What Predicts Job Quality of Vocational Rehabilitation Consumers Who Are Blind or Have Low Vision?

Abstract

The U.S. Workforce Innovation and Opportunity Act emphasizes promoting high-quality, competitive employment for people served by vocational rehabilitation (VR), but few studies have assessed VR consumers’ job quality. The purpose of this study was to investigate job quality and factors that predict the job quality for VR consumers with blindness or low vision (B/LV), taking into consideration their employment status at application. We utilized RSA-911 data of VR consumers with B/LV who were closed in competitive employment during 2015, creating two separate hierarchical linear models to predict job quality for VR consumers (a) who were employed at application and (b) who were not employed at application. We investigated individual-level (consumer personal characteristics and VR services) and state/agency-level predictors. Job quality and some predictors of job quality differed by employment status at application, although the strongest predictors (education level at application, gender, benefit receipt at application, receipt of a bachelor’s or higher degree) were consistent across the models. While several additional individual-level variables were significantly associated with job quality, their effect sizes were very small. With the exception of advancing education to a bachelor’s degree or higher while receiving services, consumer characteristics at application were the primary determinants of their job quality.

Keywords

career/vocational blind or visual impairment

A large, persistent gap in employment rates exists between people with blindness or low vision (B/LV) and the general U.S. population (Lauer et al., 2020; McDonnall & Sui, 2019). In 2008–2017, employment rates were 33 to 37 percentage points lower for people with B/LV compared with people without disabilities (McDonnall & Sui, 2019). Similar gaps between these groups are evident in other facets of employment. According to recent estimates from the American Community Survey, full-time/full-year employment rates and median annual earnings of full-time/full-year employees were substantially lower for people with B/LV compared with the general population (Erickson et al., 2020). These discrepancies underscore the need to examine aspects of employment beyond basic dichotomous employment rates—such as quality of employment—for people with B/LV, as full-time employment and higher income jobs are closely related to higher job quality (Rothwell & Crabtree, 2019).

Job quality is a growing topic of interest among policymakers and researchers from multiple disciplines (Burchell et al., 2014; Howell & Kalleberg, 2019), and conceptualizations of job quality vary across disciplines and studies. Job quality measures are typically multidimensional and may include various combinations of objective and subjective indicators, such as earnings, health insurance, retirement benefits, job security and flexibility, job satisfaction, career advancement opportunities, and psychological well-being (Findlay et al., 2013; Howell & Kalleberg, 2019; Rothwell & Crabtree, 2019). Research indicates that job quality is associated with quality of life (Rothwell & Crabtree, 2019) in the general population, and the Workforce Innovation and Opportunity Act (WIOA, 2014) emphasizes promoting high-quality, competitive employment for people with disabilities. However, job quality has been minimally explored in studies focusing on people with B/LV (Lund & Cmar, 2019b, 2020).

Research on factors associated with employment outcomes for people with B/LV has been synthesized in four systematic literature reviews (Goertz et al., 2010; Lund & Cmar, 2019a, 2019b, 2020). These reviews and articles published following the reviews (Giesen & Lang, 2018; Zapata, 2020) provide a holistic view of three decades of literature on employment predictors for the B/LV population across various data sources and samples, including vocational rehabilitation (VR) consumers. The most common outcome measure in these studies was a dichotomous employment indicator (e.g., employed or not), typically conceptualized as competitive employment in Rehabilitation Services Administration Case Service Report (RSA-911) analyses. Few studies included outcomes such as earnings and job quality. Socioeconomic factors—particularly education level and previous work experience—were the most prominent predictors of employment outcomes in this body of literature. Education was positively associated with employment (Goertz et al., 2010; Lund & Cmar, 2019a, 2019b, 2020; Zapata, 2020), job quality and higher-level jobs (Lund & Cmar, 2019a, 2020), and earnings (Giesen & Lang, 2018; Lund & Cmar, 2019a, 2019b). Variables representing work experience, including working since the onset of disability, earnings at VR application, and self-support at VR application, were positively associated with employment (Lund & Cmar, 2019a, 2019b, 2020) and earnings above substantial gainful activity levels for Social Security Disability Insurance (SSDI) recipients (Giesen & Lang, 2018).

Other socioeconomic and demographic predictors had less consistent evidence or were studied infrequently; however, several patterns emerged. Supplemental Security Income (SSI) or SSDI receipt predicted lower odds of competitive employment for VR consumers in several studies (Lund & Cmar, 2019b, 2020), and higher SSDI amount predicted better employment outcomes for SSDI recipients (Giesen & Cavenaugh, 2013; Giesen & Lang, 2018). Male gender was positively associated with employment in some studies (Goertz et al., 2010; Lund & Cmar, 2019a, 2019b, 2020); associations between male gender and higher earnings were more consistent (Giesen & Lang, 2018; Lund & Cmar, 2019a, 2019b). Legal blindness (vs. less severe visual impairment) was negatively associated with employment in several studies of VR consumers (Lund & Cmar, 2019b, 2020), but other studies yielded mixed findings (Goertz et al., 2010; Lund & Cmar, 2019a, 2020). Associations between secondary disabilities and employment outcomes were typically either negative (Giesen & Lang, 2018; Goertz et al., 2010; Lund & Cmar, 2019b, 2020) or nonsignificant (Lund & Cmar, 2019a). Analyses of age, race, and ethnicity yielded mixed findings across studies (Giesen & Lang, 2018; Goertz et al., 2010; Lund & Cmar, 2019a, 2019b, 2020; Zapata, 2020).

VR consumers who received rehabilitation technology, on-the-job supports, job placement assistance, and job search assistance had better employment outcomes than consumers who did not receive those services (Cimera et al., 2015; Giesen & Cavenaugh, 2012; Giesen & Hierholzer, 2016; Giesen & Lang, 2018). Moreover, technology training was associated with employment in higher level jobs after referral for vocational placement (Leonard et al., 1999). One job-related service, job readiness training, has evidence for a negative relationship with employment (Cimera et al., 2015; Giesen & Hierholzer, 2016; Giesen & Lang, 2018). Receipt of college or university training was positively associated with employment outcomes for transition-age youth (Lund & Cmar, 2020); however, Capella-McDonnall (2005) found that this service was only related to competitive employment for VR consumers when it resulted in the receipt of a degree or certificate. Other VR services have limited evidence for their relationship with employment outcomes.

