WtW enrollees who left the first job they held on or following program entry for another job often received additional fringe benefits on the new job. They less frequently received a higher wage or worked longer hours. The overall pattern of these findings suggests that job turnover can be conducive to job advancement by at-risk adults.(52)
In nearly half of the 11 study sites, WtW enrollees who left their initial post-enrollment job received more fringe benefits on the most recent job they held during the year following program entry. Appendix Exhibit B.14 shows that job changers in Chicago, Ft. Worth, Philadelphia, West Virginia, and Yakima were significantly more likely to have an additional fringe benefit available to them on their most recent job than on their initial job post-enrollment in WtW. In no site did job changers experience a significant reduction in the availability of any one of the six fringe benefits considered in this evaluation.(53) Exhibit IV.10 displays the findings pertaining to participation in an employer's health insurance plan on the initial and most recent jobs.(54)
In general, job changes did not result in a higher wage or more work hours; however, some study sites did deviate from this pattern. Exhibit IV.11 shows that job changers experienced statistically significant increases in weekly hours of work only in Nashville and Philadelphia, and in the hourly wage only in Philadelphia, West Virginia, and Yakima.
These findings show that job turnover, rather than impeding enrollees' labor-market success, often resulted in a job that provided more fringe benefits. It less often led to a job that provided longer hours or higher wages. Fringe benefit availability, work hours, and wages were almost never lower on the new job. This does not mean that all job turnover is conducive to positive outcomes. Nor does it mean that turnover is the only or the most efficient route to a better job; enrollees who remained on the initial job may have experienced growth in hours, wages, and benefits, but the available data cannot support an investigation of whether this occurred.(55)
Percentage of WtW Enrollees Who Were Employed Sometime During the Year after Program Entry
Mean Number of Months until WtW Enrollees' First Job after Program Entry
Mean Percentage of Time with a Job by WtW Enrollees During the Year After Program Entry
Employment Status of WtW Enrollees at the Time of Program Entry and One Year Later
Hours of Work by WtW Enrollees on All Jobs Held One Year after Program Entry,
Assessed Relative to the Tanf Work Requirement
Hours of Work and Wage Rate on the Principal Job Held by
WtW Enrollees One Year after Program Entry
Percentages of WtW Enrollees with Fringe Benefits on the Principal Job
Held One Year after Program Entry
Rate of Departure by WtW Enrollees from the Initial Job Held after Program Entry
Circumstances of Departure by WtW Enrollees from the Initial Job Held After Program Entry
Job Advancement During the Year after Program Entry: Percentage of Employed Enrollees
Who Participated in Employer's Health Insurance Plan
Job Advancement During the Year after Program Entry
34. The JHU model targeted individuals who were already employed. However, some enrollees were not yet working or had been working but had recently lost their jobs at the time of program entry. In these relatively unusual cases, program staff began serving the enrollees prior to their employment and the services included assistance in finding jobs.
35. The Milwaukee site was an exception to this pattern. There, the NOW program served an especially hard-to-employ population of noncustodial parents who had been incarcerated or were on probation or parole when they enrolled in WtW. The Milwaukee enrollees who were not employed at program entry required nearly six months, on average, to obtain their first jobs.
36. Especially in the extensive pre-employment services sites, the measured duration until the first job may have been distorted by the difficulty that some enrollees had in distinguishing between subsidized employment (a common service in these sites) and unsubsidized employment.
37. If enrollees who were never employed during the follow-up year are excluded from the analysis, Exhibit IV.3 shows that the average proportion of the year that the remaining enrollees were employed is about 10 to 20 percentage points higher than for all enrollees in the non-JHU sites.
39. Hamilton (2002) reports a finding from the National Evaluation of Welfare-to-Work Strategies that, among programs with a strong focus on employment, those that emphasize job search while also providing a menu of short-term pre-employment services are more effective than those that offer primarily job search. An important dimension of the enhanced effectiveness of the mixed approach is an increase in the stability of employment.
40. The incidence of job loss during the follow-up year was moderate in four of the five "extensive pre-employment services" sites approximately 50 to 60 percent in Boston, Chicago, Nashville, and West Virginia. In contrast, the incidence of job loss was high, 74 percent, in Philadelphia. The latter may have been due in part to a program design that entailed placing enrollees in transitional subsidized jobs as a precursor to paid employment.
