act (BBA) authorized the U.S. Department of Labor (DOL) to award $3 billion in Welfare-to-Work (WtW) grants to states and local organizations. These grants support efforts, over a three-year period, to help the hardest-to-employ recipients of Temporary Assistance for Needy Families (TANF), as well as certain noncustodial parents, to prepare for employment, find jobs, and stay employed. The WtW grants program builds on the earlier enactment, in 1996, of the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA), which created the work-focused, time-limited TANF program. Whereas PRWORA was designed to increase movement off the welfare rolls into employment, WtW grants provide targeted resources for state and local efforts to help individuals facing the most serious challenges make that transition.

This report documents the continuing progress of the WtW grants program's implementation, as part of a comprehensive congressionally mandated evaluation being conducted by the U.S. Department of Health and Human Services (DHHS).⁽¹⁾ In part, the evaluation focuses on a set of in-depth study sites where detailed analyses of program operations and participant outcomes are being conducted.⁽²⁾ In addition, the evaluation includes a broad survey of all WtW grantees designed to provide an overall description of program structure, sponsorship, target populations, services provided, scale of operations, outcomes achieved, and challenges encountered.⁽³⁾ To capture changes that unfold as implementation advances, the survey has been conducted twice. Results of the first survey, conducted in late 1998, have already been presented (Perez-Johnson and Hershey 1999). This report is based on the second survey, conducted in late 1999.

The second grantee survey found signs of implementation progress. It also found, however, that the original WtW eligibility criteria continued to constrain enrollment, since congressional action to expand the eligible population had not, at the time of the survey, taken effect. Other findings suggest some changes in the ways grantees are operating, but confirm many of the findings of the first survey. The main findings from this second survey are as follows:

Summary Findings from the Second Survey of WtW Grantees
WtW program implementation has advanced, but participation levels still lag. Most grantees, except recently funded ones, are now delivering services and are operating at a somewhat larger scale than observed in the first survey a year earlier. However, restrictive eligibility rules still in effect in late 1999 continued to impede enrollment; as a result, the average pace of enrollment has not increased. Placements in employment activities are also slower than projected. Organizations that emphasize serving noncustodial parents appear to be having special recruitment challenges.
The scale at which WtW programs are projected to operate remains modest. Although, on average, second survey respondents have received more WtW funds than respondents a year ago, they have formulated more conservative participation targets. They are also less likely to combine WtW funds with funds from other sources to operate their programs — especially JTPA and TANF. These adjustments appear to reflect mainly the enrollment difficulties encountered to date; despite the declines in TANF rolls and WtW enrollment difficulties, survey respondents perceived no decline in overall need for WtW services.
Grantees emphasize unsubsidized employment but set realistic placement goals. An unsubsidized job is the ultimate outcome grantees want for all participants, but grantees clearly expect some program attrition and have some reservations about availability of jobs; they anticipate placing somewhat less than half of all enrollees in unsubsidized employment. Supported work activities of various types are envisioned by grantees as an interim step for roughly two-thirds of the participants they ultimately expect to place in unsubsidized jobs.
Most placements to date have been in low-wage, service occupations. Grantees have moved expeditiously to place WtW participants once they are enrolled. Grantees have made about a quarter of their projected placements — over 50,000 — in various work activities. Most placements are in service and administrative support positions that participants can get even with limited skills and poor work history. However, participants' placement wages average just $6.81 per hour, and opportunities for advancement are often limited.

The remainder of this report presents the survey findings in greater detail. Chapter II updates earlier findings on grantees' organizational and programmatic structure, the implementation status of WtW grantee programs, and the types of services they provide. Chapter III describes the population being served by WtW programs, how they are being recruited, and grantees' success in fulfilling their enrollment goals. Chapter IV focuses on the work activities of participants and the pace at which grantees are progressing toward their placement targets. Chapter V summarizes the views of grantees about the progress they are making and the challenges they are facing, as expressed in their survey responses. As background for presentation of these findings, the remainder of this introduction first summarizes the policy and economic context for the WtW program, changes in the WtW program itself, the overall evaluation design, and the survey that serves as the main data source for this report.

[ Go to Contents ]

A. Context of the Welfare-to Work Grants Program

Welfare is not what it used to be. Three dramatic shifts have occurred:

1. In the policy framework of assistance programs and their purpose,
2. In the organizational roles of public agencies, and
3. In the overall economy and scale of welfare programs.