Few studies included state- and agency-level predictors of employment for people with B/LV. Results have been inconsistent, but some state-level economic indicators have been significant predictors of employment: State unemployment rate was negatively associated with employment for SSDI recipients (Giesen & Cavenaugh, 2013; Giesen & Lang, 2018), and the employment–population ratio was positively associated with employment for consumers who were not employed at application (McDonnall, 2016). These findings support the importance of accounting for state economic conditions in analyses of employment outcomes. Considering the structure of the VR service-delivery system in the United States, agency type has been a factor of interest in several RSA-911 studies. Some states have a separate agency that serves only consumers with B/LV and a general agency that serves consumers with other disabilities (including some who have mild visual impairments or additional disabilities). Other states have one combined agency that serves all consumers in that state. Although findings regarding agency type have varied (Lund & Cmar, 2019b), receiving services from a separate agency (compared with a general or combined agency) was associated with better employment outcomes in some studies (e.g., Giesen & Cavenaugh, 2013; Giesen & Lang, 2018; McDonnall & Cmar, 2019).

Findings from this body of literature highlighted areas of need for future employment research, some of which are particularly relevant to VR consumers with B/LV. First, the limited research on job quality for adults with B/LV identified through the systematic literature reviews corroborates the critical need for research in this area (Lund & Cmar, 2019a, 2019b). Because competitive jobs do not necessarily equate to high-quality jobs for people with disabilities (Heyman et al., 2016), distinguishing between the outcomes of competitive employment and job quality is vital. Second, employment status at VR application has received little attention in previous B/LV RSA-911 studies, although some researchers accounted for previous employment or earnings. Considering that job retention is a role of VR, and employment status at application was a strong predictor of competitive employment for consumers with sensory disabilities (Dutta et al., 2008), differentiating between employed and unemployed applicants is warranted.

Only a few studies have focused on VR consumers with B/LV who were employed at application, although employed applicants comprised approximately one third of B/LV case closures in fiscal year (FY) 2015 (Crudden, Giesen, & Sui, 2018; Crudden, McDonnall, & Sui, 2018). Crudden, Giesen, and Sui (2018) documented several notable differences between competitively employed and unemployed B/LV applicants in characteristics and service receipt. Compared with unemployed applicants, employed applicants were more likely to be older and White, not have a noncognitive disability, have higher education levels, and not receive SSI or SSDI (Crudden, Giesen, & Sui, 2018). Furthermore, employed applicants were more likely to receive certain VR services, including on-the-job supports–short-term, on-the-job supports–supported employment, and rehabilitation technology (Crudden, Giesen, & Sui, 2018). In another study of employed B/LV applicants, Crudden, McDonnall, and Sui (2018) identified several predictors of job retention, including being male, not having a secondary disability, having at least a bachelor’s degree, being a Business Enterprise Program vendor at application, having a previous successful VR closure, and working more hours at application.

No research has focused on predictors of job quality for the broader population of VR consumers with B/LV (Lund & Cmar, 2019b), although one RSA-911 analysis of job quality for transition-age youth was conducted (Cimera et al., 2015). Another job quality study included VR consumers with all types of disabilities (Chan et al., 2016). Two subsequent analyses of job quality focused on small subpopulations of individuals with B/LV: VR consumers who are deaf-blind (McDonnall & Cmar, 2019) and VR consumers with combined traumatic brain injury and B/LV (McDonnall et al., 2020). Across the studies, personal characteristics were the primary predictors of job quality, with education level and male gender predicting higher job quality and receipt of cash benefits (including SSI or SSDI) predicting lower job quality (Chan et al., 2016; Cimera et al., 2015; McDonnall & Cmar, 2019; McDonnall et al., 2020). Conversely, relationships between job quality and state-level, agency-level, and service-related predictors varied across studies (Chan et al., 2016; Cimera et al., 2015; McDonnall & Cmar, 2019; McDonnall et al., 2020).

Despite the history of employment disparities between people with B/LV and the general U.S. population, and the recent policy emphasis on high-quality employment for all people with disabilities, few studies have focused on job quality for people with B/LV. Research on B/LV VR consumers’ job quality is limited to transition-age youth and small subpopulations. Focusing on job quality aligns with the goals of WIOA and can provide a more comprehensive understanding of consumers’ employment outcomes following VR case closure. Accordingly, the purpose of this study was to investigate job quality among adult VR consumers with B/LV and identify factors that predict high-quality, competitive employment for those consumers who enter VR with a job and those who enter VR without a job.

Method

Sample

We utilized FY 2015 RSA-911 data for this study. Our sample consisted of 8,723 consumers with a primary disability of “blindness” or “other visual impairment” who received services from separate, general, or combined VR agencies whose cases were closed with competitive employment and who were between the ages of 18 and 67 years at case closure. Competitive employment was defined as working and earning at least the federal minimum wage and included consumers closed with or without supports in an integrated setting, as self-employed, and in the Business Enterprise Program. We excluded consumers who were served by agencies in the U.S. territories. We divided our sample into two subsamples for the analyses: consumers who were competitively employed at application (n = 4,133) and those who were not competitively employed at application (n = 4,590).

A slight majority (53.9%) of consumers were male, and a slight majority (52.7%) had visual impairments less severe than legal blindness. Just more than a third (34.9%) had another disability in addition to visual impairment or blindness. The majority (71.5%) of consumers in the sample were White, 23.8% were African American, 2.2% were Asian, 1.4% were of mixed race, <1% were American Indian, <1% were Pacific Islander, and 11.5% reported Hispanic ethnicity. In terms of education level at application, 18.4% of the sample had less than a high school education, 32% had a high school diploma or equivalent, 20% had education beyond high school but no academic degree, 8.7% had an associate degree, 13.6% had a bachelor’s degree, and 7.3% held an advanced degree. Additional descriptive information about the sample as a whole, and information by subsample, is provided in Table 1.

Table 1.

Descriptive Statistics for Individual-Level and Outcome Variables.

