National trends in vocational rehabilitation: a comparison of individuals with physical disabilities and individuals with psychiatric disabilities.

Each year, the state-federal rehabilitation system supports vocational rehabilitation (VR) services for hundreds of thousands of individuals with disabilities. During the years 1977-1982, there was a substantial decrease in the purchasing power of rehabilitation expenditures, after adjustment for inflation; in the 1983 and 1984 there were modest increases, but in the latter year inflation-adjusted expenditures were still 28.3% below 1977 levels (Mars, unpublished tabulations).

Data on individuals with disabilities in the state-federal VR system are maintained by the federal Rehabilitation Services Administration (RSA). A recent analysis of these data indicated that total case closures, an index of volume of applicants and clients in the system, have declined markedly since 1977 (RSA, 1988 information memorandum). This decrease appears to be a function of at least two factors. The first is the effect of the decrease in purchasing power of VR budget allocations. The second factor is the long-term impact of the Rehabilitation Act of 1973 and its amendments, which mandated that VR services be made more available to individuals with severe disabilities. Individuals with severe disabilities require, on average, more costly services than individuals with non-severe disabilities (Kent & McLaughlin, 1984); time spent by rehabilitation counselors in evaluation and counselling is also likely to be greater for clients with severe disabilities. Thus the total number who can be served with a given budget allocation decreases as the percentage of individuals with severe disabilities in the system increases.

While these two factors affected the system as a whole, previous research indicated that groups composed of individuals with specific disabilities or demographic characteristics have had very different experiences in the state-federal system over time. For example, among individuals with multiple sclerosis, there was a 17.3% increase during the period 1977-1984 in the number of applicants for rehabilitation services in the state-federal system, and striking 62.4% increase in the number who achieved employment in the competitive labor market after receiving VR services (Kallos, Genevie, Andrews & Struening, 1989). These increases for individuals with multiple sclerosis were much greater than for individuals with any of the other physical or psychiatric disabilities examined. The authors attributed these finding to a 1978 cooperative agreement between the National Multiple Sclerosis Society, RSA, and the Council of State Administrators of Vocational Rehabilitation, in which a commitment was made to improve availability of VR services to individuals with this particular disability. On the other hand, in a retrospective study of gender equity in vocational rehabilitation (Menz, Hansen, Smith, Brown, Ford & McCrowey, 1989), no improvement in either access to rehabilitation services or in rehabilitation outcomes among women found, despite legislative directives barring discrimination on the basis of sex.

Information on changes over time in the relative success of disability-defined subgroups in achieving vocational rehabilitation is essential for administrators, planners, policy makers and advocacy groups. In the research described in this paper, we examined the record of the state-federal VR system in serving individuals with psychiatric disabilities and individuals with physical disabilities during the period 1977-1984. Within each of these disability categories, we compared subgroups defined on the basis of severity of disability. By deriving indicators at key points in the vocational rehabilitation process, we also sought to identity factors inside and outside the VR system which may have brought about the observed trends.

Methods

Sample Selection

Information is collected by the state agencies on a standard reporting form (R300/911) for each individual who is referred for services in the state-federal system. At the end of the fiscal year, data on each case closed during that year are forwarded to RSA by each state agency. Currently, the total number of closures is greater than 500,000 per year; in the late 1970's, total closures exceeded one million per year.

Data tapes were requested from RSA for all cases closed in fiscal years 1977 through 1984. To facilitate data processing and reduce costs in analyses involving several years of data, a 10% uniform sample, stratified by year of closure and state agency, was created. The sample consists of 545,746 closures during the period 1977-1984. Those individuals whose cases were closed from referral status (code 00) were eliminated prior to sample selection for the following reasons: a) those closed at referral are often |paper referrals' and never have direct contact with state agency staff; b) many agencies do not report information on a referred individual until he or she actually applies for services; c) a medical review of the individual's records, resulting in assignment of a |major disability' category does not occur until a formal application for services is made; the nature of the disability for those closed at referral is indicated only by less reliable data item, |disability as reported.' Thus, with the exceptions indicated below, the population of interest consisted of all individuals closed from applicant status (code 02), extended evaluation (code 06), and all those accepted for services who were either rehabilitated (code 26) or not rehabilitated (codes 28 and 30).

Two disability categories, psychiatric and physical, were defined, based on the |major disability' code present in R300/911 data. All individuals coded as |psychotic' por |psychoneurotic' were selected (N=76,701). Those with a major disability code of 'alcoholism' (N=30,532), |drug addiction' (N=8,633) or |other character, personality or behavior disorder' (N=40,826), well as |mental retardation' (N=53,532) were excluded from the analysis. Those in the |other character, personality or behavior disorder' category were eliminated in the advice of RSA staff, because this category is poorly defined and because there are indications that many individuals assigned to this category are not truly physchiatrically disabled. All other closures were assigned to the physical disability category (N=333,651). These include a wide range of disabilities, such as persons who were disabled due to blindness, deafness, orthopedic impairment, malignant neoplasm, polio or heart disease. In addition to major disability, which is the basis for the above criteria, the data system also contains information on secondary disability, where one exists. No selection criteria in the current research involved secondary disability.

