Skip to main content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.


The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Building an Employment Focused Welfare System

Publication Date



Building an Employment Focused Welfare System:



Work First and Other Work-Oriented Strategies in Five States



Pamela A. Holcomb

LaDonna Pavetti

Caroline Ratcliffe

Susan Riedinger

June 1998

This report was prepared for the U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation, by the Urban Institute under contract number HHS-100-95-0021, Delivery Order 10.




Chapter 1: Toward an Employment Focused Welfare System: Work First and Other Strategies

  •  The Context for Reform Employment Focused Welfare Reform Strategies in Five States A Framework for Understanding State Work-Oriented Welfare Reform
    • A Fundamental Distinction: Two Approaches to Work First Beyond Work First: Diverse Strategies to Promote Employment

    Study Objectives, Methodology and a Roadmap to the Rest of the Report

Chapter 2: Work First: The Core of Current Efforts to Reform Welfare to Work Programs

  •  What are Recipients Required to Do?
    •  How Work First, Work Mandate States Define Participation How Work First, Participation Mandate States Define Participation

    Job Search/Job Readiness

    •  When are Individuals Expected or Required to Conduct a Job Search? What Types of Job Search Assistance are Provided?

    Unpaid Work Experience and Subsidized Employment

    •  Unpaid Work Experience Subsidized Employment

    Establishing Stronger Linkages with Employers

    •  Involving the Employer Community Targeted Job Development and Job Retention Using Employers as Guest Speakers at Group Job Search Sessions

    Education and Training in a Work First Environment The Harder-to-Serve: A Critical Yet Largely Unaddressed Challenge

Chapter 3: Program Coverage and Exemptions: Who Must Participate?

  •  Program Coverage: A Broad Range of Choices Broad and Narrow Exemption Policies: Basic Trade-Offs
    •  Exemptions and Child Care Issues Exemptions and the Harder-to-Serve Complexities Added to Exemption Choices by PRWORA

Chapter 4: Raising the Stakes: The Role of Sanctions and Time Limits

  •  Sanctions: Raising the Stakes for Noncompliance
    • Three Types of Penalties for Noncompliance Issues Related to Avoiding Sanctions and Coming into Compliance

    Time Limits: Raising the Stakes for Reducing Welfare Dependency

    •  Key Characteristics of Time Limits Time Limits Present Complex Interactions and Trade-offs

Chapter 5: Financial Policies to Encourage and Support Work: The Role of Earned Income Disregards

  •  Key Features of Earned Income Disregards Income Disregard Policies Impact on Recipients’ Ability to Combine Welfare with Work

Chapter 6: Up-Front Diversion Strategies

  • Diversion through One-Time Financial Assistance Diversion through Identification and Use of Alternative Resources Diversion through Up-Front Applicant Job Search Diversion Policies: A Wide Range of Implications

Chapter 7: Organizational Strategies to Support Employment Focused Welfare Reform

  •  A New Administrative and Service Delivery Structure: The Wisconsin Experience
    •  The Basics: One Stop Delivery, Consolidated Funding, Integrated Staff Teams

    Integrated Case Management: New Roles for Income Maintenance and Welfare-to-Work Staff

Chapter 8: Recipient Experiences In Work First Programs

  •  How Many Recipients Found Jobs and What Did They Earn? What Share of Recipients Who are Subject to Mandatory
    •  Participation Requirements Engaged in Program Activities?

    How Many Mandatory Recipients Participated in Job Search or Other Activities? How Quickly Did Recipients Move into Program Activities? What Happened to Recipients Over the Course of 12 Months? How Many Recipients Were on Welfare One Year Later?

Appendix A: AFDC Caseload Trends, Economic Indicators and Socio-Economic Characteristics in the Five Study States

Appendix B: Description of State Administrative Data


  •  1.1 Selected Characteristics of the Five Study States 1.2 Work-Related Welfare Reform Efforts in the Study States 1.3 Key Characteristics of Study States’ Strategies to Create an Employment Focused Welfare System 3.1 Study State Exemption Policies 5.1 Treatment of Earned Income in the Five Study States 5.2 Monthly Break-Even Wages: Before and After Earned Income Disregard Policy Reform 5.3 Earned Income Policies’ Impact on AFDC/TANF Grants by Level of Wages and Hours Worked 8.1 Employment and Wages of Mandatory Recipients 8.2 Participation in Welfare-to-Work Program Activities by Mandatory Recipients 8.3 Participation in Specific Welfare-to-Work Program Activities 8.4 Time from Program Entry to First Activity: Percent of Mandatory Who Entered an Activity 8.5 Percent Distribution of Month by Month Activities of Massachusetts’ Mandatory Recipients A.1 AFDC Caseload Trends and Economic Indicators for the Total U.S. and the Study States A.2 Selected Demographic Characteristics of AFDC Families for the Total U.S. and the Study States A.3 Selected Characteristics for the Total U.S., Study States, and Local Study Sites B.1 Characteristics of Sample Mandatory Recipients  



Chapter 1


Toward an Employment Focused Welfare System: Work First and Other Strategies


The Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996 made unprecedented changes to the nation’s welfare system. Based on the goal of creating a work-oriented, transitional assistance program for families in need, PRWORA eliminated the former Aid to Families with Dependent Children (AFDC) program, an entitlement program which provided monthly cash assistance to families with little or no income. The new law also eliminated the Job Opportunities and Basic Skills Training (JOBS) program, a federally mandated welfare-to-work program for AFDC recipients designed to help families move off welfare and attain self-sufficiency.

PRWORA replaced these programs with a single block grant—the Temporary Assistance for Needy Families (TANF) program. The TANF block grant gives states far greater flexibility to design their own cash assistance and welfare-to-work programs than was permitted in the past. States are presently involved in translating PRWORA’s goal of a work-oriented, transitional assistance program into an operational reality, a process that promises to be ongoing given the inherent complexity and enormous challenge it entails. Many states had already begun this process through welfare reform waiver demonstration projects.1

Despite the overall flexibility afforded by the TANF block grant (as well as, at least initially, increased resources), the law does include some important restrictions on how block grant funds are to be spent and contains several mandates concerning the work-related aspects of TANF. In particular, PRWORA:

  • requires states to achieve considerably higher participation rates in work or work-related activities for recipients than in the past;2
  • calls for participation rates to be calculated in ways that are likely to increase participation in work activities, especially in subsidized employment and/or unpaid work experience, and limit participation in education or training;3 and,
  • imposes a work requirement on recipients after two years of benefit receipt, a requirement that states may opt to impose sooner if they so desire.

In addition, PRWORA eliminated the open-ended entitlement to cash assistance, which was the crux of the former AFDC program, and replaced it with a five year maximum lifetime limit on a family’s receipt of federally-funded TANF benefits. States have the option of imposing a shorter lifetime limit and may also, at their own discretion, continue to provide cash assistance to recipients beyond five years through state general revenues.4

The PRWORA work-related requirements and lifetime limit on benefits, coupled with increased flexibility and freedom from federal regulations, present states with a wide range of program design issues, policy choices and trade-offs that touch upon all dimensions of the current welfare system. Whether states are in the early stages of implementing changes designed to make their welfare systems more employment focused in response to PRWORA or simply on the lookout for new ideas and practices, the experiences of other states provide valuable information and lessons that can inform policy makers and administrators as they confront and reassess these issues in their own states.

Employment Focused Welfare Reform Strategies in Five States

In order to encourage and stimulate the cross-fertilization of ideas across states, the U.S. Department of Health and Human Services asked the Urban Institute to document key practices and strategies states have used thus far to make their welfare systems more employment focused, particularly with respect to strategies emphasizing quick entry into the labor market. Six local sites in five states were selected for intensive examination:

  • Indiana: Indianapolis (pop. 817,604) and Scottsburg (pop. 22,528)
  • Massachusetts: Worcester (pop. 718,858)
  • Oregon: Portland (pop. 614,104)
  • Virginia: Culpeper (pop. 30,528)
  • Wisconsin: Racine (pop. 182,982)5

These states were chosen for in-depth analysis because they exemplify a mix of different strategies to achieve the common goal of increasing employment among welfare recipients. The states vary in terms of the average cash payment they provide recipients—Indiana and Virginia are fairly low grant states while Massachusetts, Oregon and Wisconsin provide relatively high grants.6

In recent years, all of the study states have experienced significant declines in their cash assistance caseloads that are well above the national average, low unemployment and strong economies (see Table 1.1; more detailed characteristics on the study states and local sites can be found in Appendix A). 


Table 1.1

Selected Characteristics of the Five Study States


  U.S. Indiana Massachusetts Oregon Virginia Wisconsin
AFDC Caseload  % Decline  1993-1996 11.40% 31.10% 25.00% 27.20% 15.30% 30.30%
Unemployment  Rate 1996 5.4 4.1 4.3 5.9 4.4 3.5
Job Growth Rate  1992-1996 9.1 9.2 7.9 14.0 9.0 9.0


Work-oriented reforms in place at the time of this study were implemented at different points between 1993 and 1996 (see Table 1.2). Since the passage of PRWORA, Indiana and Wisconsin both implemented new work-oriented reforms while Virginia, Massachusetts, and Oregon have made few changes.

Thus, while this study captures state experiences at one point in time, it also reflects states at different stages in their own evolution toward a more employment focused welfare system. It is also important to note that this study took place too soon after TANF went into effect to fully capture the implications and impact of the new federal welfare reform law (e.g., progressively steeper participation rate requirements, lifetime limit on benefit receipt).


In recent years states have implemented a broad range of programmatic, policy and organizational changes to create assistance systems that emphasize work rather than cash assistance. The primary welfare-to-work strategy used by states to create a more work-oriented, transitional assistance system is “Work First,” a philosophy and program strategy that emphasizes helping recipients find unsubsidized employment as quickly as possible. In addition to Work First programs, states have implemented a variety of other policies that supplement and reinforce the emphasis on work.

Work First programs share a common philosophy regarding work: any job is viewed as a good job and program efforts should be geared toward helping recipients enter the paid labor force as quickly as possible. Thus, Work First programs generally emphasize participation in structured or semi-structured job search programs, rather than education and training. 


Table 1.2

Work-Related Welfare Reform Efforts in the Study States



Project Name

Project Coverage

Start Date

Indiana          Impacting Families  Welfare Reform  (IMPACT) 

IMPACT  Modification





Massachusetts          Welfare Reform '95     

Welfare Reform '95  Modification





Oregon          JOBS Waiver Project 

JOBS Plus  (Subsidized Employment) 

Oregon Option


6 counties 





Virginia    Virginia Initiative for  Employment Not  Welfare (VIEW) Statewide,  two year  phase-in 4/95   
Wisconsin          Work Not Welfare 

Self-Sufficiency First/  Pay for Performance 

Wisconsin Works (W-2)

2 counties   







A Fundamental Distinction: Two Approaches to Work First

All of the five states included in this report implemented Work First programs. Broadly speaking, these states implemented two distinct approaches to Work First. The primary difference between the two approaches pertains to what happens to recipients who are unsuccessful in obtaining employment after a specified period of time.

  • Work First, Work Mandate models include a firm and relatively short (i.e., a few months) amount of time during which a recipient can collect cash assistance before a strictly defined work mandate goes into effect. If recipients have not been successful in finding a job or other means of support that enables them to go off welfare within this specified time frame, they must combine welfare with unsubsidized employment or work in either a subsidized employment or unpaid work experience position in order to receive cash assistance.
  • Work First, Participation Mandate models do not impose a strictly defined work mandate and instead concentrate resources on helping recipients obtain private sector employment for the entire time a recipient is unemployed and receiving welfare. Mandatory participation is a central feature of this model as well. However, participation is enforced within a broader range of possible activities—such as additional job search, education, training or community work experience.

As of January 1997, Virginia and Massachusetts operated Work First, Work Mandate programs and Indiana, Wisconsin and Oregon administered Work First, Participation Mandate programs. Since then, Wisconsin has embarked on a radical restructuring of its entire welfare system, including changes which make it more closely resemble a Work First, Work Mandate model.

It is important to note that while the Work First, Work Mandate model is common among the five states in this study, it is atypical of current state work-oriented programs nationwide. At the time our study took place, Massachusetts and Virginia were the only two states in the country which imposed such a strict work mandate on recipients so soon (just 2-3 months) after going on welfare. Another five states required recipients to participate in employment or community service after a set period of time ranging from 18 to 30 months.7

Since the TANF work requirement on recipients after two years of benefit receipt has yet to be reached, it is still too early to know the degree to which these two program models will be maintained in the future. States may in the future move to a strict definition of work as found in the Work First, Work Mandate study states or they may continue to take advantage of the law’s relatively loose definition of “work,” which may include activities other than unsubsidized employment, subsidized employment and community work experience. States may also choose to apply different program approaches to different populations.

Recognizing at the outset that the distinctions between the two models are likely to evolve and change over time, we reference them throughout the report merely as a way to provide a framework for categorizing and understanding key differences between approaches adopted to make state welfare systems more employment focused. Furthermore, as discussed in greater detail in subsequent chapters, there are important differences between states in the specifics of their design and implementation of either model.

Beyond Work First: Diverse Strategies to Promote Employment

In addition to the two approaches to Work First noted above, there are several other key dimensions of state work-oriented policies which cut across these program models and play an important role in shaping the overall design of their programs. Table 1.3 summarizes key strategies used by the study states to shift to a more work-oriented system. These include:

  • Fewer exemptions from required participation in welfare-to-work program activities and rules;
  • More stringent sanctions for noncompliance with work program mandates to reinforce the importance of employment;
  • Diversion efforts including one-time cash payments, pursuance of alternative resources and applicant job search requirements to discourage families from applying for cash assistance;
  • More generous earned income disregards to support recipients’ efforts to work, especially if they can only find part-time or low-wage work; and,
  • Time limits on the receipt of cash assistance to encourage recipients to seek employment as quickly as possible and to end long-term welfare dependency;
  • Organizational changes including shifting more (or all) responsibility for program management and operations to Workforce Development agencies, devolving more responsibility to local offices, relying more on performance based contracts, and changing the roles and responsibilities of eligibility and welfare-to-work staff.

The two Work First, Work Mandate states—Virginia and Massachusetts—are the most similar of the study states. In addition to a strict work mandate, both states have implemented stringent sanctions for noncompliance, relatively short time limits with no or limited extensions, and generous earned income disregards. Virginia’s Work First, Work Mandate program is targeted to all able-bodied household heads without a child under the age of 18 months. Massachusetts’ program is targeted only to able-bodied household heads without a child under the age of six. Although Massachusetts has begun to combine the eligibility and welfare-to-work responsibilities previously held by two workers into one position, organizational strategies have not been central to reform in either state.

There is far more diversity among the three Work First, Work Participation states. Oregon’s approach to reform is characterized by a nearly universal participation requirement, stringent penalties for noncompliance with procedures to protect the well-being of children, substantial emphasis on efforts to divert applicants from the welfare system, and development of alternative program activities to help harder-to-serve recipients successfully make the transition to employment. Organizational strategies, including an increased emphasis on work-oriented performance standards, have been central to the state’s reform efforts.

An increased reliance on performance based contracts and job placement goals for local offices has also played an especially important role in Indiana’s approach to reform. Other key features of Indiana’s reform strategy include greater reliance on penalties for noncompliance, time limits, and a two-track program approach that make job ready recipients subject to different requirements than recipients determined not yet job ready.8 


Table 1.3

Key Characteristics of Study States' Strategies to Create

and Employment Focused Welfare System

(As of July 1997)


  Work First, Work Mandate States Work First, Participation Mandate States
  Massachusetts Virginia Indiana Oregon Wisconsin (pre W-2)a
Target Groups      Non-disabled  recipients without a  child under age 6 Non-disabled  recipients without  a child under age  18 months Non-disabled  recipients  without a child  under age 3 Recipients  without a child  under age  12 weeks Recipients without a  child under age  12 weeks
Penalty for   noncompliance        Progressively steeper  penalty ---> loss of all  cash assistance      Loss of all cash  assistance        Grant reduction  (adult portion)        Progressively  steeper penalty  ---> loss of all  cash assistance    Determined by  amount of participation;  maximum penalty is  loss of all cash and  Food Stamp benefits  , except for $10
Diversion   Strategies    None      Lump sum  cash payment    Pursuance of  alternative  support Pursuance of  alternative  support, applicant  job search Pursuance of alternative  support, applicant  job search
Treatment of   Earnings for   Eligibility/Benefit   Determination $30 and 50%  of earned income  disregarded All income  disregarded  up to the Federal  Poverty Line Standard AFDC  earned income  disregards  ($30 and 1/3) 50% of earned  income  disregarded Standard AFDC  earned income  disregardsb  ($30 and 1/3)
Time Limit Benefit  elimination Benefit  elimination Benefit  reduction Benefit  elimination Noneb
Organizational   Strategies      Integrated case  management      No major  changes      Performance-  based  contracting    Integrated case  management,  performance-  based  outcomes Transfer of  administrative authority  to Work Force  Development,  one-stop service delivery



    • a Under W-2, effective September 1997, eligibility for assistance is determined on the basis of income and employability. There is a 24 month benefit elimination time limit and no earned income disregard. b Wisconsin's two Work Not Welfare waiver demonstration counties had a $30 and 1/6 earned income disregard and a 24-month time limit

    At the time of our site visit, Wisconsin’s approach to reform was characterized by intensive efforts to divert recipients from the welfare system, stringent penalties for noncompliance, a nearly universal participation requirement, and large-scale organizational restructuring. Although not the focus of this study, Wisconsin has since adopted even more significant reforms through the statewide implementation of W-2 in September, 1997.

    Finally, it needs to be underscored that the move to make welfare an employment focused, transitional assistance system in these five states is taking place when the economy is strong and the low-wage labor market is generally expanding. Consequently, these states provide examples of the types of strategies being implemented and employment outcomes being achieved in a strong economic environment that facilitates and supports efforts to move recipients into the labor market as quickly as possible.

    A downturn in the economy could cast a very different light on the strategies currently being considered and implemented and call for a different or modified set of policy responses. However, this report illustrates how states have taken advantage of a strong economy to make dramatic changes to the face of welfare, both in making fundamentally different assumptions about what should and can be expected of recipients, and then restructuring the system to support and reinforce these assumptions.


    The purpose of this study is (1) to broaden understanding of employment focused welfare reform policies by describing Work First program approaches and other policy strategies used by the five study states; and (2) to discuss some of the interactions and trade-offs involved related to their implementation.

    The information included in this study report is primarily drawn from site visits conducted to state and local welfare offices between January and March of 1997. Discussions were held with: (1) state welfare administrators and/or senior program staff; (2) local welfare administrators and front-line eligibility and welfare-to-work staff; (3) staff from other relevant local agencies and organizations such as employment and training service providers; and (4) employers participating in subsidized employment programs and/or unpaid work experience programs. Information obtained through these discussions is supplemented with administrative data on mandatory recipients’ participation in welfare-to-work program activities over a 12 month period (see Appendix B).

    The remainder of the report is structured as follows:

    • Chapter 2 describes the study states’ Work First programs in greater detail, including the mix and use of job search, subsidized employment and/or unpaid work experience, and education and training;
    • Chapter 3 provides an overview of program coverage issues and exemption policies;
    • Chapter 4 examines penalties for noncompliance and time limits;
    • Chapter 5 addresses earned income disregard policies used to encourage and support recipients’ efforts to engage in unsubsidized employment;
    • Chapter 6 examines up-front diversion policies and practices;
    • Chapter 7 highlights some key organizational strategies used to support the shift to a more employment focused assistance system; and,
    • Chapter 8 concludes the report with a summary of recipient participation patterns.

    Chapter 1 Footnotes

    1. Prior to the enactment of PRWORA, 43 states were granted permission by the U.S. Department of Health and Human Services to implement changes in their AFDC and welfare-to-work (i.e., JOBS) programs under the authority provided through Section 1115(a) of the Social Security Act. Most of these waivers were granted between 1993 and 1996. Some were fairly limited in scope while others were quite comprehensive; some took place on a statewide basis while others were limited to selected areas.

    2. Throughout this report we use the term “work activities” to refer to unsubsidized or subsidized employment, unpaid community work experience and on-the-job training programs. The term “work-related activities” is used to refer to activities designed to prepare recipients to find employment, such as job search or vocational education and training.

    3. Although PRWORA explicitly defines the activities in which recipients must be involved to count towards a state’s participation rate, it does not preclude states from requiring or allowing recipients to participate in activities other than those that count towards the participation rate (e.g., substance abuse treatment). States that expect to exceed the required work participation rates could also place a greater percentage of their total caseload in education and training activities.

    4.States are also permitted to exempt up to 20 percent of their TANF caseloads from the lifetime limit on benefits.

