Abstract
Background
Vocational Rehabilitation (VR) Returners are individuals who receive services, have their case closed, and later return for additional services. This understudied group is critical for improving service delivery and resource allocation.
Objective
We examined demographic and case factors predicting return to VR services, compared employment outcomes between returners and one-time consumers, and identified VR services linked to competitive employment outcomes.
Method
Using Rehabilitation Services Administration (RSA-911) data (2017–2024), we identified 110,061 returners (6.8%). Logistic regression models examined predictors of return and service types associated with competitive employment; linear regression models compared continuous employment outcomes (weekly hours and wages) between returners and one-time consumers.
Results
Individuals receiving public benefits showed higher return rates, with benefit receipt accounting for initial racial disparities. Black and American Indian individuals were more likely to receive benefits; Hispanic individuals less likely. Returners achieved higher employment rates but worked fewer hours at lower wages. Job Placement Assistance and Diagnosis and Treatment were positively associated with employment outcomes.
Conclusion
VR return patterns reflect intersections of benefit receipt, disability characteristics, and structural constraints. Returners’ higher employment rates but lower wages and hours suggest strategic workforce participation, highlighting the need for longitudinal, career pathway approaches in VR service delivery.
Introduction
The federal–state Vocational Rehabilitation (VR) program served 872,460 people with disabilities between 2022 and 2024, making it the largest public workforce development system for this population (Thomas et al., 2025). VR agencies under the Rehabilitation Act provide a wide range of individualized services—including counseling, training, education support, assistive technology, and workplace accommodations—to help people with disabilities prepare for and obtain competitive integrated employment. This service delivery approach has evolved significantly in recent years. The Workforce Innovation and Opportunity Act (WIOA) of 2014 modified the Rehabilitation Act of 1973 by making competitive integrated employment the main priority for vocational rehabilitation services. WIOA established competitive integrated employment as its main priority through which people with disabilities work with non-disabled colleagues at market wages while having access to career progression opportunities. The WIOA transitioned from labor exchange to career pathways which marked a fundamental shift because it recognizes that people require continuous assistance during their working lives when job markets transform or personal situations shift or career progression becomes available.
Understanding VR Returners
VR programs have long served as a critical support system for individuals with disabilities seeking to enter or re-enter the workforce (Kendall & McLennan, 2021; Nivorozhkin, 2019). While much of the existing research and policy focus has centered on first-time participants, a growing and important subset of VR consumers, known as returners, have received comparatively little attention. Returners are individuals who re-engage with VR services after a prior case closure, often seeking additional support to navigate new or evolving employment challenges (Gavin et al., 2022). This changing environment has produced a vital yet insufficiently studied pattern of people who seek additional VR services after their previous case ended, commonly referred to as “returners”. The return population makes up a substantial percentage of annual service recipients, but researchers have not extensively studied their population features, service usage, and employment results (Escorpizo et al., 2011).
The reasons individuals return to VR are multifaceted and often reflect the dynamic nature of the modern labor market, personal circumstances, and long-term career development. Multiple complex factors influence the decision to re-engage with VR services, and these factors often intersect in ways that are unique to each individual's life and work experience. For some, returning to VR is prompted by significant life changes, such as the onset of a new disability, the progression of an existing condition, or a change in functional capacity, that directly impact their ability to perform job duties or maintain employment. In these cases, VR services may be sought to reassess vocational goals, secure new accommodations, or identify alternative employment options that align with their current capabilities.
Others may return due to external labor market shifts. The rapid pace of technological advancement, automation, and industry restructuring has rendered some occupations obsolete while creating demand for new skills and competencies. Individuals who were once successfully employed may find themselves needing to upskill or reskill in order to remain competitive. These changes suggest that returning to VR services should not be viewed as a failure or setback, but rather as a natural and appropriate response to the evolving needs of individuals navigating complex and often nonlinear career trajectories. VR services can provide access to training programs, certifications, and career counseling that help individuals navigate these transitions. Job loss is another common driver of return. Economic downturns, company closures, or layoffs can disproportionately affect workers with disabilities, who may already face barriers to employment. In such cases, returning to VR offers a structured pathway to re-enter the workforce, often with enhanced support tailored to the individual's evolving needs. As the workforce continues to change and as individuals with disabilities pursue greater autonomy, the role of VR as a recurring system is becoming increasingly relevant.
