The Parents’ Fair Share (PFS) Demonstration:
Matching Opportunities to Obligations

Chapter 4:
Program Scale and Participation Patterns

[ Main Page of Report | Contents of Report ]

Contents

1. Program Scale
2. Participation Rates
   1. Overall Participation Rates and Levels
   2. Participation by Component
   3. Program "Coverage"
   4. Assessing the Participation Rates
3. Conclusions

Endnotes

This chapter examines the scale of the PFS pilot programs and the rates of participation in program activities among noncustodial parents who were referred into PFS during the first 7 to 10 months of pilot operations. Both of these measures are important to assessing the feasibility of the PFS approach and the appropriateness of implementing a random assignment evaluation to test its impacts.

Overall, the data in this chapter suggest that most of the PFS sites have been able to devise systems to identify and refer large numbers of noncustodial parents for PFS services. Just as important, a large fraction of these referred parents -- about two-thirds -- actually participated in a PFS activity within four months of the referral. This compares favorably with the participation rates measured in welfare-to-work programs for AFDC recipients, which also attempted to impose participation requirements on large numbers of individuals who may not have sought these services on their own. Finally, the data also indicate that most of the noncustodial parents who were referred to PFS but failed to participate were either referred for further enforcement action or found jobs on their own. This suggests that the sites developed relatively B tracking and follow-up systems that allowed few parents to "slip through the cracks." All of these findings help to support the conclusion that PFS is indeed operationally feasible.

I. Program Scale

A key goal of the pilot phase was to determine whether sites could mount and operate large-scale PFS programs. As explained in Chapter 1, this is a critical factor in assessing the extent of institutional change, the quality of inter-agency linkages, and the feasibility of conducting a random assignment evaluation during Phase II of the demonstration.

Throughout the pilot, MDRC has used two broad indicators to assess the scale of the pilot programs: 1) the number of noncustodial parents "referred" to PFS, and 2) the number who "participate" in program activities. As noted earlier, during the site selection phase, MDRC stipulated that each site would be expected to build the capacity to serve at least 300 participants during the pilot phase. A "participant" was defined as an eligible noncustodial parent who attended at least one session of peer support and one session of an employment and training activity (although substantially more participation was expected). This definition was later expanded to include individuals who did not meet the definition of a participant but did find and report employment to program staff.

There is no specific target for referrals; sites were expected to refer as many noncustodial parents as necessary to obtain at least 300 participants. However, referrals are an important measure of the sites' ability to identify potentially eligible noncustodial parents and process their cases. This is especially important because, as described in Chapter 1, many CSE agencies did not routinely "work" the cases targeted for PFS prior to the start of the demonstration. It was expected that sites would need to identify many more than 300 noncustodial parents in order to achieve the participation goal; it was assumed that there would be attrition at several points in the identification and referral process, as described in previous chapters.

Table 4.1 shows the total number of noncustodial parents who were referred to PFS and the number who participated in both peer support and employment and training or reported employment in each site through June 1993. The referral figures indicate that several of the sites have been quite successful in identifying and contacting eligible participants. Overall, more than 3,400 noncustodial parents had been referred to PFS by that time, an average of nearly 400 per site. This projects to a total of more than 5,000 by the end of the pilot phase. The Ohio site had referred well over 600 noncustodial parents and the Michigan site more than 500; several other sites had referred more than 300 parents.

TABLE 4.1

Two issues should be kept in mind in assessing these figures. First, in most sites, the number of referrals represents only a fraction of the total number of noncustodial parents who were identified as potentially eligible for PFS. As noted earlier, some of those who were identified were never scheduled for hearings or appointments with CSE staff and some of those who were scheduled failed to show up. Others showed up but were not referred to PFS, usually because they admitted that they were employed or were found to be ineligible for another reason (e.g., they were disabled). PFS certainly affected the behavior of some of these individuals, in some cases by persuading them to find or admit to employment without being referred into the program. ⁽¹⁾

Second, the differing intake procedures described in Chapter 2 mean that the definition of the term "referral" differs from site to site. The most important difference is that in Missouri (and, to some extent, New Jersey), the number of referrals actually reflects the number of noncustodial parents who voluntarily appeared for a recruitment meeting and requested PFS services; in most other sites, it reflects the number who were ordered or referred into PFS during a court hearing or an appointment with CSE staff.