    5. Population figures are for the county in which the local site are located.

    6. As of 1995, the national monthly average AFDC benefit for a family of three was $389. The maximum benefit was $265 in Virginia, $288 in Indiana, $460 in Oregon, $517 in Wisconsin, and $565 in Massachusetts.

    7. A strictly defined work mandate goes into effect after 60 days in MA, 90 days in VA, 18 months in CA, 24 months in DE, MT, and RI, and 30 months in VT.

    8.This two-track approach was eliminated as of June, 1997.


    Chapter 2


    Work First: The Core of Current Efforts to Reform Welfare to Work ProgramsEmployment focused welfare reform initiatives that emphasize the need for recipients to gain quick exposure to and entry into the labor force are often referred to as “Work First” programs. The underlying philosophy of these programs begins with the expectation that most recipients are capable of finding work and assumes that the best way to succeed in the labor market is to join it. It is believed that job advancement and higher wages will come from the experience of working, rather than from first building skills through education and/or training. Hence, employment is both the goal and the expectation, even if the only jobs that can be obtained pay low wages and lack benefits.

    The term Work First neatly captures a shift in both philosophy and practice from the focus on education and training that was commonly found in traditional JOBS programs. Over time, the JOBS program focus on basic education and further skill building through other types of education and training came under criticism. It appeared that engaging in these activities had too often, in practice, become ends unto themselves and that only limited emphasis was placed on connecting participants with jobs. Additional work-oriented activities, most notably job search and unpaid work experience, were permitted but neither (particularly work experience) was used to any significant degree and they typically followed, rather than preceded, participation in education activities.1

    From a programmatic perspective, a Work First philosophy suggests that the first welfare-to-work activity should be a job search that occurs either shortly before or after an individual begins receiving assistance. The increased emphasis on moving recipients off welfare and into jobs as quickly as possible also suggests the need for limiting the amount of time clients spend in activities and closely monitoring their progress. Beyond making job search the first and central activity, a Work First philosophy offers little guidance on what additional strategies should be used to serve clients who are unable to find or retain employment.

    In general, states which have adopted a Work First philosophy continue to utilize the same core components that were offered under the former JOBS program—job search, unpaid work experience/ subsidized employment, education and training—but have made adjustments to how they are used. Common examples of how this is done include: sequencing activities so that job search precedes other types of activities; making greater use of unpaid work experience (or subsidized employment); and limiting education and training activities and/or requiring they are combined with more work-oriented activities (e.g. job search).

    As discussed at the outset of this report, the study states have combined these practices using two basic program approaches: (1) a Work First, Work Mandate approach which encourages recipients to find unsubsidized employment and requires unpaid work experience for those who do not find jobs soon after initial receipt of benefits and (2) a Work First, Participation Mandate approach which uses job search to move those who are job ready into employment and requires those who do not find jobs to engage in other types of program activities designed to make them job ready. Not surprisingly, the degree of program variation increases upon a closer look at Work First programs.

    This chapter provides examples of how Work First actually works from a programmatic perspective at the local level.2 The distinction between Work First, Participation Mandate and Work First, Work Mandate approaches is used to help sort the variation in programs across the five study states. A basic overview of participation requirements in the five study states is presented first, followed by descriptions of the structure, content, and mix of work and work-related activities, efforts to strengthen linkages with employers and views about harder-to-serve recipients. Subsequent chapters address policies that are often administered in conjunction with Work First programs—sanctions, time limits, earned income disregards and up-front diversion.


    All five study states require mandatory recipients to begin participating in program activities far sooner than the two-year maximum allowed under PRWORA. They have also defined what activities are required of recipients in a variety of different ways.

    How Work First, Work Mandate States Define Participation

    In general, the Work First, Work Mandate states focus on moving recipients who do not find jobs within a short time period into unpaid work experience. In order to continue to receive benefits beyond 60 days, Massachusetts requires mandatory recipients who have not found at least a part-time unsubsidized job (i.e., a minimum of 20 hours per week) to engage in 20 hours per week of unpaid work experience or subsidized employment. Before the 60-day work requirement takes effect, recipients are provided the opportunity, but not required, to participate in job search, education or training. Recipients may also participate in these activities after the work requirement takes effect but only in addition to meeting the 20-hour work obligation.

    Virginia requires mandatory recipients to engage in job search immediately after benefits are authorized. If the recipient has not found at least a part-time unsubsidized job (i.e., a minimum of 20 hours per week) within 90 days after benefits are authorized, she must work in an unpaid work experience or subsidized employment position for 30 hours per week in order to continue to receive benefits. If an unpaid work experience position cannot provide 30 hours of work, the balance must be made up by simultaneously engaging in job search, education or training.

    How Work First, Participation Mandate States Define Participation

    The Work First, Participation Mandate states provide somewhat greater flexibility than Work First, Work Mandate states in terms of what activities are required of recipients to meet the participation mandate. Additionally, their program designs do not call for the swift imposition of a strictly defined work mandate.

    Indiana requires a minimum of 20 hours of participation per week. Prior to June 1997, participation requirements varied according to whether an assessment determined a recipient was “job ready.” This distinction was phased out after the time of our site visits. All recipients are currently required to meet the participation requirements formerly reserved only for job ready recipients.

    Job ready recipients must first participate in job search. If the initial job search is not successful, the 20-hour participation requirement can be met by participating in unpaid work experience, additional job search, or a combination of these two types of activities. Job ready recipients can also engage in education or training activities but only if these activities are performed in addition to engaging in 20 hours of more work-oriented (i.e., job search/unpaid work experience) activities.

    Prior to June 1997, recipients whose assessment determined they were not yet job ready had to meet the 20-hour participation requirement by: (1) engaging in job search and/or unpaid work experience or (2) combining education or training activities with at least eight hours per week of job search and/or unpaid work experience. These recipients were reassessed every six months to determine if they had become job ready.

    Oregon requires recipients to engage in an up-front job search for 40 hours per week; part of this job search must occur before benefits will be authorized. Those who do not find a job within 30 days must engage in work or work related activities for at least 20 hours per week (full-time participation is encouraged and preferred). Oregon also permits recipients with substance abuse or mental health problems to fulfill the 20-hour minimum participation requirement by participating in treatment activities, a practice that is unique among the five study states and uncommon among states in general.

    Wisconsin (pre W-2) required all mandatory recipients to engage in an up-front job search prior to the authorization of benefits. This up-front requirement consisted of 60 hours of job search over a 30 day period. For those who did not find a job after this initial job search, the number of hours and type of activity varied by the age of the youngest child. Custodial parents with a child between 12 weeks and nine months of age were required to participate a minimum of 20 hours per week in some combination of parenting classes, job readiness activities, and basic education (job search was not required but also not discouraged). Those with a child between nine and 12 months of age were required to participate in at least 20 hours of job search/job readiness. Those with children over age one were required to participate a minimum of 20 hours but generally closer to 30-35 hours a week in program activities. For the vast majority, this activity was job search/job readiness but recipients were also permitted to participate in short-term training, education and unpaid work experience.

    The program approaches and participation requirements used by the five study states originated in the context of pre-PRWORA waivers. Like most state waiver demonstrations, high priority was placed on increasing participation in employment and work-related activities. At the same time, efforts to increase participation were not shaped by or measured against the higher and more narrowly defined participation requirements required under PRWORA. Thus, while the study states provide examples of program practices and approaches that may generally result in higher rates of participation than were typical of traditional JOBS programs, they do not necessarily reflect what may be needed to meet PRWORA’s participation rates.

    The remainder of this chapter describes various components of the study states’ Work First programs. Job search is examined first because of the central role it plays in Work First programs. State experiences with unpaid work experience and subsidized employment programs are described next, followed by a section that highlights efforts to strengthen linkages with employers. We then turn to a discussion of the role of education and training in a Work First environment. The last section presents views on harder-to-serve recipients in the context of a Work First program approach and some steps taken by a few study states to address the needs of this population.


    Job search programs can be structured in a variety of ways. There are no fixed rules regarding when a job search should be initiated, what activities should be included in it, how much recipients should be responsible for directing their own job search versus having it structured and directed for them, or how closely recipients’ job search activities should be monitored. Each of these dimensions of job search is examined separately below.

    When Are Individuals Expected or Required to Conduct a Job Search?

    In keeping with the general premise that the first activity should be job search, most states which have embraced a Work First philosophy have resequenced program activities so that individuals are generally either required or encouraged to engage in a job search before being assigned to any other activity. However, as reflected in the five study states, the exact point at which individuals are required or encouraged to engage in a job search still varies. Of special note is the practice that requires otherwise eligible applicants to conduct a job search before benefits will be authorized, thereby taking the notion of an up-front job search to its ultimate conclusion (see Chapter 6 for additional discussion of applicant job search in the context of diversion policies).

    Two of the three Work First, Participation Mandate states in our study—Wisconsin (pre W-2) and Oregon—have implemented an applicant job search requirement as a condition of eligibility, although the requirement is structured differently. Compared to Oregon, Wisconsin’s up-front job search period lasts longer (30 days compared to 5-10 days) and requires recipients to engage in a more self-directed job search that places greater emphasis on making contacts with employers (see box for additional description of these two states’ applicant job search requirement).

    At the time of our site visit, the other Work First, Work Participation state—Indiana—did not have an applicant job search. In fact, Indiana differed from the rest of the study states in that it deferred assignment to any activity until after a recipient completed a formal, structured client assessment to determine job readiness.3

    The Work First, Work Mandate study states—Virginia and Massachusetts—do not impose an applicant job search requirement. Although there is no reason why this model could not include such a requirement, these states rely on the swift imposition of the work mandate to provide applicants and new recipients with a powerful incentive to engage in a job search immediately.

    In fact, Massachusetts does not formally require individuals to engage in a job search even after benefits are authorized. Workers do, however, contact recipients several times to impress upon them the importance of finding a job and encourage them to make use of available job search assistance services. The work mandate is stressed in these discussions to help convince recipients’ that undertaking a job search resulting in a paid job is the preferred course of action.

    Virginia does reinforce the work mandate’s implicit job search incentive with a formal job search requirement (including penalties for noncompliance) that begins immediately after benefits are authorized. 



    Applicant Job Search in Oregon and Wisconsin

    Key Characteristic: Both states require applicants to engage in a job search before their applications for cash assistance will be approved. Although many states have or are in the process of implementing a job search requirement for applicants, not as many go so far as to delay or deny the initial approval of benefits contingent upon the applicant’s participation in a job search. 

    Up-Front Applicant Job Search in Wisconsin ( pre-W-2) 

    • Between March 1996 and September 1997, Wisconsin required all cash assistance applicants, with the exception of parents of children 12 weeks of age or younger or otherwise exempt, to engage in 60 hours of job search activities within the 30-day application processing period. Applications were terminated for those who did not complete the requirement. 
    • At least half of the 60 hours of required job search had to consist of verified employer contacts. An employer contact was counted as one hour of job search. Activities required to fulfill the remaining hours of job search and the degree to which these activities were structured varied by county. 

    Up-Front Applicant Job Search in Oregon 

    • In mid-1996, Oregon implemented a four week up-front job search which includes an up-front applicant job search. Between 1-2 weeks of the total four week up-front job search requirement must be fulfilled before benefits will be authorized. Each week of job search consists of 40 hours of participation in a combination of independent job search and structured job readiness activities. 
    • During the first few days of the applicant job search, applicants complete a literacy test and an assessment to determine whether they need substance abuse treatment. If either of these tests indicate that job search is not the appropriate starting point for that applicant, job search can be replaced with other more appropriate activities (e.g., substance abuse treatment, adult basic education). 


    What Types of Job Search Assistance are Provided?

    Job search components range on a continuum from little more than telling applicants to look for work and monitoring their progress to providing recipients highly intensive and structured assistance in preparing for and finding a job. A review of past research indicates that a job search component that combines a recipient’s actual job search with group classroom instruction on job search and job readiness is more effective than one that solely relies on self-directed and unsupervised job search.4 Some, but not all, of the study states have adopted this mixed approach. Key characteristics and differences between job search components in the five study states are described next.

    Regardless of whether recipients are required to engage in a structured job search class or are expected to direct their own job search, bringing an assortment of job search resources together is a basic and important way to facilitate a job search. All of the sites we visited provided “resource rooms” to assist recipients in their job search which contained, at the very least, phones, typewriters and/or computers, fax machines, and listings of job opportunities.

    Beyond making resource rooms available to recipients, common types of job search assistance include providing instruction on: (1) practical skills directly tied to finding a job and (2) job readiness skills designed to facilitate recipients’ job search and better prepare them for the world of work. Practical job search skills include hands-on instruction on resume and cover letter preparation, completing employment applications, identifying job leads, and interviewing skills. Typical job readiness activities include instruction on proper work habits and attire, family life skills (e.g. parenting, time management, financial management and budgeting), goal-setting and problem solving.

    The local sites in the Work First, Work Mandate states offered recipients very little by way of structured assistance in practical job search skills or job readiness skills.

    • In Worcester (MA), recipients are encouraged (but not required) to attend a 1-2 hour group session which serves the dual purpose of a general program orientation and a mini-job club session on how to conduct a job search. Employment workers from the Department of Employment Services (DES) are co-located in local welfare offices to provide job search assistance on a individualized basis, but it is up to the initiative of the recipient to seek their help.5 Also, on their own initiative, recipients in Worcester may go to the DES to obtain information on job openings.
    • In Culpeper (VA), a week long, half-day workshop is held once a month that provides instruction on job search and job readiness. In addition to this relatively minimal amount of assistance, recipients can receive more one-on-one job search guidance when attending required weekly meetings with their welfare-to-work case manager.

    The Work First, Participation Mandate sites place greater emphasis on providing job search and job readiness assistance to recipients, although the degree to which these services are delivered in a structured setting varies. Portland (OR) and Indianapolis /Scottsburg (IN) provide extensive and well-developed job clubs that include instruction on job search and job readiness skills, combined with a guided and monitored job search. By contrast, Racine (WI) does not hold a job club specifically for welfare recipients, but does make a rich array of job search assistance resources and job readiness workshops available to recipients.

    • Indiana contracts with a variety of service providers, thus the specifics of recipients’ job search assignment vary somewhat across providers. Recipients stay with the same provider for the entire job search period. A recipient’s job search assignment in Indianapolis (IN) and Scottsburg (IN) typically starts with one or two weeks of group job club which includes structured half day job readiness/practical job search sessions combined with actual job search. This is followed by a highly monitored full-force individual job search. One provider noted that above all else, structured group job clubs help motivate recipients, break down negative attitudes, and build confidence. Discussions with several providers indicated that a high degree of emphasis was placed on practical job search skills (e.g., interviewing), proper workplace behavior, developing a job search strategy, motivational exercises, participants sharing job leads with each other, realistic goal-setting and problem solving (e.g., arranging for back-up child care in advance).6
    • Oregon also contracts with different service providers for welfare-to-work program activities. In Portland (OR) these services are provided by the local community college and include: (1) ongoing, open-entry job search and job readiness workshops and (2) networking sessions which provide an opportunity for participants to share their job search experiences with each other and receive one-on-one job search counseling.
    • Recipients’ job search in Racine (WI) is significantly less structured than the other two Work First, Participation Mandate states. This is due primarily to the adoption of a more universal rather than targeted approach to service delivery, including the centralization of all employment and training services within a single Workforce Development Center.

    As part of this reorganization, the traditional practice of holding separate structured job clubs for welfare recipients was eliminated. Instead, recipients have access to the same centralized resources and attend the same job search/job readiness workshops as all other job seekers. (Job readiness workshops covering different topics are offered daily.) Recipients are required to engage in a specified number of hours of actual job search and job readiness activities, but it is left up to the client to choose which workshops they attend and how intensively they want to take advantage of other job search assistance resources available through the Workforce Development Center. Local staff reported that most recipients responded positively to the idea of “being treated just like everyone else” and adapted easily to the less structured job search format. There was some concern that the least motivated or confident recipients found this new approach more intimidating and difficult to navigate than the previous targeted and structured job search component.



    Monitoring Recipients’ Job Search

    Monitoring recipients’ job search activities can take a variety of forms and serve both as a mechanism for tracking their progress as well as ensuring that recipients are in fact engaging in required activities. Monitoring practices in the five study states, as of January 1997, are summarized below. 

    Wisconsin (pre W-2): In Racine, workers monitored compliance with the job search requirement on a monthly basis. Recipients submitted business cards and/or copies of completed applications as verification of employer contacts. Certificates of completion were used to verify participation in job readiness workshops. Recipients recorded the time spent in job search assistance resource rooms through a punch-in/punch-out time clock. 

    Oregon: In Portland, workers monitored compliance with the job search requirement on at least a weekly basis. Recipients submitted employer business cards and/or copies of completed applications as verification of employer contacts. Participation in formal job readiness workshops was monitored through attendance records kept by the job readiness provider. Other job search/job readiness activities were tracked through activity logs maintained and submitted by recipients. 

    Virginia: In Culpeper, workers monitored compliance with the job search requirement on a weekly basis. Recipients submitted a self-reported list of five employer contacts on a weekly basis. There was no attempt to verify recipients’ reported activities. 

    Indiana: Contract service providers were responsible for monitoring recipients’ attendance in job readiness/job search classes and general job search activities. This information was forwarded to caseworkers on a weekly basis. Attendance reports were kept on recipients’ participation in formal job readiness/job search classes. Methods used to monitor and verify employer contacts ranged from requiring recipients to bring in documentation signed by the employer to relying on recipients’ descriptions of their contacts with employers. Some providers required daily reports from recipients while others met weekly or biweekly with recipients. 

    Massachusetts: As there is no formal job search requirement in this state, no effort was made to monitor nor verify recipients’ job search activities. Recipients choosing to receive job search assistance from employment staff were asked to report on their job search activities so that staff could provide ongoing input and guidance. 


    Our discussions with staff did not produce information that would suggest that any one of these approaches to job search is particularly more promising or effective than another. The prevailing view was that local economies were so strong and low-wage jobs so plentiful that only those with serious problems (which could not be addressed through job search components) had difficulty finding a job. In conjunction with a strong economy, resequencing job search so that it was typically the first activity, emphasizing the importance of finding any job as quickly as possible, and imposing stricter participation requirements and stiffer penalties for noncompliance were typically viewed as the key ingredients to a successful job search.


    PRWORA has generated much interest in unpaid work experience and subsidized employment because these activities provide an acceptable means to meet the new federal work participation requirements.7 Both unpaid work experience and subsidized employment are considered alternative forms of work. Ideally, participation in these programs will teach participants who lack prior work experience about basic work habits, familiarize them with the “world of work,” and provide them with skills and experience that can place them in a better position to transition off welfare and into unsubsidized employment.8 Unpaid work experience is also considered a way to make welfare less attractive compared to paid employment.

    Unpaid work experience is typically structured to provide participants the opportunity to work for public and nonprofit employers in exchange for welfare benefits. Subsidized employment, on the other hand, is typically structured to provide participants the opportunity to work for private employers. Whereas unpaid work experience is based on a non-financial exchange of work for benefits, employers pay subsidized employment recipients regular wages which are financed out of diverted welfare funds.

    These types of programs present significant administrative and operational challenges—finding interested businesses and organizations with positions that can be filled by welfare recipients, monitoring recipients’ participation, etc. Additionally, there is always the danger that when actually implemented, these activities fall far short of the ideal. With respect to unpaid work experience, there is particular concern that these positions end up being little more than a place for recipients to mark time in a “make work” job. In the case of subsidized employment, there is concern that these programs simply give employers a financial bonus for hiring someone they would have hired anyway.

    As stated previously, most states did not focus much attention or resources on these types of activities under the former JOBS program. Hence, most states have little experience with developing and operating these programs and no experience with operating them on a large-scale basis. This study’s site selection process was shaped in part by the desire to include at least a few states that had program approaches specifically designed to include well-developed unpaid work experience or subsidized employment components. Massachusetts and Virginia were selected for inclusion in part because of their Work First, Work Mandate program approach. Oregon was selected in part because it has a relatively well-developed, statewide subsidized employment program for welfare recipients.

    Overall, it appeared that Massachusetts and Virginia experienced much success in expanding their unpaid work experience component to meet the increased need for these activities resulting from the adoption of a Work First, Work Mandate approach. In contrast, both states experienced little success in implementing their newly created subsidized employment programs. Oregon has experienced much greater success with its subsidized employment initiative and has chosen to use unpaid work experience on a very minimal basis.