In addition, responding to setbacks, many returners are motivated by aspirations for career growth and advancement. These individuals may seek VR services to pursue higher education, transition into more fulfilling or better-pay roles, or exploring new career pathways that align with their interests and strengths. VR can support these goals through individualized planning, tuition assistance, and coordination with employers to ensure appropriate accommodations are in place. Lastly, some returners may have previously exited VR services with short-term or entry-level employment outcomes and later recognize the need for more sustainable, long-term career development (Young, 2010). As VR system increasingly emphasize career pathways and competitive integrated employment, returning participants may find new opportunities to build on their prior progress and achieve greater economic self-sufficiency.
Literature on VR Returners
Research specifically examining VR returners is extremely limited. Only two studies have explicitly addressed repeat use of state VR services (Marsh et al., 2022; Russell, 2021). Marsh et al. (2022) conducted a study of VR returners in Minnesota, analyzing 788 individuals who underwent four or more service sequences. Their findings revealed that individuals who are deaf or hard of hearing, have serious mental illness, intellectual disability, or autism are more likely to return for multiple VR sequences, with 72% of high-sequence users receiving Supplemental Security Income/Social Security Disability Insurance. The time between service sequences decreased with each return (from 765 days at the first return to 532 days by the fourth), while employment outcomes remained steady at approximately 51% across first and repeat sequences.
Russell (2021) conducted a series of t-tests using Wisconsin state VR data merged with Rehabilitation Services Administration (RSA-911) case records, examining 741 consumers with supported employment plan types who exited successfully from 2014 to 2017. The study found that 25.91% (192 individuals) returned to VR for further services. Key findings included that returners had significantly shorter periods between employment and case closure (207.63 days vs. 252.18 days for non-returners) and lower hourly wages ($8.33 vs. $8.62). Several services were significantly associated with return, including On-the-Job Training for consumers with psychosocial impairments, suggesting that consumers with significant disabilities may need additional or continuous support following successful case closure. While these studies provide valuable initial insights, they are limited to specific states and disability populations. This scarcity of research represents a significant knowledge gap, particularly given that understanding patterns of return service use is essential for developing data-informed strategies that enhance sustained employment success and career advancement for individuals with disabilities.
Tracking Returners to VR Services
The mechanism to track clients of VR services is the national administrative dataset referred to as the Rehabilitation Services Administration Case Service Report 911 (RSA-911). This rich dataset contains over 15,000 annual entries representing youth and adults with disabilities who received VR services while pursuing employment or postsecondary education goals. While the RSA-911 data is intended to track service-delivery to clients as they progress through the VR system, we know very little about the changes in service delivery over time for individual clients, and how these changes may impact career pathways.
In a recent exploratory study, Lombardi et al. (2025) developed a process to identify “returners” in the RSA-911 data. The focus of this study was on transition-age youth (14–24) who qualified for Pre-Employment Transition Services (Pre-ETS) and their postsecondary education pursuits. Specifically, cases were identified as “returners” if more than one case had been opened for an individual between the ages of 14 and 24 and after they had been found eligible to receive Pre-ETS. After applying these inclusion criteria, 159 out of 159,968 total disabled youth (less than 1%) were deemed “returners”. Importantly, this study showed very few disabled youth returned for services between the ages of 14 and 24; however, we anticipate this pattern will be different if the age range is expanded and represents the adult life span beyond age 24.
Better understanding of VR returners could significantly inform policy development, case management practices, and service planning strategies, particularly given RSA's emphasis on performance metrics and sustained employment outcomes. The federal VR program operates under a performance accountability framework that emphasizes not only successful case closures but also long-term employment sustainability. Understanding patterns of return service use is therefore essential for developing data-informed strategies that enhance sustained employment success and career advancement for individuals with disabilities. Examining service patterns and employment outcomes among returners can provide critical insights into the effectiveness of VR's evolving approach to career pathway development and inform evidence-based practices that support individuals with disabilities throughout their working lives.
