Job search, job placement, and on-the-job supports are valuable services provided to many people with intellectual and developmental disabilities (IDD) to obtain work in the community. Investigating those who were unemployed at the time of service entry, this study seeks to extend understanding about the effect of services. Using extant data, a sample of 39,277 people with IDD using Vocational Rehabilitation services were studied to understand the potential cumulative effects of these job-related services and individual characteristics on job attainment. Findings showed people with IDD of different demographic groups had different outcomes. Also, those receiving three job-related services were 16 times more likely to obtain employment than the reference group. This study has wide implications for research, policy, and practice.
People with intellectual and developmental disabilities (IDD) continue to face considerable barriers to labor force participation and employment (Hartnett, Stuart, Thurman, Loy, & Carter, 2011; Nord, Luecking, Mank, Kiernan, & Wray, 2013; Sulewski, Kugler, & Kramer, 2010). Within the past decade, the Employment First movement has emerged. This grassroots response advances the philosophy and values that employment should be the first and preferred outcome of the disability support service system (Association for People Supporting Employment First, 2010; Callahan, Griffin, & Hammis, 2011; Mank & Grossi, 2013; Nord, Butterworth, Carlson, Grossi, Cohen Hall, & Nye-Lengerman, 2015; Rogan & Rinne, 2011).
In recent years, many states developed policies and directives aimed at improving the employment outcomes of people with IDD (Nord & Hoff, 2014). Research suggests that state employment-related support systems demonstrated nominal improvements in employment outcomes. National data from 1988 to 2012 suggested only minor improvements in the number of people engaged in integrated employment, while there were increased numbers of people served overall and in nonemployment (Butterworth et al., 2014).
Fundamental to obtaining employment for many people with IDD and central to improving employment outcomes within the disability support systems is having reliable access to quality supports to facilitate job acquisition and retention. Extensive research has shown the effectiveness of supported employment and customized employment models, which connect people with IDD and other disabilities to employment. Rather than require a particular skill level to obtain paid work, these models assume skills can be learned on-the-job, with the appropriate training and support services (Bellamy & Melia, 1991; Bond, Drake, & Becker, 2008; Callahan, 2002; Griffin, Hammis, Geary, & Sullivan, 2008; Office of Disability Employment Policy, 2005; Rusch, 1990; Wehman, Revell, & Kregel, 1998).
Job acquisition supports often span many different activities and distinct phases in the job attainment process. There are a number of promising practices that support systems and professionals can employ to advance integrated employment. These include: (a) individualizing supports and employment goals to respond to a job seeker's individual preferences and needs, (b) building relationships with the local business community, (c) finding jobs by way of professional and personal networks, (d) developing jobs through negotiation and customization, (e) marketing individual strengths that can address a specific employer's need, and (f) providing ongoing on-the-job supports and coaching upon hire (Butterworth, Migliore, Nord, & Gelb, 2012; Callahan, Shumpert, & Condon, 2009; Griffin, Hammis, & Geary, 2007; Hoff, Gandolfo, Gold, & Jordan, 2000; Luecking, Fabian, & Tilson, 2004).
Advancing Employment Outcomes—(VR) Services
The Rehabilitation Act 1973 (1973) has become a compelling driver of employment outcomes for people with disabilities. Title I of the Act identifies employment of people with disabilities, including those with the most significant disabilities, with a principal pathway to economic independence, full community participation, and a way out of poverty. As such, the Act lays out the formula grant program to support states to implement VR services. Over the past decade, VR experienced an increase in the numbers of people served (Butterworth, Hall, Smith, Miglione, & Winsor, 2014) and emerged as a compelling way to support people with IDD, including those with significant impairments, to obtain individual work in an integrated setting. In the 2012 fiscal year, more than 45,000 people with intellectual disabilities had their VR cases closed. Sixty-five percent of this group received some VR service, from which 52% were employed at closure (Butterworth et al., 2014). These outcome data remained fairly consistent over the past decade (Burgess & Cimera, 2014).