Variable	Employed at application	Not employed at application	Overall
Consumer personal characteristics
Age at closure	49.2 (11.8)	41.9 (14.0)	45.3 (13.5)
Education	5.7 (2.4)	4.8 (2.3)	5.3 (2.4)
Female	49.1	43.5	46.1
African American	21.6	25.9	23.9
Asian	2.2	2.2	2.2
American Indian	0.8	0.9	0.9
Multiple races	0.9	1.5	1.2
Hispanic ethnicity	11.3	11.7	11.5
Legal blindness	45.4	49.0	47.3
Additional disability	33.0	36.7	34.9
Receipt of SSDI	19.6	26.6	23.3
Receipt of SSI	6.6	19.0	13.1
Source of support (personal income)	76.4	9.4	41.2
Receipt of certificate	1.3	2.9	2.1
Receipt of associate degree	1.6	2.8	2.2
Receipt of bachelor’s or higher degree	3.4	11.5	7.7
Service-related factors
Disability-related skills training	27.2	32.5	30.0
Rehabilitation technology	60.5	60.7	60.6
Job placement assistance	9.0	32.8	21.5
Job search assistance	5.4	22.8	14.5
Job readiness training	3.3	17.5	10.8
On-the-job supports–supported employment	2.08	4.58	3.39
On-the-job supports–short term	8.66	15.32	12.16
Job Quality Index (factor score)	0.21 (1.12)	–0.19 (0.77)	0.0 (0.97)

Note. Values are percentages or means. Values in parentheses are standard deviations adjacent to means. SSDI = Social Security Disability Insurance; SSI = Supplemental Security Income. Total N = 8,723: employed at VR application n = 4,133, and not employed at application n = 4,590.

Variables and Measures

Job quality index

The dependent variable was an index created by the authors to represent job quality. The measure is comprised of three variables: (a) whether the consumers’ job provided medical insurance (yes = 1, no = 0), (b) the proportion of consumers’ weekly earnings to their state’s living wage, and (c) the proportion of consumers’ hourly wage to their state’s median hourly wage. Living wage is an established minimum income that provides for financial independence; it is higher than the poverty threshold and considers the cost of living within states (Nadeau, 2017). State median hourly earnings were obtained from the U.S. Bureau of Labor Statistics website (https://www.bls.gov/oes/tables.htm). Principal component analysis (PCA) was utilized to create the composite index variable. We included all consumers with B/LV who were closed with employment to develop the index. The principal axis method was utilized to extract the components, and ones were used as prior communality estimates. The Kaiser–Meyer–Olkin Measure of Sampling Adequacy of .54 met the minimum acceptable cutoff (Kaiser & Rice, 1974), and Bartlett’s Test of Sphericity was significant (10,206.66, p < .0001), supporting the appropriateness of our data for PCA. We determined that one factor provided an appropriate solution for the data, based on the eigenvalue >1 criteria, and all factor loadings were above .40 (i.e., .58, .93, and .88). This factor explained 66% of the total variance. The job quality index (JQI) is a standardized factor score (a linear composite of the weighted observed variables) with a mean of zero and an SD of 1. Positive scores represent job quality that was higher than the average job quality of all consumers with B/LV who were closed in 2015, and negative scores indicate job quality that was lower than average.

State- and agency-level predictors

To account for the hierarchical nature of the RSA-911 data, we included both Level 2 (state/agency) and Level 1 (individual) variables in our models. We included four state-level economic indicators to control for the economic conditions within consumers’ states: per capita personal income (SSTI, 2016), employment–population ratio (U.S. Census Bureau, 2018a), the percentage living below poverty (U.S. Census Bureau, 2018b), and the state gross domestic product (U.S. Bureau of Economic Analysis, 2020). We selected these indicators based on the hypothesis that they may be more closely related to state earning levels than other economic indicators. The four state-level variables were grand-mean centered prior to their inclusion in the statistical analyses. We also included agency type as a dichotomous variable to explore the effect of separate agencies on job quality (1 = separate agency, 0 = combined or general agency).

Individual-level predictors

Given the limited research on predictors of job quality for consumers with B/LV, we included individual-level predictors with evidence for a relationship with any type of employment outcome for this population. Variables consisted of consumer personal characteristics and specific services received or service-related factors. We also included disability-related skills training, as this service can help consumers develop important workplace skills. Personal characteristics treated as dichotomous variables included: (a) race and ethnicity, assessed with five variables (African American race only, Asian race only, American Indian race only, mixed race, and Hispanic ethnicity of any race); (b) level of vision loss (legal blindness = 1, less severe visual impairment = 0); (c) sex (female = 1, male = 0); (d) presence of an additional disability; (e) receipt of SSI at application; (f) receipt of SSDI at application; and (g) personal income as a primary source of support at application. Continuous personal characteristics were age at case closure and education level, which ranged from 0 (no formal schooling) to 12 (academic degree above a master’s). These continuous variables were grand-mean centered prior to their inclusion in the models. Service-related variables were all dichotomous and fell into two categories: (a) the receipt of 7 services (i.e., disability-related skills training, rehabilitation technology, job placement, job search assistance, job readiness, on-the-job supports–supported employment, and on-the-job supports–short-term) and (b) three variables representing advancement in education while receiving VR services (i.e., receipt of a certificate, receipt of an associate degree, and receipt of a bachelor’s or higher degree). The educational advancement variables were calculated by comparing education level at application to education level at case closure.

Statistical Analyses

We utilized hierarchical linear modeling and conducted separate, but identical, analyses with both samples to determine whether the associations between job quality and (a) personal characteristics and (b) services were different for consumers who were employed and who were not employed at VR application. We first evaluated the variability in job quality across state VR agencies by estimating unconditional two-level models, which did not include any predictor variables. Then, we entered the 5 state/agency-level and 23 individual-level variables of interest to examine the effect of each variable on the JQI. PROC MIXED in SAS Version 9.4 using maximum likelihood estimation was utilized for the analyses. Given the large sample sizes, an alpha level of .01 was used to determine statistical significance.

We utilized f ² as a quantitative measure of the magnitude of the predictor variables’ association with the outcome. We calculated this effect size for all Level 1 and Level 2 variables in the final models. This effect size measures the proportion of variance explained by a given fixed effect variable relative to the proportion of outcome variance unexplained by the final model (Lorah, 2018). A larger effect size reflects a stronger relationship between the JQI and the predictor variable. Standard conventions for the magnitude of f² are small at 0.02, medium at 0.15, and large at 0.35 (Cohen, 1992).

Results

Job Quality of VR Consumers With B/LV

The JQI had a range of −1.03 to 12.78 and a median of −0.35 for our sample. There were several outliers on the higher end of the scale: 120 people had scores three or more standard deviations above the mean. Considering the individual factors that make up the JQI to evaluate job quality, 25.7% of our overall sample had a job that provided medical insurance, 37.5% earned at or above a living wage, and 22.7% earned at or above their state’s median hourly wage. Weekly earnings ranged from 3% to 1370% of state living wage levels and hourly wages ranged from 30% to 1090% of state median wages, but most people had earnings on the lower end of the range (medians of 82.5% and 65.5%, respectively). Job quality indicators were higher for consumers who were employed at application (33.4% had medical insurance, 47.7% earned at least a living wage, 29.3% earned at least their state’s median hourly wage) than for those not employed at application (18.9% had medical insurance, 28.3% earned at least a living wage, and 16.8% earned at least their state’s median hourly wage).