The purpose of this study is to examine the demographic characteristics, service patterns, and employment outcomes of individuals who return to VR for additional services following case closure. Specifically, the study addresses the following research questions:
What demographics and case characteristics predict return to VR services? How do employment outcomes differ between VR returners and one-time service consumers? Which VR services are associated with competitive integrated employment outcomes among returners?
Methods
The data used in this study were extracted from the RSA-911 data for Fiscal Years 2017–2024. The RSA-911 database contains comprehensive information about consumers who received state VR services, including demographic information, disability characteristics, types of services received, and employment outcomes.
Sample
All cases meeting baseline criteria (approved IPE and valid closure record) were included in the sample (n = 1,613,375). Among these, additional criteria were then applied to classify participants as returners versus non-returners. Specifically, individuals with at least two application dates and at least two exit dates were identified as having completed multiple VR service episodes and were classified as returners. Using this classification, the majority of included participants (93.1%; n = 1,503,381) received VR services once and were categorized as non-returners, while 6.8% (n = 110,061) were identified as returners who completed between two and five service episodes. Among returners, 100,048 consumers had two VR service episodes, 8,959 had three, 947 had four, and 107 had five. Consumers with six or more service episodes (n = 19) were considered statistical outliers and were excluded from subsequent analyses. These cases were excluded due to their extremely small cell size and highly atypical service utilization patterns that markedly differed from the majority of consumers and could exert disproportionate leverage on statistical estimates. Demographic characteristics of the sample are summarized in Table 1.
Demographic Characteristics of VR Service Recipients by Return Status.
Note. IPE = Individualized Plan for Employment; SSDI = Social Security Disability Insurance; SSI = Supplemental Security Income; TANF = Temporary Assistance for Needy Families. Percentages may not sum to 100% within categories due to rounding or missing data.
Variables
The dependent variables, employment outcomes, were assessed using: (a) employment type at case closure, (b) hourly wage at exit, and (c) hours worked per week at exit. For analysis, employment outcomes were dichotomized to indicate attainment of competitive integrated employment versus all other outcomes. Both hourly wage and hours worked per week at exit were analyzed as continuous variables in their original numeric form to capture the intensity of employment among individuals who achieved successful case closure.
The independent variables captured participants’ demographic, disability-related, and service characteristics. Demographic variables included gender (male, female, nonbinary, or prefer not to disclose), race/ethnicity (White, Black or African American, Hispanic/Latinx, Asian, American Indian or Alaska Native, Native Hawaiian or Pacific Islander, or multiple/other), and level of education (less than high school, high school diploma or equivalent, postsecondary education, or bachelor's degree or higher). Disability-related variables included the primary type of disability, categorized into five groups: mental (psychological/psychosocial), physical (mobility or orthopedic), intellectual and learning, auditory/communicative, and visual disabilities. The dataset also captured the significance of disability, distinguishing between significant, most significant, and other disabilities not classified in these categories. Additional indicators included the primary and secondary sources of disability and whether the individual self-identified as having a disability. These variables were included to account for variations in disability characteristics that may influence participation in and outcomes of VR services.
Receipt of public benefits at application, including Supplemental Security Income (SSI), Social Security Disability Insurance (SSDI), Temporary Assistance for Needy Families (TANF), or other forms of public assistance, was included as an indicator of economic support. Medical insurance status was also recorded, categorized as Medicaid, Medicare, coverage through the Affordable Care Act (ACA), private insurance, or uninsured. These variables were incorporated to capture socioeconomic and health-related factors that may influence individuals’ engagement with and outcomes from VR services. Employment-related variables included employment status at the time of Individualized Plan for Employment (IPE) development (employed and seeking advancement, employed and maintaining employment, or unemployed and seeking new employment) and the presence of employment barriers such as long-term unemployment, homelessness, low income, criminal record, or single-parent status.
Service provision variables captured the range of VR services received by consumers during their service period. Core services included vocational counseling and guidance, college or occupational/vocational training, job readiness and placement assistance, supported or customized employment services, and work-based learning experiences. Additional services encompassed assessment and diagnostic evaluation, disability-related skills training, benefits counseling, transportation, maintenance, rehabilitation technology, and personal assistance services. Each service type was coded dichotomously (received = 1; not received = 0) based on the RSA-911 coding schema. Including this wide range of service indicators allowed examination of how specific service patterns contribute to employment outcomes and client return behaviors.
