The $3 billion Welfare-to-Work (WtW) grants program established by Congress as part of the Balanced Budget Act (BBA) of 1997 provided funds to over 700 state and local grantees. Congress appropriated funds for FY1998 and FY1999, and grantees were allowed five years to spend their funds.(1) The intent of the grants program, administered at the national level by the U.S. Department of Labor, was to supplement the welfare reform funds included in the Temporary Assistance for Needy Families (TANF) block grants to states, which were authorized under the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA).(2) WtW funds were to support programs especially those in high-poverty communities to assist the least employable, most disadvantaged welfare recipients and noncustodial parents make the transition from welfare to work.
The BBA mandated the U.S. Department of Health and Human Services to evaluate the newly established WtW grants program and report the findings to Congress. This is one of several reports on that evaluation, which Mathematica Policy Research (MPR) is conducting along with its subcontractors the Urban Institute and Support Services International. This report presents findings from the outcomes analysis component of the evaluation. It describes the characteristics and subsequent experiences of enrollees in programs funded by WtW grants in the following 11 study sites:
| · Baltimore County, Maryland(3) | · Boston, Massachusetts | · Chicago, Illinois |
| · Fort Worth, Texas(4) | · Milwaukee, Wisconsin | · Nashville, Tennessee |
| · Philadelphia, Pennsylvania | · Phoenix, Arizona | · St. Lucie County, Florida |
| · West Virginia (29 counties) | · Yakima, Washington(4) |
This report is based on information for individual WtW enrollees obtained from (1) a baseline survey of enrollees conducted by MPR in 1999-2002, (2) a 12-month follow-up survey of enrollees conducted by MPR in 2000-2003, and (3) state administrative files for Unemployment Insurance.
Low rates of WtW enrollment in the study sites precluded the implementation of an experimental design for this evaluation. Consequently, the findings presented here are descriptive in nature and should not be interpreted as the impacts of WtW.
Key Questions and Findings
The analysis underlying this report was guided by four key questions regarding the individuals who enrolled in programs funded by WtW grants in the 11 study sites. Those questions provide the framework for this summary of findings from the outcomes analysis.
Who Enrolled in Programs Funded by Welfare-to-Work Grants? (Exhibit ES.1)
Consistent with the profile of TANF recipients nationwide, WtW enrollees in most of the study sites were predominantly female, were very unlikely to be married, and were typically members of racial or ethnic minority groups. The Boston site typifies this pattern. There, 93 percent of enrollees were women, and 93 percent were also minorities. Only 5 percent of WtW enrollees in Boston were married. In sharp contrast to the typical study site, Milwaukees Nontraditional Opportunities for Work (NOW) program, which served noncustodial parents who were on probation or parole or were scheduled to be released soon from prison or jail, had a clientele that was 95 percent male. Enrollees in St. Lucie County, West Virginia, and Yakima were less likely than their counterparts in the studys more urbanized sites to be women or members of a minority group and somewhat more likely to be married. In West Virginia, only 17 percent of enrollees were minorities and more than a quarter were married.
There were roughly equal numbers of enrollees above and below age 30 in most of the study sites, and about a third of them resided with a child under the age of 3, which may have presented a barrier to employment. However, enrollees in Baltimore County and in West Virginia did not fit this pattern: two-thirds of them were over age 30 and not surprisingly, given their older ages, fewer than one in six were living with a young child.
Many WtW enrollees in the study sites faced significant barriers to employment. In most sites, more than one-third of enrollees lacked a high school diploma or GED. Furthermore, they had weak employment histories, as indicated by their low pre-enrollment earnings. The median annual earnings of enrollees in the year prior to program entry was less than $2,000 in all sites except Baltimore County, where it was $3,603. But WtW enrollees did have some human capital assets. In sites other than West Virginia, at least nine in ten enrollees had prior work experience. Enrollees in Baltimore and St. Lucie counties which hosted the Johns Hopkins University (JHU) program model had particularly strong labor market qualifications; virtually all had prior work experience and at least three-fourths had a high school diploma or its equivalent.(5)
Given the nature of the WtW grants program, most enrollees had received welfare benefits TANF or its precursor, Aid to Families with Dependent Children (AFDC) as case heads at some point in their lives. For example, in Chicago, Nashville, and St. Lucie County, virtually all WtW enrollees had received TANF/AFDC at some point. In fact, in all of the study sites except Milwaukee, 85 percent or more of WtW enrollees had received welfare benefits. However, long-term receipt was the exception rather than the rule. In most of the study sites, only about one-third of WtW enrollees reported in the baseline survey that they had received TANF/AFDC for a cumulative total of five years or more.