Employed at Application Model

In the unconditional model, the Level 2 variance component was 0.065, and the Level 1 variance component was 1.189, resulting in an intraclass correlation coefficient (ICC) of 0.052, which indicates that 5.2% of the total variance in job quality was attributed to differences among state agencies. The only fixed effect estimation in the unconditional model was the intercept of 0.267 (SE = 0. 042, df = 69, t = 6.41, p < .001), which indicates an average JQI of 0.267 across consumers in the 70 state agencies in the sample. Because this value is greater than zero, the average job quality at VR case closure for people who were employed at application was greater than the average job quality of all consumers with B/LV.

Results of the full model for the employed sample are displayed in Table 2. Job quality at VR case closure was positively associated with several personal characteristics: older age at VR case closure, education level, legal blindness, and personal income as the primary source of support at application. Job quality was negatively associated with female gender, African American race, having an additional disability, receipt of SSDI, and receipt of SSI. In terms of VR services, job quality was positively related to receiving a bachelor’s or higher degree, disability-related skills training, and rehabilitation technology but was negatively related to receiving job placement assistance. Education level and receipt of SSDI had small-to-medium effect sizes, and other variables had very small or small effect sizes. None of the state-/agency-level variables were associated with job quality. This model explained 26.8% of the variance in job quality at case closure among consumers who were employed at the application.

Table 2.

Hierarchical Linear Model Predicting Job Quality for Vocational Rehabilitation Consumers Who Were Employed at Application.

Variable	Coefficient	SE	t value	p	f ²
Intercept	0.209	0.067	3.12	.003
State/agency level
Per capita personal income	0.005	0.007	0.69	.495	0.001
Employment-population ratio	0.003	0.006	0.41	.685	0.000
Below poverty rate	0.019	0.019	1.02	.312	0.000
Gross domestic product	0.000	0.000	–1.21	.231	0.002
Agency type	–0.003	0.066	–0.04	.968	0.000
Individual level
Age at closure	0.004	0.001	3.39	.001	0.002
Education	0.153	0.007	23.12	<.001	0.135
Female	–0.280	0.030	–9.37	<.001	0.019
African American	–0.130	0.038	–3.45	.001	0.004
Asian	–0.058	0.103	–0.56	.573	0.000
American Indian	0.119	0.169	0.71	.479	0.000
Multiple races	–0.171	0.159	–1.07	.284	0.000
Hispanic ethnicity	–0.082	0.052	–1.58	.115	0.001
Legal blindness	0.156	0.038	4.13	<.001	0.007
Additional disability	–0.105	0.033	–3.19	.001	0.003
Receipt of SSDI	–0.576	0.044	–13.06	<.001	0.042
Receipt of SSI	–0.425	0.063	–6.72	<.001	0.011
Personal income as primary source of support	0.252	0.041	6.16	<.001	0.006
Receipt of certificate	–0.037	0.133	–0.28	.782	0.000
Receipt of associate degree	0.021	0.119	0.18	.859	0.000
Receipt of bachelor’s or higher degree	0.544	0.083	6.54	<.001	0.013
Disability-related skills training	0.147	0.038	3.84	.000	0.004
Rehabilitation technology	0.183	0.037	4.99	<.001	0.009
Job placement assistance	–0.179	0.059	–3.04	.002	0.003
Job search assistance	–0.056	0.074	–0.76	.447	0.000
Job readiness training	–0.186	0.088	–2.10	.035	0.000
On-the-job supports–supported employment	–0.041	0.116	–0.35	.725	0.000
On-the-job supports–short-term	0.024	0.062	0.39	.696	0.000

Note. n = 4,133 at the individual level; n = 70 at the state/agency level; f² = effect size measure; Intraclass correlation coefficient, ICC = 0.036 / (0.036 + 0.882) = 0.040. SSDI = Social Security Disability Insurance; SSI = Supplemental Security Income.

Not Employed at Application Model

In the unconditional model, the estimated Level 2 variance component of 0.020 and the Level 1 variance component of 0.579 yielded an ICC of 0.033, indicating 3.3% of the total variance in job quality was between agencies. The average JQI at VR case closure was −0.189 (SE = 0.023, df = 73, t = −8.13, p < .001) for people who were not competitively employed at VR application across the 74 state agencies in the sample. This negative value indicates that job quality for this sample was lower than the average job quality of all consumers with B/LV at VR case closure.

Statistical results of the final model for the not employed at application sample are displayed in Table 3. Regarding personal characteristics, job quality at closure was positively associated with education level and legal blindness, while job quality at closure was negatively associated with female gender, African American race, having an additional disability, receipt of SSDI, and receipt of SSI. In terms of VR services, receipt of a certificate, receipt of an associate degree, and receipt of bachelor’s or higher degree were positively associated with job quality at VR case closure, while receipt of job readiness training and on-the-job supports–supported employment were negatively associated with job quality. Education level and receipt of bachelor’s or higher degree had small-to-medium effects, and other variables had very small effects. None of the state/agency-level variables were associated with job quality. This model explained 18.7% of the variance in job quality at case closure among consumers who were not employed at the application.

Table 3.

Hierarchical Linear Model Predicting Job Quality for Vocational Rehabilitation Consumers Who Were Not Employed at Application.