    Unpaid Work Experience

    Based on discussions with local staff and a limited number of unpaid work experience site providers, the most noteworthy findings regarding states’ experiences with unpaid work experience can be summarized as follows:

    • Local staff in all study states viewed unpaid work experience as a valuable way to prepare not yet job ready recipients for the world of work. Perhaps equally important, they found that telling recipients that they would be placed in an unpaid work experience position (even if they rarely actually did so) served as a very effective incentive to motivate recipients to seek and obtain unsubsidized employment.
    • Since shifting to a Work First, Work Mandate approach, Massachusetts and Virginia had significantly expanded reliance on unpaid work experience but had not found it necessary to mount massive, large-scale unpaid work experience programs.9 Staff in all states, including Massachusetts and Virginia where demand for work experience slots was highest, reported that they had not encountered difficulty in recruiting sufficient numbers of organizations willing to take on work experience participants.
    • Several factors appeared to diminish the need for a large unpaid work experience component in Massachusetts and Virginia. First, in the case of Massachusetts, a broad up-front exemption policy guarantees that only a relatively small proportion of the caseload at any given time will be subject to the work requirement. Additionally, in both states:
    • the economy is good and substantial numbers of recipients gain employment or simply close their cases prior to the imposition of the work requirement;
    • a subset of recipients who do not initially find jobs also appear to be leaving welfare not long after being assigned to a work experience position for unsubsidized employment or other reasons;
    • significant numbers of families lose their grant due to sanctions for noncompliance; and,
    • there are generous earned income disregard policies in place that make it easier for recipients to combine welfare with unsubsidized employment.
    • Many of the administrative and operational challenges commonly associated with operating unpaid work experience programs (e.g., developing slots, keeping slots filled, monitoring participation and serving as a liaison between recipients and site providers) had not been experienced to any significant degree in the states we visited. Racine (WI) and Worcester (MA) found it helpful to dedicate certain staff to the tasks of developing and monitoring unpaid work experience slots, including serving as a liaison between recipients and site providers. In Massachusetts, recipients were permitted and encouraged to develop their own unpaid work experience positions, thereby generating additional slots and giving recipients flexibility to work out assignments that better matched their own interests or schedules.10
    • The handful of unpaid work experience providers we spoke with did not find the program to be overly administratively burdensome and were generally pleased with their experiences with participants. In general, these providers viewed community service as a mutually beneficial experience and were willing to “go the extra mile” to help recipients succeed in their assignment. Several providers also noted that recipients generally cut their assignment short and rarely remained in a position beyond one to two months. Therefore, these providers felt that taking on unpaid work experience participants was best suited for organizations that could afford to be flexible in terms of providing types of work that required little to no in-house training and tasks that could be easily dropped or transferred to someone else.

    The experiences of the two Work First, Work Mandate states suggest that even in states which opt to define work activities solely or primarily in terms of unpaid work experience (or subsidized employment), there may not be as much need for recipients to be placed in these component activities as might be anticipated—at least when the economy is strong and entry level jobs are plentiful. Whether the same holds true in a period of economic decline is another matter.

    Subsidized Employment

    To date, subsidized employment programs for welfare recipients remain an even more infrequently used work alternative than unpaid work experience. To the extent that subsidized employment is used, it is often limited to a few select areas where an arrangement has been worked out with a few employers.

    Of the states featured in this study, only Oregon had made serious inroads in building a subsidized employment program for welfare recipients. Wisconsin had not placed much emphasis on subsidized employment for welfare recipients in the past but planned to greatly expand the use of subsidized employment under W-2. Although there was no statewide effort to rely more on subsidized employment in Indiana, a variation of a subsidized employment program was operated in Indianapolis (IN) by America Works, a private for profit organization which had obtained state and local funding to serve low-income individuals, including TANF and Food Stamp recipients. Virginia and Massachusetts originally envisioned implementing a large subsidized employment program but their attempts to do so have been stymied due to common implementation barriers.

    According to interviews with local staff, the major operational challenges and barriers to successful implementation of these kinds of programs include:

    • general lack of awareness among employers about the possibility of and financial incentives for participating in subsidized employment;
    • reluctance on the part of employers to participate in these programs even though they receive a subsidy, due to the administrative burden of meeting program rules as well as the stigma that is still attached to hiring welfare recipients;
    • insufficient investment of staff resources and time spent on aggressively marketing this feature to employers, maintaining regular contact with participants and employers to ensure the placement goes smoothly, and alleviating the administrative burden on employers; and,
    • the difficulty of finding a “good fit” between recipient needs and employer needs—if recipients have sufficient skills to be considered good candidates for subsidized employment, they are generally able to find unsubsidized employment.

    In contrast to Virginia’s and Massachusetts’ experience, Oregon has enjoyed considerable employer, legislative, and staff support for and commitment to its “JOBS Plus” subsidized employment program. The remainder of this section focuses on the aspects of JOBS Plus that have facilitated Oregon’s ability to overcome many implementation problems common to subsidized employment programs.

    • Designed by the Private Sector and Supported by Employers. The JOBS Plus program is different from typical subsidized employment programs for welfare recipients in that it was the brain child of a respected member of the private sector who took employers’ needs into account when designing its major features. From the start, employers were generally more receptive to the program because it was designed by one of their own and viewed as “user-friendly.”

      Consistent and Aggressive Marketing. When first implemented, thousands of employers received recruitment letters from the private business whose president was responsible for the creation of JOBS Plus. This initial recruitment method was successful, in part because it was spearheaded by the private sector rather than a public agency. To maintain a large and current pool of employers, welfare-to-work job development staff constantly market the program by sending information to employers, making presentations to large groups of business people, and seeking out referrals from other employers. The local JOBS Implementation Council, which includes local business representatives, also markets the program to the employer community.

      Administratively Easy for Employers to Participate. The program is designed so that it does not present an administrative burden or hassle for employers. To participate, employers simply sign an agreement acknowledging they understand and agree to the program’s rules. The subsidized position may last up to six months. Employers are not required to hire recipients at the end of this period, although most are prepared to do so. Employers must also agree to provide participants a mentor. Participants are treated the same as any other temporary employee in the company—they receive a paycheck, unemployment compensation, worker’s compensation, sick leave, and vacation time.11 Employers submit monthly timesheets on participants to the state and are reimbursed out of a funding pool of diverted TANF and Food Stamps benefits.

      Additional Work and Education Supports for Participants. The JOBS Plus program is designed to ensure that participants are never any worse off financially for taking a JOBS Plus position. If participants are paid less than what they would have received from TANF and Food Stamps, the welfare office issues them a supplemental check. Moreover, recipients receive an array of additional supports for participating in JOBS Plus.

      Prior to starting a placement, the welfare office pays for any special work clothing that may be needed. While participating in a JOBS Plus position, participants receive day care, transportation assistance and maintain their eligibility for Medicaid. The gross income limit used to determine eligibility is increased to 130 percent of the Federal Poverty Line. They also receive the full amount of any child support collected and are encouraged to apply to receive the Earned Income Tax Credit on a monthly (as opposed to annual) basis through their employer. In Portland (OR), rent does not increase for participants living in Section 8 housing.

      After 30 days of participation with the same employer, the employer must contribute $1 to an “individual education account” for every hour the participant works. This money can be accessed by participants for education and training purposes after they leave welfare.


    Changing the welfare system to be more employment focused has accentuated the need to expand working relationships with employers in order to support efforts to place recipients in unsubsidized employment and work activities. Despite the key role employers play in the welfare-to-work equation, welfare-to-work programs traditionally have not focused as much attention or energy on building ties with the employer community as might be expected.

    In part, this is because the focus of these programs was on longer-term education and training rather than employment and/or unpaid work experience. To the extent that obtaining employment was emphasized, welfare-to-work programs tended to focus on helping recipients attain good job search skills and identifying promising job leads rather than actively recruiting employers to hire welfare recipients or forging ongoing partnerships with the employer community.

    On the flip side, employers traditionally have not sought out welfare-to-work programs as a source of labor to fill job openings, relying instead on more standard venues for recruitment and hiring (e.g., word of mouth, newspaper ads). Moreover, staff frequently report that employers, as a rule, are reluctant to hire welfare recipients. From an employer perspective, welfare recipients represent a group that lacks requisite skills, work experience, and familiarity with workplace expectations, and thus, are not generally perceived to be particularly promising employees.

    In shifting to a more work-focused welfare system, states and localities have begun to confront more explicitly the challenge of bridging the gap between employers and welfare-to-work programs. This section highlights three types of efforts observed in the five study states to strengthen linkages with employers: reaching out to involve employers and obtain their input, targeted job development and job retention, and using employers as guest speakers at job search sessions.

    Involving the Employer Community

    Soliciting the involvement of the employer community can take a variety of forms and may occur at both the state and local level. It was common for welfare staff to report that it was often difficult to elicit employers’ involvement and to overcome their reservations about hiring welfare recipients. There is also a strong sentiment that the private sector has a vital role to play in the success of welfare reform initiatives, a responsibility to be involved, and much to gain by being more involved. Bringing employers “to the table” was viewed by staff as key to making employers feel vested in the process and outcome of welfare reform.

    Indiana established local welfare reform taskforces that sought to bring together various agencies, community leaders and employers to solicit their views about what was needed to reform welfare at the local level. In most areas, these taskforces were disbanded after welfare reform officially went into effect. However, in Scottsburg (IN), the Mayor used the taskforce as a springboard for forging an ongoing and active partnership between local government, the welfare office and local, private sector employers. In Indianapolis, an employer board, with active support from local government, has met for several years on issues pertaining to welfare reform and low-wage workers.

    In assessing how best to forge a partnership that benefits employers and welfare recipients alike, members of the Scottsburg taskforce stressed that communication and coordination must be a two-way street—employers need to become better informed about the barriers recipients face in making a successful transition to employment, and welfare offices need to become better educated about what employers need and look for when hiring and retaining employees. It was further noted that employers needed to be willing to assume some flexibility and risk when hiring and employing former welfare recipients. For example, some local employers waived their requirement that new hires must have a high school diploma or equivalent, contingent upon the welfare recipient continuing to pursue obtaining a high school equivalency degree while employed.

    Prior to welfare reform, the welfare office in Culpeper (VA) used the Chamber of Commerce to serve as a liaison between themselves and employers in the community. In the wake of welfare reform, the relationship was further developed so that the local welfare office could better tailor its Work First, Work Mandate model to be more responsive to employer needs. Local businesses were surveyed by the Chamber of Commerce on the types of characteristics and skills they sought in employees as part of this endeavor.

    Wisconsin’s W-2 program design calls for the establishment of Community Steering Committees in each county, a mechanism for community involvement and collaboration that was first tried in two counties under the Work Not Welfare pilot demonstration. These committees are made up of representatives from different sectors of the community and are responsible for supplementing W-2 agency efforts to form linkages with employers by: (1) creating and identifying job opportunities, (2) developing employment strategies and promoting entrepreneurship, (3) providing mentorship opportunities, and (4) helping the W-2 agency create subsidized and community service jobs.

    Targeted Job Development and Job Retention

    One way to increase recipients’ chances of finding a job is for staff to make contact and develop relationships with employers so that they can keep abreast of job openings and direct recipients to them. This practice, known as “job development,” gives recipients who may not be able to break into the job market on their own the additional boost they need to obtain employment. “Job retention” services refer to efforts to help recipients keep jobs once they find them. Both job development and job retention are not without costs as each require additional staff resources or diverting staff from helping recipients acquire jobs on their own. However, both can also increase employers’ willingness to hire welfare recipients and encourage employers to view welfare-to-work programs as a hiring resource.

    Overall, job development and job retention services were not widely provided among the five study states. To the extent that efforts were made to secure employment for recipients through job development, these efforts tended to focus on recruiting employers to take on unpaid work experience or subsidized employment participants or working with recipients while they were in these positions. Given the strong economy, there was generally less perceived need to focus job development efforts on moving recipients into unsubsidized jobs.

    Exceptions to these general observations were found in Portland (OR) and Scottsburg (IN). In Portland, full-time job developers were assigned to market the welfare-to-work program and its subsidized employment component to employers. In Scottsburg, staff had built good working relationships with the personnel staff of major companies in the area. Staff were routinely notified when job openings became available and regularly referred recipients who best fit the employers’ hiring needs.

    It was very common for staff to express concern over the high propensity of recipients to lose jobs and return to welfare. They typically saw a much stronger need for job retention and reemployment services than job development services. However, for the most part, this type of supportive service was also generally absent from the mix of program services.

    Using Employers as Guest Speakers at Group Job Search Sessions

    Some job search programs, such as those observed in Portland (OR) and Indianapolis (IN) regularly enlist employers to make presentations to recipients at their job search sessions. Several positive outcomes were associated with this practice. It gave recipients the opportunity to hear from and interact with an employer in an informal setting. In this setting, they could learn first hand about what an employer looked for and the types of behavior and skills they valued in an employee. Although the same material might be covered by staff in other sessions, the “real life” dimension of having it communicated directly by employers seemed to make it more meaningful to recipients. It also provided the added benefit of developing additional linkages with employers and increasing their awareness of the fact that “welfare recipients” were really job seekers who wanted to work.


    With its emphasis on quick entry into the workforce, there is no predefined role for education and training in Work First models. In general, the study states have continued to include education and training in the overall mix of program services but have sharply curtailed the use of these activities in favor of job search and, in some cases, work activities.

    Some of the more common ways states have tailored education and training to fit within the framework of a Work First approach are: (1) permitting recipients to engage in education and training only after completing a job search that has proven unsuccessful, (2) permitting participation in education and/or training only when it is coupled with more work-oriented activities (i.e., job search or unpaid work experience) and (3) providing more education and training activities on a more flexible, open-entry basis so that participants do not have to wait for long periods of time before starting a work-related activity. In addition, greater effort is made to keep recipients from drifting in and out of an education component. Recipients who are not attending classes regularly or not making progress are reassigned to to more work-oriented activities.

    Even when education and/or training is permitted, the focus has shifted from longer-term basic education or “career-oriented” education programs toward more short-term education and training. At the same time, many respondents noted that there was a lack of such programs currently in existence. At the time of our site visits, the study states were still in the early stages of exploring how education and training could best be restructured to complement a Work First approach, as opposed to simply de-emphasized.

    For example, in Racine (WI), a staff person from the local technical college is under a full-time contract to work on developing new short-term training courses (usually 16 weeks or less) that respond to the needs of the labor market. This endeavor has required developing stronger linkages with local industry as well as working closely with a co-located labor market specialist. Staff in Portland (OR) developed their own short-term clerical and electronics training courses because there were no short-term training courses available that met the needs of their recipients.

    Developing and implementing short-term education and training programs brings its own set of challenges. Not only must current linkages with educational agencies and training providers be restructured, these agencies and providers may be resistant, if not outright opposed, to the underlying goals and philosophy of a Work First approach. In addition, although coupling education and/or training with more work-oriented activities is one way to ensure that the focus on employment and work is not diluted, it may also mean that fewer recipients in practice will opt to engage in these activities.

    Many recipients in a Work First environment reportedly choose not to engage in education and training simply because they have no desire to do so. However, according to staff, others do not because they feel it is beyond their ability to balance the demands of participating in more than one activity and meeting their family obligations (e.g., responsibilities associated with raising children, etc.). On the other hand, supporters of this approach emphasized that this is no different from what non-welfare working individuals must do if they are interested in pursuing additional education and training for their own job advancement purposes.


    As states have moved to program approaches that incorporate a Work First philosophy, most of the focus and emphasis to date has been placed on the front-end of Work First—that is, on job search. While the economy plays a pivotal role in the ease or difficulty recipients face in finding employment, those who do not obtain or retain employment often possess multiple barriers that make it especially difficult for them to participate in required program activities and successfully transition off welfare. Key characteristics of those considered harder-to-serve include some combination of the following: little to no previous work experience, low basic skills, mental health problems, substance abuse, domestic violence, child behavior problems, and legal problems.

    In general, the staff with whom we spoke favored the strong focus on employment under Work First. At the same time, they also expressed the need for more education and short-term training when it seemed appropriate, flexibility to work more intensively with some recipients who need more personal attention, and additional tools and resources to address the needs of hard-to-serve recipients.

    Subsidized employment or unpaid work experience provides one strategy for recipients with little work experience or low basic skills. However, participation in these activities may prove too difficult for some recipients with particularly severe personal or family problems. Others may be able to handle an unpaid work position but the experience will not lead to unsubsidized employment unless other supports are made available.

    Officials and staff also noted that the need to develop more intensive and effective strategies to address the needs of the harder-to-serve has steadily increased since moving to a Work First approach. They believe this is due to the success experienced in moving more job ready recipients into employment. Many also feel that the population left behind is increasingly composed of recipients for whom a simple Work First strategy does not work. Time limits and strict participation requirements have also contributed to a greater sense of urgency for turning more attention to this population.

    Among the study states, Oregon placed the most emphasis and program resources on dealing with harder-to-serve recipients. For example, in Portland (OR):

    • Some staff case managers are assigned to work more intensively to identify and resolve barriers for those who do not find jobs;
    • All welfare-to-work recipients are required to attend an “addictions awareness” session where they are screened for potential chemical dependency;
    • Mental health counselors are located on-site to provide ongoing counseling to recipients, as well as consultations with case managers who may be experiencing difficulty working with a particular recipient; and,
    • Recipients who are identified as having mental health or substance abuse problems may be required to participate in treatment or counseling as a condition of receiving benefits. Hours spent in these activities count toward meeting the participation requirement.

    The average cost of serving welfare-to-work recipients in Oregon has increased, a fact that program staff attribute to working more intensively with recipients facing far more barriers to self-sufficiency than in the past. Staff also generally perceived that their investments in harder-to-serve recipients have paid off.12 Moreover, these costs are not as high as might be expected, primarily because a substantial share of the costs of the more expensive specialized services such as mental health, substance abuse or vocational rehabilitation services are covered by other agencies and/or funding streams.

    Oregon recently asked local offices to develop strategies that will help recipients who do not find jobs make the transition from welfare to work. Since our site visits, Indiana made a similar request to local offices and set aside additional money to fund special “barrier busting” initiatives. Indiana has also increasingly placed greater priority on having job search service providers work more intensively on job search/job readiness skills with recipients who have still not found a job toward the end of their job search assignment. Although not yet implemented at the time of our site visits, Wisconsin’s W-2 program features different levels of employment options, including community service jobs for those who need to build work experience and a “Transition” component that is reserved for those unable to perform independent, self-sustaining work.13

    Moving the harder-to-serve off welfare and into jobs presents a difficult challenge for many reasons. States that are interested in identifying effective or promising strategies to deal with the harder-to-serve have relatively little experience or research on which to draw. For the most part, designing and implementing additional strategies and supports for the harder-to-serve is still in the very early stages of planning and development.

    Because the harder-to-serve is not a homogeneous population, the type and intensity of additional resources, supports and strategies need to vary depending on the particular combination of barriers that recipients face. How best to incorporate new strategies to deal with this population without diluting a Work First message and environment also creates a balancing act for states. According to administrators and staff, a key implementation challenge when first shifting to a Work First philosophy was changing the mindset and expectations of staff about recipients’ potential employability. Staff were encouraged to stop focusing so heavily on recipients’ barriers and adopt the philosophy that recipients can in fact find jobs. While states do not want to lose the overall Work First focus of moving recipients quickly into employment, there is a growing recognition that Work First is the first step and not the final answer to helping recipients make the transition from welfare to work.

    Chapter 2 Footnotes

    1. In FY 1994, based on average monthly data, 42.7 percent of JOBS participants engaged in educational activities, 15.8 percent engaged in training, 10.7 percent engaged in job search/job readiness and 4.3 percent engaged in community work experience or similar activities.

    2. Although not addressed in this study, child care assistance and other supportive services also play a key role in states’ efforts to move recipients into work.

    3. Indiana is currently preparing to implement an up-front applicant job search component. In contrast to Oregon and Wisconsin but like many other states which have an applicant job search, Indiana will not make application processing or benefit authorization contingent upon the fulfillment of this up-front requirement.11Recipients typically quit their subsidized employment job before the end of the six month period and thus few JOBS Plus placements turn into permanent jobs. Between 30-40 percent of placements last for only two months. Recipients who quit one JOBS Plus assignment are permitted one additional placement.

    4. Amy Brown, Work First: How to Implement an Employment - Focused Approach to Welfare Reform, New York: Manpower Demonstration Research Corporation (1997).

    5. In other areas across the state, there is a move to consolidate all employment related program staff and resources into one-stop career centers that are available to the general public. In these areas, welfare recipients must go to the career centers to receive employment services. DES staff are still co-located in the welfare office but only for the purpose of providing welfare recipients information about services available through the career centers and helping them navigate the new system.

    6. Recipients are encouraged to treat their job search like a job and providers use the structured job club as a way to familiarize recipients with common workplace expectations. For example, America Works, a contracted service provider in Indianapolis (IN), reinforces the importance of punctuality and attendance by requiring those who arrive late, miss a class or are not dressed in proper attire to start their job club assignment over again the following week.

    7. Unpaid work experience is also often referred to as community work experience, community service employment or workfare. Subsidized employment for welfare recipients is also commonly called grant diversion or work supplementation.