Variables

The indicators selected for this study were derived from two R300/911 variables: closure code and work status at closure. The closure code indicates the stage of the rehabilitation process at which an individual was closed from the active caseload status. For purpose of the analyses performed for this study, the relevant closure categories are: (a) Closed at application (code 02), that is, not accepted to receive VR services; (b) Closed, accepted to receive services, but without achieving rehabilitation (including both those who actually received services [code 28] and those who were closed after acceptance but before starting a rehabilitation program [code 30]); (c) Closed, rehabilitated after successfully completing a rehabilitation, a minimum of 60 days of employment (code 26).

For those who meet the criterion for rehabilitation, the variable work status at closure indicates the type of employment achieved. For the purpose of the current study, each rehabilitant was assigned to one of two categories: (a) Competitive labor market, involving work for wages, salary or commission; (b) All other employment outcomes, including sheltered workshops conducted by a nonprofit organization which provides work under special conditions for individuals who are physically or mentally disabled, state-agency-managed business enterprises, keeping house (|homemaker' status) and working without pay on a family farm or in a family business.

One set of outcome indicators in this study consisted of the percent change in the number of closures in each category. In tracking change in the number of individuals who apply for services, who are accepted and who are eventually rehabilitated, the first year for which data were available to this study, 1977, was used as a baseline. Change was expressed as a percentage: the number of closures in that category in a given year, relative to the number in 1977. A positive percentage indicates an increase in the category compared to 1977, a negative value represents a decrease.

In addition to changes in total numbers of individuals in certain closures categories, three outcome rates were defined:

Rate of acceptance: the number of individuals closed after having been accepted for services divided by the total number of individuals closed after having applied for services. The denominator includes all individuals in the database for a given year. The numerator includes only those individuals from the denominator who were accepted for services.

Rehabilitation rate: the total number of individuals closed after achieving vocational rehabilitation divided by the total number closed after acceptance for services.

Competitive employment rate: the total number of individuals closed as rehabilitated who achieved competitive employment divided by the total number closed as rehabilitated.

Note that each of the above rates is computationally independent of the other two. Thus a change in the rate of acceptance for a particular subgroup will have no mathematical impact on the rate of rehabilitation or the competitive employment rate for that subgroup. By deriving these independent indicators of the system's performance, it is possible to pinpoint the stage or stages in the vocational rehabilitation process at which changes have occurred over time.

In examining changes over time in the outcome indicators an important consideration is that an individual can remain on the active caseload of an agency for more than one year. Information about this individual becomes part of the R300/911 database only in the fiscal year in which his or her case is closed. Thus the acceptance rate for a given year is not the percent accepted for services among those who applied in that year; rather, it is the number of individuals whose cases were closed in that year at some point after being accepted for services, among all those whose cases were closed in that year. In this paper, the term 'total applications' in a particular year will be used as shorthand for |total applicants from the current and previous years whose cases closed in this year.'

In addition to type of disability (psychiatric versus physical, as described above), closure code and employment status, the sample was across-classified on the basis of severity of disability. An individual is categorized as severely disable in the R300/911 reporting system if any one of the following conditions is met (RSA, 1974): 1. The client is assigned to any one of approximately 30 disablility

categories which are thought to indicate severe disability.

Examples are multiple sclerosis and accidents and injuries

involving the spinal cord. For some disability categories,

certain additional criteria must be met. For example, a

psychotic disorder is considered severe if the individual currently

requires institutional care in a mental hospital or psychiatric

ward of a general hospital or has a history of being

institutionalized for treatment for three months or more. 2. The client at any time in the VR process, has been a recipient

of Social Security Disability Insurance (SSDI) or a recipient

of Supplemental Security Income (SSI) by reason of

blindness or disability. 3. There is documented evidence of loss and limitation meeting

certain physical or behavioral criteria (Functional Limitation

Factors) and the individual requires multiple services

over an extended period of time. Examples are individuals

who are unable to make use of public bus or train service,

individuals who are unable to climb one flight of stairs or

walk 100 yards on the level without pause, those who have

a poor work history, and individuals who exhibit inapropriate

social behavior.

Differences between key rates for a given subgroup at different points in time and differences between subgroups in percent change in the number of individuals in certain closure categories over time were evaluated statistically using methods described by Fleiss (1981) for determining the significance of a difference between two independent proportions. The year 1977 is the baseline from which all change is assessed; 1977 baseline values for all key indicators used in this study shown in Table 1. [TABULAR DATA OMITTED]

Results

Changes in inflation-adjusted total expenditures for vocational rehabilitation during the period 1977-1984 are displayed in Figure 1. Adjustments were based on the Consumer Price Index, using 1967 as the basis year. There was a decline in expenditures in each year during the period 1977-1982. The slope of the line in the figure indicates that the declines were greatest in 1980, 1981 and 1982. In each of these years, total expenditures with respect to 1977 levels dropped by more than 10%. The first year since 1977 in which total expenditures increased with respect to the previous year was 1983, and another increase occurred in 1984. However, total expenditures in 1984 were still substantially below 1977 levels.