Variable	Coefficient	SE	t value	p	f ²
Intercept	–0.099	0.035	–2.87	.006
State/agency level
Per capita personal income	–0.004	0.004	–0.93	.357	0.001
Employment-population ratio	0.008	0.004	2.18	.033	0.001
Below poverty rate	0.025	0.011	2.29	.025	0.002
Gross domestic product	0.000	0.000	–0.21	.832	0.000
Agency type	0.027	0.039	0.71	.482	0.000
Individual level
Age at closure	–0.002	0.001	–2.22	.026	0.002
Education	0.078	0.005	16.03	<.001	0.059
Female	–0.130	0.021	–6.18	<.001	0.008
African American	–0.085	0.025	–3.40	.001	0.003
Asian	0.155	0.072	2.17	.030	0.001
American Indian	0.119	0.108	1.10	.272	0.000
Multiple races	–0.106	0.085	–1.25	.211	0.000
Hispanic ethnicity	–0.040	0.035	–1.14	.254	0.000
Legal blindness	0.072	0.025	2.89	.004	0.003
Additional disability	–0.073	0.022	–3.25	.001	0.003
Receipt of SSDI	–0.180	0.026	–6.90	<.001	0.011
Receipt of SSI	–0.175	0.028	–6.28	<.001	0.007
Personal income as primary source of support	0.042	0.036	1.17	.243	0.001
Receipt of certificate	0.202	0.062	3.24	.001	0.002
Receipt of associate degree	0.221	0.064	3.48	.001	0.003
Receipt of bachelor’s or higher degree	0.684	0.035	19.73	<.001	0.086
Disability-related skills training	0.016	0.026	0.60	.546	0.000
Rehabilitation technology	0.055	0.025	2.19	.029	0.002
Job placement assistance	–0.061	0.026	–2.39	.017	0.002
Job search assistance	–0.063	0.029	–2.20	.028	0.002
Job readiness training	–0.097	0.031	–3.18	.002	0.001
On-the-job supports–supported employment	–0.134	0.052	–2.57	.010	0.001
On-the-job supports–short-term	–0.040	0.033	–1.23	.219	0.000

Note. n = 4,590 at the individual level; n = 74 at the state/agency level. f² = effect size measure; Intraclass correlation coefficient, ICC = 0.011 / (0.011 + 0.476) = 0.023. SSDI = Social Security Disability Insurance; SSI = Supplemental Security Income.

Discussion

Little is known about the job quality of individuals with B/LV or the factors that are associated with job quality for this population. To address this gap in the literature, we investigated job quality and predictors of job quality for VR consumers with B/LV who obtained competitive employment, taking into consideration their employment status at application. Typically, almost a third of VR consumers with B/LV are employed at the time they apply for services, and just less than half of the consumers in our sample who achieved competitive employment in FY 2015 were employed at application. Accounting for employment status at application was important as our samples differed at entry and exit from VR services. A majority (76.4%) of consumers who entered VR with employment reported personal income as their primary source of support as application, and this percentage increased to 90.1% at case closure. While only a small portion (9.4%) of those not employed at VR application were self-supporting prior to receiving services, 81.9% reported personal income as their primary source of support at case closure.

We found that the job quality of VR consumers varied widely, and individuals who came to VR with employment had higher average and median job quality at case closure than those who were not employed at application. Reasons for seeking VR assistance may differ for these groups, as do the services they receive (Crudden, Giesen, & Sui, 2018), and therefore, the association between services received and outcomes may differ. We documented in this study that the association between job quality and services received did differ among those who entered VR with employment and those who entered VR without a job. Of note is that the model for consumers who entered VR with employment explained a greater amount of variance in job quality than the model for consumers who entered VR without a job.

Several variables demonstrated a consistent relationship with job quality across the two models. Education level at entry to VR was a significant predictor of higher job quality in both models as found in studies of job quality for the general U.S. population (Brucker & Henly, 2019; Rothwell & Crabtree, 2019) and VR consumers (Chan et al., 2016), including subpopulations of VR consumers with B/LV (Cimera et al., 2015; McDonnall & Cmar, 2019; McDonnall et al., 2020). Education level was the strongest predictor of job quality for VR consumers with B/LV who were employed at application and the second strongest predictor for consumers who were not employed at application. Obtaining a bachelor’s or advanced degree was associated with higher job quality for both groups of consumers. This variable explained more variance in job quality for consumers who were not working at VR applications and was the strongest predictor for this group.

Gender also had a significant association with job quality in both models, with males having higher job quality than females. Similar findings regarding gender and job quality have been documented in the general population (Brucker & Henly, 2019) and other populations of VR consumers (Chan et al., 2016; Cimera et al., 2015; McDonnall & Cmar, 2019; McDonnall et al., 2020). White race was associated with higher job quality in other populations (Brucker & Henly, 2019; Rothwell & Crabtree, 2019), but only African American consumers had significantly lower job quality than consumers of other races in our models, and the effect size was very small. Despite the negative associations between legal blindness (vs. less severe visual impairment) and employment outcomes for VR consumers documented in the literature (Cavenaugh & Rogers, 2002; Cimera et al., 2015; Darensbourg, 2013; Giesen & Cavenaugh, 2012, 2013; McDonnall, 2016), legal blindness was associated with higher job quality in our models. The positive relationship between legal blindness and job quality is supported by Leonard et al.’s (1999) finding that blindness (vs. low vision) was associated with higher level jobs. In both models, SSDI recipients had significantly lower job quality than nonrecipients as found in previous studies (McDonnall & Cmar, 2019; McDonnall et al., 2020). SSDI receipt was the second strongest predictor of job quality for consumers who were employed at application. A potential explanation for this finding is that SSDI recipients limited their earnings to avoid losing benefits.

Although obtaining a bachelor’s or advanced degree was associated with higher job quality for all consumers, obtaining an associate degree or a certificate were only significant predictors of higher job quality for consumers who entered VR without a job. Perhaps this difference is due to the fact that consumers employed at application began VR with a higher quality job, and this type of educational advancement was not able to add significantly to existing job quality. In addition, certifications were strongly associated with job quality only for people in the general population who did not have postsecondary education (Rothwell & Crabtree, 2019), so consumers’ existing education level may determine the impact of obtaining a certificate. While consumers without a job at VR entry were much more likely to obtain an associate degree or certificate, only a small portion of either group obtained these credentials (less than 3%, see Table 1). Associate degrees and certificates offer labor market returns for the general population, although the amount of return varies by field of study and its match to local labor market conditions (Carnevale et al., 2020; Holzer & Dunlop, 2013). Earnings associated with some certificates and associate degrees surpass those for 4-year degrees, and an approximately equal number of these credentials and bachelor’s degrees are awarded each year (Carnevale et al., 2020). Given the value of advancing education but knowing that not all consumers will be interested in or capable of completing a bachelor’s degree, VR counselors should be aware of the opportunities for employment and higher job quality that some certificates and associate degrees afford.

Differences in the associations between specific services received and job quality were observed for the two groups. In general, the job-related services either did not have a relationship with job quality or had a small, negative relationship. Job placement was associated with lower job quality only for consumers employed at application, and job readiness and on-the-job supports-supported employment were associated with lower job quality only for consumers who were not employed at application. These findings are consistent with other studies in which a negative relationship between job-related services and job quality was found (Chan et al., 2016; McDonnall & Cmar, 2019; McDonnall et al., 2020). While job-related services have consistently been associated with obtaining employment, these negative relationships observed in this and other studies indicate that consumers who require assistance from VR with preparing for or locating employment, rather than being able to prepare for and find it on their own, are more likely to obtain lower quality jobs.