Vocational Rehabilitation (VR) offers services with the intent that people who participate will either maintain their current employment or enter into an integrated job. These services center on providing supported employment, which seeks to match competitive jobs with an individual's “strengths, resources, priorities, concerns, abilities, capabilities, [and] interests” (Rehabilitation Act Amendments, 1998, sect. 7). Professional use multiple methods to accomplish integrated work. Job search entails matching an individual's interests, preferences, and capabilities with paid work rather than a protracted assessment, training, and counseling period (Becker, Swanson, Bond, & Merrens, 2008; Bond, 2004). Job placement then involves connecting the person with a disability to available jobs. A third component is the provision of individualized, on-the-job supports. Best practices suggest supports should be customized based on client needs to promote success in the work setting (Bond, 2004; Rogan, Banks, & Howard, 2000).
Understanding Employment Characteristics
Research shows that successful employment outcomes of VR service users with disabilities relate to some individual characteristics. Personal attributes associated with paid work at VR closure include: gender, race, age, level of disability, higher levels of education, employment at the time of application, and number of jobs since disability onset (Chan et al., 2014; Dutta, Garvey, Chan, Chou, & Ditchman, 2008; Xu & Martz, 2010; Yamamota & Alverson, 2013). Dutta and colleagues (2008) found that the receipt of cash benefits had a deleterious impact on employment, with those receiving assistance least likely to find work.
Examples of service characteristics linked with successful job placement include supported employment and an assessment of the level of disability. Wehman, Chan, Ditchman, and Kang (2014) found that supported employment for transition-age youth with IDD related to employment at closure when compared with young adults who did not receive supported employment. When people with a traumatic brain injury (TBI) received diagnostic and treatment services, they were more likely to have significantly higher wages at case closure, whereas job placement and job search services correlated to significantly lower wages at closure for this population (Schonbrun, Sales, & Kampfe, 2007).
Using Rehabilitation Service Administrative data from FY 2005, Dutta et al. (2008) found several employment predictors. These components included holding a job at the time of VR application, and receipt of job placement services, on-the-job training, and further job maintenance supports (Dutta et al., 2008). Of note, these outcomes studies included people with a job at the time of VR application. Previous research has also not considered the impact of receiving multiple services (i.e., a combination of job search, job placement, and on-the-job supports) on obtaining competitive employment.
Researchers continue to call for additional studies on the effectiveness of employment supports to build evidence for the effectiveness of various intervention methods (Bond, 2004; Nord, Barkoff, Butterworth, Carlson, Cimera, Fabian . . . Wohl, 2015; Wehman et al., 2016). Past research on integrated employment included people who already have jobs in the sample. Since people who enter VR without a job have different VR experiences than those with a job, as highlighted by the greater likelihood of unemployment at the time of discharge from VR, the purpose of this article is to examine which supports most strongly relate to the promotion of integrated employment among people with IDD who were unemployed. Also, previous studies considered each service type individually and without considering the potential aggregate effect of receiving more than one service.
The aim of this study was to extend what we know about the effects of personal characteristics and service usage on employment system outcomes related to employment achievement for those who access VR without a job at intake. Specifically, this study intends to fill a gap in the literature regarding job-related services and employment outcomes of people IDD who enter VR services unemployed. This study sought to answer the following research questions: What effects do individual characteristics and job-related services have on employment attainment at VR closure for those people with IDD who are unemployed at VR intake? How does the provision of more than one service affect employment outcomes for people who enter VR unemployed?
Data Source and Sample Selection
This study drew subjects from the Rehabilitation Services Administration Case Record Report (RSA-911) dataset fiscal year 2011. The RSA-911 dataset is a rich administrative dataset that includes individual-level data of VR recipients that experienced a case closure, as collected by state VR agencies. The data set includes demographic information, benefit usage across various public programs at program entry and closure, VR service provision, and employment-related outcomes.
The inclusion criteria for this study permitted selection of cases with IDD from all general VR agencies in the 50 states and Washington, D.C., only. In that the purpose of this study was to evaluate the relationship of job acquisition services on unemployed applicants, it was necessary to isolate those who were unemployed at application. As such, all VR service users with a job at application were eliminated to remove a potentially confounding factor to the analysis. The resulting sample included 39,277 subjects or 56% of the initial 69,614 people with disabilities who had case closures in the fiscal year 2011.