What Services Did Enrollees Receive? (Exhibit ES.2)
PRWORA emphasized the rapid employment of TANF recipients who are capable of working. Data from the evaluations 12-month follow-up survey shed light on whether WtW enrollees in the study sites received services consistent with this emphasis. In most of the study sites, 80 percent or more of WtW enrollees received some type of employment preparation service during the year following program entry. Such services are typically designed to quickly address barriers to employment and move them into jobs. They are distinct from skill enhancement services, which generally provide longer-run solutions to human capital deficits.
The employment preparation services most frequently received by WtW enrollees were job readiness training and job search/placement assistance, each of which was received by more than half of enrollees in seven of the study sites (Boston, Chicago, Nashville, Philadelphia, Phoenix, West Virginia, and Yakima). The Phoenix and Yakima enrollees typically received brief job readiness training followed by job search assistance a combination of services that was highly consistent with rapid transition to employment. In contrast, the enrollees in Boston, Chicago, Nashville, Philadelphia, and West Virginia typically received extended job readiness training (or, in the case of Nashville, skill enhancement services), followed by job search assistance. It generally took enrollees in the latter sites longer to become employed than their counterparts in Phoenix and Yakima.
Less than half of the enrollees in the remaining four study sites (Baltimore County, St. Lucie County, Milwaukee, and Ft. Worth) received job readiness training and job search/placement services. Most of the enrollees in the two JHU sites were already employed and therefore had less need for these services, but they did have relatively high rates of receipt of counseling and mediation services. Many of the ex-offender/noncustodial parents served by Milwaukees NOW program received services that were less common in the other sites: peer support/discussion groups, legal assistance, and substance abuse treatment. Low rates of receipt of job readiness training and job search/placement assistance by enrollees in Ft. Worth were not offset by other services, resulting in the lowest rate of receipt of any employment preparation services among the study sites (68 percent).
Skill enhancement services (education and training) were not prominent features of programs funded by WtW grants. Enrollees in Baltimore County, St. Lucie County, and Nashville were more likely than those in the other sites to have received these services; yet even in these sites, fewer than half of enrollees received them. The federal legislation that initially authorized the WtW grants program (the BBA) permitted skill enhancement services to be provided to enrollees only after they had obtained jobs. Subsequent amendments eased this restriction by allowing such services prior to employment for a maximum of six months.(6) Still, it is clear that federal policymakers intended for most investment in human capital under WtW grant-funded programs to occur after, rather than before, an enrollee obtained employment.(7) However, in more than half of the study sites, most of the enrollees who did receive skill enhancement services had not obtained a job prior to the commencement of those services.
Did Enrollees Achieve Success in the Labor Market? (Exhibit ES.3)
In only 3 of the 11 study sites (Baltimore County, St. Lucie County, and West Virginia) were most enrollees employed one year after entering WtW; however, in almost all of the sites they were much more likely to be employed at that time than when they entered the program. In the non-JHU sites, no more than about one-fourth of WtW enrollees were employed when they entered WtW. A year later, about four in ten were employed.(8) All of these increases in employment rates were statistically significant.(9) The enrollees in the Transitional Work Corporations WtW program in Philadelphia typify this pattern: just 7 percent of enrollees were employed at the time of enrollment, but 36 percent were employed one year later.
