Individuals receiving AFDC in the comparison group were not subject to participation requirements and could not receive any intervention services apart from the supportive services, though they could still participate in education and employment-related activities available in their communities. However, FIA referred about 24 percent of the comparison group to Work First, which might have reflected staff confusion over when to make referrals.
MOST followed by Work First cost $4,354 (in 1999 dollars) per program participant, which included Work First costs and the costs of MOST to the welfare department and to education and training providers. This was $2,053 more than the cost of similar services that comparison group members accessed on their own. A cost-benefit analysis examined costs and benefits to the participant and the government:
- The participant. MOST followed by Work First resulted in a gain of $262 to the average participant; this means that the sum of increases in earnings and fringe benefits was estimated to be $262 higher than the sum of increases in tax payments, decreases in AFDC payments and Food Stamps, and costs associated with losing Medicaid coverage.
- The government. MOST followed by Work First resulted in a loss of $329 to the government per participant; this means that the sum of all costs associated with providing employment services and related support services paid by the Detroit FIA and other government agencies was estimated to be $329 higher than any monetary benefits. Monetary benefits included the sum of increases in payroll taxes and income and sales tax and decreases in costs associated with cash assistance payments, Food Stamps, Medicaid, and transfer administration.
MOST launched statewide in Michigan in 1988; Work First gradually replaced MOST beginning in October 1994 and was still active when study findings were released. Enrollment in the study began in May 1992 and continued until June 1994. The study followed participants for five years. Study enrollment ended prior to the transition to Work First, but some participants were still receiving services at the time of transition and were referred to Work First.
The MOST intervention was based on a skill-building philosophy. After participants attended an orientation, their designated MOST case manager assigned them to activities in the following categories:
- Education. Case managers often referred those without a high school diploma or GED to take courses to earn a credential, English-as-a-second-language classes, or basic education classes from the Detroit Public Schools. Participants with a basic education credential could also take courses for credit toward a two- or four-year degree at local community colleges.
- Occupational training. Participants had access to training options in various fields, including nursing, business and clerical occupations, computer programming, cosmetology, and child care. MOST also offered a 20-week basic skills and occupational training program for participants without a high school diploma or GED.
- Individual job search. Case managers assigned some participants to search for a job on their own and log their weekly number of employer contacts. This assignment was usually used after participants had completed an education or training course.
- Job search assistance. MOST referred some participants to external job search programs.
- Work experience. Case managers assigned some participants to work experience, which could include unpaid work, on-the-job training, and paid work, which most frequently took the form of college work-study. Less than 2 percent of participants took part in this activity.
Case managers also provided supportive services, including child care payments, referrals to child care providers, and transportation assistance. They monitored participants’ engagement and decided whether to sanction those who failed to comply with participation requirements, though case managers rarely imposed sanctions.
Beginning in October 1994, MOST case managers sent intervention participants to Work First if they were currently receiving AFDC and not engaging in training or education. This included about one-quarter of the evaluation’s sample. The FIA continued to manage benefits administration and sanctions, as well as the intervention’s supportive services, but MOST case managers were no longer involved in activity assignment. In contrast to MOST, Work First was based on the premise that individuals receiving benefits best learn work-related skills and habits in the workplace and emphasized fast placement into jobs. Work First required newly referred participants to engage in a job search independently for 30 days before Work First case managers conducted an intake to Work First, which included testing and assessment. If the participant had not yet obtained employment, the case manager assigned them to a contracted service provider for job search assistance or, more rarely, training or educational activities or community work experience.
During the first year of Work First, the case managers were Detroit Michigan Works Agency employees, but in following years, this role was also contracted out to the organizations providing services to intervention participants. These organizations had flexibility in how to structure their services, though they generally included classroom instruction and independent job search. In 1997, Work First began to encourage clients working at least 20 hours a week to attend an education or training program in addition to their job.