The participant totals are somewhat easier to interpret. Since the pilot began in mid-1992 and was expected to last roughly 18 months, one would expect that sites should have achieved roughly two-thirds of the overall goal by June 1993 -- in other words, to have served at least 200 participants -- by this point. The data in Table 4.1 illustrate that four of the pilot sites (Ohio, Tennessee, Michigan, and Minnesota) ⁽²⁾ were on this pace, and two others (Missouri and Massachusetts) were quite close. New Jersey had only recently initiated a court-based referral process, so its totals, while low overall, were rising quickly. The other two sites, Alabama and Florida, have had more difficulty building effective referral systems. The nine-site total, 1,855, represented more than two-thirds of the overall goal of 2,700 participants.

[Go To Contents]

II. Participation Rates

Participation rates -- defined as the proportion of those referred to PFS who participate in program activities within a specified period -- provide important evidence of the sites' ability to move noncustodial parents from the referral stage into program activities. They may also point to the strength of monitoring and noncompliance follow-up procedures. Moreover, a relatively high participation rate among those randomly assigned to the program group is assumed to be a precondition for measuring impacts during Phase II of PFS. ⁽³⁾

A. Overall Participation Rates and Levels

In the pilot phase, the point of referral to PFS (and BIF completion) is the closest available approximation to the point of random assignment in an experimental evaluation. Table 4.2 indicates the fraction of those referred to PFS who participated in any program activity within four months of the referral. ⁽⁴⁾ This figure, which includes all noncustodial parents referred to PFS by the end of February, 1993, is approximately 67 percent. The participation rates for the two main PFS components -- employment and training and peer support -- were approximately 50 percent and 60 percent, respectively. ⁽⁵⁾

TABLE 4.2

The table also shows that the overall participation rate varied considerably across sites, from 50 percent in Michigan to 90 percent in Missouri. This disparity is primarily attributable to differences in intake procedures and service populations. For example, one would expect a high participation rate in Missouri because the pilot program primarily served volunteers. Thus, the population of "referred" individuals in this site includes only noncustodial parents who appeared at a recruitment meeting and requested services; many others were recruited and never showed up, but they are not counted in the base. In most other sites, the base includes all noncustodial parents who were referred or ordered into PFS during a court hearing or meeting with child support enforcement staff. The nature of this message (i.e., the level of "mandatoriness"), the authority of its deliverer (i.e., a judge or hearing officer versus a CSE staff person), the site's ability to follow up on those who fail to respond, and the real or perceived consequences for noncooperation, all vary across sites and probably affect the participation rate. ⁽⁶⁾ For example, if noncustodial parents are ordered into PFS by a judge who has a reputation for jailing people, this may have a different effect than if the order is issued by a CSE staff person.

The bottom panel of Table 4.2 shows the total number of PFS activity sessions attended within four months of referral by those who participated in any PFS activity. ⁽⁷⁾ The definition of a "session" depends on the activity, ranging from a 90-120 minute peer support meeting to a full day of work in an on-the-job-training position. ⁽⁸⁾ The table shows that the overall average number of sessions is nearly 30. More than 70 percent of all participants attended 10 or more activity sessions, and nearly one-fourth attended 40 or more sessions. Again, the patterns vary across sites, and are affected by the program model. As will be discussed in subsequent chapters, the types of activity assignments and the expectations for participants varied from site to site and affect how many sessions participants are required to attend.

Other figures (not shown in the table) suggest that the relatively short four-month follow-up period does not detract much from the overall participation rates reported earlier. In other words, relatively few individuals began their PFS participation in the 5th month or later. For example, for an early cohort referred to PFS by the end of December 1992, the rate of participation within six months of referral is 66.3 percent, compared to a rate of 65 percent within four months of referral.