41. We caution the reader not to interpret the differences in employment rates over the year following WtW enrollment (Exhibit IV.4) as impacts of the local programs. While the programs probably contributed to these positive differences, other factors were no doubt at work. One of those factors is the well-documented tendency for individuals to enter employment and training programs shortly after experiencing a dip in their employment. Even in the absence of the programs, many of these individuals could have had better employment outcomes over the succeeding months as a consequence of their own efforts and better luck.
42. Loprest (2003) reports a 42 percent employment rate among adults who had left TANF during the two years prior to the 2002 Survey of America's Families. Some of these individuals had returned to TANF by the time of the survey interview.
43. PL 104-193, section 407, subsections (c) and (d). The 30-hour requirement became effective in fiscal year 2000.
44. The figures presented in Appendix Exhibit B.8 can be transformed (by dividing the sum of the percentages in rows two and three by the sum of the percentages in rows one through three) to obtain the proportion of enrollees who were not satisfying the TANF 30-hour work requirement because they were employed but their hours of work were insufficient. This proportion was 26 percent in St. Lucie County and was lower in all of the other study sites. For example, it was just 5 percent in Phoenix.
45. In four of these five sites (West Virginia, Baltimore County, Ft. Worth, and Nashville), WtW enrollees had relatively high rates of work-limiting health problems at the time of enrollment (Exhibit II.2).
46. This evaluation also examined the availability of dental insurance, paid holidays, and paid vacation leave. Appendix Exhibit B.12 presents findings for all of the six types of benefits that were examined.
47. Household poverty rates for employed enrollees in all of the study sites were simulated under the assumptions of 40 hours of work per week and 4.3 weeks per month at the actual wage rate on the primary job, no income from government programs (including the EITC), and whatever other income their households actually received in the month prior to the month of the survey interview. The simulated poverty rates are as follows (in percentages): Baltimore County 20, Boston 20, Chicago 60, Ft. Worth 44, Milwaukee 37, Nashville 42, Philadelphia 64, Phoenix 35, St. Lucie County 42, West Virginia 71, and Yakima 40. Alternatively, if the measure of income during the month prior to the month of the survey interview is assumed to include the enrollee's actual earnings during that month, no income from government programs, and whatever other income the household actually received, then the simulated poverty rates are higher, often substantially so: Baltimore County 42, Boston 58, Chicago 82, Ft. Worth 65, Milwaukee 59, Nashville 70, Philadelphia 82, Phoenix 76, St. Lucie County 63, West Virginia 87, and Yakima 61. An important factor contributing to the differences between these two sets of poverty rates is the lack of consistent full-time work by employed WtW enrollees over an entire month.
48. The definition of availability of a benefit varies depending on the type of benefit. Availability of health insurance refers to active participation in an employer's health insurance plan. Availability of paid sick leave and a pension refers to the potential for the enrollee to participate in these plans one year after enrolling in WtW (i.e., at the time of the interview), whether or not the enrollee actually participated in or received benefits from the plans.
49. Two study sites present exceptions to the positive relationship between wages and fringe benefits. In St. Lucie County, employed enrollees had relatively low wages but relatively high rates of availability of fringe benefits. Many of these enrollees were already employed when they entered the WtW program, thus affording them the opportunity to accrue greater tenure on their jobs, and perhaps satisfy the tenure requirements that many firms impose for access to fringe benefits. The opposite pattern prevailed in Yakima, where wages were high and the availability of fringe benefits was low. We can only speculate that this reflected the local economy and the types of jobs available and/or the job-placement strategy pursued by the Yakima WtW program.
50. The measure of jobs held during the year following enrollment in WtW includes all jobs without regard for the degree of temporal overlap in the jobs.
51. The ending of a work period encompasses three similar reasons for departure from a job: (1) a layoff, (2) the ending of a work period, and (3) the ending of a period of self-employment.
52. The findings on job advancement are based on comparisons between the most recent job held during the year after enrolling in WtW and the initial job for individuals who changed jobs during that time.
53. In a few of the study sites, the estimated rates of availability of some fringe benefits are lower on the most recent job than on the first job; however, none of the difference in these estimated rates are statistically significant.
54. The improvement in fringe benefits (and wages) may be inflated in Philadelphia and possibly other study sites as well by the fact that the local WtW program placed large proportions of enrollees in an initial job that was temporary and subsidized. Thus, by design, the initial job typically offered few or no benefits and paid a low wage.
55. The 12-month follow-up survey gathered data on job characteristics at one point in time for each employer. Consequently, the data cannot support an analysis of job progression while an enrollee remained with an employer.