Table of Contents | Previous | Next |
Chapter 1. Introduction
In the post-welfare reform world, an important policy question has taken new prominence: how to improve employment prospects for the millions of Americans who face serious obstacles to steady work. These individuals, including long-term welfare recipients, people with disabilities, those with health or behavioral health problems, and ex-prisoners, often become trapped in costly public assistance and enforcement systems and find themselves living in poverty, outside the mainstream in a society that prizes work and self-sufficiency.
The Enhanced Services for the Hard-to-Employ Demonstration and Evaluation Project is sponsored by the Administration for Children and Families and the Office of the Assistant Secretary for Planning and Evaluation in the U.S. Department of Health and Human Services (HHS), with additional funding from the U.S. Department of Labor.1 It is evaluating four diverse strategies designed to improve employment and other outcomes for low-income parents and others who face serious barriers to employment:
-
A comprehensive employment program for ex-prisoners in New York City;
-
A two-generation Early Head Start program in Kansas and in Missouri that provides enhanced self-sufficiency services and skills training to parents, in addition to high-quality child care;
-
Two alternative employment strategies for long-term welfare recipients in Philadelphia: one that emphasizes services to assess and treat recipients’ barriers to employment and another that places them in paid transitional employment; and
-
An intensive telephonic care management program for Medicaid recipients in Rhode Island who are experiencing serious depression.
MDRC, a nonprofit, nonpartisan social and education policy research organization, is leading the effort to test these four programs using a random assignment research design, the “gold standard” of program evaluation. The research team also includes the Urban Institute, the Lewin Group, Group Health Cooperative, and United Behavioral Health. Over the next several years, the project will generate a wealth of data on the implementation, effects, and costs of these promising approaches.
This first report in the evaluation describes the origin of the project and the rationale for the demonstration, the study design, the four programs and the characteristics of their participants, and identifies some early lessons about the challenges of designing and operating programs that target the hard-to-employ.
Why Focus on the Hard-to-Employ?
For at least three decades, policymakers, researchers, and program operators have developed and studied strategies to improve employment outcomes for people who face serious obstacles to steady work. Interest in the hard-to-employ surged in the 1990s, when the strong economy, rising employment, and dramatic declines in the welfare caseload all combined to focus a spotlight on groups who had been left behind. For the first time on a large scale, welfare agencies began developing or brokering services for recipients with mental health conditions, substance abuse problems, disabilities, and other serious barriers to work. Parallel changes were occurring in other systems: Criminal justice officials began to focus on the daunting problems facing prisoners returning to their communities, and the rapid growth of disability programs led policymakers to look for ways to encourage work among beneficiaries.
By definition, the hard-to-employ need special assistance to find and keep jobs. Their characteristics — disabilities, unstable behavioral health problems, very low skills, criminal records — place them at the back of the queue in a competitive labor market. To succeed, they may need special training, assistance in accessing health services or searching for a job, or other services.
There are at least four compelling reasons to invest in improving the employment prospects of those who face serious barriers to steady work. First, from a taxpayer’s perspective, it is costly to support individuals who, with assistance, could work. And, in fact, Americans have demonstrated that they are typically willing to spend more in the short run to increase self-sufficiency in the long run. Second, there may be benefits for society when hard-to-employ people are able to work steadily — for example, beneficial effects on public safety, family structure, and child well-being. Third, many believe that the retirement of the baby boom generation will produce tighter labor markets in the not-too-distant future, making it critical to take the best advantage possible of our nation’s human resources. Finally, many of the hard-to-employ very much want to work, and most Americans strongly believe that all individuals deserve the opportunity to make the most of their skills and ambitions.
What Is Known About the Hard-to-Employ?
The challenges faced by the hard-to-employ are clear, even if the strategies for addressing them are less understood. The barriers that prevent individuals from working can be grouped into three broad — and sometimes overlapping — categories:
-
Human capital deficits, including very low basic skills, limited English proficiency, and lack of work experience.
-
Health problems, including disabilities, behavioral health conditions (depression, substance abuse), and chronic physical health problems (hypertension, obesity) that can affect employability.
-
“Situational barriers,” a catch-all category that includes such problems as a lack of transportation and the need to care for a disabled dependent. One of the most important situational barriers, however, is a criminal record. Convicted felons are considered highly undesirable by employers and, in fact, are legally barred from many occupations in growing employment sectors.
Classifying barriers to employment in this way is useful, because different types of barriers require different kinds of services or supports. For example, disabilities may require workplace accommodations and special job search assistance, whereas a lack of work history may be overcome by providing work experience in a supportive setting. Individuals with health problems may need care management to ensure consistent and quality treatment.