    8. Evaluation research on small-scale workfare programs in the 1980s found that welfare recipients’ participation in unpaid work experience program did not lead to an increase in participants’ employment or earnings. Thomas Brock, David Butler, and David Long, Unpaid Work Experience for Welfare Recipients: Findings and Lessons from MDRC Research, New York: MDRC (1993).

    9. Our analysis of participation patterns among recipients subject to the work requirement in Worcester (MA) and Culpeper (VA) revealed that only one-fifth to one-quarter of recipients were in either subsidized employment or community work experience at any time over a 12 month period (see Chapter 8).

    10. In Massachusetts, eligibility staff responsible for monitoring recipients during the initial 60 day time frame encourage recipients who are approaching the end of this period to look for a community service experience position that best suits their schedule and interests. Recipients who have not found an unsubsidized job or community service position by the end of the 60 days must take whatever community service assignment is available. Although there is an attempt to take the recipients’ interests and scheduling constraints into consideration when making an assignment, there is no guarantee this will happen.

    11. Recipients typically quit their subsidized employment job before the end of the six month period and thus few JOBS Plus placements turn into permanent jobs. Between 30-40 percent of placements last for only two months. Recipients who quit one JOBS Plus assignment are permitted one additional placement.

    12. The costs of serving welfare-to-work participants increased from an average of $92.82 per case per month in FY 95-96 to $162.46 per case in FY 96-97. See State of Oregon, Adult and Family Services Division, Public Assistance Data Charts, Statewide Data, June 1997.

    13. W-2 Transition participants are required to participate up to 28 hours per week in work or other developmental activities up to their ability, and up to 12 hours per week in education or training.


    Chapter 3


     Program Coverage and Exemptions: Who Must Participate?Program coverage and exemptions are critical dimensions of welfare-to-work programs. In the past, a significant proportion of the adult AFDC caseload did not participate in welfare-to-work programs because they were granted an exemption. The criteria used to define what circumstances made an individual exempt from mandatory participation were set at the federal level through statutory language. In contrast, PRWORA leaves policy decisions regarding program coverage and exemptions largely up to the states, or at state discretion, to local offices.

    Under the former JOBS program, approximately one-half of adults receiving AFDC were exempt from mandatory participation.1 The JOBS exemption that affected the largest proportion of recipients was the exclusion of adult primary caretakers with children under 3 years of age (or under age one, at state option). In addition, recipients caring for an ill or disabled household member, working 30 or more hours per week, living in a remote area, or experiencing a permanent or temporary medical or psychological problem that affected their ability to work were not required to participate in JOBS.2

    Many states began to narrow exemption criteria prior to PRWORA’s enactment through waivers or the state option under JOBS. The most common change was to require parents of young children (as young as 12 weeks) to participate in JOBS. This trend has continued under PRWORA.


    Expanding participation by narrowing exemptions is potentially one of the most significant changes a state can make in the course of reshaping its welfare-to-work program to emphasize employment. As shown in Table 3.1, the study states have made very different choices regarding who should be exempt from participating in work or work-related activities. With the exception of Massachusetts, all moved toward narrower exemption policies, although the specifics of these policy changes varied. Those with the narrowest exemption policies —Wisconsin and Oregon—tightened exemptions in progressive stages over a longer period of time. Due to the different exemption policies adopted by the study states, the program strategies, approaches and requirements discussed in this report roughly apply to anywhere between 20 and 90 percent of a given state’s TANF caseload.

    • Massachusetts’ Work First, Work Mandate program is targeted to able-bodied recipients whose youngest child is over the age of six, covering only about 20 percent of the total caseload. A very structured medical review process is used to determine whether a recipient should be exempt because of a medical or psychological problem.
    • Indiana and Virginia both implemented exemption policies that were narrower than under the JOBS program, although they narrowed the exemptions in different ways. During the first two years of its welfare reform initiative, Indiana eliminated exemptions for recipients living in remote areas, but maintained the standard exemption for those families caring for a child under the age of three. Between June 1997 and December 1998, however, exemptions based on the age of the youngest child will gradually be lowered to 12 weeks. Virginia lowered the age of the youngest child exemption to 18 months, but maintained most other JOBS exemptions.
    • Oregon was one of the first states to eliminate nearly all of the JOBS exemptions. Only VISTA volunteers, women in their last trimester of pregnancy and recipients caring for a child who is 12 weeks or younger are categorically exempt from participation. Other exemptions can be granted by a case manager, but must be mutually agreed to in writing by the case manager and the recipient.
    • Wisconsin lowered the age of the youngest child exemption to one year in March 1996. Then, beginning in January 1997, Wisconsin reduced the youngest child exemption to 12 weeks and eliminated almost all other exemptions.


    Table 3.1

    Study State Exemption Policies

    (as of January 1997)


      Exempt while   Youngest Child   is Less than-- Other Types   of Exemption
    Broad Exemption Policy
    Massachusetts          6 yearsa            Disabled (as determined  by a medical review team);  caring for a child on SSI or  disabled spouse or relative;  last four months of pregnancy;  teenagers attending school full-time
    Narrower Exemption Policy
    Indiana 3 yearsb Disabled; sole care giver of  disabled household member
    Virginia          18 months          Disabled or the sole care giver  of a disabled household member;  last two trimesters of pregnancy;  teenagers attending school or  enrolled in vocational training  full-time
    Very Limited Exemption Policy
    Oregon    12 weeks    Last month of pregnancy; other  exemptions can be granted on a  case-by-case basis
    Wisconsin 12 weeks Last two trimesters; teenagers  subject to Learnfare




    a In Massachusetts, recipients with children between the ages of 2 and 6 (about one-third of the adult TANF caseload) are encouraged to participate in work-related activities but do not face penalties if they choose not to do so. Recipients with children under the age of two (slightly less than half of the adult TANF caseload) are altogether exempt.   b Effective June 1997, the age of the youngest child exemption was lowered in Indiana from age 3 and younger to age 2 and younger. It will be further lowered to age 1 (effective 12/97) and then to 12 weeks (effective 12/98).



    The remainder of this chapter examines issues and trade-offs relating to narrow versus broad exemption policies. The specifics of the study states’ exemption policies are used to highlight some of these trade-offs.


    In the context of welfare-to-work initiatives, program coverage shapes the degree to which the welfare system can potentially be transformed from an income maintenance based system to a work-based system. Universal program coverage is ideologically consistent with the principle that all recipients should be expected and/or required, as a condition of eligibility for benefits, to engage in efforts that will reduce or eliminate their need for cash assistance. Exemption policies run counter to the underlying rationale for universal program coverage. On the other hand, categorical exemption policies are based on the recognition that it is more difficult and costly for certain types of recipients to participate in program activities.

    For example, individuals with disabilities may need additional attention and services to enable them to participate in activities. These activities may be more costly than traditional welfare-to-work program activities. Likewise, mothers with very young children have traditionally been targeted for exemption in large part because there are significant child care costs associated with requiring mothers with young children to work, particularly those with infants. Yet exempting this population comes with a high trade-off since it affects a substantial share of the TANF caseload and places these families at higher risk of staying on welfare longer and experiencing difficulty in moving off welfare.

    Setting up-front exemption criteria makes it possible to screen out these types of recipients immediately, thereby avoiding the additional costs and difficulties (both for the state and the clients themselves) associated with requiring their participation. By exempting a specific group of recipients from the work requirement, the states relieve themselves of the need to build in additional supports and strategies to accommodate the special needs of certain populations (e.g., sufficient child care assistance, services targeted to help disabled populations). However, exempted individuals may lose out on the positive benefits associated with participating in activities designed to lead to employment.

    Thus, states face several trade-offs when designing exemption policies. If they exempt large numbers of recipients, the degree to which they can alter the culture of welfare and affect a wholesale shift toward an employment focused welfare system is weakened. However, a broader exemption policy enables welfare-to-work program strategies to be more uniformly applied in a more concentrated fashion, thereby increasing the likelihood that the smaller pool of mandatory participants will receive the benefit of full exposure to services. If it is determined that exemptions should be limited to a very small number of individuals, states must then spread resources over a larger pool of recipients, address the needs of a much more diverse group of recipients, and forego an up-front mechanism for controlling program costs.

    In practice, these trade-offs play out in different ways, reflecting differences in priorities, financial considerations, and program philosophies. Concerns over the cost and supply of child care weigh heavily on choices regarding how broadly or narrowly to define exemptions relating to families with young children. Similarly, concerns over the costs and complexities of helping harder-to-serve recipients make the transition to employment heavily influence whether states opt to exempt individuals with physical or mental disabilities and how narrowly or broadly to define the exemptions.

    Exemptions and Child Care Issues

    Wisconsin and Oregon, the two study states that require participation of recipients with children older than 12 weeks, both significantly increased financial resources for child care assistance and have also made child care assistance a near-guarantee for the working poor. In Oregon, no families were on a waiting list for child care assistance and case managers reported that most families were able to find a child care provider. Wisconsin more than doubled child care funding since 1994 and recent infusions of funding eliminated a waiting list of over 6,000 working poor families seeking child care assistance.

    Child care considerations influenced Indiana’s decision under its initial waiver to keep the JOBS exemption which excluded adults with children under age three from mandatory participation requirements. The state chose to focus its efforts initially on increasing actual participation among its non-exempt caseload rather than spreading limited resources over a broader pool of recipients who would need child care. Having made significant inroads toward achieving its initial goals, the state has since begun to gradually phase-in lowering the age of the youngest child exemption down to 12 weeks.3

    Massachusetts is an example of a state which, due in part to concerns over child care costs, chose to couple its Work First, Work Mandate program with a broad exemption policy—the work mandate applies only to families headed by an able-bodied adult with school-aged children. Recipients are expected to complete their work requirement while their children are in school, thus child care assistance is not provided for time spent in unpaid work experience. The state does, however, provide assistance to families during the summer months when their children are not in school. Although Massachusetts has substantially increased funding for child care, it is not currently planning to expand the work mandate to families with younger children. Instead, this additional funding is being targeted to families who are seeking employment on their own.

    Exemptions and the Harder-to-Serve

    Although some types of harder-to-serve recipients traditionally have been exempted from participating in welfare-to-work programs, Oregon and Wisconsin now require these recipients to participate in welfare-to-work program activities.

    To address the needs of a more diverse group of recipients, Oregon expanded the range of welfare-to-work program activities in which recipients can be required to participate. Since 1993, local offices have had the authority to mandate participation in substance abuse treatment or mental health counseling. Program staff expect that a greater share of their future financial resources and program development efforts will be targeted to recipients who in the past have either been exempt from participation or tended to perform poorly in traditional welfare-to-work programs and activities.

    As of January 1997, Wisconsin was still in the early stages of developing specific approaches to working with harder-to-serve recipients. In Racine (WI), those who previously would have been exempt due to a disability were being referred to the Division of Vocational Rehabilitation for assessment and assistance with developing an activity plan.

    Complexities Added to Exemption Choices by PRWORA

    The PRWORA work requirement, lifetime limit on benefits, and work participation rates add another dimension of complexity to the trade-offs involved in determining who, if anyone, should be exempt from participating in work or work-related activities. Although states now have the flexibility to design their own exemption policies, these provisions under PRWORA are likely to encourage states to move toward narrower exemption policies, especially regarding the age of the youngest child.

    PRWORA requires that all adult recipients participate in work activities within two years and that states meet work participation rates that started at 25 percent of the TANF caseload in FY ‘97 and increase to 50 percent in FY 2002. Single parents of children under the age of six who are unable to obtain child care are exempt from the two-year work requirement, but are still included in the calculation of the work participation rate. In addition, although PRWORA allows states to exempt 20 percent of their caseload from the lifetime limit on benefits, no such exemption exists for the work requirement. With the exception of single parents with children under the age of one, all families are included in the calculation of a state’s work participation rate.

    An analysis of states’ initial decisions regarding exemptions from work requirements suggests that most states are lowering the youngest child exemption in response to PRWORA.4 Twenty-six states have lowered the age of the youngest child exemption to one year. An additional 18 states adopted a youngest child exemption that is lower than one year, provide for no exemptions based on the age of the youngest child or make decisions on a case by case basis. Only seven states (including Virginia and Massachusetts) have maintained youngest child exemption policies that are above age one.

    While it is clear that states are moving to expand participation to single parents with younger children, information is not yet available to determine whether states will continue to exempt recipients facing physical or mental difficulties or will expand program coverage to include them as well. It is possible that states will maintain current exemption policies for these families but will tighten the process for determining who is actually exempt from participation.

    Over time, if states are successful in reducing their caseloads by expanding participation to a greater share of the TANF caseload, there may be greater movement towards including recipients who need more assistance to make the transition from welfare to work. The extent to which, and the speed at which, this expansion will occur is likely to be dependent on whether states believe they have the program knowledge and resources to help formerly exempt recipients successfully make the transition to employment.

    In sum, the related issues of program coverage and exemptions represent a critical design issue which is influenced by ideological, cost, capacity, and child care considerations as well as federal requirements. In general, states have narrowed exemptions as part of the larger effort to move toward a more work-oriented welfare system but most have stopped short of committing to the development of full-scale programs that require universal participation.

    Chapter 3 Footnotes

    1. Individuals who met JOBS exemption criteria were permitted to participate in the JOBS program on a voluntary basis. Policies regarding the treatment of volunteers (e.g., priority for receipt of services) varied by state.

    2. Under JOBS, states were also required to grant exemptions to recipients who were more than three months pregnant or full-time VISTA volunteers. Teenagers with a child under age 3, but without a high school degree (or equivalent), were required to participate in an educational activity.

    3. Wisconsin and Oregon also lowered the age of the youngest child exemption in successive stages. The phase-in approach allowed each of these states to more effectively manage cost and participation issues pertaining to child care.

    4. Keith Watson, Jerome Gallagher, et. al. Temporary Assistance for Needy Families (TANF) One Year After Federal Welfare Reform: A Description of State TANF Decisions as of October 1997, Washington D.C.: Urban Institute, 1998.


    Chapter 4


    Raising the Stakes: The Role of Sanctions and Time LimitsSanctions and time limits raise the stakes for welfare recipients by creating short-term and long-term negative financial consequences for those who do not follow participation requirements and move off welfare. States have long had the ability to impose sanctions. In contrast, time limits represent a new and significant departure from traditional welfare policy in that the termination of cash assistance after a predetermined number of months of benefit receipt guarantees that cash assistance will be temporary, rather than an ongoing entitlement.

    Sanctions (i.e., fiscal penalties on recipients’ grants) are imposed in response to specific actions committed by an individual that constitute noncompliance (e.g., failing to attend scheduled activities, failing to engage in a required job search or an unpaid work assignment). They are intended simultaneously to hold recipients accountable for their actions while encouraging them to follow program participation requirements when they might not have done so otherwise.

    Time limits, by contrast, set an arbitrary deadline on the length of time that recipients can collect benefits. States may choose to provide temporary extension policies in recognition of the fact that some recipients may have justifiable reasons for still being on welfare upon reaching the time limit. In principle, however, a time limit makes no distinction between those who “play by the rules” (i.e., follow program requirements) and those who do not.

    The subject of time limits evokes much interest and concern. The increased use and severity of sanctions in many states has received far less attention, even though some state sanction policies have an equal or even greater immediate impact on families than time limits. For example, while time limits will cause some unknown number of families to be terminated from assistance in the future, families in many states have already been cut off welfare due to sanctions for noncompliance. Thus, at least in the short term, sanctions have a much more tangible impact on welfare families than time limits.

    In this chapter, we examine policies relating to sanctions and time limits in the five study states. Sanction policies are given greater attention than time limits on benefit receipt in large part because they are much more directly tied to states’ efforts to make welfare more work-oriented and are intertwined with the other short-term steps taken to move recipients into employment.1 Although their implications are far-reaching, time limits have thus far generally played a relatively minor role in the day-to-day effort to move welfare recipients into jobs.

    As time passes and more recipients reach the maximum time limit on benefits imposed by their state, it is likely that (1) the abstract and distant nature of time limits will diminish, (2) related policies concerning extensions and hardship exceptions will assume greater importance, and (3) time limits will play a larger role in motivating recipients to find employment. At the time of our site visits, however, little had happened at the local level with respect to time limits beyond informing recipients about the time limit. Consequently, the discussion of time limits is confined to a general description of these policies in the five study states and their interaction with other policies designed to support and reinforce work.


    In recent years, many states have moved from invoking sanctions as a warning to treat program mandates seriously to actually imposing them when noncompliance occurs. Stricter enforcement has often been accompanied by stiffer penalties in many states. These concurrent trends first occurred under waivers and have continued under PRWORA, which gives states almost complete authority over the design and implementation of sanction policies.

    The most dramatic change in state sanction policies is the increased use of “full-family” sanctions, a term used to describe sanction policies which eliminate a family’s entire grant during periods of noncompliance. Currently, about half of all states have policies in place that allow full-family sanctions. Many adopted tougher sanctions in response to the common perception that the sanction policy under the former JOBS program was not strong enough to motivate recipients to adhere to participation requirements and did not exact a severe enough penalty on those who did not comply.2

    Three Types of Penalties for Noncompliance

    A useful way to distinguish key differences between state sanction policies is by the amount or severity of the penalty imposed. Broadly speaking, the sanction policies used in the five study states provide examples of three different types of fiscal penalties currently used across the country to address the issue of noncompliance.

    • Adult-Only Sanctions, as found in Indiana, remove the portion of the TANF grant covering the needs of the noncompliant adult (roughly one-third of the total grant) but continue to provide assistance to cover the needs of the child(ren) in the assistance unit; the length of the sanction period varies depending on how many times the recipient has exhibited noncompliant behavior.3
    • Full-Family Sanctions result in the elimination of the entire grant, thereby ensuring that no cash assistance is provided to a family in sanction status. Some states, such as Virginia, rely solely on a full-family sanction while others resort to a full-family sanction only after less severe sanctions have been imposed and failed to result in a change in the recipient’s behavior.

    For example, Massachusetts imposes an adult-only sanction for the first three months of noncompliance before imposing a full-family sanction; Oregon imposes a $50 monthly sanction for the first two months of noncompliance, then an adult-only monthly sanction ($140, on average) for the third and fourth months of noncompliance and finally, a full-family sanction for any additional months of noncompliance.

    • Pay-for-Performance, an innovative alternative to traditional sanctions developed in Wisconsin, is designed to replicate as closely as possible the experience of the working world. It ties the amount of the penalty for noncompliance directly to the number of hours an adult recipient fails to participate in required activities—just as a regular employee does not get paid for hours of missed work, recipients’ benefits are reduced for each hour of missed participation.4 Thus, depending on how fully a recipient complies, the amount of the penalty (or alternatively, the amount of the grant) could vary substantially from one month to the next. The maximum penalty for noncompliance is the most severe of the five study states (and states in general) because Food Stamp benefits are also subject to reduction for missed hours of participation.

    It is difficult to obtain consistent data on how often sanctions are imposed, in part because a recipient can be sanctioned more than once for different reasons and because sanction data is tracked in different ways across states. Although it appears that sanction rates vary considerably across states, they are, nevertheless, used quite extensively in some localities and/or states. For example, in Indianapolis (IN), 45 percent of a sample of mandatory recipients were sanctioned for noncompliance over a 12-month period.

    Issues Related to Avoiding Sanctions and Coming into Compliance

    How sanction policies are implemented affects how many recipients are actually sanctioned and how many come into compliance after the sanction takes effect. Implementation practices often reflect differences in how states view and balance the dual intent of sanctions both to motivate recipients to participate in program activities and to hold recipients accountable for their actions. Implementation is also shaped by administrative factors such as the degree to which the sanction process is automated, processing procedures, and how much discretion staff are given to determine whether and when a sanction should be imposed.5

    If the primary intent of using sanctions is to motivate recipients to comply with program mandates, states may focus staff resources on minimizing the actual enforcement of sanctions by actively encouraging noncompliant recipients to come into compliance and making additional efforts to address any barriers to participation. If, however, the primary intent is to make it clear that recipients will be held accountable for their actions and that there are unambiguous negative consequences for failing to comply with program rules, states may favor policies that promote swift enforcement and leave responsibility for initiating compliance up to the recipient.

    All of the states featured in this study send written notices to noncompliant recipients indicating that their benefits will be cut and/or terminated if they do not contact their case manager to establish “good cause” or begin to comply with program requirements. Beyond this standard practice, however, how the dual goals of sanctions are translated into practice varies across states.

    Oregon is the only study state that makes an extensive effort to encourage recipients to comply prior to imposing a sanction and takes steps to safeguard the well-being of children in families whose grants are terminated for noncompliance. Before initiating a sanction, a case manager must contact the recipient to determine whether she has either “good cause” for not participating in program activities or barriers that can be addressed. Further, before the full-family sanction can be imposed, case managers are required to:

    • conduct an intensive case management session with the noncompliant adult to determine if there are barriers to participation that can be addressed;
    • make a home visit and develop a plan with other community agencies involved with the family (including child protective services) that addresses the safety of the child(ren). The plan may include follow-up home visits, protective services, or other community involvement;6 and,
    • obtain a management review and approval for both the decision to impose a full-family sanction and the safety plan for children.