Measures and Definitions
The outcome of interest was whether people with disabilities obtained employment at the time of case closure. This variable was binary and defined as either employed (coded as 1) or unemployed (coded as 0). In total, 20,187 subjects (51.4%) were unemployed and 19,090 subjects (48.6%) were employed at closure. Independent variables were chosen to control for key demographics highlighted in past research (Chan et al., 2014; Dutta, Garvey, Chan, Chou, & Ditchman, 2008; Wehman et al., 2014; Xu & Martz, 2010; Yamamota & Alverson, 2013).
Making this study unique was the job-related services variable. This dummy variable was constructed using three items that denoted the combined use of job search services, job placement services, and on-the-job supports. This variable resulted in eight mutually exclusive service categories. For example, a person receiving no job-related support was designated as “none,” whereas a person who received job search and job placement services was designated as “job search & placement” and so on. The group coded “none” was the reference group. This structure allowed the research team to isolate the combined effects of service combinations and compare them to the no-job-related-service group. Refer to Table 1 for more details regarding definitions and coding structures.
Summary statistics were computed, and logistic regression analysis conducted. This modeling approach was used to evaluate the odds of achieving an employment outcome at closure (coded 1) or being unemployed at closure (coded 0). The logistic regression model accounted for the effects of several personal characteristics and to model the effects of job-related services. Hosmer-Lemeshow Goodness of Fit test was to evaluate the fit of the final model.
Table 2 provides the descriptive statistics for the sample. On average, the sample was 26.3 years old, and all participants held a special education certificate of completion. The majority were male (60.1%), White (68.5%), with a significant disability (98.3%), and with a cognitive impairment (68.6%). Before VR closure, more than a third of all subjects received no service, a third accepted a single service, 22% received two services, and 12% obtained all three services. The three types of VR services offered included job search, job placement, and on-the-job support.
The logistic regression model evaluated the likelihood of exiting VR with an employment outcome. This analysis yielded a significant model χ2 (15) = 8,134.19, p < .001. The Nagelkerke R2 was computed to be .25 and the classification accuracy of this model was 68.7%.
As seen in Table 3, nearly all independent variables were found to have a significant effect on the likelihood of obtaining work at VR closure. Specifically, subjects who are White had slightly better odds of exiting VR with a job (OR = 1.28) than people from minority groups. People who entered VR with higher education levels had slightly better odds of exiting VR with a job (OR = 1.13) than people who had lower levels of education. Being female (OR = 0.78) or older (OR = 1.00) indicated slightly lower odds of exiting VR employed. However, having a significant disability (OR = 0.38) is associated with significantly lower odds of an employment outcome than those without a significant disability. Compared to the reference group, people who had sensory/communication disorders, people with physical impairments (OR = 0.77), and those with psychosocial and other mental impairments (OR = 0.81) had moderately lower odds of exiting VR with a job. The odds of people with cognitive impairments exiting with a job did not significantly differ from the reference group.
In considering job-related services, all service arrangements were associated with significantly greater odds of exiting VR with a job, compared to the reference group, those that received no job-related service. Specifically, those with job search, job placement, or on-the-job support services had between 1.39 to 3.75 greater odds (job search OR = 1.39; job placement OR = 3.75; on-the-job support OR = 3.40) of having a job at closure, compared to the reference group. These service effects grew enormously in magnitude for those receiving two and three job-related services. Compared to the reference group, those receiving job search and placement services had 4.03 greater odds. Those granted job search and on-the-job support had 7.43 greater odds. People with job placement and on-the-job support had 9.26 greater odds of exiting with a job. Finally, those provided job search, placement, and on-the-job support had 16.39 greater odds of leaving VR with a job, compared to the reference group.