However, the 6-month data for this early cohort do suggest that many participants remain active in PFS activities beyond the fourth month; the average number of sessions attended rises from 26 within 4 months to nearly 33 within 6 months. This suggests that the figures in the bottom panel of Table 4.2 would be larger if additional follow-up data were available. This is also indicated by the bottom row of the table, which shows that 42 percent of the participants were still attending program activities in the fourth month after referral; many of these individuals continued to participate in months 5 and 6, and beyond.

B. Participation by Component

Table 4.3 shows the rates of participation in the key PFS components. This table includes only those noncustodial parents who participated in a PFS activity -- the same group that was the focus of the bottom panel of Table 4.2. Once again, the table examines the first four months following the referral.

TABLE 4.3

These data show that nearly all noncustodial parents who participted in PFS activities attended peer support. As will be discussed further in Chapter 5, peer support has successfully engaged the participants in almost every site. The average number of peer support sessions attended for those who participated in this activity is nearly 12; a surprisingly high figure given initial skepticism about men's interest in this type of activity. ⁽⁹⁾

Participation is employment and training activities was somewhat less consistent. Still, about three-fourths of the participants attended at least one employment and training activity within four months. The average number of sessions for those who participated was nearly 25. Once again, there is considerable variation across sites owing in part to differing patterns of scheduling and participation in specific types of employment and training activities. This is discussed further in Chapter 7.

The last row of Table 4.3 shows that very few participants attended mediation. This was a voluntary activity that usually required participation by custodial parents. As discussed further in Chapter 8, most sites found it difficult to obtain this cooperation.

C. Program "Coverage"

The participation figures discussed above can only be fully assessed by considering the activities of the nonparticipants; many of those who did not participate in PFS activities may have been engaging in activities -- such as employment -- that are consistent with the goals of PFS. Others may not have engaged in such activities, but may have been promptly identified and referred back to the CSE agency for further enforcement action. ⁽¹⁰⁾

Table 4.4 indicates that about 90 percent of those referred to PFS either participated in program activities, reported employment, or were referred back to the CSE program for noncooperation. The figure is high in all sites where complete data are available. This indicates that the number of people who "slipped through the cracks" was relatively small and suggests that effective monitoring and follow-up procedures are in place. ⁽¹¹⁾

TABLE 4.4

D. Assessing the Participation Rates

There is no obvious benchmark for assessing the PFS participation data described above. However, the best available comparison is probably to earlier welfare-to-work programs evaluated by MDRC. Like PFS, these programs operated through large bureaucratic systems and attempted to mandate participation in employment-related activities by individuals who may not have done so otherwise.

Although direct comparisons are problematic for several reasons, it appears that the rates of participation in PFS activities are somewhat higher than the rates measured in previous evaluations of mandatory welfare-to-work programs that produced statistically significant earnings and employment impacts. For example, in an evaluation of California's Greater Avenues for Independence (GAIN) program, MDRC found that about a third of those who were referred to GAIN by welfare eligibility workers actually participated in a GAIN activity within 6 months. ⁽¹²⁾ This rate was somewhat higher, about 46 percent within twelve months, in a pre-GAIN work/welfare program in San Diego. ⁽¹³⁾ In PFS, the rates of participation within four months of referral in sites that served primarily mandatory clients (i.e., excluding Missouri) ranged from 50 to 81 percent. ⁽¹⁴⁾

There are several factors that should be considered in making this comparison. First, many of those who did not participate in the welfare-to-work programs were deferred or excused for program-recognized reasons or left welfare. In contrast, there are relatively few reasons why a noncustodial parent referred to PFS would not be expected to participate.⁽¹⁵⁾ Second, the penalties for non-cooperation are quite different. In welfare-to-work programs, noncooperative individuals risked reductions in their welfare grants. In PFS, nonparticipants could, at least in theory, be incarcerated, although the likelihood that this would actually occur varied from site to site.⁽¹⁶⁾