    The rest of the study states send warning letters before imposing a sanction and provide recipients the opportunity to establish a “good cause” exception. However, workers are not required to make any special effort to contact families before the sanction goes into effect.7 In Indiana and Massachusetts, recipients are provided an opportunity to come into compliance before an adult-only sanction will be imposed whereas in Virginia and Wisconsin (unless good cause has been established), no such opportunity is provided and the sanction goes into effect the following month.

    Two strategies—minimum sanction periods and demonstrating “good faith”—were used by the study states to minimize incidences in which sanctions were lifted only to be imposed again. This “revolving door” situation reportedly was a common problem under the conciliation process required under the former JOBS program and was thought to dilute the seriousness of sanctions in the minds of recipients and create an excessive administrative burden on staff.

    Wisconsin’s pay-for-performance model provides the flexibility to recalculate grants (or alternatively, penalties for noncompliance) on a monthly basis and therefore the minimum sanction period is just one month—as long as the recipient fully complies in subsequent months. In contrast, Virginia and Indiana have minimum sanction periods that are longer with each incidence of noncompliance.8 In order to get their benefits restored, recipients must comply with program rules but still receive the penalty as long as the minimum sanction period is in effect.

    Progressively longer minimum sanctions reinforce the principle that there are consequences for noncompliance. They can also reduce the incentive to come into compliance. For example, on the second occasion that a recipient in Virginia receives a full-family sanction for not complying with program rules, benefits will not be restored for three months; under similar circumstances in Indiana, a recipient would have to wait for a full year before the adult portion of the grant could be restored.

    Massachusetts and Oregon do not have minimum sanction periods but do require noncompliant recipients to show a good faith effort to comply with program requirements before benefits are reauthorized. In Massachusetts, a recipient who is not employed within 60 days is given a 10-day grace period to begin working in a unpaid work experience position before an adult-only sanction is initiated. If the recipient is still noncompliant at the end of the third month of benefit receipt, a full-family sanction is imposed and the case is closed.9 In order to get their benefits restored, recipients must comply with participation requirements for two weeks. In Oregon, the specific activity required to restore benefits is decided jointly between the case manager and the recipient and must be an activity that can be completed within a 24 hour period.

    As this overview of sanction policies in the five states suggests, state sanction policies reflect varying trade-offs between achieving administrative efficiency, sending and maintaining a firm and unambiguous message that there are real consequences for failing to comply, and minimizing the use of sanctions when possible. These trade-offs are particularly difficult to assess because it is not clear what proportion of sanctioned individuals fail to comply because they have other sources of income available to them and/or do not think compliance is “worth the hassle” versus those who would participate if given additional encouragement or if more effort was made to uncover and address barriers. Greater understanding of why families are noncompliant could lead to more informed judgements on how these various considerations should be balanced against one another.


    Time limits are premised on the assumption that welfare recipients will make greater efforts to move off welfare if they know in advance that they will not be able to collect cash assistance indefinitely. Although evidence regarding the impact of time limits is slowly beginning to trickle in, time limits still represent largely uncharted territory in the realm of welfare reform. Critical questions remain unanswered as to how time limits affect patterns of work and welfare receipt, what happens to families whose benefits are terminated in the short and long-term, and how various subgroups respond to and are affected by time limits.10

    It is also important to note at the outset that the definition of a time limit has evolved and broadened over time, resulting in the use of the same term to describe quite different applications of the time limit concept. The work requirement that is “triggered” after two years of benefit receipt under PRWORA (or sooner, at state option) is often referred to as a time limit. For the purposes of this report, that type of time limit is considered to be one dimension of work participation requirements and, as such, is not addressed in this section. The focus here is on time limit policies which simply terminate or reduce benefits after a prescribed period of time.11

    Key Characteristics of Time Limits

    As illustrated by the following overview of key characteristics of time limits in the study states (as of early 1997), there is a fair amount of variation in how states have designed time limit policies. Each of the study states impose a 24-month time limit but within different time frames. Recipients can collect benefits for 24 months within a 84-month period in Oregon, 60-month period in Massachusetts and Virginia, 48-month period in Wisconsin and 24-month period in Indiana.

    Massachusetts is the only study state which uses different criteria to determine who is subject to the work requirement and who is subject to the benefit termination time limit. Able-bodied recipients with a child between the ages of two and six are not subject to the state’s work requirement but are subject to the time limit.

    Two important distinctions between Indiana’s time limit compared to other states are: (1) upon reaching the time limit, the family’s benefit in Indiana is reduced by the adult portion of the grant whereas benefits are completely terminated in the rest of the study states; and (2) Indiana’s original time limit counted both months on and months off assistance toward a recipient’s time limit. Thus, a recipient would reach the time limit exactly 24 consecutive months after benefits were authorized, regardless of the number of months the recipient actually received assistance. Other states, including Indiana since mid-1997, count only months on assistance toward the time limit.

    Time Limit Extension Policies. All of the study states have adopted different up-front exemptions (summarized in Chapter 3) and back-end extension and/or exception policies from the time limit:

    • Massachusetts provides no specific criteria for extensions to the time limit, although it does permit extensions to be granted on a case-by-case basis.
    • Virginia allows extensions for up to one year if the unemployment rate in the locality is greater than 10 percent or if the extension will allow the participant to complete an employment-related education or training program; extensions can also be granted for up to three months if a recipient has actively sought employment but was unable to find a job, or loses a job for reasons unrelated to job performance.
    • Wisconsin (pre W-2) permitted extensions to be granted on a case-by-case basis to recipients who: (1) had complied with participation requirements and made a diligent effort to find work but could not do so or (2) could not find work due to low achievement ability or learning disabilities, severe emotional or family problems, or local labor market conditions.12
    • Indiana permits extensions for up to one year if a recipient has substantially complied with program requirements and still is unable to find employment in spite of all appropriate efforts.
    • Oregon permits extensions for up to three months for the care of family members who suffer a serious health condition. Unlike other states, Oregon also exempts individuals from the time limit if they fully comply with program participation requirements.

    Time Limits Present Complex Interactions and Trade-offs

    Time limits create new trade-offs as states attempt to design coordinated and coherent employment focused welfare programs. Some of the interactions between time limits and other key strategies used in the five study states to make welfare more employment focused are discussed next.

    Time Limits and a Work First Approach. In a time limited environment, an advantage of a Work First approach is that welfare recipients stay on welfare for shorter periods of time and therefore do not use up the finite months of eligibility allotted to them as quickly. An alternative view is that time limited welfare calls for providing recipients more opportunity to use the time on welfare to engage in longer term skills preparation that will help them attain higher paying jobs and stay off welfare permanently. Regardless which approach is taken, time limits also accentuate the need to incorporate job retention and reemployment strategies into the overall program design.

    Time Limits and Sanctions. Interactions between state sanction and time limit policies play out in several ways that may not be fully appreciated when either is examined separately. For example, in Virginia and Massachusetts, the combined effect of full-family sanctions and time limits with limited extension policies is that most noncompliant individuals will have been cut off welfare long before reaching the time limit. Therefore, time limits will only affect those who have “played by the rules” and complied with program requirements. In Oregon, the combination of a full-family sanction policy and a time limit policy that exempts recipients who comply with program rules makes it unlikely that many recipients will actually be affected by the state’s time limit. There is no interaction between time limits and sanctions in Indiana because both result in the removal of just the adult portion of the grant. By the same token, however, the time limit also provides no additional incentive for sanctioned recipients to come into compliance.

    Full-family sanctions that do not automatically result in the closure of a case also have different implications in a time-limited environment. For example, Virginia does not automatically close sanctioned cases. Therefore, unless the sanctioned family initiates a request to have their case closed, every month of noncompliance also counts toward the time limit—even though no benefits are received.13

    Exemptions from Work Requirements and Time Limits. States may have a variety of reasons for exempting different types of recipients from work requirements. Optimally, these exemptions should mesh with time limit exemption policies. In Massachusetts, however, a substantial portion (approximately 31 percent) of the TANF caseload is subject to the time limit, but not to the work requirement. Thus, when this group of recipients hits the time limit, they will do so without ever having been exposed to a work requirement. To the extent that the state’s Work First, Work Mandate program provides recipients an opportunity to gain valuable work experience and an incentive to obtain unsubsidized employment, exempting recipients from the work requirement but not the time limit places these recipients at a comparative disadvantage.

    Earned Income Disregards and Time Limits. While earned income disregard policies are generally intended to support recipients’ efforts to work, and make it easier for families to combine welfare with work, time limits ultimately penalize those who benefit from earned income disregards. This situation arises because the more generous the earned income disregard, the more months a family can work without losing eligibility for some amount of cash assistance. However, unless a state exempts months in which recipients work, each month on assistance counts toward the time limit. Thus, when taken together, these two policies are working at cross-purposes against one another. While most states have increased their earned income disregard, very few have opted to exempt working recipients from the time limit. Among our study states, only Oregon allows working recipients to continue to receive benefits once they reach the time limit.

    Chapter 4 Footnotes

    1. Time-triggered work requirements, which also often fall under the general category of time limits, are addressed in Chapter 2.

    2. Under JOBS, the penalty for noncompliance was removal of only the adult portion of the grant (roughly one-third of the total family grant).

    3. In addition, sanctioned adults (but not children) in Indiana lose Medicaid coverage for varying periods of time (e.g., a minimum of six months for the second and third occurrence of noncompliance) when they are sanctioned.

    4. Wisconsin (pre W-2) calculated the penalty for noncompliance as follows: for each missed hour of required participation, a penalty equal to the federal hourly minimum wage was imposed, first on the cash assistance grant and then on Food Stamp benefits. Recipients who participated in their required activity for less than 25 percent of the scheduled hours lost their entire cash assistance grant and all but $10 of Food Stamps for the next month. The family’s case was closed if complete noncompliance (i.e., no hours of participation) occurred for three consecutive months.

    5. U.S. General Accounting Office, States’ Early Experiences with Benefit Termination (GAO/HEHS-97-74; May 1997).

    6. If closure of the case puts children at risk, protective payments may be issued for all or part of the grant while a long-term plan is put into place. Dual-payee checks, authorized representatives and vendor payments are all acceptable protective payment methods.

    7. Since the time of our site visits, Indiana has adopted a new policy to follow-up on sanctioned cases to determine if an individual’s noncompliant behavior is the result of any barriers to participation and, if so, to work with recipients on removing them. Wisconsin, Virginia and Massachusetts make no attempt to follow-up on sanctioned cases.

    8. In Indiana, the minimum sanction period is two months, 12 months, and 36 months for the respective first, second and third occurrence of noncompliance. In Virginia, the sanction must be in place for a minimum of one month for the first occurrence of noncompliance, three months for the second occurrence and six months for the third occurrence before a recipient’s grant can be restored.

    9. Because of the processing time involved in enforcing the adult-only sanction, it is common for noncompliant recipients to be notified that they will receive a full-family sanction before they have actually experienced the adult-only reduction in their grant. This administrative complication partially undermines the intent and rationale of the state’s progressive sanction policy which seeks to give recipients a “second chance” before completely terminating the family’s grant.

    10. For a more detailed discussion of key issues regarding time limits, see Dan Bloom and David Butler, Implementing Time-Limited Welfare: Early Experiences in Three States, New York: MDRC (1995) and Dan Bloom, After AFDC: Welfare-to-Work Choices and Challenges for States, New York: MDRC (1997).

    11. Under PRWORA, states cannot use federal block grant funds for families after 60 months of benefit receipt. States can, however, exempt up to 20 percent of the total number of families receiving TANF assistance from the lifetime limit on benefits. The design and implementation of extension policies are left to the discretion of each state.

    12. Wisconsin’s time limit was first implemented in two waiver demonstration counties and became effective on a statewide basis under W-2.

    13. Local staff in Virginia reported that priority is placed on conveying to recipients the importance of closing their case should they be sanctioned.


    Chapter 5


    Financial Policies to Encourage and Support Work: The Role of Earned Income DisregardsEarned income disregards often receive little attention in public debate on how to promote and reinforce recipients’ work efforts, in part because these policies require greater understanding of how eligibility and benefits are determined. Yet, the impact of these policies on family income and ability to combine work and welfare can be quite significant, especially in a Work First environment where recipients are encouraged to take any job, even if it is part-time and pays low wages.

    PRWORA gives states total flexibility to determine how to treat earnings when calculating eligibility and benefit levels for applicants and recipients. In contrast, the former AFDC program applied a uniform set of rules regarding the treatment of earnings in conjunction with eligibility and benefit determination. These rules were commonly criticized for not being generous enough to provide recipients sufficient reason to seek or maintain employment. The vast majority of recipients relied solely on welfare for support; those who did try to combine welfare with work were often barely better off than those who did not.

    In response to these shortcomings, most states with pre-TANF waivers modified existing AFDC earned income disregard policies. The new earned income disregard policies are designed to increase the attractiveness of work relative to relying solely on welfare for support.1

    This chapter examines earned income disregard policies in the five study states (see Table 5.1). Our discussion makes the following key points:

    • by allowing states to design their own earned income disregard policies as opposed to using the same policy across all states, a state’s benefit level alone no longer determines whether a working recipient is better off in one state versus the next;
    • the Work First, Work Mandate study states—those with the most rigorous work requirement as well as severe penalties for noncompliance—coupled this approach with very generous earned income disregards that made even part-time, low-paying unsubsidized employment substantially more financially attractive than relying solely on welfare; and,
    • strict work requirements, stiffer sanctions and benefit time limit policies change the potential impact of earned income disregards and the role they play in providing an incentive to work. On the one hand, strict work requirements and stiffer sanctions heighten the attractiveness of working in unsubsidized jobs, particularly when coupled with generous earned income disregard policies. On the other hand, the time limit on federal assistance presents a serious drawback for using generous earned income disregards because months during which recipients combine welfare with work count toward the time limit. Thus, earned income disregards cause welfare recipients who work to lose valuable months of eligibility.


    The earned income disregard policy in effect under the former AFDC program entitled recipients who found employment to:

    • a $90 work expense disregard for as long as they worked (and received welfare);
    • an additional $30 earned income disregard for the first 12 months of employment; and,
    • an additional disregard for the first 4 months of employment equal to one-third of the remainder of a recipient’s earnings.

    This policy gave recipients the greatest amount of continued financial assistance during their first four months of employment because one-third of earnings (plus $30) was not counted when calculating their eligibility and benefit levels. However, the financial advantages of working diminished significantly after that point because every dollar of income earned above $120 per month resulted in a dollar reduction in the total amount of the grant. 


    Table 5.1

    Treatment of Earned Income

    in the Five Study States*

    (as of July 1997)


      Earned Income  Disregard Eligibility   Period Eligible   Groups Special   Provisions
    Standard Policy   under Former   AFDC Program $30 and 1/3      $30 and 1/3  for 4 months;  $30 for 8  additional months All working  applicants and  recipients None     
    Indiana          $30 and 1/3          $30 and 1/3  for 4 months;  $30 for 8  additional months    All working  applicants  and recipients      Working recipients  still considered  technically eligible  if initial earnings  do not exceed  Federal Poverty Line
    Massachusetts              $30 and 1/2             

    $30 and 1/3

    As long as  eligible for  assistance     

    As long as  eligible for  assistance

    Applicants and  recipients subject  to the time limit 

     Recipients not  subject to time  limit who received  TANF in previous  4 months

    Earned income disregard  extension was  accompanied by a  2.75 percent reduction  in benefits       
    Oregon          50 percent          As long as  eligible for  assistance      All working  applicants  and recipients      Gross income limit  based on a state defined  standard— roughly  one-third less than  income limit typically  used by other states
    Virginia                $30 and 1/3                  $30 and 1/3  for 4 months;  $30 for 8  additional  months          Recipients who  are subject to  the time limit  only            Disregards are used  to determine net  earned income.  Recipients receive  full benefit as long as  net earned income  and the benefit  amount do not  exceed the Federal  Poverty Line
    Wisconsin (pre W-2)      $30 and 1/6      As long as  eligible for  assistance Applicants and  recipients in  two demonstration  counties None     




    * All states use a $90 work expense disregard


    Among the 32 states that obtained waivers to change the rules governing earned income disregards, the most common policy change was to replace the time limited aspect of the earned income disregard with a permanent disregard that did not vary over time. While many states eliminated the time limit on the disregard and made it more generous, some states chose to offset the costs associated with this change by: (1) coupling it with an overall reduction in the amount of income that could be disregarded, (2) restricting its use to just a portion of the caseload or (3) concentrating it on those with very low earnings.

    For example, prior to the implementation of W-2, Wisconsin eliminated the time limit on the earned income disregard but also reduced the amount of the disregard to one-sixth of total earnings plus $30 per month.2Massachusetts and Oregon took a slightly more complicated approach. To finance the earned income disregard, Massachusetts reduced benefits for all families subject to the time limit by 2.75 percent.3 It eliminated the time limit on the disregard for all recipients who had received assistance and provided a more generous earned income disregard to families subject to the time limit ($30 and ½ of all earnings). Oregon increased the earned income disregard to one-half of total earnings but did not expand its state-defined gross income limit (which is relatively low compared to other states.) Therefore, the potential number of recipients who could benefit from this change was contained.

    Among the five study states, the earned income disregard changes in Virginia and Massachusetts—both Work First, Work Mandate states—were the most generous. As discussed below and summarized on Table 5.2, these changes produce varying impacts on the amount of earnings recipients can have and still remain eligible for some level of cash assistance.



    Making Work Pay in Virginia

    Virginia’s generous earned income disregard provides continued cash assistance to full-time working recipients earning up to $1110 per month. Recipients who work do not lose any cash assistance until their total net income raises them out of poverty, as defined by the Federal Poverty Line (FPL). 

    • The state continues to use a $90 work expense disregard and the 30 and 1/3 earned income disregard (including the four month limit on the latter) to determine countable income for eligibility purposes. However, instead of retaining the practice of a dollar for dollar reduction in benefits for all non-disregarded earnings, recipients continue to receive a full grant so long as the combined total of benefits and earnings do not exceed the FPL. 
    • In the absence of any earned income disregard, the maximum amount of income a recipient can earn and still be eligible to receive benefits in Virginia is much lower than the FPL. The new earned income disregard policies give recipients’ greater opportunity to combine work with welfare and, in so doing, significantly raise their family income. Once a recipient’s net earned income and benefits exceed the FPL, the benefit payment is reduced until total income equals the FPL. 
    • The combination of Virginia’s earned income disregard policy, Work First, Work Mandate program approach and full-family sanction policy makes work in unsubsidized employment by far the most attractive option. A recipient with two children who works 20 hours per week at a minimum wage job would receive the equivalent of $905 per month when Food Stamps, earnings (before taxes) and cash assistance are taken into account compared to just $606 per month if the recipient simply complied with the work requirement by working in a unpaid work experience position. 





    Under the former AFDC earned income disregard policy, the degree to which the policy permitted welfare recipients to combine earnings with benefits depended completely on a state’s benefit level. Recipients residing in high benefit states could earn more and still remain on welfare than recipients in low benefit states. For example, under the old AFDC rules, a working recipient in a 3-person case in Massachusetts (a relatively high benefit state) could have still been eligible to receive some cash assistance until monthly earnings reached $990 while a working recipient in Virginia (a relatively low benefit state) would have lost all assistance when their monthly earnings reached $560.

    The flexibility provided first through waivers, and now through PRWORA, allows states to take their benefit payment and need standards into account when designing earned income policies, thereby giving states total control over determining how much recipients can earn before they no longer qualify for cash assistance. As Table 5-2 illustrates, this flexibility has allowed a low benefit state like Virginia to continue to provide cash assistance to working recipients at much higher incomes than was possible under the former AFDC program.

    Under the old AFDC rules, a Virginia recipient working more than four months would have had her assistance terminated if she earned more than $380 per month. Under Virginia’s current earned income disregard policy, the same recipient can earn almost triple that amount ($1110 per month) before completely losing benefits. 


    Table 5.2

    Monthly Break-Even Wages:

    Before and After Earned Income Disregard Policy Reform*

    (As of July 1997)


      First 4 months of Employment After 12 months of Employment
    Study State (payment standard) Pre-Waiver Post-TANF Pre-Waiver Post-TANF
    Virginia ($291) $560  $1,110  $380  $1,110 
    Massachusetts ($565) $990  $1,050  $670  $1,050 
    Indiana ($288) $550  $550  $380  $380 
    Oregon ($460) $620  $620  $550  $620 
    Wisconsin ($517) $900  $740  $610  $740 




    * Calculations assume a 3-person family working 40 hours per week and uses 1996 payment standard information. Wisconsin figures reflect the earned income disregard in effect in two demonstration counties prior to statewide implementation of W-2. Virginia figures represent the break-even wages for the largest portion of the state’s population.