Most WtW enrollees were employed sometime during the year following program entry, even if they were not employed at the time of entry or the follow-up survey. For example, 67 percent of the Phoenix enrollees were employed sometime during the subsequent year, even though their employment rates were only 3 percent at entry and 33 percent one year later. Across the 11 study sites, about two-thirds or more of the enrollees worked sometime during the subsequent year. On average, it took those who achieved employment about 4 to 5 months to find a job if they did not have one at the outset. The Phoenix and Yakima enrollees moved rapidly into jobs, respectively taking just 3.8 and 4.3 months on average. In contrast, enrollees in Milwaukee, many of whom had severe barriers to employment, were slowest to obtain jobs, taking an average of 5.8 months. In general, the lags in finding initial jobs and the low persistence of employment to the end of the year resulted in enrollees being employed for only about one-third to one-half of the year, except in the JHU sites, where they were employed for nearly three-fourths of the year.
WtW enrollees who were employed one year after program entry worked full-time, or nearly so, on their principal job. The mean weekly hours of work did not vary greatly across the study sites, ranging from 32 to 37. There was greater variability in the mean hourly wage rate, which ranged from a low of $5.75 in West Virginia to $9.08 in Baltimore County and $9.82 in Boston. It was in the $7-to-$8 range in the other eight sites. The proportion of enrollees who received health insurance benefits on their principal job was less than 20 percent in all but three of the sites and exceeded 30 percent only in the Baltimore County site.
For enrollees who were employed, low wages were a barrier to escaping poverty. This conclusion is based on a simulation analysis of poverty rates for employed enrollees. The assumptions underlying the simulations were that these individuals were consistently working 40 hours per week at their actual wage rates on their principal jobs and had no income from government programs but did have income from other sources (such as the earnings of other household members). Even with the assumption of full-time work over an entire month, the simulations showed that these employed enrollees and their households would have experienced high poverty rates, ranging from 20 percent in Baltimore County and Boston to 64 percent in Philadelphia and 71 percent in West Virginia. This finding is based on a measure of income that does not include the earned-income tax credit (EITC).
How Were Enrollees Faring One Year After Entering WtW? (Exhibit ES.4)
Welfare dependence among WtW enrollees fell sharply during the year following program entry. In all of the study sites except Baltimore County, St. Lucie County, and Milwaukee, most enrollees were on TANF when they entered WtW. For example, 85 percent of enrollees in Yakima were on TANF when they entered WtW. One year later, rates of receipt of TANF were significantly lower in all but three sites (Baltimore County, Milwaukee, and Nashville), as again exemplified by Yakima, where the rate of TANF receipt was cut nearly in half, to 44 percent. The lower rates of TANF receipt did not necessarily mean that most enrollees were self-sufficient; in all but the two JHU study sites, no more than four in ten enrollees were employed and off TANF after one year.
End-of-year household incomes were low and poverty rates were high for WtW enrollees in all of the study sites. Enrollees in Philadelphia and West Virginia had mean monthly incomes of less than $1,200 and the highest poverty rates at least 85 percent among enrollees in all of the study sites. Enrollees in Baltimore County had the highest mean monthly income, $1,611, and the lowest poverty rate, which was nevertheless high in an absolute sense, at 49 percent.
Although poverty was pervasive among WtW enrollees one year after program entry, its incidence was lower among those who were employed, typically by about 20 percentage points relative to the rate for enrollees who were not employed. The difference in end-of-year poverty rates between employed and not-employed enrollees was greatest in sites like Baltimore County and Boston where wage rates were high, and smallest in sites like St. Lucie County and West Virginia where wage rates were low (mean wage rates are displayed in Exhibit ES.3). The fact that poverty rates even among employed enrollees were high in an absolute sense (above 50 percent in all sites except Baltimore County) is a reflection of both low wages and the lack of consistent full-time employment over the course of a month.
Conclusions
WtW enrollees were much more likely to receive employment preparation services than skill enhancement services. Consistent with the legislation that authorized the WtW grants program, more than two-thirds of enrollees in each of the 11 study sites received employment preparation services designed to get them ready for and move them into jobs. There was considerable variability across the sites in the types and duration of these services. With the exception of Baltimore County, Nashville, and St. Lucie County, no more than about one-third of enrollees received skill enhancement services to increase their human capital so that they could qualify for better jobs. The relatively few enrollees who did receive skill enhancement services typically had not obtained employment prior to the commencement of those services.
