Enhanced Services for the Hard-to-Employ


Fueled by a strong economy and passage of the 1996 federal welfare law, which imposed new work requirements and time limits on cash benefits, welfare caseloads declined precipitously during the 1990s. Between 1993 and 2000, the number of families on welfare dropped 56 percent nationally, with individual states experiencing reductions ranging from 20 percent to more than 90 percent. Yet more than two million families were on welfare at the end of this period, and many others had replaced welfare with employment that later dried up or had left the rolls without any other source of income.

As states consider how to deal with families who reach welfare time limits or do not receive welfare but have no other means of support, it is essential to find out who the hard-to-employ are and what services can help them find and keep jobs. Led by MDRC in collaboration with several partners, Enhanced Services for the Hard-to-Employ was the first comprehensive attempt to understand this diverse low-income population and to test interventions aimed at the most common barriers to their employment.


Agenda, Scope, and Goals

Many hard-to-employ families with low incomes face one or more of the following obstacles to finding and sustaining employment: low basic skills, limited work experience, health problems, or a criminal record. The Hard-to-Employ demonstration was designed to field and evaluate a variety of innovative ways to boost employment, reduce welfare receipt, and promote well-being in this population.

Each of the sites in the demonstration targeted a segment of the hard-to-employ population, such as individuals leaving prison, those in need of mental health treatment, or those who would benefit from a “two-generation” intervention. The following four sites participated in this project:

  • Center for Employment Opportunities (CEO), New York City. Parolees who were mandated to work as a condition of parole were placed in paid transitional employment at one of several dozen worksites around the city for two to three months, followed by job placement into unsubsidized jobs. The program also included a fatherhood program, employer-driven skills training programs, a greater focus on postplacement retention services, and job coaching.
  • Rhode Island Mental Health Evaluation. The target population for this intervention was working-age parents who were on Medicaid and had undiagnosed depression. An enhanced treatment group received intensive telephonic outreach and follow-up from managed care case managers to encourage participation in mental health treatment. The program provided access to employment services as well.
  • Kansas and Missouri Early Head Start (EHS). Aimed at poor families who had children four years old or younger or a baby on the way, and who were interested in receiving EHS, this “two-generation” intervention sought to enhance the link between the TANF employment and EHS child development programs. The children in the program group were enrolled in EHS services and the parents received home visits to establish and work toward self-sufficiency goals.
  • Alternative Employment Strategies for Long-Term Welfare Recipients in Philadelphia. This study tested two employment strategies — paid transitional employment (provided by the Transitional Work Corporation) and upfront barrier removal services (provided by Jewish Employment and Vocational Services).

Design, Sites, and Data Sources

The evaluation had three components:

  • The implementation and process analysis examined how the programs operate, based primarily on site visits and interviews with program staff and administrators.
  • The impact analysis used a rigorous research design to measure the programs’ effects on outcomes including employment, welfare use, and family functioning. Random assignment of study participants was completed in 2006. The outcomes for both groups were followed for at least three years, using public administrative records and surveys of study members. In addition, follow-up surveys were conducted at 15 and 36 months after random assignment in most sites.
  • The cost analysis compared the financial costs of the interventions.