    The benefits available to recipients at various combinations of hours and earnings further illustrate how working recipients fare in the five study states (see Table 5.3). Although Virginia and Massachusetts both implemented more generous earned income disregards, they work in different ways. Recipients working part-time (i.e., 20 hours per week) at minimum wage receive more assistance in Massachusetts than in Virginia—$387 per month versus $291 per month. However, because benefits phase out at a slower rate in Virginia than Massachusetts, recipients with higher wages and/or working full-time fare better in Virginia than in Massachusetts. For example, recipients working full-time at minimum wage in Virginia still receive $291 per month while similarly situated recipients in Massachusetts receive just $149 per month.

    The only recipients in Indiana, Oregon and Wisconsin who benefit from their states’ earned income policies are recipients working part-time and at low-wages and none receive as much support as working recipients in the Work First, Work Mandate states.

    • In Indiana, earned income disregards provide little opportunity for recipients to combine work with welfare and thus little financial incentive to support recipient efforts to seek and maintain employment. Recipients who earn minimum wage and work part-time are eligible for just a $50 monthly grant for only four months; recipients earning higher wages or working more hours lose all their benefits immediately.
    • In Oregon, earned income disregards primarily help recipients working at minimum wage. A recipient in a part-time job receives a monthly grant of $222. However, if the same recipient earned $7.50 an hour in a part-time or full-time job, she would no longer be eligible to receive any assistance.
    • In Wisconsin (pre-W-2, effective only in two demonstration counties), increased earned income disregards primarily helped recipients working limited hours and at low earnings. However, benefits were available to recipients over an extended period of time rather than during just their initial transition into employment.

    In sum, the earned income disregards in effect in the two Work First, Work Mandate states—Massachusetts and Virginia—support and reward recipients’ efforts to enter the paid labor force and stay employed to a far greater extent than the Work First, Work Participation states. Thus, while Massachusetts’ and Virginia’s program model calls for recipients that have not found jobs to be engaged in work programs soon after benefits are authorized, these states also give recipients much more financial incentive to obtain unsubsidized employment and provide continued financial support for those that do work. 


    Table 5.3

    Earned Income Policies' Impact on AFDC/TANF

    Grants by Level of Wages and Hours Worked


      Indiana Massachusetts Oregon Virginia Wisconsin
    First 4 months of Employment
    20 hours per week, 4.33 weeks per month
    $5.50 per hour $50  $387  $222  $291  $220 
    $7.50 per hour 0 $300  0 $291  $76 
    $10.50 per hour 0 $170  0 $291  0
    40 hours per week, 4.33 weeks per month
    $5.50 per hour 0 $149  0 $291  0
    $7.50 per hour 0 0 0 0 0
    $10.50 per hour 0 0 0 0 0
    After 12 months of employment
    20 hours per week, 4.33 weeks per month
    $5.50 per hour 0 $387  $74  $265  $220 
    $7.50 per hour 0 $300  0 $265  $76 
    $10.50 per hour 0 $170  0 $265  0
    40 hours per week, 4.33 weeks per month
    $5.50 per hour 0 $149  0 $248  0
    $7.50 per hour 0 0 0 0 0
    $10.50 per hour 0 0 0 0 0


    The ability of earned income disregards to produce the desired effect on recipients’ work efforts depends largely on whether recipients understand how they can affect their benefits and total income if they get a job. Given the general complexity of earned income disregards, it is not surprising that workers report that they tend to confine discussion about the earned income disregard to general statements such as “you may still be able to receive some benefits if you get a job.” More detailed explanations of how the earned income disregard policies could work in the recipient’s favor frequently go unaddressed. Relative to other types of information imparted to recipients about program rules and policies, earned income disregards generally tend to receive little priority or attention.

    Among our study sites, the exception to this general observation can be found in Virginia. Virginia’s earned income policy is both generous and relatively easy to explain and understand. Workers reported they used the earned income policy as a tool to motivate recipients to seek employment, emphasizing how the policy worked and how it benefited recipients. Workers in Culpeper (VA) also made a point of informing recipients about the Earned Income Tax Credit, another means of financial support for low-income families.

    As stated elsewhere in this report, earned income disregards also carry different trade-offs in the context of a time-limited environment. Since every month a welfare recipient is able to combine work with welfare counts toward the time limit, recipients who benefit from the additional support made possible by earned income disregards run the risk of reaching their time limit sooner than if benefits were terminated when they started working.

    A final point of note is that the environment in which states are implementing more generous earned income disregards is quite different from that of the past. For recipients who are capable of obtaining employment, stricter participation requirements and sanctions for noncompliance make unsubsidized employment a more attractive option than welfare, even in the absence of earned income disregards. Therefore, the degree to which earned income disregard policies provide an incentive to seek and obtain income may be diminished as other incentives to become employed have increased. At the same time, given that a Work First approach encourages recipients to take any available job no matter how little it pays, the ability of earned income disregard policies to reduce poverty by supplementing the earnings of low-wage working families assumes greater importance.

    Chapter 5 Footnotes

    1. As of October 1997, only 15 states still used the earned income disregard policy associated with the former AFDC program. See Keith Watson and Jerome Gallagher, et. al., Temporary Assistance for Needy Families (TANF) One Year After Federal Welfare Reform: A Description of State TANF Decisions as of October 1997, Washington, D.C.: The Urban Institute (1998).

    2. Prior to statewide implementation of W-2, this earned income disregard policy change was implemented only in the two Work Not Welfare waiver demonstration counties. There is no earned income disregard under W-2.

    3. In addition to helping finance the earned income disregard, state officials also hoped the grant reduction would encourage time limited recipients to find employment.




    Up-Front Diversion StrategiesIn addition to changing the program focus and requirements of their welfare-to-work programs, some states have concurrently implemented strategies which aim to deter those who may not need cash assistance on an ongoing basis from applying for benefits or completing the application process. These strategies are commonly grouped under the collective heading of “diversion.” Diversion policies have sparked much interest as well as concern, in part because so little is known about how many families are actually diverted, what happens to these families, and whether their needs are being sufficiently met through alternative sources of support. Although lack of data on these issues leaves critical questions unanswered, our study collected information on how diversion policies are implemented and viewed by staff at the local level in the study states.

    Diversion policies can be grouped into three different categories:

    • One-Time Financial Diversion Assistance
    • Identification and Use of Alternative Resources
    • Applicant Job Search

    Financial diversion assistance is a straightforward alternative form of public assistance which is uniquely different from regular cash assistance because it provides a one-time lump sum payment rather than an ongoing source of public support. Its relevance may be limited to a relatively small proportion of applicants but it is arguably a more appropriate form of assistance for individuals whose financial needs are of an immediate and short-term nature.

    Identification of alternative resources and encouraging individuals to reconsider their intention to apply in light of these alternatives, is a much less structured form of diversion than one-time financial diversion assistance. As such, its implementation may vary in terms of intensity, tone, and the overall message conveyed to clients. Local staff must strike a balance between strongly encouraging clients to rely on alternative sources of support and not conveying this message in a way that intimidates needy families from applying for cash assistance. To be done well, this form of diversion requires workers to have good interviewing and communication skills, knowledge of what resources and referral systems exist in the community, and sufficient time allotted to carry out the interview.

    Up-front job search, also described in Chapter 2 of this report, is targeted to individuals for whom the first two types of diversion policies may not be appropriate but who may be able to find employment quickly if given sufficient incentive to do so. Unlike those who are diverted before ever applying for benefits, the “diversion” is accomplished during the application processing period and before benefits are actually approved. Up-front job search requirements, especially those which make benefits contingent upon fulfilling the requirement, also appear to have an additional diversionary effect in that they may deter some applicants from even submitting an application.

    The remainder of this chapter provides a fuller description of these three “diversion” policies, followed by a discussion of some implications and concerns associated with their use.


    Financial diversion assistance provides individuals seeking cash assistance with a one-time lump sum cash payment in lieu of monthly benefits. This diversion method is specifically targeted to individuals who are eligible to receive ongoing cash assistance but are really just in need of short-term temporary assistance. Likely candidates for diversion assistance are individuals who have lost employment but who may already have another job lined up to start in the near future or who need a short-term infusion of cash to cover an unanticipated financial emergency (e.g., car repair, medical bills).

    Only a relatively small proportion of applicants may fit the circumstances which make financial diversion assistance an appropriate alternative. At the same time, this type of diversion policy does provide a less expensive alternative to welfare by reducing the likelihood that what ostensibly should have been a short-term stay on welfare turns into a long-term proposition.

    Currently, financial diversion assistance is offered in about a dozen states, one of which is Virginia.1 Each have slightly different policies regarding its use. In Virginia, all TANF-eligible individuals can receive a lump sum financial diversion assistance payment equal to 120 days of assistance in lieu of receiving a monthly welfare check. This payment can be provided only once over a sixty month period and bars an individual from receiving TANF for 160 days.

    Individuals seeking cash assistance in Virginia rarely opt to receive financial diversion assistance; only about two percent of recipients accepted it in favor of regular cash assistance. Local staff attributed the low utilization rate to the fact that the vast majority of applicants in fact needed steady and ongoing cash assistance. At this relatively early juncture in implementation, it is difficult to ascertain how accurately staff perceptions reflect the extent to which diversion payments may be a viable and/or attractive alternative for recipients. The state’s low utilization of diversion assistance may also be due to reluctance on the part of staff to aggressively “market” this option because they are unconvinced about its value.2


    Discussing alternative forms of support with applicants in an interview that occurs prior to eligibility determination is a more broad-based method of diversion. The up-front interview is conducted with all cash assistance applicants, not just those subject to welfare-to-work program requirements. In contrast to traditional welfare intake practices which concentrated on ensuring that eligible applicants signed up for the full range of benefits provided through the welfare system, diversion interviews focus on matching client needs with the least level of support needed and, if possible, drawing on community resources.

    Among the five study states, three—Indiana, Oregon, and Wisconsin—had implemented an up-front interview which was either explicitly or implicitly designed to deter some individuals from applying for cash assistance. Although each of these states shared the same goal of encouraging clients to consider welfare only as a last resort when all other means of support had been exhausted, the goal was interpreted and articulated by staff somewhat differently across states:

    • Local staff in Oregon viewed the goal of its up-front “self- sufficiency interview” as one of helping applicants realize that other options may be available to meet their needs, thereby rendering the need for ongoing cash assistance unnecessary.
    • In Wisconsin, local staff stated that the goal of its up-front “self- sufficiency planning interview” was to prevent any applicants who could get by without going on assistance from doing so by helping them to reconsider their situation in the context of any and all other available resources.3
    • Indiana couched its version of this policy in language that emphasized that workers should not interpret the interview as a mechanism for coercing individuals seeking assistance from applying. Staff in Indiana were somewhat ambivalent as to whether the true goal of its “up-front needs assessment” was to divert clients from applying or simply to make them aware of alternative resources.

    Although these implementation differences are nuanced and difficult to capture, they suggest that the underlying tone of diversion interviews can contribute to potential differences in the overall message conveyed to applicants. Hence, the implementation of such an interview should be accompanied by a clear and consistent statement of its goal, more intensive staff training and regular monitoring to ensure that diversion interviews are implemented as intended.

    In comparison to Indiana, local staff in Oregon and Wisconsin spent more time with applicants on: (1) learning about their financial circumstances, (2) helping them identify alternative sources of support, and (3) encouraging them to use community-based resources and apply only for noncash assistance (e.g., Medicaid, Food Stamps, child care assistance). Wisconsin’s diversion interview included providing individuals with information on financial planning and management in conjunction with what they had learned about the applicants’ existing budget and expenses. Local staff in Oregon could authorize child care immediately for families who could avoid applying for welfare if they received child care assistance.

    Staff interviewed in Oregon and Wisconsin viewed the diversion interview as an effective practice that worked as intended, while local staff in Indiana were much more skeptical about its value and indicated that it had no discernible impact. Local staff in Oregon and Wisconsin also stressed that the diversion interview was particularly effective when combined with a careful explanation of the participation and job search/work requirements the individual would need to fulfill both before and after benefits were authorized. As Indiana did not have an up-front job search requirement and staff did not stress program requirements during the diversion interview, this may account for their less favorable view about the effectiveness of the up-front interview.


    The applicant job search requirement implemented in Oregon and Wisconsin reportedly served its intended purpose of moving some portion of applicants into jobs, with the end result of diverting some from going on welfare. The anecdotal evidence, particularly in Wisconsin, is that applicant job search is quite effective in diverting clients from going on welfare. It is not known what proportion of applicants find jobs versus fail to complete the application process because (1) they are already receiving unreported income or find alternative sources of support or (2) they simply find or perceive that the requirement is beyond their capacity to fulfill.

    It stands to reason that stringent up-front requirements, particularly those which are a condition for eligibility, may have a diversionary effect. The stricter the up-front requirement, the more likely individuals will respond by deciding either not to apply or to withdraw their application.4 Thus, while the magnitude of the impact of an up-front applicant job search requirement is unknown, its diversionary effect may result from reasons other than fulfilling its intended purpose of moving applicants into jobs.


    This overview illustrates that “diversion” can be approached from different angles which correspond to differing circumstances and levels of need among applicants. States can include any one or all of these diversion policies in the up-front portion of the eligibility process. Although the separate and cumulative impact of these policies is not known, our discussions with local staff suggest that combining a diversion interview with an up-front applicant job search may have a more powerful diversionary effect than just a diversion interview. For example, local staff in Wisconsin reported that they saw a drop in applications after the diversion interview was implemented but that the number of diversions skyrocketed after an up-front applicant job search (as a condition of eligibility) was implemented six months later.

    While diversion policies have been implemented during a period of declining applications for assistance, insufficient data prevents drawing conclusions regarding what, if any, relationship exists between the two. It may be that the declining applications for assistance are attributable to reasons that have nothing to do with the extent to which diversion policies directly or indirectly serve to divert families from welfare.

    For example, a strong economy diminishes the need for cash assistance and is likely to play an important role in the overall downward trend in applications. In addition, the implementation of stricter participation requirements and sanctions may deter those who do not really need cash and/or employment assistance from applying for benefits. On the other hand, anecdotal accounts from local staff in both Oregon and Wisconsin suggest that their up-front diversion policies do in fact have an impact on individuals’ decisions to submit an application and/or complete the application process.

    There is substantial need for further research on the impact of diversion policies on caseload decline and family well-being. Our examination of diversion also indicates the need to consider the following key questions:

    • To what extent are families being diverted because parents find that they can in fact get by with the help of other forms of publicly funded assistance (e.g., Medicaid, Food Stamps, child care)? To what extent do diverted families rely on alternative community-based resources? What proportion of diverted families apply for assistance at a later date?
    • To what extent are families in states with up-front applicant job search requirements being diverted because they are finding jobs as opposed to giving up on proceeding with their application because they are or at least perceive themselves to be incapable of fulfilling these requirements?
    • To what extent are families shying away from seeking assistance —never even exploring whether they might be eligible for assistance—because they feel intimidated and/or are misinformed about various program requirements and conditions of eligibility? Is this more likely to occur among some types of individuals than others? What safeguards can states incorporate into their diversion policies to ensure that needy families are not improperly diverted?

    Diversion policies can provide useful alternative approaches to reducing welfare dependency. At the same time, they also warrant some degree of caution. The challenge for states is to design and implement policies that serve the intended purpose of diverting families who do not really need to receive ongoing cash assistance but do not prevent truly needy families from applying for and completing steps required to obtain benefits.

    Chapter 6 Footnotes

    1. Among our study states, Wisconsin recently implemented a variation of financial diversion assistance under W-2 which provides “job access loans” to individuals experiencing a period of temporary financial crisis or in need of short-term assistance to enable them to obtain or continue employment. Unlike other financial diversion assistance programs, the job access loans must be paid back in cash or a combination of cash and volunteer in-kind community work.

    2. Utah’s financial diversion assistance component was also initially met with resistance by workers. However, its use has grown over time; up to about 20 percent of eligible TANF applicants receive diversion assistance. See U.S. General Accounting Office, Welfare Reform: Three States’ Approaches Show Promise of Increasing Work Participation GAO/HEH-97-80, May 1997.

    3. This section describes diversion interview practices in Wisconsin prior to the implementation of W-2. The up-front interview was modified to reflect W-2’s program design.

    4. For example, Wisconsin’s pre-W-2 up-front job search requirement was much stricter than Oregon’s in terms of total hours of job search required before benefits could be authorized. It is likely that this requirement contributed to Wisconsin’s dramatic drop in applications for cash assistance.




    Organizational Strategies to Support Employment Focused Welfare Reform For many states, a particularly critical and challenging dimension of moving to a more work-oriented welfare system is determining how best to reshape the structure and culture of welfare offices so that work replaces income maintenance as the primary mission. Wisconsin has made significant broad-based organizational reforms in this area. Although less comprehensive in scope, administrative and staffing changes also figure importantly in many states’ overall efforts to make welfare more employment focused. Many changes are designed to infuse more emphasis on employment into the eligibility process, thereby altering the “front-door” of the welfare office (see box).

    In this chapter, we highlight strategies of interest in two broad areas: service delivery and staffing. A description of the organizational and service delivery structure in Racine (WI) is provided to illustrate how Wisconsin has sought to redesign the way welfare-to-work services are delivered by integrating welfare into the larger workforce development system. Efforts by some of the study states to consolidate eligibility work with welfare-to-work case management functions under a single staff position are then described.


    Wisconsin provides the most dramatic example of how the traditional benefit and service delivery structure can be changed to transform welfare into a work-based system. At the state level, administrative responsibility for all aspects of cash assistance and Food Stamps was placed under a newly created Department of Workforce Development which is also responsible for all employment and training programs and workforce issues in general. At the local level, Job Centers are being expanded to serve as: (1) the physical and administrative locus for comprehensive employment and training services for all job seekers and (2) the intake point for those seeking cash and employment assistance. The following describes how this reorganization of the service delivery system was implemented in Racine (WI) as of January 1997. 



    Changing the Front Door of the Welfare Office

    Historically, families’ first interactions with the welfare office concerned eligibility and focused on what individuals needed to do to get and stay on welfare — not what they could do to avoid going on welfare or what could be done to help them get off welfare. In an effort to make welfare more oriented around issues of work and self-sufficiency rather than eligibility and money, local offices have made numerous changes to the front door of welfare—specifically, the intake and eligibility determination process. These include: 

    • informing and impressing upon applicants the importance of obtaining employment as quickly as possible and providing a clear explanation of welfare-to-work program rules and expectations early in the application process (this can include requiring clients to sign a “personal responsibility” agreement during the eligibility process as a condition of full receipt of benefits); 
    • changing the physical environment where individuals apply for benefits to promote the work focus; 
    • requiring applicants to engage in an up-front job search and/or register for work with the local employment service office; and, 
    • moving employability planning activities into the up-front eligibility process; combining responsibility for eligibility work and management of the clients’ welfare-to-work activities under the same worker and/or transferring some welfare-to-work activities (e.g., monitoring compliance with up-front job search requirements) over to eligibility workers. 


    The Basics: One Stop Service Delivery, Consolidated Funding, Integrated Staff Teams

    In order to co-locate and integrate services under one roof, it was necessary for Racine (WI) to construct a new Workforce Development Center. Employment and training providers (e.g., Job Training Partnership Act, (JTPA) Job Service, the local technical college) located throughout various parts of the city then co-located some or all of their staff within the new Workforce Development Center. In addition, this effort is different from other states which are moving to consolidate workforce programs under one roof because the TANF and Food Stamp programs are also co-located within the Center.

    One-Stop Service Delivery. The Workforce Development Center is currently the entry and exit point for all “job seekers”—those requiring only general information and resources on job openings as well as those in need of specialized services. It is the only place where individuals can apply for public assistance. In the context of the Workforce Development Center’s service delivery structure, public assistance recipients are just one type of job seeker and the benefits they receive constitute one of a variety of specialized forms of assistance available to help support individuals while they prepare for and seek employment.

    The first floor of the Workforce Development Center is open to the public at large and contains an employment resource area, a career development center, an academic improvement center, interviewing rooms and an area to hold workforce development workshops for employers and job seekers. Specialized employment and training programs (e.g., JTPA, Food Stamp Employment and Training, JOBS) and public assistance staff are located on another floor. Interactive touch-screen self-service kiosks are stationed in the lobby to inform individuals (with assistance from the receptionist, if needed) of the various services available to them through the Center. These kiosks are programmed to lead individuals interested in obtaining public assistance or child care assistance through an automated eligibility prescreening process and generate intake appointment dates.