In some sites, employment preparation services were more consistent with rapid job entry than in other sites. WtW enrollees in Phoenix and Yakima received employment preparation services that were highly consistent with rapid transition to employment. Those services consisted primarily of brief job readiness training followed by assisted job search. In contrast, enrollees in Boston, Chicago, Nashville, Philadelphia and West Virginia typically received extended job readiness training (or, in the case of Nashville, education and training), followed by job search assistance. The amount of time that it took enrollees to become employed was generally lower for the former group than the latter.
Most enrollees were employed sometime during the year after they entered WtW, but their employment tended to be unstable. With the exception of enrollees in the two JHU sites that primarily served employed persons, very few enrollees were employed when they entered WtW. But most about two-thirds or more were employed sometime during the subsequent year. However, that employment often proved to be unstable; only about 40 percent of enrollees were employed at the end of the year (except in Baltimore County, St. Lucie County, and West Virginia, where year-end employment rates exceeded 50 percent).
Enrollees who were employed one year after program entry typically worked a lot of hours but received low wages and few fringe benefits. Enrollees who were employed one year after entering WtW worked nearly full-time, on average. However, their hourly wage rate tended to be low, averaging only $7 to $8 per hour in eight study sites and just $5.75 in West Virginia. In all but one site, only about one in five employed enrollees was participating in an employer-sponsored health insurance plan at the end of the year.
The incidence of poverty was high among WtW enrollees one year after program entry, but it was lower among those who were employed. The year-end poverty rate for all enrollees exceeded 60 percent in every study site except Baltimore County, where it was 49 percent. However, the rate for employed enrollees was 20 to 30 percentage points lower than for those who were not employed. Nevertheless, the generally low wages earned by employed enrollees and their lack of consistent full-time employment over an entire month meant that even for this group the incidence of poverty was high in an absolute sense 50 percent or more in every site except Baltimore County.
| Baltimore Co. | Boston | Chicago | Ft. Worth | Milwaukee | Nashville | Philadelphia | Phoenix | St. Lucie Co., FL | West Virginia | Yakima, WA | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Female | 92 | 93 | 97 | 94 | 5 | 99 | 99 | 96 | 92 | 79 | 80 |
| Racial or Ethnic Minority | 79 | 93 | 98 | 78 | 95 | 89 | 98 | NA | 67 | 17 | NA |
| Married | 6 | 5 | 3 | 9 | 5 | 3 | 3 | 7 | 13 | 27 | 15 |
| Less Than 30 Years Old | 31 | 56 | 51 | 59 | 54 | 47 | 45 | 60 | 50 | 39 | 45 |
| Child Under Age 3 in Household | 11 | 28 | 35 | 40 | 13 | 24 | 23 | 42 | 37 | 16 | 22 |
| High School Diploma or GED | 80 | 62 | 58 | 54 | 51 | 57 | 49 | 41 | 75 | 66 | NA |
| Median Earnings (000s of dollars)a | 3.6 | NA | 0.3 | 1.6 | 0.8 | NA | 0.4 | 1.4 | 1.3 | 0.0 | 0.5 |
| Ever Employed | 99 | 93 | 93 | 97 | 96 | 98 | 90 | 96 | 100 | 84 | 92 |
| Ever Received TANF/AFDC | 94 | 95 | 98 | 97 | 14 | 99 | 97 | 94 | 98 | 87 | 91 |