Welfare Now Defined as Short-Term Step Toward Employment. Unlike the Aid to Families with Dependent Children Program (AFDC) that preceded it, TANF is explicitly defined as short-term assistance with an emphasis on preparation for employment. TANF recipients are required to work as soon as they are job-ready or have received assistance for 24 months, and most can receive federally funded TANF for only 60 months during their lifetime. States can impose even tighter time limits and penalties. Congress underscored its emphasis on work as the goal for TANF recipients by requiring states to meet steadily increasing requirements for the percentage of their TANF caseload that must be engaged in unsubsidized employment or other work activities. In fiscal year 2000, states must have 40 percent of their caseload in work activities; this requirement increases to 45 percent in fiscal year 2001, and to 50 percent in 2002. If states fail to meet there standards, they face financial penalties. Most TANF programs therefore stress job search assistance and encourage or require recipients to find employment rapidly, rather than promote participation in extensive education and training. At the same time, most states have chosen to reinforce work requirements and incentives by disregarding larger fractions of recipients' earnings in benefit calculations as a way of making work pay, and by dedicating increased resources to child care and transportation assistance to help offset the cost of working.

Organizational Roles Have Changed. PRWORA and the BBA have given states, and even localities, increased control over their strategies for moving welfare recipients into employment. PRWORA gives states a total of about $16.5 billion annually in block grants from DHHS through FY 2002, and establishes a broad policy framework for TANF programs, but leaves states great discretion in defining the combination of financial assistance and employment and support services they offer. The BBA gave primary responsibility for WtW programs to states' workforce development agencies; although they must use WtW funds only for allowed uses, they have considerable latitude in defining ways to promote job entry, retention, and advancement. States must pass 85 percent of their WtW funding to local private industry councils (PICs) or the equivalent agencies.⁽⁴⁾ They also have great flexibility in program design. In effect, at the local level, the job of moving welfare recipients into employment is now shared between human services agencies, responsible for TANF and its work programs, and the workforce development system with its responsibility for WtW programs.

WtW Emphasizes Employment for the Most Disadvantaged. In concept, WtW programs were intended to complement the "work first" programs that states establish with TANF funds. TANF work programs would encourage entry into the labor market by recipients who were better prepared in their education, work history, and personal skills to find and succeed in employment. WtW grants would help states and localities focus special resources and program strategies on people who were particularly disadvantaged and were likely to have the greatest difficulty finding and holding a job.

Congress sought to concentrate WtW resources on those with the greatest need in two ways. First, WtW funding allocation formulas favor areas with the greatest need by incorporating measures of the concentration of poverty and benefit receipt. Second, Congress established spending rules that required grantees to serve primarily individuals who exhibited several specific indicators of disadvantage in the labor market. As originally enacted, the BBA required that WtW grantees spend 70 percent of their grant funds on (1) long-term TANF recipients or recipients within a year of reaching a TANF time limit, who also have two of three specific problems affecting employment prospects; or (2) noncustodial parents of children in a long-term TANF case, who themselves face two of the three specified problems.⁽⁵⁾ The three problems specified in the original language of the BBA were (1) lack of a high school diploma or GED and low reading or math skills; (2) a substance abuse problem requiring treatment; and (3) a poor work history. The remaining 30 percent could be served if they met a set of less stringent criteria.⁽⁶⁾

Welfare Rolls Have Declined Dramatically. The welfare rolls, which began to shrink before passage of PRWORA and the BBA, have continued to decline in the first few years since passage. From January 1994 through June 1999, the total number of AFDC (and then TANF) cases declined by almost 50 percent, from 5.05 million to 2.54 million (U.S. Department of Health and Human Services 2000). This decline, in the view of most researchers, is due in part to the waiver-based experiments of many states in the early 1990s and, subsequently, to the new PRWORA policies, as well as, in large part, to the strength of the U.S. economy. Substantial fractions of those exiting the assistance rolls are going to work.⁽⁷⁾

Declining caseloads are leaving TANF agencies serving individuals with a greater concentration of employment barriers. Most TANF agencies visited in the early stages of this evaluation reported that many of the more educated, skilled, and motivated recipients have left assistance, and that the remaining caseload consists largely of individuals who face the kinds of multiple barriers the WtW program is intended to address, even if they do not always meet the eligibility criteria originally set forth in the BBA. Data from a recent New Jersey survey illustrate the differences between those who have left and those who remain on assistance. Of those in the survey sample who had left assistance for employment, 66 percent had high school diplomas, and 68 percent had worked in the past two years; of those who remained on TANF, the corresponding rates were only 48 and 46 percent, respectively (Rangarajan and Wood 1999).