Rates of acceptance, rehabilitation and competitive employment for the entire eight-year period are presented in Table 2. The rate of acceptance during this period was nearly identical for individuals with severe physical disabilities (66%) and individuals with severe psychiatric physical disabilities (65%), but for those not severely disabled, the acceptance rate was more than 10% higher for individuals with psychiatric disabilities (55%) than for individuals with physical disabilities (45%). Rehabilitation rates were much higher for individuals with physical disabilities than for individuals with psychiatric disabilities, both among individuals with severe and non-severe disabilities. Those not severely disabled had higher rehabilitation rates than those who were severely disabled. Individuals with non-severe disabilities also achieved much higher rates competitive employment than individuals with severe disabilities. However, individuals with psychiatric disabilities, despite a lower rehabilitation rate than individuals with physical disabilities, had a higher rate of competitive employment. Among those who were closed rehabilitated, nearly 90% of individuals with non-severe psychiatric disabilities and more than 75% of individuals with severe psychiatric disabilities achieved competitive employment. [TABULAR DATA OMITTED]

Among applicants, there was a statistically significant increase in the percentage of individuals with severe disabilities over the period 1977 to 1984, among both individuals with physical disabilities (p. 01) and among individuals with psychiatric disabilities (p .01); the year-to-year changes are illustrated in Figure 2. The degree of severity among individuals with psychiatric disabilities has always been greater than among individuals with physical disabilities, but this disparity has decreased somewhat over time. The percentage increase in severe disability for the period 1977-1984 has been greater for individuals " with physical disabilities (8%) than for individuals with psychiatric disabilities (5%). By 1984, 67% of individuals with psychiatric disabilities and 48% of individuals with physical disabilities were categorized severely disabled.

After an increase from 1977 to 1978, the total number of cases closed per year in the VR system declined from 571,820 in 1978 to 432,260 in 1984, a decrease of 24%. The decline in this indicator, which reflects the number of applicants for rehabilitation services, holds both for individuals with psychiatric disabilities (a 27% decline) and physical disabilities (a 24% decline). Within each of these disability groups, the drop in applicants was much greater among those who were not severely disabled. Figure 3 shows the change over time, relative to 1977 figures, for each of the four disability-severity groups. Note that while there was a decrease of only 3% from 1977-1984 in the number of applicants who were severely physically disabled, the number of applicants who were severely psychiatrically disabled dropped 18% during this period. The difference in percent decrease between these two groups is statistically significant (p .01). The decrease for applicants who were not severely disabled was also greater among individuals with psychiatric disabilities (35%) than among individuals with physical disabilities (30%); this difference is also statistically significant (p .01).

The acceptance rate for each subgroup is shown in Figure 4. Between the years 1977 and 1984, the ccceptance rate increased significantly for both individuals with severe psychiatric disabilities (p .05) and individuals with severe physical disabilities (p .01), and has declined significantly for both individuals with non-severe psychiatric disabilities (p .01) and individuals with non-severe physical disabilities (p .01). Among individuals who were severely disabled, the increase in acceptance rate was greater among individuals with physical disabilities than among individuals with psychiatric disabilities: in 1977, 66% of individuals with psychiatric disabilities were accepted, as opposed to 64% of individuals with physical disabilities. By 1984, a higher percentage of individuals with physical disabilities (69%) than of individuals with psychiatric disabilities (68%) was being accepted. Among those not severely disabled, the rate of decline was comparable for individuals with physical disabilities and individuals with psychiatric disabilities, but at any given point in time during the study period, individuals with psychiatric disabilities were much more likely than individuals with physical disabilities to be accepted for services. In 1984, 53% of individuals with non-severe psychiatric disabilities, but only 42% of individuals with non-severe physical disabilities were accepted for services.

Yearly rehabilitation rates for the four groups are shown in Figure 5. For individuals with severe psychiatric disabilities, individuals with severe physical disabilities and individuals with non-severe psychiatric disabilities, rehabilitation rates remained fairly constant over time; there is no statistically significant difference between the 1977 and the 1984 rates for any of these groups. However, for individuals with non-severe physical disabilities, the rehabilitation rate declined significantly from 76% in 1977 to 70% in 1984 (p .01). Despite this decline, individuals with non-severe physical disabilities still had the highest rehabilitation rate of any of the groups in 1984. Individuals with severe psychiatric disabilities had the lowest rehabilitation rate in 1984, roughly 50% in each year.