Data Analysis
All data pre-processing, sample selection, and statistical analyses were executed in R version 4.4.1 (R Core Team, 2024). To answer the research questions at hand we fitted a series of regression models using the stats package in R. Given an interest in both categorical and continuous dependent variables, we provide details below depending on the nature of the outcome variable.
When the dependent variable was dichotomous (e.g., returner versus non-returner) it was treated as a Bernoulli distributed variable and connected to the explanatory variable(s) via the logit link allowing us to estimate the probability of success. The first model fitted was an unconditional model in which only an intercept was estimated where its resulting point estimate corresponds to the log-odds of success. By exponentiating the estimated log-odds, the odds of success is returned which in-turn informs the probability of success: odds / (1 + odds). Afterwards, we inserted predictor variables by category (e.g., barriers) to model the partial log-odds where the intercept corresponds to the log-odds of success for the reference group (i.e., the estimate is conditional on the modeled predictors). To better understand the impact of the explanatory variables we provide odds ratios as our effect size and further, present model-implied probabilities to enhance interpretations. When the dependent variable was continuous, we elected to estimate linear regression models. In a similar fashion, we estimated an intercept only model to return an unconditional expected value of the outcome (e.g., wage and weekly hours).
Results
To estimate the base probability of returning to VR services, an intercept-only logistic regression model was fitted using the “returner” indicator as the dependent variable. The log-odds of return were −2.614 (SE = 0.003, p < .001), corresponding to an odds ratio of 0.073 and an overall return probability of 6.82%. This sample therefore represents a large and diverse population of VR participants, with approximately one in fifteen consumers re-engaging with VR services following a previous case closure. Next, we examined whether specific individual characteristics were predictive of returners.
Predictors of Return to VR Services (RQ1)
To address RQ1, we fitted a series of logistic regression models. Based on the results of the preliminary models, we examined interaction effects for selected variables. Model-implied probabilities of returning to VR services are summarized in Table 2, followed by detailed results for the preliminary and interaction models.
Model-Implied Probabilities of Returning to VR Services.
Race & Ethnicity
Indicators for all non-White race and ethnicity groups were inserted into a logistic regression. Significant effects were observed for Black (B = 0.136, SE = 0.007, p < .001; OR = 1.146), Asian (B = 0.149, SE = 0.020, p < .001; OR = 1.161), and Hispanic (B = −0.170, SE = 0.009, p < .001; OR = 0.844) groups. Therefore, Black and Asian individuals were more likely to return compared to White individuals; whereas, Hispanic individuals were less likely to return than White individuals.
Gender
Females were slightly less likely to return (B = −0.020, SE = 0.006; p < .05; OR = 0.980) than males.
Barriers
Relative to individuals who experienced no employment barriers, each of the 12 barriers were found to have a significant positive relationship with returning for VR services. EXTANF and LIS had odds ratios of 1.980 and 1.954; while Displaced Homemaker had the smallest OR at 1.165; with all other barriers possessing ORs between 1.204 and 1.533. Therefore, experiencing barriers was predictive of re-engaging with VR.
Veteran Status
We observed that veterans were less likely to return (B = −0.376, SE = 0.021; p < .001; OR = 0.687) than non-veterans.
Public Benefit Receipt
Individuals receiving public benefits at application were significantly more likely to return. Significant effects were observed for SSDI (B = 0.940, SE = 0.007; p < .001; OR = 2.560), SSI (B = 0.807, SE = 0.007, p < .001; OR = 2.241), and TANF (B = 0.523, SE = 0.021, p < .001; OR = 1.686).
Medical Insurance
Individuals with medical insurance at application were more likely to be returners. Those covered by employer plans were 1.984 times more likely to return; those covered on the Affordable Care Act were 1.862 times more likely to return; those with public insurance were 1.721 times more likely to return; and lastly, those with private insurance were 1.260 times more likely to return than those without medical insurance.
Educational Attainment
Relative to individuals without a postsecondary degree, license, or certificate individuals with such education were more likely to return for VR services, with the exception being those with a degree above a Masters (B = −0.090, SE = 0.051, p = .084; OR = 0.915). Odds ratios ranged between 1.230 (Bachelors’ degree) and 1.564 (Vocational Training Certificate).