[ Go to Contents ]

B. Changes in the Welfare-to-Work Grants Program

Evidence accumulated quickly, as the WtW grants program was implemented, that congressionally defined eligibility criteria were slowing enrollment and limiting participation. Ninety percent of respondents to the first evaluation grantee survey agreed moderately or strongly that eligibility rules were excluding people from the program who were truly among the hard-to-employ but who did not meet the criteria as specified in the WtW statute. The restrictive eligibility criteria were an important factor in the slow startup of many programs. In this evaluation, exploratory visits were made to 23 grantees as part of the search for in-depth study sites, and most of those grantees had found the eligibility criteria to be a serious constraint on their ability to meet their enrollment targets (Nightingale et al. 2000). Some of the grantees visited had already been compelled to turn away or place on a waiting list individuals who met criteria for the "30 percent category," but not the "70 percent category," rather than risk audit exceptions and financial penalties if they failed to comply with the requirements for serving mostly individuals in the latter category. A similar finding emerged from another study of first-year operations, based on discussions with and visits to the first 11 states that received WtW formula funds (Nightingale, Trutko, and Barnow 1999). The early exploratory visits that were part of the national WtW evaluation also suggested that grantees who intended to serve noncustodial parents were finding it difficult to recruit them, in part, because they failed to meet the stringent criteria pertaining to employment barriers or because the custodial parents of their children were not long-term TANF recipients.

Such concerns have led to amendments in the eligibility criteria for the WtW grants program. In the FY 2000 budget appropriations act for the departments of Labor, Health and Human Services, Education, and related agencies, passed in November 1999, several provisions make it easier for both TANF recipients and noncustodial parents to qualify for WtW services. The amendments left intact the requirement that 70 percent of WtW funds be spent on a defined category of participants, but broadened this category in two ways:

1. TANF Participants Qualify Simply by Being Long-Term Recipients. The amendments remove the requirement that long-term TANF recipients exhibit additional barriers to employment. TANF recipients are eligible under the amended criteria if they have received assistance for at least 30 months, are within 12 months of reaching a time limit, or have exhausted their TANF benefits due to time limits.
2. Noncustodial Parents Qualify Under Less Restrictive Rules. Under the amended criteria, noncustodial parents are eligible if: (1) they are unemployed, underemployed, or are having difficulty making child support payments; (2) their minor children are receiving or eligible for TANF, or received TANF in the past year, or are receiving or are eligible for assistance under the Food Stamp, Supplemental Security Income, Medicaid, or Children's Health Insurance programs; and (3) they make a commitment to establish paternity, pay child support, and participate in services to improve their prospects for employment and paying child support.⁽⁸⁾

The definition of the 30 percent category was also broadened to include youth who have received foster care in the past, custodial parents with incomes below the poverty level, and TANF recipients who face barriers specified by the local WIB. Other program changes were also made: (1) allowing WtW funds to be used for pre-employment vocational education and job training for up to six months; (2) allowing competitive grantees other than WIBs to provide job readiness, placement, and post-employment services directly rather than only through contracts or vouchers; (3) streamlining reporting requirements; and (4) permitting child support enforcement agencies to share information on noncustodial parents with WIBs, to help carry out WtW programs.

These amendments expand the population that WtW grantees might enroll in their programs, but the effect of the change will occur in stages. Beginning January 1, 2000, competitive grant funds may be expended on participants eligible for WtW services under the new rules. For formula grantees, the new rules are in effect beginning July 1, 2000, and newly eligible participants can be enrolled. However, federal funds cannot be used until October 1; state matching funds can be used for newly eligible participants beginning July 1, 2000.⁽⁹⁾ The amendments thus take effect well after the period of operations about which grantees could report in the second evaluation survey in late 1999.

