Moving People from Welfare to Work. Lessons from the National Evaluation of Welfare-to-Work Strategies.. Participation Standards: What does it take to engage a substantial proportion of people in welfare-to-work program activities?


Despite the fact that participation in welfare-to-work activities is generally required in exchange for welfare receipt, welfare agencies often have a difficult time engaging a large share of their caseloads in program activities. Reacting in part to low participation rates in welfare-to-work programs, FSA broke new ground in requiring states to engage a specified -- and gradually increasing -- proportion of welfare recipients in work-related activities each month or to face possible reductions in federal funding.

The 1996 welfare law continued down this path but went still further, raising the percentage of welfare recipients who had to engage in work or work activities, narrowing the set of activities that could count in participation rates, and increasing the required number of hours of participation per week. States could lower their required participation rate by reducing their TANF caseload (relative to their 1995 caseload) and, owing to the welfare reform message's taking hold and unprecedented economic expansion, targets have been relatively easy to meet. But the recent economic downturn, which is likely to increase TANF caseloads, may make meeting participation standards more difficult. Furthermore, welfare time limits have increased the pressure on welfare-to-work programs to help all welfare recipients become self-sufficient before exhausting their welfare eligibility, thus magnifying the importance and urgency of increasing program participation.

The NEWWS programs, which began operating under FSA, did not put time limits on welfare receipt, generally exempted more people from participation than is now the case, imposed less severe financial penalties for nonparticipation than are now enforced in many states, and did not have enhanced earned income disregards -- which allow more people to fulfill their participation requirement by working while on welfare -- as many states now do. Nonetheless, the programs all shared PRWORA's goals of requiring welfare recipients to do something in exchange for their welfare benefits, and the NEWWS findings provide lessons and guidance concerning participation.

  • Three elements strongly affect a program's participation rate and in turn how program staff behave and emphasize different procedures and outcomes in order to meet participation benchmarks: the definition of a participant, the base of people mandated to participate, and the period of time over which the rate is measured.

Table 5 shows the three main inputs to participation rate calculations and how much the participation rates of six of the NEWWS programs would have varied as a function of these components. Note that all six programs were successful in the sense that, over the five-year follow-up period, they vigorously enforced the participation mandate, increased employment, and reduced welfare. If participation is calculated using method 1, then the participation is low because only a small percentage of all welfare cases in a month included someone who participated at least 20 hours or was employed at least 15 hours every week. If, however as in method 2 recipients who are being sanctioned and recipients who participated at all during a month (that is, for any number of hours) are counted and people who qualify for exemptions are removed from the base, then the participation rate increases significantly. Methods 1 and 2 compute participation rates over a one-month period. Method 3, in contrast, lengthens the period of measurement to include everyone who participated at all over a two-year period. As the table shows, this calculation yields the highest participation rates. Method 3 is the one used most often in research (and was discussed in the previous section), but the monthly methods (particularly method 1) correspond more closely to that outlined in the 1996 federal welfare law.

Table 5.
Methods of Calculating Participation Rates: How Participation Standards Are Defined Strongly Affects What Rates Can Be Achieved
Method Definition of a Participant Population Base Measurement Period Participation Rate Range
  Statuses Counted Level of Involvement Required      

SOURCE: Hamilton et al., 1995.

NOTE: The participation rate ranges shown were calculated using data for the LFA and HCD programs in Atlanta, Grand Rapids, and Riverside.

1 Participated in JOBS activities or was employed while receiving AFDC Participated at least 20 hours per week during every week or was employed at least 15 hours per week during every week All AFDC cases in the month Month 5%-10%
2 Participated in JOBS activities or was employed while receiving AFDC or sanction was imposed or requested Participated at all or was employed at all or sanction was imposed or requested at all All JOBS-mandatory individuals in the month Month 35%-44%
3 Participated in JOBS activities Participated at all All JOBS-mandatory individuals entering program in a year Two years after program entry 44%-74%
  • Achieving high rates of participation in welfare-to-work program activities takes time and resources. In general, the higher the required participation rate, the more money is needed.

Greatly increasing program participation requires committing considerable staff time and resources. Effort must be made to reach out to potential participants, assign them quickly to program activities, monitor their participation and progress, and reengage those who drop out. In addition, penalties must be enforced in a timely manner for those who do not comply with the participation mandate. Finally, the more participants a program has, the more resources are needed to pay for program activities and support services such as child care, transportation, and other ancillary expenses. The NEWWS findings show that programs must work with almost all the welfare recipients targeted by a mandate in order to have a sufficiently large proportion of them count as participants in any given month. In fact, staff often spend almost as much time (in some cases, even more time) trying to engage nonparticipants as working with participants.

  • Even if funding is plentiful and case management is flawless, in any given month some people will not count as participants for unavoidable reasons.

Figure 10 shows the participation statuses in a typical month of enrollees in the six NEWWS programs represented in Table 5. Averaged across the six programs, only a small proportion of recipients required to participate actually met FSA's definition of a participant in a typical month. What were the other recipients doing? As the figure indicates, somewhat more than one-third of them had not yet attended a program orientation; about one-fifth had attended an orientation but were not active in that particular month; and one-third participated in a program activity, but not for enough hours to meet the participation threshold. Some of these reasons for not meeting the definition of participation may be preventable, particularly through greater staffing: Tighter case monitoring and shorter gaps between activities can encourage people to enter activities more quickly and minimize the amount of time that people wait for new activities to begin. Yet other reasons, such as recipients' becoming ill or having to care for an ill child, lie outside the purview of welfare-to-work programs.

  • By themselves, rates of participation may not be as critical to increasing job finding and welfare exit as are the design and implementation of program activities.

Several factors may be more important than participation rates in realizing the goals of welfare reform -- for instance, whether the types of activities in which people participate are the ones that can best help them to become self-sufficient, what people are doing while in those activities, and who is participating. In other words, there is little gain to having a lot of people participating in activities if the activities themselves are ineffective or inappropriate for those who participate in them.

Figure 10.
Reasons for Nonparticipation in a Typical Month:
Some Types of Nonparticipation Are Preventable, but others are Not

Reasons for Nonparticipation in a Typical Month: Some Types of Nonparticipation are Preventable, but others are Not.

SOURCE: Hamilton, 1995.

NOTE: The percentage shown are averages for the LFA and HCD programs in Atlanta, Grand Rapids, and Riverside.