The yearly changes in number of individuals rehabilitated in each subgroup with respect to the baseline year, 1977, are shown in Figure 6. For individuals with severe physical disabilities, there was a gain of 4% in the number of rehabilitants in 1984 with respect to 1977. For individuals with severe psychiatric disabilities, there was a 14% decrease in number of rehabilitants. This difference between percent change for individuals with severe physical disabilities and individuals with severe psychiatric disabilities is statistically significant (p .01). Individuals with non-severe disabilities fared much worse than either group of individuals with severe disabilities: between 1977 and 1984 there was a 36% drop for individuals with psychiatric disabilities and a 42% drop for individuals with physical disabilities in terms of number of rehabilitants. Note that for both groups of individuals with severe disabilities, there were increases in the number of rehabilitants in 1978 and again in 1979 which preceded a period of decline; for the non-severely disabled, there was little or no increase in the late 1970's, and the decline in number rehabilitated began sooner and was more precipitous (see Figure 6).

Competitive employment rates among those who were rehabilitated are shown in Figure 7. For all groups, except individuals with non-severe psychiatric disabilities, there was a statistically significant increase in this indicator over the study period (p .01 for each of the three groups for which there was a significant effect). However, individuals with non-severe psychiatric disabilities had the highest absolute rate of rehabilitation to competitive employment in each year. Thus the increases experienced by the other groups have closed the gap between them and individuals with non-severe psychiatric disabilities with respect to this index. In 1984, individuals with non-severe physical disabilities had nearly the same rate of competitive employment (88%) as individuals with non-severe psychiatric disabilities (89%). Next were individuals with severe psychiatric disabilities (78%) followed by individuals with severe physical disabilities (71%).

With respect to percent change in the number who entered or returned to competitive employment each year (Figure 8), individuals with severe physical disabilities experienced a substantial increase over the period 1977 - 1984 (19.9%), individuals with severe psychiatric disabilities experienced a small decrease (3.4%) and there were large decreases for individuals with non-severe psychiatric disabilities (33%) and individuals with non-severe psychiatric disabilities (36.5%). The difference between individuals with severe physical and severe psychiatric disabilities is statistically significant (p .01); the difference between the two groups of individuals with non-severe disabilities is not significant.

The total number of rehabilitants produced by the state-federal VR system and any change which occurs in this indicator from year to year (Figure 6) is a function of three factors measured in this study: the number of applicants (Figure 3), the acceptance rate (Figure 4), and the rehabilitation rate (Figure 5). The number of individuals who return to competitive employment and any change which occurs in this indicator from year to year (Figure 8) is a function of the same factors as well as of the rate of competitive employment among rehabilitants (Figure 7). To compare the cumulative effects of these factors we subtracted the percent change between 1977 and 1984 in number of applicants among individuals with psychiatric disabilities from the percent change in number of applicants among individuals with physical disabilities. This process was repeated for percent change in total number of rehabilitants and percent change in total number returned to competitive employment; separate calculations were performed for individuals with severe and non-severe disabilities. The results, shown in Figure 9, indicate that among those who were severely disabled, the disparity between individuals who were physically disabled and individuals with psychiatric disabilities in terms of percent change in total rehabilitants was greater than the disparity in terms of percent change in total applicants. Since rehabilitation rates for the two groups were stable over the period 1977-1984 (Figure 5), the increased discrepancy must be due to the fact that improvement in the acceptance rate was not as great for individuals with severe psychiatric disabilities (2%) as for individuals with severe physical disabilities (5%; see Figure 4). Among individuals with non-severe disabilities, there is a reversal (Figure 9): the percent decrease in applicants with psychiatric disabilities between 1977 and 1984 was greater than for applicants with physical disabilities but the decrease in number of rehabilitants was greater for individuals with physical disabilities. Since the decline in acceptance rates was comparable for the two groups (Figure 4) the reversal must be due to the fact that the rehabilitation rate of individuals with non-severe psychiatric disabilities remained stable over the study period while the rate for individuals with non-severe physical disabilities declined (Figure 5).

With respect to number of individuals entering competitive employment, the discrepancy in percent change over time between individuals with severe physical disabilities and individuals with severe psychiatric disabilities is even greater than the discrepancy in percent change for number of rehabilitants (Figure 9). This is because the improvement in competitive employment rate was greater for individuals with severe physical disabilities than for individuals with severe psychiatric disabilities (Figure 7). Among individuals with non-severe disabilities, the competitive employment rate remained relatively constant for individuals with psychiatric disabilities but improved slightly for individuals with physical disabilities (Figure 7). Thus the difference favoring individuals with non-severe psychiatric disabilities as opposed to non-severe physical disabilities is somewhat smaller with respect to competitive employment than it is with respect to total number rehabilitated (Figure 9).

Discussion

Trends in Service Delivery to Individuals with Severe Disabilities

Our findings indicate that during the period 1977-1984, the percentage of individuals with severe disabilities in the state-federal VR system rose dramatically, that the acceptance rate for individuals with severe disabilities increased, and that among those rehabilitated, the percentage of individuals with severe disabilities who achieved competitive employment increased markedly. These trends characterized both individuals with severe psychiatric disabilities and individuals with severe physical disabilites. Among individuals with non-severe disabilities, there were either marked declines or much smaller improvements on each of these indicators.