    No work or work-preparation services offered through the Workforce Development Center are specifically targeted only to public assistance recipients. Instead, all activities and services provided to recipients can be accessed through the following resources located on-site:

    Career Development Center. Career counselors are available to help job seekers evaluate their interests, skills and aptitudes and to make career decisions based on their strengths and preferences. There is also a twelve session (9 hour/3 credit) career development course offered on-site by the local technical college.

    Resource Area. The general resource area includes a library containing a wide variety of materials on all aspects of seeking, obtaining and keeping a job; ten computer workstations to access JOBNET; and separate offices that are reserved by employers for job interviews. The JOBNET posts available jobs arranged by region and job category including entry level, part-time, technical, skilled trades, professional and managerial. A description is included of each job and its requirements, as well as how to apply for the position. Computers with resume writing software packages are also available to create resumes. On-site telephones and fax machines are also available. A punch-in/punch-out time clock is discreetly kept in one corner of the resource room which welfare recipients use to document the time using these resources.

    On-Site Child Care. The YWCA operates a small on-site child care center in the Workforce Development Center for anyone using the Center. The maximum amount of time a child can stay in the “child waiting center” is 3.5 hours at a time.

    Academic Improvement Center. An academic improvement center staffed with instructors is located on-site, adjacent to the general resource area. This center is available to any and all individuals interested in upgrading their academic skills in order to earn a GED or High School Equivalency Diploma, get a better job, or apply for advanced education. Many self-paced computer instructional programs are available and English as a Second Language (ESL) evening classes are available on a regular basis. On-site GED testing is offered once a month.

    Workshops for Job Seekers and Employers. The Workforce Development Center offers, on average, 10 workshops each week on a variety of subjects related to finding and keeping a job, as well as workshops specifically targeted to employers. All workshop participants receive an embossed certificate of completion. For welfare recipients, these certificates serve as documentation of time spent in a job search/job readiness activity.

    Consolidated Funding and Integrated Staff Teams. Racine’s Workforce Development Center moves beyond simply co-locating services in that it has a consolidated funding base and has reorganized most services and staff along functional lines that cross traditional program lines.

    The consolidated funding base is made up of various funding streams including: JTPA, veterans and dislocated worker programs, Wagner-Peyser, TANF, the Child Care Development Fund, adult basic education, and local property taxes. Consolidating funding from these various sources has reportedly played an enormously important role in facilitating the Workforce Development Center’s ability to achieve staffing and service integration.

    Staffing and service integration has allowed the Center to design a comprehensive menu of services for both job seekers and employers that ranges in intensity from self-service to very specialized services. Integration has also alleviated many of the coordination problems and turf issues that hinder effective and efficient service delivery.

    For example, job development activities are consolidated under a single employer marketing team made up of staff from various programs and organizations. By centralizing job development under one integrated team, referrals to employers can be streamlined—employers are not “hit up on from all sides” and staff from various programs no longer “bump over each other.” An integrated Job Readiness Team is responsible for administering and operating all activities designed to prepare and assist job seekers, including services offered through the on-site Career Development Center and the on-site Academic Improvement Center. A single child care team handles child care assistance eligibility determination and processing for all programs and eligible individuals. An integrated case management team is cross-trained to work with JTPA, Food Stamp Employment and Training, Veteran’s and Dislocated Worker program participants.

    INTEGRATED CASE MANAGEMENT: NEW ROLES FOR INCOME MAINTENANCE AND WELFARE-TO-WORK STAFF The effort to infuse a greater focus on employment and self- sufficiency into the eligibility process has led some states to broaden the role of income maintenance workers. Traditionally, income maintenance staff have focused exclusively on the financial aspects of an individual’s case—eligibility determination, issuance of benefits, redeterminations and general ongoing case maintenance.1 Their interaction with the welfare-to-work program was minimal, consisting primarily of determining whether a recipient should be referred to the welfare-to-work program, adjusting benefits levels in response to sanction enforcement requests and, in some offices, processing child care payment authorizations.

    Eligibility workers in all the study states now emphasize the importance of getting a job and moving off welfare in their interactions with clients and convey the message that welfare should be viewed as a temporary source of assistance. As discussed in the preceding chapter on diversion practices, eligibility workers in some of the study states also spend time with applicants exploring alternative sources of support and monitoring up-front job search.

    Beyond focusing more attention on these types of issues, some states have formally combined eligibility and some welfare-to-work functions (typically “case management” which includes developing employability plans, monitoring participation and enforcing sanctions) into a single staff position. This approach has both advantages and disadvantages.

    The primary benefits associated with integrating responsibility for eligibility and self-sufficiency related activities is that it ensures that the focus on employment is integrated into the eligibility process, allows recipients to report to and interact with a single person who understands all aspects of the case, and eliminates the need to coordinate welfare-to-work program rules that affect benefits across different types of workers (e.g., enforcing sanctions).

    However, this expansion of responsibilities creates a challenge for workers in large part because welfare-to-work case management and eligibility-related work require different skills. Many staff find it difficult to excel in both types of work as well as meet the differing workload demands their consolidated position entails.2 Therefore, an integrated case management model can also contribute to a loss of efficiency and expertise associated with specialization.

    Oregon was the only study state that had several years of experience in using an integrated case management model. Since making this initial change, local offices have continued to experiment with different staffing patterns within this broader model so as to use staff more efficiently and better meet the differing needs of clients. For example, all case managers in Portland (OR) handle both financial and welfare-to-work aspects of a case but their case management responsibilities are further specialized by the specific type of welfare-to-work services needed (see box for a more detailed description).

    As of January 1997, Massachusetts and Wisconsin had also expanded eligibility workers’ responsibilities and were planning to move to a fully integrated case management model.3 In Racine (WI), eligibility workers were responsible for conducting the up-front diversion interview and monitoring the up-front job search requirement in addition to eligibility determination. Welfare-to-work staff were responsible for imposing sanctions directly on recipients who failed to meet participation requirements without having to coordinate this action with the eligibility workers.4

    Eligibility workers in Worcester (MA) had assumed responsibility for explaining the new Work First, Work Mandate program approach to applicants and contacting or meeting with recipients at specific points in time over the initial 60 day period to track their status and encourage them to find a job or, as the work requirement deadline neared, to find a community work experience position. Eligibility staff were also preparing to assume responsibility for authorizing transportation and child care assistance, a change that was greeted with reservations on the part of eligibility staff given the demands of their existing caseload. An advantage of shifting responsibility for these activities from welfare-to-work staff to eligibility staff was that it would free up the former to spend more time recruiting employers to participate in the subsidized employment and unpaid work experience programs.

    While these two examples of organizational and administrative reforms highlight ways in which states are fundamentally restructuring how eligibility and employment services are delivered to families, they by no means capture the range of strategies being explored and implemented by states to transform the message, culture and organization of welfare. 



    Portland, Oregon: 

    Specialization within an Integrated Case Management Model

    Oregon has integrated eligibility and welfare-to-work case management functions into a single staff position. At the same time, different staff are assigned to work with individuals at different points in their path to employment. Local staff noted that this tailored case management staffing model worked particularly well because it allowed them to better balance the needs of recipients entering an up-front job search with the needs of longer-term clients for whom a relatively short-term and fast-paced job search was not sufficient. 

    • Specialized Up-Front Self-Sufficiency/Intake Workers. One set of workers screens all incoming applicants and conducts self-sufficiency planning interviews to assess their needs, explores possible alternative sources of support that might enable applicants to avoid going on welfare, and initiates the eligibility determination process. 
    • Specialized Job Search /Case Management Workers. Some staff are responsible for completing the eligibility determination process and continuing to work with these recipients on a short-term basis while they are participating in an up-front job search. These staff focus on monitoring compliance and ensuring recipients are provided short-term supportive services needed to find employment. Job search assistance services (e.g., job clubs, etc.) are provided by contracted service providers. 
    • Specialized Harder to Serve/Case Management Workers. Some staff are dedicated to working with recipients who still have not found employment after the initial job search period. Their work entails handling the financial aspects of the case, working more intensively with recipients to identify barriers to employment and developing alternative activity plans to assist them in becoming more employable. 
    • Other Specialized Staff. Some case managers are dedicated to work solely with subsidized employment participants and others are dedicated solely to handling transitional benefits for recipients who have found work and are no longer receiving cash assistance. 


    For example, some states are devolving more responsibility down to the local offices for the design and operation of programs and focusing more on employment-related outcome measures. Among our study states, Indiana transferred responsibility for contracting with service providers to the local level, thereby allowing local office administrators to decide, within parameters set by the state, which providers in the community could best serve their needs. To shift the message and focus to the primary goal of employment, Indiana also implemented performance based contracting with heavy emphasis on employment outcomes and established monthly job placement goals for each county office. To keep staff focused on increasing participation and helping clients find employment, Oregon relies on performance measures and outcomes first implemented in 1991. Local offices are given the flexibility to design specific program strategies and approaches they believe are needed to best meet performance goals.

    A common theme running through discussions with program administrators and staff was that shifting to a more work focused, transitional assistance system necessitated a change in the mindset, expectations, and tasks set forth for staff and welfare recipients alike. For Wisconsin, this has meant embarking on a path that entirely replaces the traditional staffing and organizational structure of welfare. Other states are attempting to redesign the eligibility and service delivery structure within the existing organizational framework.

    Chapter 7 Footnotes

    1. These activities are sometimes handled by the same worker and sometimes divided across workers so that one set of income maintenance staff handles intake and eligibility determination and another set of staff handles ongoing case maintenance.

    2. Dan Bloom and David Butler, Implementing Time Limited Welfare: Early Experiences in Three States, New York: Manpower Demonstration Research Corporation (1995).

    3. As of January 1997, the other two states featured in this study—Indiana and Virginia—had retained the traditional staffing pattern of dividing eligibility and welfare-to-work case management activities between different staff. In response to more recent statewide changes in staffing classifications and allocations, it is anticipated that at least some portion of Indiana’s eligibility and welfare-to-work staff will assume integrated case management responsibilities which encompass both eligibility and welfare-to-work case management functions.

    4. Staff responsibilities have been further restructured under W-2 to achieve a more integrated case management staffing model. A single “Financial Employment Planner” position was created that is responsible for case management, eligibility and employability determination, development of a client’s employability plan, and all other services available to a participant in a W-2 position (e.g., job access loans, determining eligibility for child care, Food Stamps, Medicaid). Other staff are assigned to work solely with W-2 clients in need of more intensive employment assistance.




    Recipient Experiences In Work First ProgramsThe policies described in this report set the framework in each of the five states for the way in which Work First programs are implemented at the local level. Because the implementation of complex programs depends on so many factors, even the best programs often fall short of the ideal. Based on discussions with state and local staff, it appears that the local offices examined for this study had by and large successfully made the shift to a Work First approach.

    To understand better how closely the experiences of recipients reflect stated policies and practices, we examined administrative data pertaining to a cohort of non-exempt applicants and recipients in four of our local sites over a 12-month period.1 Given the program designs and goals of the five study states, the primary objectives guiding this analysis were to determine:

    • if recipients were in fact finding jobs and, if so, how much these jobs paid;
    • if recipients who were required to participate in the states’ Work First program were actually doing so;
    • if the central activity was in fact job search and to what extent recipients engaged in other activities, particularly unpaid work experience;
    • if recipients moved quickly into program activities after the initial authorization of benefits;
    • what happened to recipients over the course of 12 months with respect to participation in activities, employment and welfare receipt; and,
    • what proportion of recipients were on welfare one year later.

    While our analysis sheds light on each of these key issues, data limitations prevent us from being able to fully address each for every state, and comparisons between states should be treated with caution. Thus, this analysis is best viewed as a means to further supplement and enrich findings obtained through other sources of data (e.g., discussions with staff) presented in the preceding chapters. The remainder of this chapter summarizes the key findings of our analysis of the administrative data.

    How Many Recipients Found Jobs and What Did They Earn?

    As shown in Table 8.1, the fraction of recipients who were reported as having obtained unsubsidized employment at some point during the 12 month period ranged from 36 percent in Worcester (MA) to 66 percent in Culpeper (VA). Rates of unsubsidized employment in Indianapolis (IN) and Portland (OR) fell more in the middle, with 45 percent of recipients with a record of employment in Portland (OR) and 55 percent in Indianapolis (IN). The average starting hourly wages ranged from a low of $5.37 per hour in Culpeper (VA) to a high of $6.52 in Worcester (MA)


    Table 8.1

    Employment and Wages of Mandatory Recipients


      Indianapolis   Indiana   (n=402) Worcester   Massachusetts   (n=134) Portland   Oregon   (n=529) Culpeper   Virginia   (n=134)
    Record of  Employment 55% 36% 45% 66%
    Average Wage $5.90  $6.52  $6.37  $5.37 


    In general, the employment rates reported here are likely to be lower than actual rates of employment achieved because administrative data maintained by welfare offices do not necessarily include information on every client who finds a job.2

    What Share of Recipients Who Were Subject to Mandatory Participation Requirements Engaged in Program Activities?

    Participation rates must be interpreted carefully because participation can be defined and measured in many ways.3Table 8.2 shows participation rates over the course of 12 months using different definitions of participation. The term “recipient” always refers only to those who did not meet up-front exemption criteria and were subject to participation requirements (i.e., mandatory recipients).4

    If participation is broadly defined to include an initial assessment/orientation, in addition to standard program activities—job search, unpaid work experience/subsidized employment, education, and training—the local sites achieved high rates of participation. Over 90 percent of the mandatory recipients in Indianapolis (IN), Portland (OR) and Culpeper (VA) and almost 75 percent in Worcester (MA) were engaged in at least one of these activities. 


    Table 8.2

    Participation in Welfare-to-Work Program Activities

    by Mandatory Recipientsa


      Indianapolis   Indiana  (n=402) Worcester   Massachusetts  (n=134) Portland   Oregonb  (n=529) Culpeper   Virginia  (n=134)
    Any program activity plus  orientation/assessment 93% 72% 100% 93%
    Any program activity  plus employment 67 66 81 93
    Any program activity 32 64 75 89




    a Program activities are job search, unpaid work experience/subsidized employment, education, and training.  b Orientation/assessment is integrated into the eligibility process. Therefore, the Oregon sample reflects 100% participation in orientation/assessment.


    If participation is defined to exclude assessment/orientation but include unsubsidized employment, participation rates are somewhat lower but still quite high. Two-thirds of mandatory recipients in Worcester (MA) and Indianapolis (IN) were engaged in a program activity (beyond assessment) or were in unsubsidized employment at some point over the 12 month period and participation rates using this definition were even higher in Portland (OR) and Culpeper (VA)—81 percent and 93 percent, respectively.

    If, however, participation is narrowly defined as being engaged in a standard program activity specifically designed to lead to employment (i.e., excludes both assessment/orientation and unsubsidized employment), then rates of participation vary considerably and more sharply reflect differences in program design across states. Culpeper (VA), which requires recipients to participate in job search and then (if necessary) community work experience, achieved the highest level of participation using this narrow definition—89 percent of mandatory recipients were engaged in an activity. The vast majority of mandatory recipients in Portland (OR) and Worcester (MA) also participated using this definition—75 percent and 64 percent respectively. In contrast, Indianapolis (IN) had a far lower participation rate using this narrow definition, with only 32 percent of mandatory recipients participating during the 12-month period.

    Indianapolis (IN), presents an interesting case in that relatively few mandatory recipients move beyond assessment to other program activities. Yet, mandatory recipients in Indianapolis (IN) participate in assessment or unsubsidized employment at rates that equal or exceed the other local sites. As discussed below, a far lower percentage of mandatory recipients participate in a job search component in Indianapolis compared to the other sites, which explains much of the difference in the participation rates using the narrowest definition. It is possible that the relatively high rate of employment despite low participation in program activities in Indiana reflects the power of a strong economy. Although this question is beyond the scope of this analysis, it seems clear that other factors beyond participating in program activities contribute to welfare recipients’ success in finding jobs.

    How Many Mandatory Recipients Participated in Job Search or Other Activities?

    This section provides a more detailed breakdown of mandatory recipients’ participation in four types of program activities—job search, “work activities” (i.e., unpaid work experience/subsidized employment, but not unsubsidized employment), education and training. Table 8.3 presents these data in two different ways. We present the data these two ways because, as shown in Table 8.2, the degree to which mandatory recipients actually engaged in activities varies so substantially across the four states. The first panel of the table shows the breakdown of participation by activity for only active mandatory recipients—that is, those recipients who reportedly engaged in one of the four activity categories. (This is the group participating according to the narrowest definition described above.) To provide a broader picture of overall participation, the bottom panel of the table shows the fraction of all mandatory recipients who participated in any of the four activity categories.

    Participation in Job Search. The states featured in this study all place high priority on job search and this is reflected in the recipients’ activity patterns—the most common activity among mandatory recipients was job search. Among active mandatory recipients, participation in job search was nearly universal (95 percent or higher) in Portland (OR) and Culpeper (VA) and very high in Indianapolis (IN) and Worcester (MA)—75 percent and 78 percent, respectively.

    When all mandatory recipients are taken into consideration, however, the proportion of those engaged in job search looks very different. Of all mandatory recipients in the samples, the fraction that participated in job search was: 87 percent in Culpeper (VA), 71 percent in Portland (OR), 50 percent in Worcester (MA) and 24 percent in Indianapolis (IN).

    The comparatively low level of participation in job search in Indianapolis (IN) again reflects the fact that approximately two-thirds of all recipients never entered a program activity after completing the initial assessment, but it is likely that many conducted an independent job search and started working. Similarly, in Worcester (MA), recipients may independently search for work but are not formally required to participate in a job search component or report on their job search activity before the work requirement goes into effect. Therefore, job search activity rates for Worcester (MA) and Indianapolis (IN) underrepresent the fraction of recipients who actively seek employment during the first quarter of the 12-month period. 


    Table 8.3

    Participation in Specific Welfare-to-Work Program Activities


    Active Mandatory Recipients

      Indianapolis  Indiana  (n=127) Worcester   Massachusetts  (n=86) Portland   Oregon  (n=380) Culpeper   Virginia  (n=119)
    Job Search 75% 78% 95% 98%
    Training 12 6 11 1
    Education 25 2 16 15
    Work Activitiesa 3 44 11 21

    All Mandatory Recipients

      Indianapolis   Indiana  (n=402) Worcester   Massachusetts  (n=134) Portland   Oregon  (n=529) Culpeper   Virginia  (n=134)
    Job Search 24% 50% 71% 87%
    Training 4 4 8 1
    Education 8 1 12 13
    Work Activitiesa 1 28 8 19




    a Again, it is important to recall that data on “work activities” reported here do not include employment in unsubsidized jobs.


    Participation in Work Activities. As would be expected, local sites in the Work First, Work Mandate states—Virginia and Massachusetts—had a higher share of mandatory recipients engaged in “work activities” (i.e., subsidized employment and unpaid work experience, but not unsubsidized employment) than local sites in the Work First, Participation Mandate states.

    Despite their similar program approaches, more mandatory recipients participated in “work activities” in Worcester (MA) than in Culpeper (VA). Of all mandatory recipients, a little over one-quarter (28 percent) participated in a work activity at some point during the 12 month period in Worcester (MA), compared to 19 percent in Culpeper (VA). Among active mandatory recipients, the difference is even greater—44 percent of mandatory recipients in Worcester (MA) had engaged in a work activity as compared to 21 percent in Culpeper (VA).

    Worcester (MA) may have higher rates of participation in work activities than Culpeper (VA) because Virginia requires mandatory job search prior to the imposition of their work requirement, while Massachusetts does not. Virginia’s job search requirement, coupled with a severe penalty for noncompliance, may motivate more recipients to actively seek and find jobs. In addition, a lower proportion of individuals may need to be assigned to a work activity in Virginia because some portion will have already lost benefits as a result of noncompliance.5

    Participation in Education and Training. The study states’ Work First programs sought to limit the use of education and training, often by permitting mandatory recipients to pursue education and training only if it was coupled with more work-oriented activities (i.e., unpaid work experience, job search, subsidized or unsubsidized employment). Participation patterns reflect each state’s objectives to curtail participation in these activities. Relative to job search and work activities, participation in education and training was uncommon in each of the four sites.

    Given its more limited role in the context of overall patterns of participation, there was still a great deal of variation across states in the extent to which education and/or training was used. Among mandatory recipients who engaged in any activity (i.e., active recipients), participation in education was the lowest in Worcester (MA) and the highest in Indianapolis (IN). Just two percent of the active mandatory recipients in Worcester (MA) engaged in education. In contrast, one-quarter of active mandatory recipients in Indianapolis (IN) participated in education at some point over the 12-month period. Higher rates of participation in education and training in Indiana reflect the state’s two-track program design, in effect over the sample period, provided more opportunity for certain clients to engage in education or training than is the case in the other states.6

    Indianapolis (IN), along with Portland (OR), also had the highest participation in training among active mandatory recipients—12 percent and 11 percent recipients. Almost no active mandatory recipients in Culpeper (VA) participated in training (1 percent), but the rate of participation in education was much higher (13 percent).