Another service difference was the value of receiving disability-related skills training and rehabilitation technology: These services were significantly associated with higher job quality only for consumers who were employed at application. Neither of these services was a significant predictor of job quality in prior studies of subpopulations of consumers with B/LV, which included all consumers (those who came in with employment and those who did not) in one model (McDonnall & Cmar, 2019; McDonnall et al., 2020). Individuals who have a job at application may be seeking VR services for different reasons than those who are not employed. Although the RSA-911 data do not allow us to determine whether employed consumers were seeking assistance with job retention, it is likely that many who enter VR with a job need specific help to maintain their current job (e.g., additional technology, assistance with other accommodations, training in blindness skills). Assistance in these areas would help them maintain an existing job, which is likely to have higher job quality than a new job obtained by someone without a recent work history. It is also possible that people who enter VR with a job have recently experienced vision loss and need initial training in blindness skills and assistance with technology to continue working in their current position. Given that the employed at application group was older and had a higher education level, this group is likely to have more years of work experience and a higher level of earnings.

Only a small amount of variability in our job quality measure was associated with Level 2 factors, indicating that the vast majority of the variability in job quality is attributable to individual-level rather than agency- or state-level factors. None of the state- and agency-level variables in these models were significant predictors of job quality as found in a study of consumers with combined B/LV and traumatic brain injury (McDonnall et al., 2020). Chan and colleagues (2016) did identify state- and agency-level variables that predicted their job quality measure for all consumers served by VR, and they also documented a greater, although still relatively small (i.e., 7%), amount of variance explained at the agency level. One potential reason for the lack of significant state-level economic indicators in our study is that the job quality measure already took into account differences in cost of living and state earnings levels based on how the factor score was developed, while Chan and colleagues’ measure did not. We included only one agency-level variable in our model, agency type. Several studies have investigated the association between agency type and B/LV consumers’ outcomes, and findings have been mixed (Lund & Cmar, 2019b). This study adds to the small number of studies that have documented a lack of meaningful differences in earnings or job quality based on agency type (Capella, 2001; Estrada-Hernandez, 2008; McDonnall et al., 2020).

Limitations and Future Research Directions

Job quality measures commonly include both objective and subjective indicators due to the multidimensional—and often subjective—nature of this construct; however, only objective indicators were available in the RSA-911 dataset. Given this limitation of the data, our job quality measure included objective criteria with a focus on earnings, which may not capture all aspects of job quality that are important to B/LV consumers. For example, some people may not need or want to work full-time or earn a living wage and may seek work that provides intrinsic, lifestyle, or social benefits (Cooke et al., 2013). Some jobs (e.g., self-employment) do not provide insurance and would result in lower job quality in our index, but these jobs may provide other benefits such as flexible work hours. It would be valuable to include subjective indicators (e.g., flexibility, job security, and job satisfaction) in future investigations of job quality for people who are B/LV. Other potential individual-level predictors of job quality that were not available in the RSA-911 dataset would be important to consider in future job quality analyses. One example is self-reported health status, which was associated with job quality for people with and without disabilities in a previous study (Brucker & Henly, 2019).

Several other limitations associated with the data should be acknowledged. RSA-911 is data coded by VR counselors which may be incomplete, similar to self-report data, and it is archival data. It is not possible to establish a causal relationship between variables with this type of data. Another limitation is the age of the data. Because the FY 2015 RSA-911 dataset reflects pre-WIOA services and outcomes, this study does not capture the potential impact of WIOA on B/LV consumers. Furthermore, the data do not reflect the dramatic social and economic changes resulting from the COVID-19 pandemic. Therefore, our results may not generalize to consumers who received VR services in the years following WIOA and COVID-19. Despite this limitation, this study is the first known investigation of job quality for VR consumers with B/LV, and it provides a baseline for comparison with future post-WIOA and post-COVID-19 studies. Future research that uses more current RSA-911 data from multiple fiscal years to investigate the relationship between key demographic and service variables and job quality for B/LV consumers would extend our findings.

Implications for Practice

Several predictors of job quality for consumers with B/LV identified in this study have implications for VR agencies that serve this population. First, our findings extend the literature documenting the importance of educational advancement in improving job quality for VR consumers (Chan et al., 2016; McDonnall et al., 2020) to the B/LV population. The most relevant step that VR counselors can take to increase B/LV consumers’ job quality is to support their obtainment of a degree, particularly a bachelor’s degree or higher. Obtaining an associate degree or a certificate may also improve job quality, although the return on investment can vary. Vocational rehabilitation counselors should be aware of the differences in return on investment for associate degrees and certificates for various fields of study in their local areas and provide guidance to consumers. Counselors should encourage consumers to consider these options and explain that advancing their education can help them secure higher quality jobs.

An additional consideration related to educational advancement and job quality is the supply and demand for various fields of study. An increase in education may not be associated with greater job quality if demand is low and high-quality jobs are not available in that field (Findlay et al., 2017). Howell and Kalleberg (2019) emphasized the need to focus on areas in which demand is projected to increase by obtaining education and upgrading skills for “high-quality jobs of the future” (p. 43). Vocational rehabilitation professionals can use labor market information to learn about high-demand fields, jobs in those fields, education and training requirements, and characteristics of jobs (e.g., salary and working conditions), taking into consideration the labor market in the consumer’s area. Sharing this information with consumers can help them make an informed choice and pursue education in fields of interest that offer suitable jobs that match their definition of quality.

Our findings also indicate that consumers with B/LV who receive SSDI benefits or job-related services may be at risk of low job quality. Regardless of employment status at VR application, SSDI receipt was associated with lower job quality at closure, which substantiates the importance of benefits counseling. Considering that SSDI beneficiaries often lack knowledge about work incentives (Olney & Lyle, 2011), VR counselors can provide education about disability benefits to ensure consumers are aware of and understand the intricacies of work incentives. Benefits counseling may be particularly helpful for consumers who are reluctant to exit the SSDI program but have the skills and qualifications to obtain higher-quality jobs. Employer-provided health insurance may be a critical element of job quality for individuals who fear losing the medical benefits associated with SSDI (Olney & Lyle, 2011). Receipt of job-related services, including job placement, is associated with lower job quality, although receipt of these services likely indicates a need for these services. Compared with consumers who require assistance from the rehabilitation system to find a job, consumers who find a job on their own may be more likely to negotiate higher pay and benefits or have the option of being more selective in their job choices. They may also have a larger social network or be more aware of how to use their social network to assist in obtaining a job. Therefore, teaching and encouraging proactivity and independence in job-seeking could be another avenue for improving consumers’ job quality.