The increased percentage of individuals with severe disabilities in the state-federal rehabilitation system over the period 1977-1984 (Figure 2) is most likely due to the long-term consequences of the mandate to serve individuals with severe disabilities contained in the Rehabilitation Act of 1973. However, a skeptic might wonder whether compliance with the mandate is genuine, or whether counselors are simply assigning the |severe' designation to more applicants in order to comply with the letter of the law. While there is no direct evidence available on possible shifts in counselor classification patterns, there is an indication in the data that suggests such shifts were not widespread. Rehabilitation rates of individuals without severe disabilities are markedly higher than for individuals who are severely disabled (Table 2). If over time, an increasing proportion of those termed severely disabled did not truly meet the RSA criteria for severity, then the rehabilitation rates of individuals with |severe' disabilities in our analyses should have improved over time, reflecting the presence in this group of an increasing proportion of less severely disabled individuals. As shown in Figure 5, this was not the case: rehabilitation rates for individuals with severe disabilites remained relatively stable during the study period. These results support the position that counselors have been applying the criteria for severity consistently and that there has been a genuine increase in the percentage of individuals with severe disabilities in the state-federal VR system.

Disparate Trends and Possible Causes

The 1973 Rehabilitation Act specifies an "Order of Selection" of clients to be served, ensuring that individuals who have severe disabilities will be given first priority, whenever all eligible individuals who apply cannot be served. The definition of severe disability in the "Order of Selection" specifically includes mental as well as physical disabilities that cause substantial functional limitation. To our knowledge there is no legislative mandate in the Rehabilitation Act of 1973 or its amendments which assigns a higher priority to individuals with severe physical disabilities than to individuals with severe psychiatric disabilities. Despite the absence of such a formal mandate, our results indicate that both in terms of percent change in total applications, percent change in total individuals rehabilitated and percent change in total number who return to competitive employment, individuals with severe psychiatric disabilities fared much worse in the state-federal system during the period 1977-1984 than did individuals with severe physical disabilities (Figure 9).

The most striking disparity is in percent change in yearly application (Figure 3). The decline between 1977 and 1984 was much greater among individuals with severe psychiatric disabilities than among individuals with severe physical disabilities. One possible explanation is that the incidence and/or prevalence of severe mental illness is decreasing over time, but there is no evidence to support this idea (Surles, 1986). Another possible explanation is that during the study peiod, VR sevices for individuals with psychiatric disabilities became more available outside the state-federal system. A direct test of this hypothesis would require"a nationwide retrospective survey of all providers of vocational rehabilitation sevices. However, if there had been a substantial increase in availability of alternative vocational rehabilitation programs for individuals with psychiatric disabilities, the decline in applicants in the state-federal system with psychiatric as opposed to physical disabilities should be comparable among individuals with severe and non-severe disabilities. Instead, our analysis indicates that the relative decline is much greater for individuals with severe psychiatric disabilities (Figure 9). Therefore, the factors causing a disparity between individuals with physical disabilities and individuals with psychiatric disabilities appear to be specific to those whose disabilities are severe.

Yet another possible explanation for the marked decline in number of applications by individuals with severe psychiatric disabilities during the study period is that the characteristics and needs of individuals with severe mental illness have been changing over time in such a way that for an increasing number of these persons, traditional VR services are no longer appropriate. In the past decade, a number of subgroups of individuals characterized by serious mental illness have emerged and gained national attention: individuals who are homeless (Baxter & Hopper, 1985), individuals who are young, uninstitutionalized and develop severe and persistent mental disabilities, often before entering or becoming established in the work force (Pepper, Ryglewicz & Kirshner, 1982), and individuals with a |dual diagnosis' - substance abuse in addition to a psychiatric disability (Brown, Ridgely, Pepper, Levine and Ryglewicz, 1989). Individuals in each of these groups require a broad range pof educational, medical and social services. The mix of services currently available in the state-federal VR system and collaborating local, state and federal agencies, may not have kept pace with the changing needs and characteristics of the population of individuals who are seriously mentally ill. Additional research would be necessary to establish whether the gap between needs and available services has in fact increased over time, and whether this factor is related to the sharp decline in the total number of applications for VR sevices by individuals with severe psychiatric disabilties.

Another possible explanation of this decline is that there has been a drop-off in referrals to the state-federal VR system by state psychiatric inpatient facilities. VR units have been housed in many of these facilities since the 1950s as part of cooperative agreements between state-level mental health systems and the state-federal VR system (Barker, 1988). Because the inpatient population has been declining at these large state facilities as part of the shift to cummunity-based outpatient care, referrals to VR from large state facilities have no doubt decreased. We know of no data that would indicate whether the community-based mental health centers which are now providing the bulk of treatment for individuals with serious mental illness have developed patterns of referral to the VR system that would compensate for the loss of referrals from the large state inpatient facilities.