As the individual-level logistic regression findings highlight, not all unemployed people with IDD have the same experiences with the VR system. Work attainment disparities exist across all demographic characteristics evaluated, which are consistent with past employment research (Chan et al., 2014; Hasnain & Balcazar, 2009; Leonard, D'Allura, & Horowitz, 1999; Xu & Martz, 2010; Yamamota & Alverson, 2013). Although the disability and impairment-related demographic variables are unable to pinpoint the severity of one's disability and the effect it has on one's life, these findings indicate that those with significant disabilities and a physical or psychosocial impairment were significantly less likely to obtain employment. These service outcome inequities are likely due to multiple factors, both internal and external to the VR system, and they are critical to recognize and balance at the programmatic and practice levels.
Past research has shown a link between specific job-related service and employment outcomes. These studies used broader samples that included people that entered VR services with jobs (Dutta et al., 2008; Moore, Harley, & Gamble, 2004; Schonbrun, Sales, & Kampfe, 2007; Wehman et al., 2014). This study hones in on people with IDD seeking VR services that enter the system unemployed, thus eliminating a critical confounding variable to the job-related service usage variables. It is the first to investigate the combined of effects job-related services, which found that people with IDD who received any job-related services were significantly more likely to obtain employment compared to those with no service at all. The most striking finding was the magnitude of increased odds of employment for those receiving two or three job-related services, ranging from 4 for those receiving job search and placement services to over 16 for those receiving job search, job placement, and on-the-job support. This study is unique in reporting these findings, as previous studies have not included the combined impact of receiving combinations of services.
An unexpected result was the comparatively minor effects of on-the-job supports alone, considering that a person with IDD receives these services after obtaining a job. Though these effects remain considerably higher than the reference group, compared to service mixes that include on-the-job support and other services, the odds of employment for those just receiving on-the-job support alone are modest. This finding suggests that, though significantly correlated, on-the-job support alone does not maximize the chance of retaining employment for at least 90 days, defined as a positive VR closure. A possible explanation might be that the job search and placement processes are key elements for connecting job seekers with IDD to work that might more closely match their interests and abilities. Future research would benefit from a more in-depth understanding of the connectedness of vocational rehabilitation service delivery to better articulate the facilitators of employment outcomes.
The RSA-911 dataset provides a robust individual-level understanding about the VR service users, provision, and outcomes. The focus and findings of this study related to employment attainment outcomes, demographics, and job-related service usage are widely applicable across state VR programs across the country. However, results gleaned from the RSA-911 dataset and inferences made from analyses are limited to correlational findings. Thus, causal statements cannot be made. Additionally, it would be informative to evaluate these outcomes over time, to determine whether these patterns are consistent across years.
Though limitations to the findings include lack of accounting for the level of support, the practices used, and the informal supports accessed by this population, these findings do have wide applicability and demonstrate that the full suite of job-related support is most advantageous. Future research on the practices and support strategies used to achieve these employment outcomes for people with IDD would be beneficial, so as to isolate and advance the most effective practices employed under the suite of job-related services. Also, it would be worthwhile to understand how professionals and people with IDD opt to deploy and use various service options. For example, does VR staff tend to encourage the full suite of services for people with IDD they view as “most employable”? Or are people with IDD more likely to elect one type of support over others? These findings also provide some early indications of a potentially effective and concrete system strategy regarding service access targeting unemployed people with disabilities who become involved in the VR system. Specifically, VR should consider prioritizing the full array of job-related services to all people with disabilities who enter VR unemployed so as to maximize the likelihood of a positive outcome.
As people with IDD in the United States continue to demand greater access to services and supports that successfully facilitate individual employment in the community, VR services and other formal support systems will be looked to as critical partners. Thus, it is necessary for these systems to understand that there are sub-groups of people with IDD that experience services differently and achieve lower outcomes as a result. It is also critical for these systems to understand the importance of the different types and mix of supports used to make work a reality. Through this more sophisticated understanding of service usage and outcomes, employment systems can adapt to better monitor and serve job seekers with disabilities. Looking to the future, researchers must address two objectives: Consider the ongoing system's ability to respond to the needs of different demographic groups and seek to identify effective system and professional strategies to equal the playing field for all people with disabilities.
This publication is supported by grant CFDA #90-DD0708 from the Administration on Intellectual and Developmental Disabilities and cooperative agreement #H133B080005 from the National Institute on Disability Independent Living and Rehabilitation Research (NIDILRR), U.S. Department of Health and Human Services.