[ Go to Contents ]

C. Changes in the Evaluation Design

The overall evaluation of the WtW grants program was designed to address five key questions:

1. What types and packages of services do WtW grantees provide? How do they compare to services already available under TANF or Job Training Partnership Act (JTPA) funding?
2. What are the net impacts of various WtW program approaches on employment and on families' well-being?
3. What challenges do grantees confront as they implement and operate WtW programs?
4. Do the benefits of WtW programs outweigh their costs?
5. How well do PICs and other non-TANF organizations — the primary vehicles for funding and operating WtW programs — meet the challenge of implementing WtW programs for those hardest to employ?

The design of the evaluation has evolved since its inception, because of the slow pace of enrollment in WtW programs. This has resulted in a narrowing of the questions the evaluation can address. Under the modified design adopted by DHHS, the evaluation plan now includes two main components:

A Descriptive Assessment of All WtW Grantees. The mail survey of all grantees in 1998 and 1999 is providing an overview of program designs and activities, target populations, characteristics of participants, and, to the extent they are available, placement outcomes. Visits to several dozen grantees before the first survey helped develop a fuller understanding of program variations and provided a basis for selection of in-depth study sites. This report is based on the second survey.

In-Depth Process and Implementation Study. In 1999-2000, site visits are being conducted in 12 to 15 grantee sites, selected because of their innovative approaches, settings, or target groups or because they are typical of the most common WtW interventions. These visits include discussions with staff of WtW programs and related agencies, focus groups with participants, and program observation. The aim is to identify implementation issues and challenges, as well as lessons on program implementation. In most of these sites, follow-up data will be collected through surveys and administrative data and used for analysis of participants' activities in the programs and their employment outcomes. Analysis of program costs will also be conducted. A process and implementation report will be issued in spring 2000.

The original evaluation design called for impact and cost-effectiveness analysis to be conducted in about 10 of the in-depth study sites, based on a random assignment experimental design. This design would have made it possible to determine the difference WtW programs make in employment and family well-being outcomes. It now appears, however, that this component of the evaluation will be feasible in at most a few sites. The main barrier to conducting the impact study is the difficulty finding grantees that are identifying more eligible candidates than they can serve; such "excess demand" is a necessary precondition for the use of random assignment. However, in lieu of impact estimates, the "enhanced process analysis" described above will yield systematic follow-up data on the employment and social outcomes of participants in most of the in-depth study sites. Such findings will be reported in stages — in mid-2001, late 2002, and mid-2003.

In addition to the components of the core evaluation, a special process and implementation study focuses on documenting welfare and employment systems operated by American Indian and Alaska Native grantees, the supportive services they provide, and how these tribal grantees integrate funds from various sources to move their members from welfare to work. Results of the tribal program evaluation will be reported in fall 2000.

[ Go to Contents ]

D. The Grantee Survey

The second survey of WtW grantees used the same instrument as the first survey, with a few minor exceptions. In the second survey, two separate versions of the questionnaire were used. One version, for grantees previously included in the first survey sample, focused the wording of questions on actual implementation progress, enrollment, and placements, since these grantees could generally be expected to have passed the start-up phase. A second version was used for Round 3 competitive grantees, which had received grants less than a month before the survey was mailed. This version retains, in the wording of many questions, an option allowed in the first survey for grantees to report on their program plans and projections if the program had not yet begun. Other minor refinements were made (in both versions) to clarify question intent and expand the topics on which respondents' views were solicited.

The second grantee survey is, in large part, a repeat survey for early WtW grantees, but it also includes newer grantees for the first time. The second survey began with a sample of 681 formula and competitive grantees, compared to 598 in the first survey. The sample increased because additional competitive grants were awarded in November 1998 and September 1999.⁽¹⁰⁾

An overall response rate of 71 percent was achieved, from respondents who approximately mirror the composition of the overall sample.⁽¹¹⁾ Formula grantees, competitive grantees, and organizations with both types of grants accounted for roughly the same proportions of the overall survey sample and the survey respondents (Table I.1). Survey respondents and nonrespondents are also roughly comparable in the size of the grants they received (not shown in table).