The shift toward community-based outpatient care of individuals with severe mentally illness may have reduced the number of applicants with severe psychiatric disabilities, for another reason as well. For individuals with psychiatric disabilities, one set of criteria for severity involves current or frequent hospitalization for mental illness (see Methods). It follows that as the tendency to care for those with serious mental illness on an outpatient rather than an inpatient basis increases, the proportion of individuals with psychiatric disabilities who can meet the hospitalization criteria for severe disability in the state-federal VR system decreases. Interestingly, hospitalization is not a criterion for severe disability among individuals with physical disabilities.

Another criterion for severity which may have become more difficult to meet for individuals with a psychiatric disability involves beneficiary status for SSI of SSDI (see Methods). Because individuals with psychiatric disabilities were particularly hard hit by the widespread termination and deinal of SSI/SSDI benefits which occurred in the early 1980's (Goldman & Gattozzi, 1988), it is likely that many fewer aplicants for VR services who had psychiatric disabilities during that period were eligible for SSI or SSDI benefits at the time of application. Since eligibility for SSI or SSDI results in an automatic designation of 'severe disability' in the VR system, the disproportionate loss of eligibility by individuals with psychiatric disabilities in the Social Security system may have had the unexpected consequence of reducing the likelihood that an individual with a psychiatric disability would be categorized as severely disabled in the VR system. And as documented in this paper, the likelihood of obtaining access to services in the state-federal VR system is much greater for an individual categorized severely disabled.

Another set of factors to consider in explaining the marked reduction in applications for VR services by individuals with severe psychiatric disabilities involves possible changes in the deployment and evaluation of rehabilitation counselors in the state-federal system. There is general agreement that persons with severe psychiatric disabilities are best served by counselors with specialized training and caseloads (Tashjian, Hayward, Stoddard & Kraus, 1989). However, there are no readily available data indicating how these factors have changed over time. Two key state-level indicators to examine would be: change in the annual budget allocation for training counselors to work with individuals who have psychiatric disabilities, and change in the proportion of all VR counselors who are specialists in psychiatric disability. Such information is critical for assessing the role the state agencies may have played in the decline of service delivery to individuals with severe psychiatric disabilities.

In addition to problems associated with lack of specialized training, there may be disincentives for counselors with non-specialized caseloads to work with individuals with severe psychiatric disabilities. The rehabilitation rate for individuals with severe psychiatric disabilities is nearly 15% lower than for individuals with severe physical disabilities (Table 2). And while long-term follow-up data are lacking, it is probable that individuals with severe psychiatric disabilities, whose mental illness is often episodic in nature, are on average less likely to have long periods of continuous employment than are individuals with severe physical disabilities. Together, these two factors may attach a stigma to individuals with severe psychiatric disabilities within the vocational rehabilitaion system. Non-specialist rehabilitaion counselors - especially those who must achieve a yearly qouta of successful case closures - may actively avoid caseloads that include individuals with severe psychiatric disabilities.

While the greatest source of disparity between individuals with severe psychiatric disabilities and individuals with severe physical disabilities was the percent change over time in number of applications, the analyses indicate that there were also differences between the two groups which occurred after individuals applied for services. The acceptance rate increased significantly less over time for individuals with severe psychiatric disabilities than for individuals with severe physical disabilities (Figure 4) and the rate of competitive employment has increased less for individuals with severe psychiatric disabilities (Figure 7). While these two latter differences are relatively small, both are in the same direction as the percent change in number of applications and thus have an additive effect, so that the disparity between the two groups grows over the course of the rehabilitation process. Thus the original difference between individuals who are psychiatrically disabled and individuals who ar physically disabled in percent change with respect to number of applicants - approximately 14% - increases to 17% with respect to number rehabilitated and increases again to 23% with respect to number competitively employed (Figure 9). Future research focused on determining the causes of the disparity between the two groups must therefore consider not only factors which affect application to the system, but also factors which differentially affect individuals in the system, in terms of the likelihood of acceptance for services and the likelihood of obtaining competitive employment.

While the focus of this paper has been on idividuals with severe disabilities, there are two findings with regard to individuals with non-severe disabilities which are of interest. First, there have been dramatic decreases in individuals in the VR system with non-sevgere disabilities in terms of total applications (Figure 3), rehabilitations (Figure 6) and competitive employment (Figure 7). Second, the percent decrease over time for individuals with psychiatric and physical disabilitities was comparable among those whose disabilities were not severe (Figure 9). Given the mandated priority on serving individuals with severe disabilities, individuals with non-severe physical disabilities appear to have borne the brunt of the budget cuts which characterized the study period (Figure 1) to the same extent as individuals with non-severe psychiatric disabilities.

