Executive Summary

disability; however, no data addressed the question of why some never applied.

Answering the third question, GAO found most VR clients received only
modest services: Less than half received any type of education or training
services, the total value of purchased services averaged only $1,573 per
client, and just under half received purchased services costing less than
$500. The appropriateness of services for each individual is the key
question; RSA data did not allow GAO to reach conclusions, but GAO did find
that states purchased more services for clients with physical than with
mental disabilities, more for clients with severe than with non-severe
disabilities, and more for white clients than for black, Hispanic, or
American Indian clients.

On the fourth question, GAO concluded that evidence on VR results was
mixed. In contrast to the short-term gains typically reported by the
program, GAO's evaluation of long-term outcomes found that rehabilitants'
gains in employment and earnings from time of referral to their
case-closure year of 1980 faded after about 2 years. The fraction working
shrank steadily. By 1988, the last year examined, 61 to 66 percent of
rehabilitants (depending on type of disability) had some earnings;
however, this was either no better than or below the pre-program level
(depending on type of disability), and only a third had worked
continuously since 1980. Conversely, rehabilitants did do better than
dropouts on all measures of work and earnings, even after statistical
analyses controlled for some pre-program differences between the groups.
Extensive VR efforts were not uniformly effective, however, as shown by
the GAO finding that the group of clients who received significant services
but were not rehabilitated (21 to 36 percent of those served, depending on
type of disability) did no better in later employment and earnings than
dropouts who never got any services after the initial evaluation.

Principal Findings

Who Is Potentially Eligible and Who Accepted?

The 14 to 18 million figure of those potentially eligible represented an
upper limit of the population in the mid-1980's, since the data allowed
estimation of those eligible on only two of the three statutory criteria.
(Employability cannot be determined using extant data.) Now, however,
this figure is a more appropriate estimate for the future since the

Executive Summary

What Services Are
Provided?

What Is Their Effect on
Employment and
Earnings?

modification of the employability criterion in the Rehabilitation
Amendments of 1992. About 65 percent of those served in 1988 had severe
disabilities, which is comparable to the 69 percent of the national
work-disabled population who have severe disabilities. Applicants,
however, were much less likely to be older (over age 45), to be female, or
to have disabilities such as orthopedic impairments or chronic health
conditions, than were persons in the national work-disabled population.
These findings raise questions about why some disabled-population
subgroups may not have sought VR services.

Beyond the two initial services (diagnosis and evaluation, and counseling), about half of VR clients also received some type of skill-enhancing service, such as education or training. Smaller percentages of clients received other services targeted on difficulties associated with their specific disabilities. No data were available to allow GAO to evaluate whether the disparities in purchased services noted previously were appropriate.

By RSA's definition, about 60 to 70 percent of clients accepted for services were rehabilitated—that is, they completed the planned services and then held a job for at least 60 days. Some clients held a job before they were referred for VR, of course, but more worked for wages immediately after closure than before (from 8 to 18 percentage points more, depending on the type of disability). Although this wage-earning group shrank in subsequent years, average earnings did rise, and rehabilitants continued to do better than dropouts.

GAO'S statistical analyses to control as much as possible for prior
differences showed statistically significant positive effects for
rehabilitants, when compared with dropouts, with rehabilitants more
likely to be employed and have higher earnings at the 5-year point across
all three disability groups. Specifically, rehabilitated clients with physical
disabilities were 12 percentage points more likely to be employed and
earned about $2,000 more per year; rehabilitants with emotional
disabilities were 15 percentage points more likely to be employed and
earned about $1,600 more; and those with mental retardation were
19 percentage points more likely to be employed and earned about $1,000

more.

In contrast, clients who were not rehabilitated had long-term economic outcomes very similar to those for clients who dropped out. This raises

Executive Summary

questions about the program's impact because this group, on average, remained in the program for as long as rehabilitated clients and received up to two thirds of the VR agency-purchased services received by a fully rehabilitated client.

Recommendations

GAO recommends that the Commissioner of RSA begin the review, authorized in the 1992 amendments to the Rehabilitation Act, of the adequacy of existing VR data for various users, with particular emphasis on measures of the VR referral process and of the cost, intensity, and frequency of services. In addition, GAO recommends that RSA determine why disparities exist in the cost of purchased services for clients of different races. To better evaluate the economic impact of the VR program, RSA should continue its commitment to a longitudinal study of the VR program, and the Secretary of Education and the Secretary of Health and Human Services should negotiate an agreement to produce updated computer matches of client and earnings data. Finally, to explore the broader issues of who can be served, at what intensity, and with what results, GAO recommends that the Secretary of Education take steps to establish the National Commission on Rehabilitation Services authorized by the 1992 amendments. The Commission can review GAO's findings and other up-to-date information on VR outcomes in order to derive recommendations for the future direction of the program, particularly for the next reauthorization.

Agency Comments

Responsible officials of the Department of Education provided oral comments on the findings, conclusions, and recommendations in this report. Although they raised a number of issues about GAO's analysis, in general they agreed with GAO's recommendations.