Disability Characteristics
Those with more significant disabilities, on average, were more likely to return for VR services compared to those without the most significant disabilities. Those with the most significant disabilities were 1.315 times more likely to be returners (B = 0.274, SE = 0.015, p < .001); while those with significant disabilities were 1.08 times more likely to be returners (B = 0.080, SE = 0.016, p < .001). To better understand how different disability types were related to returning for services, the sample was organized into five categories: mental (33%), physical (17%), intellectual and learning disability (33%), auditory/communicative (12%), and visual (5%). Within this structure of disability types, individuals with auditory (B = 0.188, SE = 0.010, p < .001; OR = 1.21) or visual (B = 0.131, SE = 0.014, p < .001; OR = 1.14) disabilities were more likely to return for VR services compared to those with mental disabilities; whereas, individuals with physical (B = −0.199, SE = 0.010, p < .001; OR = 0.82) or intellectual and learning disabilities (B = −0.164, SE = 0.008, p < .001; OR = 0.85) were less likely to return for VR services compared to those with mental disabilities.
Interaction Effects
Race/Ethnicity by Disability Benefits (SSDI/SSI)
Based on our preliminary model results, we constructed follow-up models that incorporate interaction effects using selected variables. First, a logistic regression model was fitted with ‘returner’ as the dependent variable and included a benefits indicator (where if the individual received either SSDI, SSI, or TANF it was coded as ‘1’, ‘0’ otherwise), along with indicators for the various race/ethnicities [white individuals who received no benefits are the reference]. Controlling for receipt of benefits, the odds ratio for Black individuals is 1.02 (down from 1.14 without controlling for benefits); the odds ratio is similar for Asian individuals (1.15 versus 1.16); finally, the odds ratio for Hispanic increased from .84 to .90. White individuals who received benefits are 2.48 times more likely to return versus White individuals who did not receive benefits.
Next, race/ethnicity was crossed with the benefits indicator. A significant interaction resulted for the BlackXDisabilityBenefits (B = −0.060, SE = 0.015, p < .001). When looking at the model predicted probabilities, it appears that receiving benefits is a stronger predictor than race/ethnicity. For example, White individuals who received benefits have an estimated probability for returning of .115 which matches the probability for Black individuals who also received benefits.
In addition, another logistic regression model was fitted in which the sample of returners, where race/ethnicity indicators were predictors and the received benefits was the dependent variable. Significant effects were observed for Black (OR = 1.58), American Indian (OR = 1.26) and Hispanic (0.73); therefore, Hispanic individuals are less likely to receive benefits relative to White individuals, whereas Black and Native American individuals are more likely to receive benefits.
Race/Ethnicity by Medical Insurance
After coding whether an individual had insurance of some sort ‘1’ or not ‘0’; a logistic regression model was fitted with ‘returner’ as the dependent variable and included the main effects of insurance and race/ethnicity and all interaction terms and yielded the following: Hispanic individuals with insurance were 1.20 times more likely of returning versus 0.79 for those without insurance relative to White individuals; Black and Asian individuals with insurance were found to be more likely to return relative to White individuals with insurance.
Employment Outcomes of Returners vs. Non-Returners (RQ 2)
Employment at Exit
To examine the probability of employment over the entire sample, an intercept-only logistic regression model was fitted. The log odds was found to be significant with an estimate of −0.181 (SE = 0.002, p < .001) which translates to an odds of 0.835. This translates to a model- implied probability of employment at exit of 0.455 for the sample. Controlling for returners, resulted in a significant nested likelihood ratio test (ΔDdf = 1 = 40,480; p < .001) thus favoring the more complex model. The log-odds estimate for returners was significantly different from zero (B = 1.329, SE = 0.007; p < .001; OR = 3.781). Therefore, returners were 3.8 times more likely to exit VR Services with employment relative to those who were one-time service recipients. The model-implied probabilities of employment at exit were 0.43 and 0.74 for one-time and returners, respectively.
Hourly Wage and Weekly Hours
Of those who exited with employment, the expected wage was 15.760 (SE = 0.02) dollars an hour. Next, we examined whether the difference in hourly wage was significantly different from zero between one-time recipients and returners. On average, returners are expected to earn 4.657 dollars less than those who were one-time recipients. The expected number of hours worked at exit, across all VR consumers, was 29.444 (SE = 0.14) a week. When examining for any differences in hours worked between one-time recipients and returners, we observed a significant difference of 8.641 (SE = 0.042) fewer hours per week for returners.