Recommendations and Directions for Future Research

We have considered several possible explanations for the disparities in VR service delivery which have developed over time between individuals with severe physical disabilities and individuals with severe psychiatric disabilities. It should be emphasized that RSA's R300/911 database, which was the only source of information considered in this paper, does not by itself provide a basis for determining the actual cause of these disparities. To thoroughly evaluate each possible explanation proposed here, additional research is needed, integrating information from a number of different sources.

In conducting this research, it would be very useful to determine the factors which account for certain year-to-year changes in key indicators. For example, the number of applicants among individuals with severe physical disabilities and among individuals with severe psychiatric disabilities was actually greater in the years 1978 through 1981 than it was in the baseline year, 1977 (Figure 3), dispite the sharp decrease in expeditures in those years (Figure 1). And, as shown in Figure 3, the gap between percent change in applicants among individuals with severe psychiatric and severe physical disabilities which emerged in the period 1977-1984 was established early on, in 1978, when the percentage increase among individuals with severe pschiatric disabilities did not match the percentage increase among individuals with severe physical disabilities. Thus a key to explaining the disparity which developed over time for the two sub-groups lies in understanding why the increase in applications among individuals with severe physical disabilities in the late 1970s was so much greater than the increase among individuals with severe psychiatric disabilities.

For the analyses presented in this paper, many disability categories have been aggregated to produce the group we call 'individuals with physical disabilities,' and which we use to compare with a group made up of individuals with psychiatric disabilities. However, individuals with physical disabilities are by no means a homogeneous group - there are indications that considerable variability exists betweeen groups of individuals with different physical disabilities, in terms of the indicators considered in this paper (Kallos et al., 1988). Advocacy groups and administrators concerned with the vocational rehabilitation of individuals with a specific physical disability, say those who have a hearing impairment, may therefore wish to examine an analysis of trends over time in which VR performance indicators are calculated for that particular group.

While there is a need for additional research to answer many of the questions raised by the findings presented in this paper, two specific policy-related recommendations can be made based on current knowledge about the state-federa VR system. Firts, a review of the criteria for severe psychiatric disability should be undertaken, particularly with respect to hospitalization; these criteria need to be updated to take into account the trend toward treatment of individuals with severe psychiatric disabilities in community-based outpatient settings. An alternative set of criteria for severity should be established, perhaps based on level of functioning rather than on hospitalization. Second, a set of meaningful vocational rehabilitation indicators should be calculated yearly, monitored over time and disseminated on a routine basis. These indicators might include some of the rates and change scores defined in this paper, as well as additional variables drawn from RSA files and other data sources. A tabulation of indicators should be provided for each major disability-severity type, both for the nation as a whole and for each state agency. Without ready access to this kind of system-wide information, program planners, administrators, legislators and advocacy groups working at the state and national level will not be able to detect potentially significant changes in service delivery, will be hampered in evaluating the impact of recent policy and programming initiatives, and will lack a key element for informed decision-making.

Conclusions

1. The period 1977 - 1984 was characterized by a decline in expenditures

in the state-federal vocational rehabilitation

system, a dramatic decrease in individuals who make application

for vocational rehabilitation services, a shift toward

serving individuals with severe disabilities and a marked increase

in the percentage of rehabilitated indidviduals who return

to competitive employment. 2. In terms of percent change in both the number of individuals

rehabilitated each year and the number of individuals returned

to competitive employment each year. there were

substantial increases during the study period for individuals

with severe physical disabilities and decreases for individuals

with severe psychiatric disabilities. 3. The disparity between individuals with severe psychiatric

disabilities and severe physical disabilities stems mainly

from a marked decline during the study period in the number

of applications among individuals with severe psychiatric

disabilities. However, improvements in the rate of acceptance

for services and the rate of return to competitive

employment, were also greater among individuals with severe

psychiatric disabilities. 4. Among those not severely disabled, individuals with psychiatric

disabilities have fared about the same during the study

period as individuals with physical disabilities, with respect

to percent change in total individuals rehabilitated and total

individuals returned to competitive empployment. 5. Research is needed to determine the reasons for the disparities

which have developed over time between service delivery

to individuals with severe physical disabilities and individuals

with severe psychiatric disabilities. 6. The psychiatric hospitalization criteria for severity should be

reviewed. 7. A comprehensive set of change indicators should be defined

and routinely monitored on a yearly basis to facilitate the

detection of disparities such as those reported in this paper,

and to aid in evaluation and planning at the state and national

level.

Figure Captions

Figure 1. Change in total yearly expenditures for vocational rehabilitation in the state-federal system, with respect to 1977, adjusted for inflation. The percentage for each year is calculated by dividing total Consumer-Price-Index-adjusted expenditures in that year by total expenditures in the base year, 1977, and multiplying by 100. The 1977 baseline level is indicated by a horizontal line. Source of data: unpublished tables prepared by L. Mars, Senior Statistician, RSA.

Figure 2. Percentage of total applicants identified as severely disabled. See text for definitions of severity and diagnostic categories.