Conclusion

Although this study discovered several previously unidentified variables that were related to the job quality of VR consumers, including the provision of some VR services, the effect sizes for these new variables were very small. An overall conclusion of job quality studies of VR consumers with B/LV is that, with the exception of assisting the person with advancing their education to a bachelor’s or higher degree, VR services do not contribute substantially to job quality. More evidence exists to support the importance of VR services for increasing the chances of B/LV consumers obtaining a job (Cimera et al., 2015; Giesen & Cavenaugh, 2012; Giesen & Hierholzer, 2016; Giesen & Lang, 2018; McDonnall et al., 2020). However, other factors surrounding VR services that are unaccounted for may contribute to consumers’ job quality. For example, differences in counselors’ skills and counselor connections with businesses could potentially influence the job quality of the consumers they serve, but data related to these factors are unavailable in the RSA-911 data. Individual personal characteristics generally play a more predominant role in determining job quality than the available VR service-related factors. While most of these individual factors are beyond the control of VR, the impact of receipt of SSDI benefits is one factor that VR may be able to influence by providing benefits counseling.

Much of the variance in job quality of VR consumers is unexplained, and there is more unexplained variance for job quality associated with newly obtained jobs rather than job retention cases (i.e., consumers who come to VR with employment). Assuming that a majority of consumers who enter VR with a job are seeking to retain that job, their job quality is predetermined at application, although some may advance in their positions through VR services. The larger amount of unexplained variance for consumers who come to VR without employment could be associated with greater discrepancies in the amount of work experience (e.g., some likely have a significant work history and some have no work experience) as well as differences in consumers’ social networks and connections that may influence not only their ability to obtain a job but also their jobs’ wages (Cappellari & Tatsiramos, 2015).

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The contents of this manuscript were developed under a grant from the U.S. Department of Health and Human Services, NIDILRR grant 90RT5040-01-00. However, these contents do not necessarily represent the policy of the Department of Health and Human Services and should not indicate endorsement by the federal government.

ORCID iDs

Michele C. McDonnall

Jennifer L. Cmar

References

Brucker

D. L.

Henly

(2019). Job quality for Americans with disabilities. Journal of Vocational Rehabilitation, 50(2), 121–130. https://doi.org/10.3233/JVR-180994

Burchell

Sehnbruch

Piasna

Agloni

(2014). The quality of employment in the academic literature: Definitions, methodologies, and ongoing debates. Cambridge Journal of Economics, 38(2), 459–477. https://doi.org/10.1093/cje/bet067

Capella

M. E.

(2001). Predicting earnings of vocational rehabilitation clients with visual impairments. Journal of Rehabilitation, 67(4), 43–47.

Capella-McDonnall

M. E.

(2005). Predictors of competitive employment for blind and visually impaired consumers of vocational rehabilitation services. Journal of Visual Impairment & Blindness, 99(5), 303–315.

Cappellari

Tatsiramos

(2015). With a little help from my friends? Quality of social networks, job finding and job match quality. European Economic Review, 78, 55–75. https://doi.org/10.1016/j.euroecorev.2015.04.002

Carnevale

A. P.

Garcia

T. I.

Ridley

Quinn

M. C.

(2020). The overlooked value of certificates and associate’s degrees what students need to know before they go to college. https://vtechworks.lib.vt.edu/handle/10919/99010

Cavenaugh

B. S.

Rogers

(2002). Employment patterns of older workers with visual impairments. Journal of Visual Impairment & Blindness, 96(9), 655–658.

Chan

Wang

C. C.

Fitzgerald

Muller

Ditchman

Menz

(2016). Personal, environmental, and service-delivery determinants of employment quality for state vocational rehabilitation consumers: A multilevel analysis. Journal of Vocational Rehabilitation, 45(1), 5–18. https://doi.org/10.3233/JVR-160806

Cimera

R. E.

Rumrill

P. D.

Chan

Kaya

Bezyak

(2015). Vocational rehabilitation services and outcomes for transition-age youth with visual impairments and blindness. Journal of Vocational Rehabilitation, 43(2), 103–111. https://doi.org/10.3233/JVR-150760

10.

Cohen

(1992). A power primer. Psychological Bulletin, 112, 155–159.

11.

Cooke

G. B.

Donaghey

Zeytinoglu

I. U.

(2013). The nuanced nature of work quality: Evidence from rural Newfoundland and Ireland. Human Relations, 66(4), 503–527. https://doi.org/10.1177/0018726712464802

12.

Crudden

Giesen

J. M.

Sui

(2018, January 1). Contrasting competitively employed and unemployed VR applicants with visual disabilities: Characteristics and VR service delivery patterns. Journal of Vocational Rehabilitation, 49, 117–126. https://doi.org/10.3233/JVR-180958

13.

Crudden

McDonnall

M. C.

Sui

(2018). Losing employment: At-risk employed vocational rehabilitation applicants with vision loss. Journal of Visual Impairment & Blindness, 112(5), 461–474. https://doi.org/10.1177/0145482X1811200504

14.

Darensbourg

B. L.

(2013). Predictors of competitive employment of VR consumers with blindness or visual impairments. Journal of Vocational Rehabilitation, 38(1), 29–34. https://doi.org/10.3233/JVR-120618

15.

Dutta

Gervey

Chan

Chou

C.-C.

Ditchman

(2008). Vocational rehabilitation services and employment outcomes for people with disabilities: A United States study. Journal of Occupational Rehabilitation, 18(4), 326–334. https://doi.org/10.1007/s10926-008-9154-z

16.

Erickson

Lee

von Schrader

(2020). 2018 disability status report: United States. https://www.disabilitystatistics.org/

17.

Estrada-Hernández

(2008). The effects of participant and service characteristics on the employment outcomes of RSA consumers with visual impairments: A follow-up on agency-type. Journal of Applied Rehabilitation Counseling, 39(1), 28–35.

18.

Findlay

Kalleberg

A. L.

Warhurst

(2013). The challenge of job quality. Human Relations, 66(4), 441–451. https://doi.org/10.1177/0018726713481070

19.