In addition, when considering potential intervention strategies, one must take into account the public systems that interact with the hard-to-employ. For example, work-focused programs for individuals with disabilities must address the conflicting messages of a disability insurance system that makes “permanent disability” an eligibility requirement but is also trying to encourage more employment. Prisoner reentry programs must keep considerations of public safety paramount. And efforts to promote employment through public health systems may be hindered by a philosophy that favors treatment over work.
Despite the broad policy interest in serving the hard-to-employ, knowledge about effective program strategies is relatively undeveloped. Other than for welfare recipients and people with serious mental illness, there have been few rigorous experimental evaluations, and many questions remain unanswered. Here is a quick review of the best research on programs for the populations served in the Hard-to-Employ demonstration:
Long-Term Welfare Recipients
Mandatory welfare-to-work programs that include both job search assistance and short-term education or training activities appear to generate the largest impacts on employment and earnings. Strategies that combine mandatory employment services with earnings supplements have generated increases in both employment and income.2 In general, however, outcomes are much worse for the most disadvantaged welfare recipients, suggesting that other strategies, perhaps those that are more targeted and intensive, are needed to help those individuals who are hardest to employ.3 Newer approaches, not yet tested, range from models emphasizing work-focused strategies with special supports (transitional employment or versions of supported employment) to more treatment-focused services designed to address a particular barrier, typically a behavioral health problem.
Reentering Prisoners
Research has identified few examples of successful pre- or postrelease strategies for increasing employment or reducing recidivism among offenders.4 Many studies have found that in-prison vocational programs lead to lower recidivism, but their research designs are almost uniformly weak. Experts seem to agree that the most promising programs include some combination of pre- and postrelease services. Alternative sentencing is another promising strategy. There is some evidence that drug courts can reduce recidivism, although, again, most studies have used weak designs.5
Individuals with Behavioral Health Problems
Certain behavioral health problems (for example, depression, posttraumatic stress disorder, substance abuse, and domestic violence) are relatively common among low-income populations, particularly welfare recipients.6 Research suggests that current and former welfare recipients who have both physical and behavioral health problems are less likely to find and retain employment.
Although effective treatments for many psychiatric disorders do exist and have been documented in random assignment trials — consider the example of depression, for which both antidepressant medications and psychosocial treatments show promise7 — much less is known about whether employment outcomes improve as a result of successful treatment. For example, several random assignment studies of interventions to assess, recruit, and treat adults with undiagnosed depression have found that they significantly reduced depression in low-income minority populations.8 However, impacts on employment have been either short-lived or nonexistent.9 Programs that combine treatment for depression (such as intensive case management, assertive outreach, integrated behavioral health treatment, and use of specialty providers) with employment services have shown some promise, but there have been no rigorous evaluations.
Hard-to-Employ Parents and Their Children
Children and youth in hard-to-employ families face considerable risks to their cognitive and social development.10 Providing direct services to children whose parents are hard to employ may indirectly help to achieve employment outcomes for the parents. For example, preschool programs for young children have the dual benefit of taking care of the child care needs of working parents and benefiting children’s cognitive and socio-emotional functioning.11 Such intervention strategies could include both center- and home-based components, as well as a program aimed at increasing employment directly (since home- and center-based interventions for children do not always increase maternal employment).12
In short, past experience and research suggest that there is a lot to learn about which strategies are most effective in serving various hard-to-employ populations, how best to configure these strategies within programs, and how to ensure that programs for the hard-to-employ interact effectively with the systems that are already serving (or not serving) them. The Hard-to-Employ demonstration hopes to offer answers to these questions.
An Overview of the Hard-to-Employ Demonstration and Evaluation Project
In 2001, HHS selected MDRC as the prime contractor for the Enhanced Services for the Hard-to-Employ Demonstration and Evaluation Project, a nine-year study of selected programs designed to enhance employment, family functioning, and child well-being. The project was explicitly designed to build on previous research by rigorously testing a variety of innovative, policy-relevant interventions and creating an evidence base of best practices for programs.
Study Design
The evaluation includes three main components:
-
A study of the implementation of the programs, the services they deliver, and the operational challenges they encounter.
-
A study of the programs’ impacts; the key outcomes measured vary by site, but include employment, earnings, public benefits receipt, depression severity, criminal justice contacts, and others.
-
A study of the financial costs attributable to the programs.
The impact analysis uses an experimental, random assignment design, which is generally considered to be the most reliable way to assess the impact of social programs.