Figure 3. Change in total number of applications, with respect to 1977. The percentage for each year is calculated by dividing the total number of applications in that year by the total number of applications in 1977, and multiplying by 100. Psychiatric disability is indicated by squares, physical disability by triangles. The solid symbols indicate severity, open symbols lack of severity. The baseline (no change with respect to 1977) is indicated by a solid line. Each line is also labelled: SVR PHYS=severely physically disabled; SVR PSYCH=severely psychiatrically disabled; NON-SVR PHYS=non-severely psychiatrically disabled; NON-SVR PHYS=non-severely physically disabled. Because each percentage has a common denominator, the graph can be used to determine the direction of the difference in total applications from year to year as well as from each year to 1977. For example, between 1978 and 1979, total applications increased among individuals with severe physical disabilities but decreased among individuals with non-severe disabilities. Rate of change from year to year is reflected in the slope of each line.

Figure 4. Acceptance rates by year. Please see text for definition of acceptance rate and caption of Figure 3 for key to disability/severity groups.

Figure 5. Rehabilitation rates by year. Please see text for definition of rehabilitation rate and caption of Figure 3 for key to disability/severity groups.

Figure 6. Change in number of individuals rehabilitated, with respect to 1977. The percentage for each year is calculated by dividing the total number of rehabilitants in that year by the total number of rehabilitants in 1977, and multiplying by 100. Please see caption of Figure 3 for key to disability/severity groups. The baseline (no change with respect to 1977) id indicated by a solid line.

Figure 7. Rate of competitive employment by year. Please see text for definition of rate and caption of Figure 3 for key to disability/severity groups.

Figure 8. Change in number of individuals closed with a work status of competitive employment, with respect to 1977. The percentage for each year is calculated by dividing the total number competitively employed in that year by the total number competitively employed in 1977, and multiplying by 100.

Figure 9. Difference in percent change between 1977 and 1984 for physically and individuals with psychiatric disabilities, among three closure/work status categories. Each data point on the graph represents the percentage change for individuals with psychiatric disabilities less the percentage change for individuals with physical disabilities. Bars extending below the horizontal line indicate that the decrease, as measured by % change, in number of individuals in particular category was greater for individual with psychiatric disabilities than for individuals with physical disabilities during this time period or that an increase among individuals with psychiatric disabilities was less than the increase for individuals with physical disabilities. Values used to calculate these differences are taken from 1984 data points in Figures 3, 6 and 8. Please see text for additional explanation.

References

Barker, J. The coordination of efforts between vocational rehabilitation andmental health systems. In L. Perlman & C. Hansen (Eds.), Rehabilitation of Persons with Long-Term Mental Illness in the 1990s. Twelfth Mary E. Switzer Memorial Seminar. National Rehabilitation Association, 1988. Baxter, E., & K. Hopper. Troubled on the streets: the mentally disabled homeless poor. In J. A. Talbot (Ed.), The Chronic Mental Patient. Orlando, Fl.: Grune and Stratton, 1984. Brown, V. B., Ridgely, M. S., Pepper, B., Levine, I. S., & Ryglewics, H. The dual crisis: mental illness and subtance abuse. American Psychologist, 1989, 44: 565-569. Fleiss, J. L. Statistical Methods for Rates and Proportions. John Wiley, New York, 1981. Goldman, H. H., & Gattozzi A.A. Balance of Powers: Social Security and the Mentally Disabled, 1980-1985. The Milbank Quarterly, 1988, 66: 531-551. Kent, G.R,. & McLaughin, D.E. Rehabilitation expeditures: an analysis of data for the State-Federal Rehabilitation Program. Presentation at the annual meeting of the National Association of Rehabilitaion Research and Training Centers, 1984. Kallos, J.E., Genevie, L,. Struening, E.L., & Andrews, H.F. The role of advocacy groups in vocational rehabilitation service delivery to the severely disabled: a comparison between multiple sclerosis and other disorders. Journal of Neurological Rehabilitation, 1988, 2: 13-20. Mars, L. Unpublished tables. Menz, F.E., Hansen, G., Smith, H., Brown, C., Ford, M., & McCrowey, G. Gender equity in access, services and benefits from vocational rehabilitation. Journal of Rehabilitation, 1989, 55: 31-40. Mosher,L.R. The current status of the Community Support Program: a personal assessment. Psychosocial Rehabilitation Journal, 1986,9(3): 3-13.Pepper, B., Ryglewicz, H., & Kirshner, M.C. The uninstitutionalized generation: a new breed of psychiatric patient. In B. Pepper & H. Ryglewicz (Eds.), The Young Adult Chronic Patient. San Francisco: Jossey-Bass, 1982. Rehabilitation Services Manual, MT#2. Statistical Reporting System, 1974. Rehabilitation Services Administration Information Memorandum RSA-IM-88-22: Caseload statistics, state vocational rehabilitation agencies, fiscal year 1987. Surles, R.C. Changing organizational structures and relationships in community mental health. Administration in Mental Health, Symposium Proceedings, pp. 217-0227. ADAMHA. Washington, D.C., 1986.