[Go To Contents]

III. Conclusions

At the most basic level, the operational feasibility of a mandatory program may be measured by the ability of sites to identify large numbers of eligible noncustodial parents, move them into program activities, monitor their participation, and respond to noncompliance. The results presented in this chapter suggest that PFS has passed this critical test. Most of the sites have met the scale requirements set by MDRC in preparation for the project, and the rates of participation in program activities appear to be higher than those achieved in mandatory welfare-to-work programs with similar objectives. Moreover, the "coverage" data suggest that relatively few eligible noncustodial parents fell through the cracks. Subsequent chapters will examine the experiences of participants in the various program components.

Endnotes

1. In an impact evaluation, the random assignment process would be placed earlier in the intake flow than the current point of BIF completion in order to capture these effects; this means that the number of noncustodial parents entering the research sample for such an analysis would be substantially larger than the number referred to PFS during the pilot. [Back To Text]

2. In Minnesota and Ohio, totals for the two counties are combined. [Back To Text]

3. This is true both because high participation rates usually signal a B program, and because higher rates mean that a smaller research sample is needed to detect program impacts. [Back To Text]

4. This is a different definition of the term "participation" than was used in the previous section. The analysis in this section, which is consistent with those conducted in other MDRC studies of mandatory programs, follows individuals for a fixed number of months and counts as a "participant" any individual who attends at least one program activity. The previous section used a different definition of a participant and simply counted all referrals and participants recorded by the end of June 1993. [Back To Text]

5. Tables in later chapters that provide detailed information on participation in the peer support and employment and training components generally examine only those noncustodial parents who participated in the specified activity. In examining those tables it is important to remember that they include only the 50 or 60 percent of referred noncustodial parents identifed in Table 4.2 who participated in these components. [Back To Text]

6. The rate may also be affected by the no-show rate for court hearings or CSE appointments. If this rate is very high, it may be that the few noncustodial parents who do show up to court are quite likely to cooperate with the PFS order. If the rate is low, the referral population may include a larger proportion of individuals who would find employment on their own or fail to cooperate. No information is available on this issue in the pilot phase. [Back To Text]

7. As noted in Chapter 1, the MIS developed for the pilot phase did not capture all PFS activities. [Back To Text]

8. In many cases, participants attend more than one activity (e.g., peer support and job club) in the same day, which means that "sessions attended" and "days attended" are not synonymous; in fact, the figures for sessions attended represent an upper bound on the number of days a participant attended PFS activities. [Back To Text]

9. As explained in Chapter 5, expectations regarding the number of peer support sessions participants are required to attend vary from site to site. [Back To Text]

10. Most sites took immediate action in response to a failure to appear for the initial orientation; however, policies for those who began to participate and later began to miss sessions varied from site to site.[Back To Text]

11. It is important to note that delays in the CSE or court systems may delay the follow-up on cases after they are referred back for noncooperation. For example, contempt hearings may be delayed for several weeks or months in some sites. [Back To Text]

12. Riccio, et. al., 1989. [Back To Text]

13. Goldman, et. al., 1986. [Back To Text]

14. The GAIN evaluation also included data on the length of participation in program activities among those who participated. However, it is quite difficult to compare these data -- which are measured in months -- with the PFS data on sessions attended. [Back To Text]

15. A noncustodial parent's CSE case could close or, as described in the previous section, he could report employment and begin paying child support. [Back To Text]

16. I t is also important to note that, during Phase II, participation rates will include all noncustodial parents who are randomly assigned to the program group. Under current plans, this assignment will take place just before court hearings, rather than just after hearings, where the BIF has been completed during the pilot. It is expected that some noncustodial parents, when informed about the PFS participation requirement, will confess to previously-unreported employment, and thus will not be referred to the program. This will probably lead to lower participation rates than those measured in the pilot. [Back To Text]

Where to?

Top of Page
Contents

Main Page of Report | Contents of Report

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

Last updated: 04/26/01