Services Associated with Employment Outcomes Among Returners (RQ 3)
Employment Outcomes by Service Type
To examine the impact services had on exiting with employment, we inserted indicator variables corresponding to the nine most observed services into a logistic regression. Upon fitting this model, we found that returners who received Maintenance (OR = 1.259), Diagnosis and Treatment of Impairments (OR = 1.556), Job Placement Assistance (OR = 1.700), and Information and Referral Services (OR = 1.270) were more likely to exit with employment; whereas, returners who received Assessment (OR = 0.808), Vocational Rehabilitation Counseling and Guidance (OR = 0.703), Transportation (OR = 0.821), and Job Readiness Training (OR = 0.604) were less likely to exit with employment; finally, the impact of Job Search Assistance was found to be non-significant relative to the unmodeled services.
Hourly Wage and Weekly Hours by Service Type
The expected hourly wage at exit for returners was estimated to be 8.640 (SE = 0.027). Next, we inserted predictor variables corresponding to the same VR services as before to investigate differences in hourly wage. Of the nine services modeled, one was found to be nonsignificant (Maintenance); five are expected to reduce hourly wages by as much as 2.316 (SE = 0.233) dollars an hour (Job Readiness Training); and three are expected to increase hourly wage by as much as 3.245 (SE = 0.190) for Diagnosis and Treatment of Impairments.
The expected hours of work at exit for returners was estimated to be 16.185 (SE = 0.041). Controlling for the commonly observed services, we found significant differences for all modeled services. Specifically, Vocational Rehabilitation Counseling and Guidance, Assessment, Job Search Assistance, Transportation, Job Readiness Training were associated with fewer hours per week, ranging from 2.534 (Vocational Rehabilitation Counseling and Guidance) to 1.046 (Assessment); whereas, Maintenance, Diagnosis and Treatment of Impairments, Job Placement Assistance, and Information and Referral Services were all associated with more hours per week, ranging from 6.623 (Diagnosis and Treatment of Impairments) to 1.13 (Maintenance and Information and Referral Services). Table 3 presents the model-implied probabilities of employment at exit for returners and non-returners, as well as employment probabilities by service type among returners.
Model-Implied Probabilities of Employment at Exit and Employment Outcomes by Service Type.
Discussion
This study identified 110,061 VR returners, representing 6.8% of all service consumers, a substantial population that has received limited research attention. This finding alone highlights the importance of understanding return patterns, as nearly 7% of VR participants re-engage with services following case closure, suggesting that returning to VR services should not be viewed as a failure but rather as a natural response to evolving career and life circumstances.
Demographic Characteristics Predicting Return
Our findings reveal complex intersections between race, benefit receipt, and VR service utilization that challenge simplistic interpretations of racial disparities. While Black and Asian individuals initially demonstrated higher return rates compared to White consumers, this relationship changed substantially after accounting for receipt of public benefits (SSI, SSDI, or TANF). When we controlled for public benefit receipt, racial differences in return rates largely disappeared, with White and Black consumers receiving benefits showing nearly identical return patterns. This mediation effect suggests that benefit receipt status—rather than race itself—serves as the primary predictor of repeated VR service needs. The analysis becomes more complex when considering that Black and American Indian individuals were significantly more likely to receive disability benefits in the first place, while Hispanic individuals were less likely—patterns that may reflect documented disparities in benefit determination processes and access to healthcare systems necessary for disability documentation. The lower return rate among Hispanic individuals presents a unique pattern that warrants careful interpretation. Despite being less likely to receive disability benefits, Hispanic individuals also showed lower return rates. This may indicate different patterns of service engagement, reliance on informal support networks, or potential barriers to initial service access rather than reduced need for support.