| Time on TANF/AFDC Exceeded 5 Yrs | 48 | 29 | 36 | 16 | 1 | 40 | 22 | 22 | 24 | 24 | 27 |
| Source: 1999-2002 baseline survey of Welfare-to-Work enrollees. a In year prior to WtW enrollment NA = not available. | |||||||||||
| Baltimore Co. | Boston | Chicago | Ft. Worth | Milwaukee | Nashville | Philadelphia | Phoenix | St. Lucie Co., FL | West Virginia | Yakima, WA | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Employment Preparation Services Received | |||||||||||
| Job readiness training | 44 | 60 | 72 | 39 | 40 | 60 | 80 | 62 | 49 | 73 | 52 |
| Job search/placement assistance | 44 | 56 | 66 | 44 | 45 | 60 | 73 | 63 | 47 | 64 | 60 |
| Life skills or self-mgmt. training | 28 | 33 | 42 | 25 | 28 | 42 | 52 | 47 | 32 | 44 | 25 |
| Counseling | 36 | 26 | 19 | 19 | 32 | 29 | 20 | 36 | 33 | 25 | 31 |
| Peer support/discussion group | 21 | 14 | 12 | 11 | 28 | 20 | 16 | 15 | 6 | 10 | 12 |
| Treatment for chronic health cond. | 16 | 11 | 9 | 9 | 7 | 15 | 11 | 12 | 9 | 23 | 14 |
| Mediation | 31 | 8 | 5 | 7 | 14 | 13 | 6 | 12 | 29 | 13 | 12 |
| Mental health services | 14 | 11 | 4 | 8 | 7 | 14 | 7 | 11 | 5 | 12 | 8 |
| Legal assistance | 6 | 6 | 3 | 5 | 14 | 8 | 4 | 6 | 7 | 9 | 8 |
| Substance abuse treatment | 7 | 2 | 4 | 3 | 21 | 5 | 4 | 4 | 2 | 3 | 8 |
| Any employment prep. service | 73 | 79 | 82 | 68 | 80 | 85 | 89 | 83 | 80 | 87 | 86 |
| Receipt of Skill Enhancement Servicesa | 43 | 31 | 24 | 29 | 33 | 44 | 30 | 25 | 47 | 35 | 37 |
| Enrollee Obtained a Job Before Receivingt Skill Enhancement Services | |||||||||||
| Yes | 80 | 31 | 39 | 33 | 52 | 40 | 52 | 26 | 85 | 44 | 68 |
| No | 20 | 69 | 61 | 67 | 48 | 60 | 48 | 74 | 15 | 56 | 32 |
| Source: 2000-2003 12-month follow-up survey of Welfare-to-Work enrollees. a Participation in education or training programs. | |||||||||||
| Baltimore Co. | Boston | Chicago | Ft. Worth | Milwaukee | Nashville | Philadelphia | Phoenix | St. Lucie Co., FL | West Virginia | Yakima, WA | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Employment at Start and End of Year | |||||||||||
| Employed at enrollment (%) | 82 | 6 | 3 | 15 | 26 | 28 | 7 | 3 | 72 | 6 | 6 |
| Employed at end of year (%) | 72 | 43 | 41 | 41 | 41 | 43 | 36 | 33 | 73 | 57 | 49 |
| Statistical significance of difference | * | *** | * | * | * | * | * | * | * | * | |
| Employment During the Year | |||||||||||
| Employed sometime during yr. (%) | 91 | 65 | 65 | 66 | 72 | 70 | 74 | 67 | 97 | 80 | 80 |
| Mean no. of months until first joba | 4.9 | 4.7 | 4.5 | 5.2 | 5.8 | 4.6 | 4.3 | 3.8 | 4.0 | 5.1 | 4.3 |
| Mean proportion of year employed | .72 | .38 | .37 | .37 | .41 | .40 | .36 | .35 | .72 | .44 | .49 |
| Principal Job at End of Yearb | |||||||||||
| Mean hours worked per week | 36 | 34 | 34 | 35 | 37 | 32 | 33 | 37 | 35 | 34 | 32 |
| Mean hourly wage | $9.0 | $9.82 | $7.58 | $8.01 | $7.44 | $7.58 | $7.36 | $7.83 | $7.08 | $5.75 | $7.90 |
| Participated in health ins. plan (%) | 42 | 17 | 11 | 22 | 26 | 18 | 13 | 19 | 18 | 12 | 18 |
| Source: 1999-2002 baseline survey and 2000-2003 12-month follow-up survey of Welfare-to-Work enrollees. a If employed sometime during the year, but not at enrollment. b For enrollees who were employed at the end of the year. */**/*** Difference is statistically significant at the .10/.05/.01 level. | |||||||||||