In a typical random assignment evaluation, individuals who are eligible for the program are assigned, at random, to a program group, which has access to the experimental program, or to a control group that is treated as though the new program did not exist. Members of the two groups are tracked during a follow-up period and are compared on a number of relevant outcomes.13 Because the design ensures that there are no systematic differences between the members of the two groups when they enter the study, any significant differences that emerge between the groups over time can be reliably attributed to the fact that one group was exposed to the experimental program and the other was not. Such differences are known as impacts, or effects, of the program. Although all of the Hard-to-Employ programs are being evaluated using random assignment, the design must be tailored to fit each individual project. In fact, because the projects are so diverse, the evaluation can be seen as four separate but related studies.
To measure program impacts, the MDRC team will use a combination of surveys and administrative records to track the research groups over time.
Phases of the Project and Key Components
In order to structure and prioritize site development work, MDRC and its expert consultants prepared a series of papers about the implications of different targeting strategies, models, program approaches, and best practices for the evaluation design.
After discussions with HHS, the MDRC team set out to recruit programs to participate in the study. Approximately 20 were considered, nine were selected for development, and four were able to implement random assignment. Most of the potential sites that did not go forward were too small to generate the sample sizes needed for a random assignment study. In other cases, funding or management issues caused interested programs to remove themselves from consideration.
Three of the four participating programs target discrete hard-to-employ populations —welfare recipients, reentering prisoners, and Medicaid recipients with depression — while the fourth is a two-generation project in Early Head Start programs:
-
Center for Employment Opportunities, New York City. Parolees are placed in paid transitional employment at one of several dozen work sites around the city for two to three months, followed by placement in unsubsidized jobs. The program also includes a fatherhood program, postplacement retention services, job coaching, and other supports.
-
Kansas and Missouri Early Head Start. Aimed at poor pregnant women and parents with children up to 4 years old, this “two-generation” intervention provides enhanced self-sufficiency services and skills training to parents, in addition to high-quality child care. The children in the program group are enrolled in Early Head Start services, and the parents receive assistance to identify and work toward self-sufficiency goals.
-
Test of alternative employment strategies for welfare recipients in Philadelphia. Parents who have received Temporary Assistance for Needy Families for at least one year or do not have a high school degree are referred to one of two programs offering different kinds of services: (1) the Transitional Work Corporation, which places participants in temporary paid jobs and provides a range of supports and job placement assistance, or (2) the Success Through Employment Preparation program, operated by Jewish Employment and Vocational Service, which focuses on identifying and treating participants’ employment barriers before they are placed in jobs.
-
Rhode Island Working toward Wellness project. Working-age adults who have children, are on Medicaid, and are experiencing serious depression receive intensive telephonic outreach and follow-up from managed care case managers to encourage their participation in mental health treatment. The program also provides access to employment services.
Some of the participating programs — notably, the Transitional Work Corporation and the Center for Employment Opportunities — have extensive experience operating the model that is being tested. In these sites, the MDRC team worked with program staff to tailor the research design to local conditions. In sites that were starting new programs or adding significant new components to an existing model, the MDRC team also provided extensive technical assistance to develop and refine the model.
Random assignment of study participants began at a different time in each site. Approximately six months after random assignment began, MDRC visited each site to assess whether the test had been implemented as designed, to identify program challenges, and to develop recommendations to strengthen implementation.
Preliminary results from the impact analysis are expected to be available for CEO in late 2007; results from the other sites are expected in 2008 and 2009.
The Focus of This Report
The next four chapters focus on the four Hard-to-Employ programs, in each case describing the details of the program’s strategy, the evaluation’s design, and the characteristics of the program participants and the control group members. The results of the early assessments are used to describe particular challenges the sites encountered and how they are addressing them.
1 The Annie E. Casey Foundation and the W. T. Grant Foundation are providing funding for the 18-month follow-up survey to study how the model being tested in Rhode Island affects children. (back)
2 Bloom and Michalopoulos (2001). (back)
3 Bloom and Michalopoulos (2001). (back)
6 Danziger, Corcoran, and Danziger (2000). (back)
7 American Psychiatric Association (2000). (back)
8 Miranda et al. (2006). (back)
10 Duncan, Brooks-Gunn, and Klebanov (1994). (back)
11 National Institute of Child Health and Human Development Early Child Care Research Network (2000); Fantuzzo, Bulotsky-Shearer, Fusco, and McWayne (2005). (back)
12 Yoshikawa (1994); U.S. Department of Health and Human Services (2002); Werner and Smith (1992). (back)
13 The comparisons include everyone assigned to the two groups, including sample members who do not actually participate in the experimental program. (back)
| Table of Contents | Previous | Next |