Introduction Our examination of the results of the VR program is the central contribution of our report. Ours is the first study to provide national data on the long-term economic outcomes for clients who participated in the VR program following the passage of the Rehabilitation Act of 1973. There are several strengths to our approach. By using the combined RSA-SSA client and wage data, we could overcome two of the long-standing shortcomings of evaluation in the field: We could study clients over a longer term, and we could compare different client groups. In addition although not all kinds of income are included, the SSA data on earnings, which are based on employer reports, may be more accurate than data reported by clients themselves. Further, the very large size of the 1980 case-closure group in the RSA-SSA data link allowed us to look at long-term outcomes for those with different kinds of disabilities, as well as to include a number of variables of interest in our statistical analyses. There are, however, a number of limitations in our design. First, as noted previously, there are limits, imposed both by the data and by our resources, on our search for the size of the program's effects. Better data would permit better statistical controls for pre-program differences, but more complex statistical models than we could explore might possibly be helpful in this regard even with the present data. Second, we can do little to explain any outcomes-that is, to say why clients work as much as they do or earn the wages they do, particularly whether these are related to details of the VR services they received. Service data, though available in RSA's Case Service Reports (with many limitations already noted in discussing our chapter 3 analyses), were not included in the RSA-SSA data link. And no other data are available about clients' work histories that could help explain the employment and earnings figures-for example, what jobs they held or whether they worked full- or part-time—after program closure. Agency Comments Organization of the We discussed our preliminary findings and conclusions with responsible officials of the Department of Education while we were preparing our 1991 testimony, and we incorporated a number of their comments and concerns in our ongoing analysis. These officials also provided oral comments on our final findings, conclusions, and recommendations. These comments are presented, with GAO's response, in appendix I. Chapter 2 uses data from published reports to answer the first evaluation question about persons with disabilities who might be eligible for VR. Introduction Chapters 3 and 4 present our analysis of VR program data on those served in fiscal year 1988. Chapter 3 answers the second evaluation question about who is served and compares those accepted for VR services with those who were not in terms of demographic characteristics and type and severity of disability. Chapter 4 again uses RSA data on the 1988 client group to answer the third question about services received. The final chapter answers the fourth evaluation question concerning program results, using data on work and wages through 1988 for various groups of rehabilitated and other clients whose cases were closed in 1980. Chapter 2 The Population Potentially Eligible for The first evaluation question asked us to determine the size and nature of the population of people with disabilities who are eligible for vocational rehabilitation (VR) services. Not all persons with disabilities are eligible for services under the Rehabilitation Act, since the law also requires that a person's disability present a substantial impediment to employment and that the individual can benefit from VR services in terms of employability.1 The judgmental nature of these criteria makes it quite difficult to arrive at national estimates of those whom VR could be serving. To answer the question, we relied on published reports from two national surveys from the mid-1980's. Based on the data in these reports, we estimated the number of persons meeting the following two (of three) VR criteria: (1) presence of a disabling condition that (2) limits the amount and kind of work the person can do. From our review of these national survey data, we estimated that about 5 to 7 percent of the working-age population with work-limiting disabilities is served in a year by the VR program. We also found that about 69 percent of the general work-disabled population is severely disabled, which is slightly higher than the percentage of severely disabled clients (65 percent) among those served by the VR program in fiscal year 1988. In the first section of this chapter, we present estimates of the size of the work-disabled population. In the next three sections, we examine the demographic characteristics of this population, the type of conditions reported as causing work disability, and the severity of the disabling conditions. How Many Are The two surveys we examined-the 1983-85 National Health Interview The 14.3 to 18-million figure is an imprecise estimate of those eligible, however, for at least three reasons. First, eligibility is based on medical 'As noted in chapter 1, the employability criterion was modified substantially in the 1992 2Differences in the survey estimates can result from, among other things, differences in the questions' wording, the order of questions, or the time period queried. The Population Potentially Eligible for Who Are Those evidence and counselors' judgment, which will differ to an unknown degree from applicants' own views; and the surveys' self-reports of disability capture only the latter. Second, eligibility depends on a third judgment, of employability after services are received, which cannot be simulated using the survey data. Thus, there is no way to estimate whether all applicants should be viewed as some advocates did for many years and which the law does beginning with the 1992 reauthorization—as employable. Third, some people outside the 18 to 64 age group are also eligible, as are some of those beyond the household population represented by the survey data. Whatever the number eligible, it is not a definitive guide to the potential demand for VR services for an additional reason: Not all will want to enter the labor market. (Although there are other benefits and outcomes from VR services, they are primarily aimed at helping clients achieve jobs in the competitive labor market.) Table 2.1 shows that those reporting a work disability were older, less educated, and poorer than the general working-age population. The work-disabled population also differed from the working-age population, although less markedly, in race and gender. Also, in terms of education and income, persons who reported having work limitations differed considerably from the working-age population as a whole. In both surveys, about one fifth of the work-disabled group had some formal education beyond high school, compared with about two fifths of the overall working-age population. In addition, 18 percent of the work-disabled group had family incomes that fell below the poverty level, compared with 10 percent of the working-age population. |