Because this survey was the second for some respondents, the survey analysis examined changes in program implementation status for grantees who responded to both surveys. Of the 415 grantees that responded to the first survey, 314 or 76 percent also responded to the second survey. These repeat respondents thus account for 64 percent of the 487 who responded to the second survey. The changes in program implementation reported by these repeat respondents are examined later in this report, as an indicator of progress achieved among early grantees.⁽¹²⁾  

TABLE I.1 GRANTEE CHARACTERISTICS: OVERALL SAMPLE FOR SECOND GRANTEE SURVEY AND SURVEY RESPONDENTS (Percentages)
	Overall Survey Samplea (n = 681)	Survey Respondents (n = 487)
Organization Type:
JTPA SDA/PIC	77.5	72.5
Others	22.5	27.5b
Grant Type(s) Received
Formula Grant	72.5	70.6c
Competitive Grant	20.8	19.3
Formula and Competitive Grants	6.8	10.1d
Source: National Evaluation of the Welfare-to-Work Grants Program, Second Grantee Survey (November 1999 - February 2000). a This column presents information that was available on all local substate formula grantees and competitive grantees from grantee lists provided by DOL in preparation for the survey. b Other types of grantees include human services agencies, other public agencies, nonprofit community-based organizations, universities and colleges, and organizations serving people with disabilities. c Includes 296 responding grantees that received formula funding only and 48 that received formula funding and a share of their state's discretionary funding. d Includes 41 responding grantees that received formula and competitive grants, an additional seven that also received a share of their state's discretionary funding and one grantee receiving competitive and discretionary funds.

[ Go to Contents ]

Footnotes

1. The evaluation is being conducted by Mathematica Policy Research, Inc. (MPR), the prime contractor, and its subcontractors, The Urban Institute and Support Services International, Inc.

2. Details of the overall evaluation design can be found in Hershey et al. (1999).

3. We use the word "grantees" to refer to local organizations that are receiving competitive grants directly from DOL, or receiving formula funds through their state, or both.

4. Effective July 2000, the Workforce Investment Act of 1998 mandates the replacement of PICs or their equivalents with Workforce Investment Boards (WIBs), entities intended to coordinate workforce investment, adult education and literacy, and vocational rehabilitation services through One-Stop Centers.

5. The 1999 Social Security Act amendments affecting the WtW program, discussed in section B of this chapter, changed these requirements to simplify the eligibility criteria.

6. Under the original BBA specification, 30 percent of WtW funds could be spent on TANF recipients (or noncustodial parents of TANF recipients) who have characteristics associated with long-term welfare dependency, such as being a school dropout or a teen parent, or having a poor work history.

7. In New Jersey, for example, 44 percent of those on TANF in July 1997 were off assistance 12 months later; more than half of those who were off assistance were working (Rangarajan and Wood 1999).

8. Under the previous WtW eligibility criteria, noncustodial parents qualified for WtW services if they themselves satisfied the "70 percent" or "30 percent" eligibility criteria and their children were part of a long-term TANF case.

9. Grantees under the Native American WtW program could begin serving newly eligible individuals and providing job training immediately upon enactment of the amendments on November 29, 1999.

10. Although 75 new grants were awarded in Round 2, and 64 in Round 3, the overall sample did not increase by 139, for several reasons. Some of the grantees receiving new competitive grants had already received earlier formula or competitive grants. A total of 28 such cases were found in which an organization had multiple grants; they were treated as a single grantee organization for the second survey. An additional 28 original sample members on investigation proved to be no longer functioning as WtW program operators, had merged with another grantee organization, or had returned their WtW funding. The net result for the second survey was an increase in the sample of only 83.

11. Appendix Table 1 presents response rates by state. The first grantee survey also achieved a response rate of 71 percent.

12. Since repeat respondents are only part of the overall survey respondents, it is of some interest whether they can serve as a basis for judgments about overall implementation trends. As shown in Appendix Table 2, the repeat respondents are somewhat more likely to be formula grantees than are the overall respondents to the second survey. However, repeat respondents display some important changes that are similar to shifts between the first and second overall survey samples. In both cases, there is a dramatic increase in how many grantees have begun WtW service delivery, from about 50 to about 90 percent. In addition, there is a similar increase in the number of participants enrolled, although repeat respondents have increased their enrollment even more sharply. This difference is natural, since the overall second survey sample includes new grantees.

Where to Now?

Top of Page
Contents of This Chapter

Home Pages:
Human Services Policy (HSP)
Assistant Secretary for Planning and Evaluation (ASPE)
U.S. Department of Health and Human Services (HHS)

Last updated February 5, 2004