Other demographic factors also contributed to nuanced return patterns. Females were slightly less likely to return. This may reflect gendered patterns of workforce participation, caregiving responsibilities that influence career continuity, or differential experiences with workplace accommodations. Veterans also showed significantly lower return rates compared to non-veterans, although the specific factors driving this pattern—whether access to VA-specific services, different career trajectories, or other factors—require further investigation. Educational attainment presented a nonlinear pattern. Individuals with vocational certificates were more likely to return than those without such credentials, while those with graduate degrees showed no significant difference in return rates. The mechanisms driving these differences warrant further investigation. The particularly high return rates among those with vocational certificates could reflect various factors—such as the dynamic nature of technical fields, periodic recertification needs, or greater awareness of VR services—but additional research examining the specific circumstances prompting returns among credentialed individuals would be needed to understand these patterns.
Disability-related characteristics emerged as particularly strong predictors of return patterns. Individuals with the most significant disabilities were substantially more likely to return, as were those with significant disabilities compared to those with less severe disabilities. This pattern aligns with the understanding that more severe disabilities may require ongoing support as conditions evolve or workplace demands change. In addition, the analysis by disability type revealed that individuals with sensory disabilities (auditory or visual) showed higher return rates compared to those with mental disabilities (psychological or psychosocial). This finding aligns with Marsh et al.'s (2022) observation that individuals who are deaf or hard of hearing are more likely to return for multiple VR sequences. These patterns may reflect periodic needs for assistive technology updates, accommodation adjustments, or support navigating technological changes in the workplace. Conversely, individuals with physical or intellectual and learning disabilities showed lower return rates, which may suggest either more stable employment outcomes or different support needs not addressed through repeated VR engagement.
The analysis of employment barriers revealed that experiencing any of the twelve measured barriers significantly increased return likelihood. Particularly strong associations emerged for those who had exhausted TANF benefits and those identified as low-income, with return rates nearly doubling for these groups. The relationship between medical insurance and return added another layer of complexity, with those having employer-based coverage showing higher return rates—possibly reflecting the precarious nature of employment-based benefits where job loss threatens both income and healthcare access.
Employment Outcomes Among Returners
Returners demonstrated a complex employment profile that challenges simple interpretations of success. While they achieved substantially higher employment rates at case closure (74.3%) compared to one-time service consumers (43.4%), these outcomes came with important trade-offs. Returners earned approximately $4.66 less per hour and worked 8.6 fewer hours per week than one-time service consumers. These patterns align with Russell's (2021) findings and may reflect several intersecting factors. The high proportion of returners receiving disability benefits creates strong incentives for strategic part-time work to maintain benefit eligibility while avoiding benefit cliffs. Additionally, the higher rates of significant disabilities among returners may necessitate reduced work schedules to manage health conditions effectively. The employment patterns suggest that returners may be using VR services to maintain workforce attachment within the constraints of their disability-related needs and benefit structures. This interpretation is supported by Marsh et al.'s (2022) finding that employment outcomes remained steady at approximately 51% across multiple service sequences, while the time between sequences decreased with each return. Together, these patterns suggest that VR services may be functioning as intended for many returners, supporting sustained workforce participation even if that participation differs from full-time, higher-wage careers.
Service Effectiveness for Returners
The differential effectiveness of VR services for returners provides important insights into service targeting and utilization. Services directly addressing employment acquisition and health management showed the strongest positive associations with employment outcomes. Job Placement Assistance and Diagnosis and Treatment of Impairments emerged as particularly effective, suggesting these services address critical barriers to workforce entry and maintenance for returners. Conversely, services such as Assessment, Job Readiness Training, and Vocational Rehabilitation Counseling showed negative associations with employment outcomes. However, this pattern likely reflects appropriate service targeting rather than service ineffectiveness. Given that returners were more likely to have significant disabilities and face multiple barriers, these preparatory services may be appropriately directed toward individuals requiring intensive support before competitive employment becomes viable.
Russell (2021) similarly found that specific services like On-the-Job Training were significantly associated with return for consumers with psychosocial impairments, suggesting that some individuals with significant disabilities may need additional or continuous support following successful case closure. Our findings extend this observation by demonstrating how different services may serve different functions within returners’ employment trajectories, some facilitating immediate job placement while others provide necessary groundwork for future employment success.