    How Quickly Did Recipients Move into Program Activities?

    One goal of Work First is to move recipients into the labor market as quickly as possible. To do this there has been increased emphasis on moving recipients into job search or, if deemed appropriate, other program activities, even prior to the point benefits are authorized or shortly thereafter. As shown in Table 8.4, local sites in three of the four states included in the analysis—Culpeper (VA), Portland (OR) and Worcester (MA)—have been generally successful in implementing this objective. Between 86 and 88 percent of active mandatory recipients who ever participated in a program activity did so within the first month after benefits were authorized. 


    Table 8.4

    Time from Program Entry to First Activity

    Percent of Mandatory Recipients Who Entered an Activitya


      Indianapolis  Indiana  (n=127) Worcester   Massachusetts  (n=86) Portland   Oregon  (n=380)b Culpeper   Virginia  (n=119)
    1 month 36% 86% 87% 88%
    2 month 57 91 92 92
    3 month 66 91 93 94
    4 month 76 94 95 95
    5 month 81 95 96 95




    a Table reflects cumulative percentage of those with a program activity beyond assessment and orientation.  b Activity begin dates were only available for 380 of the 397 recipients who participated in a program activity beyond assessment/orientation.


    Indianapolis (IN) had a much slower “take-up” rate because their program design included an up-front mandatory client assessment that was scheduled and conducted on a group basis and had to be completed before assignment to a program activity (e.g., job search, education or training). Only 36 percent of mandatory recipients who participated in a program activity in Indianapolis (IN) were engaged in an activity within one month of benefit authorization. However, while entry into program activities began slowly in Indianapolis (IN), it reached 81 percent by the fifth month, not far behind the other states.

    What Happened to Recipients Over the Course of 12 Months?

    These programs aim not only to engage a large fraction of recipients in program activities but also to sustain participation and keep recipients from falling through the cracks. Although we were able to compile the longitudinal data necessary to determine how well this goal was achieved for just one of the local sites—Worcester (MA)—it affords a dynamic picture of the intersection between program design, implementation and client experiences.

    Our major finding is that participation patterns over time did, in fact, closely mirror the client flow and participation patterns intended to occur under Massachusetts’ Work First, Work Mandate program design. Additionally, although Massachusetts’ program design calls for participation in unpaid work experience if mandatory recipients cannot find unsubsidized employment, our analysis indicates that it was not necessary to provide many community work experience slots in any given month, both because most recipients were either combining unsubsidized work with welfare or had left the welfare rolls. Also, it should be recalled that Massachusetts exempts a very high proportion of the total caseload from work requirements. The following provides a more detailed rolling account of what happened to recipients over a 12-month period (see Table 8.5).

    The participation patterns in Worcester (MA) changed quite rapidly in the first few months, but more gradually in later months. By the second month after receiving benefits, 37 percent of mandatory recipients had already left the welfare rolls and an additional 27 percent were either working in unsubsidized employment (17 percent) or participating in a work activity (10 percent). Most of the remaining mandatory recipients were participating in job search (29 percent). Only four percent of mandatory recipients were not engaged in any activity, and it is likely that the majority of these persons were waiting for an unpaid work experience slot to be assigned to them.

    In the third month, when all mandatory recipients should either be in a work activity, unsubsidized employment, or off assistance, 28 percent were not engaged in one of these required activities. However, by the fourth month that fraction dropped considerably to just 16 percent of recipients. By the sixth month it dropped even further to just 5 percent, and remained near this level for the remainder of the year. That is, six months into the program roughly 95 percent of all mandatory recipients were off assistance or participating in a required work activity. The fraction of mandatory recipients no longer receiving assistance increased steadily until it leveled off at about 75 percent in the seventh month.

    Although 28 percent of mandatory recipients participated in work activities (see Table 8.3) over this time period, the fraction in any given month never exceeded 14 percent. During the first six months, the fraction of mandatory recipients working in unsubsidized employment exceeded the fraction participating in work activities, but this pattern was reversed in the next six months. Roughly the same fraction of mandatory recipients participated in unpaid work experience over the 12-month period, but the fraction of recipients who combined unsubsidized employment with welfare declined. That is, over time, a lower percentage of mandatory recipients received cash assistance and a lower percentage combined welfare with work. This suggests that recipients in Worcester (MA) are working their way off welfare and are no longer eligible for welfare; or, at least, that they are no longer on the caseload.

    How Many Recipients Were on Welfare One Year Later?

    While the primary goal of Work First is to move recipients into the paid labor force as quickly as possible, this approach does not focus on moving recipients off welfare permanently. As reflected in the employment rates described earlier and in the preceding analysis of recipients’ experiences in Worcester (MA), the local sites examined here have achieved considerable success in moving mandatory recipients into jobs. That is, in many important respects, local offices have been successful in implementing the goals of a Work First program approach.

    At the same time, a significant fraction of the mandatory recipients were receiving assistance at the end of a year—31 percent in Worcester (MA), 42 percent in Portland (OR) and 44 percent in Indianapolis (IN). These percentages reflect recipients who received assistance for all 12 months as well as those who left welfare but later returned. Recipients in either of these two groups may have combined work with welfare during the one year period.

    Nevertheless, the information in this chapter suggests that relatively high proportions of mandatory recipients (over 60 percent in three of the sites) actively participated in a welfare-to-work activity (see Table 8.2). High proportions also were employed at some point during the 12-month period (45 percent or more in three of the sites). However, based on the Worcester (MA) analysis, a surprisingly non-trivial proportion may still be on welfare one year after first receiving benefits, despite high rates of employment participation in “work activities.” This serves as an important reminder that the challenge of “replacing welfare with work” is an ongoing process which, even in a Work First environment, requires a wide variety and mix of policy and program responses.

    Chapter 8 Footnotes

    1. Administrative data from Wisconsin were not available for the analysis. The sample data for Indianapolis (IN) include eleven recipients from Scott County, the other local site in Indiana featured in this study. See Appendix B for further discussion of the samples and data limitations.

    2. The relatively low rate of employment in Worcester (MA) is particularly likely to be an undercount because employment outcomes are not recorded for recipients who find employment before they become subject to the work requirement unless it is combined with welfare receipt.

    3. For a fuller discussion of measuring participation, see Gayle Hamilton, The JOBS Evaluation: Monthly Participation Rates in Three Sites and Factors Affecting Participation Levels in Welfare-to-Work Programs, Washington, D.C.: U.S. Department of Health and Human Services and U.S. Department of Education (1995).

    4. It is important to recall that the proportion of total recipients who are mandatory varies across states (e.g., over 90 percent are mandatory in Oregon and Wisconsin, while only about 20 percent are mandatory in Massachusetts). Discussions of participation rates, thus, do not refer to the same portion of the total caseload in all states.

    5. Differences in the characteristics of recipients in the two samples (see Appendix B) may also account for the larger share of recipients in unpaid work experience in Worcester (MA) compared to Culpeper (VA). Sample recipients in Worcester (MA) faced more barriers to employment than their counterparts in Culpeper (VA) in that they were older and more likely to have been on assistance for longer periods of time.

    6. Indiana’s program design permitted mandatory recipients who were determined not to be job ready to fulfill the majority of the 20 hours per week participation requirement in education or training.


    Table 8.5

    Percent Distribution of Month-by-Month Activities of Massachusetts' Mandatory Recipientsa


      Month   1 Month   2 Month   3 Month   4 Month   5 Month   6 Month   7 Month   8 Month   9 Month   10 Month   11 Month   12
    Off Assistance 0% 37% 49% 57% 63% 71% 76% 75% 74% 76% 74% 69%
    On Assistance 100 63 51 43 37 29 24 25 26 24 26 31
    Unsubsidized  Employed 17 17 14 14 15 13 7 7 7 6 5 10
    Work  Activities 8 10 10 13 12 11 10 12 13 13 14 14
    Job  Search 38 29 22 10 5 2 2 2 2 2 1 2
    Education or  Training 2 1 0 0 0 0 0 0 0 0 0 0
    Enrollment/  Assessment 10 1 1 0 0 1 1 1 0 0 1 2
    Exempt 10 1 1 1 1 1 1 1 2 2 3 0
    Sanction 2 0 0 0 0 0 0 0 0 0 2 2
    No  Activity 13 4 4 5 4 1 3 2 2 1 0 0




    a These percentages represent a sample of 134 mandatory recipients in each month.





    AFDC Caseload Trends, Economic Indicators and Socio-Economic Characteristics in the Five Study States


    Table A.1

    AFDC Caseload Trends and Economic Indicators

    for the Total U.S. and the Study States


      U.S. Totals Indiana Massachusetts Oregon Virginia Wisconsin
    AFDC Average Monthly Caseloada 







    4,497,186  4,829,094  5,011,827  5,036,244  4,799,917  4,442,647 63,598  70,052  74,198  72,598  61,738  51,086 107,575  111,908  114,594  109,961  97,155  85,930 39,281  41,531  42,795  41,602  38,227  31,176 64,851  71,763  74,042  74,760  70,447  62,697 81,073  81,272  79,585  76,217  70,604  55,501
    AFDC Average Monthly Caseload  Percent Change 






    10.9%  7.4  3.8  0.5  -4.7 15.2%  10.1  5.9  -2.2  -15.0 10.9%  4.0  2.4  -4.0  -11.6 17.3%  5.7  3.0  -2.8  -8.1 14.0%  10.7  3.2  1.0  -5.8 2.3%  0.2  -2.1  -4.2  -7.4
    State Unemployment Ratesb 






    7.5  6.9  6.1  5.6  5.4 6.6%  5.4  4.9  4.7  4.1 8.6%  6.9  6.0  5.4  4.3 7.6%  7.3  5.4  4.8  5.9 6.4%  5.1  4.9  4.5  4.4 5.2%  4.7  4.7  3.7  3.5
    State Unemployment Rates  Change 1992-96 -27.0% -37.9% -50.0% -22.4% -31.3% -32.7%
    State Job Growth Rateb  1992-96 9.1% 9.2% 7.9% 14.0% 9.0% 9.0%




    a These percentages represent a sample of 134 mandatory recipients in each month.


    Table A.2

    Selected Demographic Characteristics of AFDC Families

    for the Total U.S. and the Study States


      U.S. Total Indiana Massachusetts Oregon Virginia Wisconsin
    Percent of Population  AFDC Recipientsa 4.3 2.1 3.5 2.1 2.1 2.7
    Percent of children  AFDC Recipients 11.5 5.9 9.8 5.8 6 7.4
    Maximum AFDC Benefit:b  Three Person Family (1996) $389  $288  $565  $460  $354  $517 
    RACE/Ethnicity (1994):b 





    36.4  37.4  19.9  4.2 35.1  60.9  3.2  0.3 17.0  47.9  25.0  4.9 7.4  81.0  7.1  4.3 65.1  30.7  2.7  1.5 38.2  46.4  6.7  8.2
    Age of Youngest Child:b 

    Under 2 Years (1994)

    3-5 Years

    6+ Years

    40.5  21.6  36.0 36.5  23.3  35.5 38.4  24.9  35.6 40.8  21.6  36.6 37.9  23.6  37.8 45.8  20.6  32.4




    a Department of Health and Human Services, Administration for Children and Families; Unpublished data.  b U.S. Congress, House of Representatives, Committee on Ways and Means (1996).


    Table A.3

    Selected Characteristics for the Total U.S., Study States, and Local Study Sites


      U.S. Total Indiana Massachusetts Oregon Virginia Wisconsin
    Population, 1995a        260,034,000    State: 5,803,471 

    County:  Marion 817,604  Scott 22,528

    State: 6,073,550   

    County:  Worcester 718,858

    State: 3,141,585 

    County:  Multnomah 614,104

    State: 6,618,358 

    County:  Culpeper 30,528

    State: 5,122,871   

    County:  Racine 182,982

    Unemployment Rate, 1996b    5.40%    State: 4.1%  County:  Marion 3.6%  Scott 5.1% State: 4.3% 

    County:  Worcester 4.0%

    State: 5.9% 

    County:  Multnomah 5.2%

    State: 4.4% 

    County:  Culpeper 4.8%

    State: 3.5% 

    County:  Racine 4.2%

    Per Capita Income, 1995a      22,788    State: 19,213 

     County:  Marion 22,307  Scott 15,077

    State: 24,410   

    County:  Worcester 20,738

    State: 19,437 

    County:  Multnomah 22,629

    State: 21,653   

    County:  Culpeper 18,627

    State: 19,806   

    County:  Racine 21,000

    Percent of Persons with  income below poverty  line, 1989a    13.80%    State: 10.7% 

    County:  Marion 12.1%  Scott 19.0%

    State: 8.9%   

    County:  Worcester 8.3%

    State: 12.4%   

    County:  Multnomah 13.1%

    State: 10.2%   

    County:  Culpeper 8.7%

    State: 10.7%   

    County:  Racine 10.2%

    Percent of Persons  under age 18 with income  below poverty line, 1989c 26%    State: 13.9% 

    County:  Marion 17.8%  Scott 26.7%

    State: 12.9%   

    County:  Worcester 11.6%

    State: 15.2%   

    County:  Multnomah 16.2%

    State: 13.0%   

    County:  Culpeper 8.7%

    State: 14.6%   

    County:  Racine 15.4%





    Description of State Administrative DataThe analysis of welfare recipients presented in Chapter 8 is based on state administrative data. To be included in our sample, a welfare recipient had to enter the state’s welfare-to-work program during the course of a one to two month period. Descriptive characteristics of the sample such as age, race and educational attainment are shown in Table B.1. The analysis follows each sites’ selected sample for one year, tracing their welfare-to-work activities and employment that was reported to the welfare department. The limitations of the state-specific data sets are discussed below, as well as where these limitations have the potential to bias the results.


    The sample consists of 402 adult recipients in Marion County (Indianapolis) and Scott County who became eligible for cash assistance in April 1996 and entered the welfare-to-work program by the end of May 1996. Of the 402 participants, only 11 came from Scott County. The recipients were followed from May 1996 through April 1997.

    Benefit information was available for each month; however, these data provide information on individuals’ eligibility for benefit versus actual receipt of benefit. As some recipients find employment (or other sources of income) they remain eligible for benefits, but the grant is calculated to be $0. This feature of Indiana’s welfare reform initiative applied only to recipients who were classified as job ready. Since we were unable to distinguish between individuals receiving positive versus zero grants, this analysis overstates the number of recipients who were reported as receiving assistance at years end.

    Of the 402 participants, 222 (55 percent) had a record of employment. Wage records only exist for 159 of the 222 employed (72 percent), so the reported average wage is based on these 159 individuals only. The activity records in Indiana included begin and end dates; therefore,the timing from program entry to activity is as accurate as its reporting. Program activities were combined and grouped under the four major components as follows:

    • Job Search—job development and/or job placement service, job readiness, group or individual job search, job search training.
    • Training—self-initiated training, job skills/vocational training.
    • Education—not self-initiated post-secondary education, self-initiated post-secondary education, academic training.
    • Work Activities—alternative work experience, community work experience, on-the-job training, work experience training.


    The sample consists of recipients who became subject to the work requirement in May or June 1996 and could be observed for at least 11 months. Our final sample includes 134 adult recipients in Worcester, our local site. Benefit information was available for each month the person was on assistance, which allowed us to calculate the total time on assistance and assistance by month. Employment begin dates were complete allowing for accurate reporting of the timing from program entry to employment. Employment end dates, however, were incomplete and were replaced with the date the participant went off of assistance. Employment duration, therefore, may be overstated. Similarly, program activity begin dates were complete, but end dates were often missing. The date the participant left assistance replaced the missing end dates. Thus, timing from program entry to activity is accurate, while activity duration may be overstated. Program activities were combined into four major categories:

    • Job Search—career center activity, employment assistance, DET job opportunity activity, assisted placement, job readiness, food stamp employment training.
    • Training—rehabilitation, JTPA, skills training/vocational training.
    • Education—ABE-adult basic education, English as a second language, community college participation, other college participation, full-time high school, GED.
    • Work Activities—supported work, household foster care, CWEP, subsidized employment.


    The sample consists of adult recipients of cash assistance who became subject to the Work First requirements in December 1995 and January 1996. The sample consists of 358 recipients from the East Portland Branch and 171 recipients from the Albina Branch in Portland, for a total of 529 people in the local site. This group was followed through December 1996.

    Benefit information was available at two points in time—at the beginning of the period and at the end of the period. The presence of a wage was used to determine whether or not an individual was employed. Activity begin dates were not available for all recipients who participated in a welfare-to-work activity; roughly five percent of recipients were missing this data. Therefore, timing from program entry to the first activity is not completely accurate. Program activities were combined into five major components:

    • Job Search—initial job search, recipient job search, job search (individual and group), supplemental job search, and life-skills training.
    • Training—self-initiated training, job skills training.
    • Education—ABE or acquiring high school equivalent, English as a second language, attending high school.
    • Work Activities—community work experience, work experience, on-the-job training, supported work, work supplementation, and AFDC-UP working parents.
    • Interventions—case management, drug and alcohol counseling, intervention, mental health treatment, other counseling, parenting skills.


    In Virginia, the sample consists of 134 adult recipients from Culpeper County. Culpeper implemented Work First in July of 1995, so our sample follows recipients from July 1995 through June 1996. Benefit information was available for each month the person was on assistance. Employment and activity records were complete in Virginia. Program activities were combined into four major categories:

    • Job Search—job club, individual or group job search, job readiness.
    • Training—self-initiated training, job skills training.
    • Education—English as a second language, adult basic education, GED, high school, post secondary associate, certificate or four year degree, self-initiated post-secondary education.
    • Work Activities—work experience, on-the-job training.


    Table B.1

    Characteristics of Mandatory Recipients in Sample


      Indianapolis  Indiana  (n=402) Worcester   Massachusetts  (n=134) Portland   Oregon   (n=134) Culpeper   Virginia  (n=134)





    65%  1  34  1 12%  31  57  1 18%  7  70  5 54%  0  46  0
    Gender of Head 



    93%  7 87%  13 83%  17 98%  2
    Average Age 31 34 29 29
    Marital Status 


    Never Married 




    12%  62  14  11  0 19%  59  21  1  0 NA 30%  63  5  0  2
    Type of Case 

    One Parent

    Two Parent

    95%  5 100%  0 NA 99%  1
    Percent with High School  Diploma or GED 53% 42% 54% 51%





    Bloom, Dan. (1997). After AFDC: Welfare-to-Work Choices and Challenges for States, New York: Manpower Demonstration Research Corporation.

    Bloom, Dan and David Butler. (1995). Implementing Time-Limited Welfare: Early Experiences in Three States, New York: Manpower Demonstration Research Corporation.

    Brock, Thomas, David Butler, and David Long. (1993). Unpaid Work Experience for Welfare Recipients: Findings and Lessons From MDRC Research, New York: Manpower Demonstration Research Corporation.

    Brown, Amy. (1997). Work First: How to Implement an Employment-Focused Approach to Welfare Reform, New York: Manpower Demonstration Research Corporation.

    Hamilton, Gayle. (1995). The JOBS Evaluation: Monthly Participation Rates in Three Sites and Factors Affecting Participation Levels in Welfare-to-Work Programs, Washington, D.C.: U.S. Department of Health and Human Services and U.S. Department of Education.

    U.S. General Accounting Office. (May, 1997). States’ Early Experiences with Benefit Termination (GAO/HEHS-97- 740.

    U.S. General Accounting Office. (May, 1997). Welfare Reform: Three States’ Approaches Show Promise of Increasing Work Participation (GAO/HEH-97-80).

    Watson, Keith and Jerome Gallagher, et. al. (1998). Temporary Assistance for Needy Families (TANF) One Year After Federal Welfare Reform: A Description of State TANF Decisions as of October 1997, Washington D.C.: Urban Institute.

    Acknowledgements This study could not have been completed without the cooperation and assistance of many individuals. State and local agency staff, service providers and community work experience providers were extremely generous with their time and instrumental in helping us understand their policies, programs, practices and views on a wide range of issues. We are also indebted to Steve Trost and Stacy Poulos for their valuable assistance on the administrative data analysis and to Karen Foley and Rick Scully for their contributions throughout throughout the publication process. Great Appreciation is also owed to Felicity Skidmore for her lead role in preparing the Executive Summary report. Finally, the report benefited from the careful review and suggestions of Karin Martinson, Sharon Parrot, Demetra Nightingale, Bob Lerman, Kathy Brennan and Carolyn O'Brien.

    Low-Income Populations
    Temporary Assistance for Needy Families (TANF)