Challenges. MOST was not fully staffed during the study. Authorizing payments for supportive services consumed much of case managers’ time, leaving them unable to devote much time to assessing or otherwise learning about clients. Consequently, MOST case managers did not focus on monitoring participation rates or enforcing sanctions for nonparticipation. In addition, service providers sent participant attendance data to case managers irregularly, limiting case managers’ ability to address attendance issues in a timely manner.
The study did not discuss any tools to measure fidelity to the intervention model.
MOST and Work First were part of Michigan’s Job Opportunities and Basic Skills Training program, which was funded through federal and state funds. After the Temporary Assistance for Needy Families (TANF) program replaced AFDC in 1996, Work First became part of Michigan’s TANF program. Other government entities, including the Detroit Public Schools, and private sources funded the service providers to which MOST assigned participants. Work First only used its contracted service providers.
MOST followed by Work First was implemented in Detroit, MI. Wayne County, in which Detroit is located, had relatively high unemployment early during the study period. In 1992, the county unemployment rate was 10.5 percent, though by 1996, it had fallen to 5.5 percent.
Two Detroit offices of the Michigan Family Independence Agency (FIA, known as the Michigan Department of Social Services until March 1996) implemented MOST. The Detroit office of the Michigan Works Agency implemented Work First.
A variety of community service providers delivered education and training services—and, during Work First, case management—to intervention participants.
MOST followed by Work First served individuals receiving AFDC who were at least age 18 and were single parents whose youngest child was age 1 or older. All adults receiving AFDC not exempt from AFDC work requirements were required to participate in MOST if they were not otherwise satisfying the work requirements. However, because of staffing shortages, FIA staff were unable to enroll all eligible individuals receiving AFDC in MOST, so staff prioritized those who called the AFDC office and actively expressed interest. These volunteers made up about 15 percent of the study. The study authors assumed these sample members had already decided to enroll in education or training programs and wanted to access MOST’s supportive services; about two-thirds of these volunteers were already engaged in education or training at the time of random assignment.
The majority of participants were Black or African American (87 percent) and female (97 percent). On average, participants were 30 years old and had 2 children. Few (7 percent) were employed at the time of random assignment. More than half (57 percent) had a high school diploma or GED at the time of random assignment.
Participants were required to engage in intervention services. Although intervention staff communicated to participants that non-engagement could result in sanctions to AFDC receipt, intervention staff almost never sanctioned participants.
Participants were required to engage in services while receiving AFDC, but only 33 to 47 percent of intervention participants engaged with MOST or Work First at any point during the 2 years following their enrollment. (The study used two different methods to estimate participation and presented an upper- and lower-bound estimate.) The majority of those who engaged with services within that period did so solely via MOST; though 23 percent of all participants were referred to Work First during the two years following their enrollment, only 6 percent of all participants actually took part in Work First during that period. During the 2-year period following enrollment in the study, participants took part in an activity or worked for 7.5 months on average. For participants who took part in an activity for at least 1 month, the average total length of participation was 7.9 months.
Work First monitored participants’ attendance and referred about 30 percent of participants to the sanctions process for falling below 75 percent attendance; however, MOST case managers, who were still in charge of the sanctions process, rarely imposed sanctions.
Staff included:
- MOST case managers assigned participants to services, administered supportive services, and monitored participants’ attendance to decide whether to enforce sanctions. After Work First began in October 1994, MOST case managers also referred participants to Work First and determined when to impose sanctions. All case managers held associate’s or bachelor’s degrees.
- Income maintenance workers determined eligibility for AFDC and other benefits, determined if clients were required to participate in MOST followed by Work First, and when applicable, imposed sanctions. Most income maintenance workers had not attended college.
- Staff at community organizations delivered education and occupational training until Work First began.
- Work First case managers conducted an initial intake, testing, and assessment for Work First; assigned participants to services; and monitored participants’ attendance. Originally, case managers were employees of the Detroit Michigan Works Agency; after the first year of Work First, they were contractors at outside agencies.
- Staff at service providers contracted by the Michigan Works Agency delivered job search assistance after Work First began.