Findlay

Warhurst

Keep

Lloyd

(2017). Opportunity knocks? The possibilities and levers for improving job quality. Work and Occupations, 44(1), 3–22. https://doi.org/10.1177/0730888416689813

20.

Giesen

J. M.

Cavenaugh

B. S.

(2012). Transition-age youths with visual impairments in vocational rehabilitation: A new look at competitive outcomes and services. Journal of Visual Impairment & Blindness, 106(8), 475–487.

21.

Giesen

J. M.

Cavenaugh

B. S.

(2013). Disability insurance beneficiaries with visual impairments in vocational rehabilitation: Socio-demographic influences on employment. Journal of Visual Impairment & Blindness, 107(6), 453–468.

22.

Giesen

J. M.

Hierholzer

(2016). Vocational rehabilitation services and employment for SSDI beneficiaries with visual impairments. Journal of Vocational Rehabilitation, 44(2), 175–189. https://doi.org/10.3233/JVR-150789

23.

Giesen

J. M.

Lang

A. H.

(2018). Predictors of earnings enabling likely roll departure for SSDI beneficiaries with visual impairments in vocational rehabilitation. Journal of Disability Policy Studies, 29(3), 166–177. https://doi.org/10.1177/1044207318780363

24.

Goertz

Y. H. H.

van Lierop

B. A. G.

Houkes

Nijhuis

F. J. N.

(2010). Factors related to the employment of visually impaired persons: A systematic literature review. Journal of Visual Impairment & Blindness, 104(7), 404–418. https://doi.org/10.1177/0145482x1010400704

25.

Heyman

Stokes

J. E.

Siperstein

G. N.

(2016). Not all jobs are the same: Predictors of job quality for adults with intellectual disabilities. Journal of Vocational Rehabilitation, 44(3), 299–306. https://doi.org/10.3233/JVR-160800

26.

Holzer

Dunlop

(2013). Just the facts, ma’am: Postsecondary education and labor market outcomes in the US (IZA Discussion Paper). Institute for the Study of Labor.

27.

Howell

D. R.

Kalleberg

A. L.

(2019). Declining job quality in the United States: Explanations and evidence. RSF: The Russell Sage Foundation Journal of the Social Sciences, 5(4), 1–53. https://doi.org/10.7758/RSF.2019.5.4.01

28.

Kaiser

H. F.

Rice

(1974). Little jiffy, Mark IV. Educational and Psychological Measurement, 34, 111–117.

29.

Lauer

E. A.

Boege

S. L.

Houtenville

A. J.

(2020). Annual disability statistics compendium: 2019. https://www.iod.unh.edu

30.

Leonard

D’Allura

Horowitz

(1999). Factors associated with employment among persons who have a vision impairment: A follow-up of vocational placement referrals. Journal of Vocational Rehabilitation, 12(1), 33–43.

31.

Lorah

(2018). Effect size measures for multilevel models: Definition, interpretation, and TIMSS example. Large-Scale Assessments in Education, 6(8). https://doi.org/10.1186/s40536-018-0061-2

32.

Lund

E. M.

Cmar

J. L.

(2019a). A systematic review of factors related to employment outcomes in adults with visual impairments. Journal of Visual Impairment & Blindness, 113(6), 493–517.

33.

Lund

E. M.

Cmar

J. L.

(2019b). Factors related to employment outcomes in vocational rehabilitation consumers with visual impairments: A systematic review. Journal of Visual Impairment & Blindness, 113(6), 518–537.

34.

Lund

E. M.

Cmar

J. L.

(2020). A systematic review of factors related to employment in transition-age youth with visual impairments. Rehabilitation Psychology, 65(2), 122–136. https://doi.org/10.1037/rep0000303

35.

McDonnall

M. C.

(2016). The relationship between vocational rehabilitation professional’s interactions with businesses and employment outcomes for consumers who are blind or visually impaired. Rehabilitation Counseling Bulletin, 59(4), 203–212. https://doi.org/10.1177/0034355215586389

36.

McDonnall

M. C.

Cmar

(2019). Employment outcomes and job quality of vocational rehabilitation consumers with deaf-blindness. Rehabilitation Counseling Bulletin, 63(1), 13–24. https://doi.org/10.1177/0034355218769461

37.

McDonnall

M. C.

Cmar

J. L.

McKnight

Z. S.

(2020). Service factors and personal characteristics associated with employment and job quality for vocational rehabilitation consumers with combined traumatic brain injury and visual impairment. Journal of Vocational Rehabilitation, 52(3), 223–238. https://doi.org/10.3233/JVR-201073

38.

McDonnall

M. C.

Sui

(2019). Employment and unemployment rates of people who are blind or visually impaired: Estimates from multiple sources. Journal of Visual Impairment & Blindness, 113(6), 481–492. https://doi.org/10.1177/0145482X19887620

39.

Nadeau

C. A.

(2017). Living wage calculator: User’s guide/technical notes, 2016 update. Department of Urban Studies and Planning, Massachusetts Institute of Technology.

40.

Olney

M. F.

Lyle

(2011). The benefits trap: Barriers to employment experienced by SSA beneficiaries. Rehabilitation Counseling Bulletin, 54(4), 197–209. https://doi.org/10.1177/0034355211400209

41.

Rothwell

Crabtree

(2019). Not just a job: New evidence on the quality of work in the United States. https://www.luminafoundation.org/wp-content/uploads/2019/11/not-just-a-job-new-evidence-on-the-quality-of-work-in-the-united-states.pdf

42.

SSTI. (2016). Useful stats: Per capita personal income by state, 2010–2015. https://ssti.org/blog/useful-stats-capita-personalincome-state-2010-2015

43.

U.S. Bureau of Economic Analysis. (2020). Gross Domestic Product (GDP) summary, annual by state, 2015. https://www.bea.gov/data/gdp/gdp-state

44.

U.S. Census Bureau. (2018a). American Community Survey, 2015 ACS 1-Year Estimates Subject Tables, B18120: Employment status by disability status and type. https://data.census.gov/cedsci/

45.

U.S. Census Bureau. (2018b). American Community Survey, 2015 ACS 1-Year Estimates Subject Tables, S1701: Poverty status in the past 12 months. https://data.census.gov/cedsci/

46.

Workforce Innovation and Opportunity Act of 2014. Pub. L. No. 29 U.S.C Sec 3101 (2014).

47.

Zapata

M. A.

(2020). Disability affirmation predicts employment among adults with visual impairment and blindness. Rehabilitation Counseling Bulletin. Advance Online Publication. https://doi.org/10.1177/0034355220957107