| Baltimore Co. | Boston | Chicago | Ft. Worth | Milwaukee | Nashville | Philadelphia | Phoenix | St. Lucie Co., FL | West Virginia | Yakima, WA | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| TANF Receipt at Start and End of Year | |||||||||||
| Received TANF at enrollment (%) | 24 | 56 | 91 | 92 | 1 | 64 | 91 | 67 | 41 | 78 | 85 |
| Received TANF at end of year (%) | 15 | 41 | 40 | 43 | 5 | 64 | 58 | 47 | 16 | 36 | 44 |
| Statistical significance of difference | *** | * | * | * | * | * | * | * | * | ||
| Off TANF and employed at end of | 68 | 33 | 28 | 32 | 39 | 23 | 20 | 22 | 67 | 40 | 34 |
| Mean Monthly Household Incomea | $1,61 | $1,165 | $1,247 | $1,321 | $1,505 | $1,046 | $1,141 | $1,247 | $1,397 | $1,186 | $1,546 |
| Incidence of Household Povertyb | |||||||||||
| Povertyc (%) | 49 | 74 | 86 | 77 | 65 | 81 | 87 | 82 | 62 | 85 | 68 |
| Severe povertyd (%) | 19 | 51 | 57 | 52 | 40 | 57 | 57 | 56 | 28 | 47 | 35 |
| Incidence of Household Povertyb,c by | |||||||||||
| Not employed (%) | 70 | 90 | 91 | 89 | 74 | 94 | 93 | 88 | 72 | 88 | 84 |
| Employed (%) | 42 | 54 | 78 | 62 | 50 | 64 | 78 | 69 | 58 | 83 | 54 |
| Statistical significance of difference | ** | *** | *** | *** | *** | *** | *** | *** | *** | ||
| Source: 1999-2002 baseline survey and 2000-2003 12-month follow-up survey of Welfare-to-Work enrollees. a The measure of income includes the actual dollar value of food stamps received by the WtW enrollees household during the month preceding the survey interview. b To be consistent with the standard methodology for determining poverty status, food stamps were excluded from the measure of household income for the poverty analysis. c Monthly income at end of year less than 100 percent of the federal poverty threshold. d Monthly income at end of year less than 50 percent of the federal poverty threshold. */**/*** Difference is statistically significant at the .10/.05/.01 level. | |||||||||||
Endnotes
1. BBA: Public Law 105-33, section 5001, August 5, 1997.
2. PRWORA: Public Law 104-103, section 103, August 22, 1996.
3. Baltimore County largely surrounds but does not include the City of Baltimore.
4. On this evaluation, the names used to designate two of the study sites were selected to facilitate exposition rather than to precisely identify political jurisdictions. The "Ft. Worth" site is actually Tarrant County, Texas, including the city of Ft. Worth. The "Yakima" site is actually Yakima, Kittitas, and Klickitat counties, including the city of Yakima. See Nightingale et al. (2002), Appendix A, for a detailed description of each study site.
5. The strong qualifications of WtW enrollees in Baltimore and St. Lucie counties reflect the program model, which entailed the provision of skill-enhancement services to employed persons.
6. Public Law 106-118, Title VIII, sections 801-807, November 29, 1999.
7. This interpretation is consistent with DOLs final rule for the WtW grants program (DOL 2001 page 2715, Sect. 645.220, paragraphs b and e).
8. This is similar to the 42 percent employment rate that Loprest (2003) reports for adults who have left TANF.
9. Relative to the employment rate at enrollment, the end-of-year rate was slightly lower in JHUs Baltimore County site and unchanged in its St. Lucie County site. In all of the sites, the changes in employment rates should not be interpreted as impacts of the WtW-funded programs, as external factors may have contributed to the changes.