It is important to note that the odds ratios presented above represent associations between each service type and employment outcomes in the context of other services received. Because returners typically receive multiple services during their VR episodes, these findings should not be interpreted as suggesting that any single service drives employment outcomes in isolation. Rather, VR practitioners should view these results as indicators of which service combinations and patterns are associated with superior outcomes, recognizing that effective VR service delivery typically involves a coordinated bundle of services tailored to individual needs and circumstances. For example, returners who received Diagnosis and Treatment of Impairments demonstrated stronger employment outcomes, particularly when combined with Job Placement Assistance—suggesting that addressing health-related barriers while simultaneously facilitating job acquisition may be a particularly effective service strategy. Similarly, the reduced employment outcomes associated with Assessment and Job Readiness Training services may reflect that these services are more commonly provided to returners with greater barriers or complexity, rather than indicating that the services themselves are ineffective. VR practitioners should therefore consider the full constellation of services provided and the characteristics of consumers receiving them when interpreting these findings for their own practice settings.
Implications
These findings have several important implications for VR policy and practice. The substantial proportion of returners (6.8%) indicates a need for systematic approaches to tracking and supporting individuals across multiple service episodes. The WIOA's transition from labor exchange to career pathways recognizes that people require continuous assistance during their working lives, and our findings support this framework.
Current performance accountability frameworks that emphasize case closure rates may need revision to recognize returns as potential indicators of successful long-term workforce engagement rather than service failure. The higher employment rates among returners, despite lower wages and hours, suggest that repeated VR engagement can successfully support sustained workforce participation for individuals with complex needs.
The intersection of race, disability benefits, socioeconomic status, and return patterns highlights the need for culturally responsive services that address structural inequalities. VR agencies should examine their practices to ensure equitable access and outcomes across racial and ethnic groups, while recognizing that addressing return patterns may require focusing on the underlying economic circumstances that create need for repeated support. For example, previous studies have demonstrated critical connections between disability, race, and economic hardship (e.g., Lombardi et al., 2023; Morris et al., 2022; Murray et al., 2015) with regard to how poverty has a differential and compounded negative influence for disabled adults. Supported career development that VR provides ensures disabled adults can remain in the workforce.
Given that technological advancement and industry restructuring continue to reshape labor markets, VR agencies should develop robust career pathway programs supporting skill development throughout individuals’ working lives. This is particularly important for those with sensory disabilities who may need periodic updates to assistive technologies and accommodations.
Limitations
This study has several limitations that should be acknowledged. First, the RSA-911 administrative data, while comprehensive, lacks detailed information about the quality of employment outcomes, job satisfaction, and long-term career advancement. Second, the study cannot establish causal relationships between services and outcomes due to its observational design. Third, this study period predates major economic disruptions, including the COVID-19 pandemic, which may have altered return patterns. Finally, while the RSA-911 data includes state identifiers, this analysis did not examine state-level variations in VR service delivery or local labor market conditions that may influence return patterns and employment outcomes.
Conclusion
This study examined demographic characteristics, service patterns, employment outcomes among 110,061 VR returners (6.8% of all consumers). Results showed that apparent racial disparities in return rates were largely mediated by public benefit receipt. Economic circumstances, rather than race itself, emerged as the primary driver of return patterns. However, significant upstream racial disparities in benefit receipt itself warrant attention, as Black and American Indian individuals were more likely to receive benefits while Hispanic individuals were less likely. Additional demographic patterns included lower return rates among females and veterans, higher rates among those with vocational certificates, and varying patterns by disability type. While VR returners achieved substantially higher employment rates (74.3%) compared to one-time consumers (43.4%), they worked fewer hours and earned lower wages, suggesting strategic use of VR services to maintain workforce participation within disability-related and benefit constraints. Service analysis revealed that Job Placement Assistance and Diagnosis and Treatment showed the strongest positive associations with employment outcomes. These findings challenge traditional notions of VR “success,” indicating that return to services represents adaptive career navigation rather than service failure. Future research should examine career trajectories longitudinally and explore state-level variations in return patterns.
Footnotes
Acknowledgements
n/a
Ethical Considerations
n/a
Informed Consent
n/a
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The contents of this manuscript were developed under a grant, the Vocational Rehabilitation Technical Assistance Center for Quality Employment, H264K200003, from the U.S. Department of Education. However, those contents do not necessarily represent the policy of the U.S. Department of Education, and you should not assume endorsement by the Federal government.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
