Abstract
BACKGROUND:
Customized employment (CE) is recognized in the Workforce Innovation and Opportunity Act (2014) as a strategy for promoting competitive integrated employment. However, the existing body of evidence supporting CE is mainly descriptive rather than experimental research.
OBJECTIVE:
This study examined the impact of CE on the employment outcomes, hours worked per week, and wages of transition-age youth with intellectual and developmental disabilities.
METHOD:
The outcomes of transition-age youth participating in a CE intervention were compared to those receiving treatment-as-usual using a randomized controlled trial design.
RESULTS:
Participants receiving CE were significantly more likely to secure competitive integrated employment than controls who received treatment-as-usual. Participants in the intervention and control conditions earned similar wages. Participants in the control condition worked more hours per week than those in CE.
CONCLUSION:
The findings from this study demonstrate the effectiveness of CE to assist transition-age youth with intellectual and developmental disabilities in obtaining competitive integrated employment, but future research is needed to examine factors impacting weekly hours and wages of participants in CE.
Keywords
Introduction
Customized employment (CE) has been described as a vocational rehabilitation (VR) service that facilitates competitive integrated employment (CIE) for individuals with significant disabilities [1]. CE is a process for individualizing the relationship between a job seeker with a significant disability and an employer [2]. To do this, the personal strengths, interests, and needs of a job seeker are identified and then matched with the specific needs of a business in a way that mutually benefits both parties [3]. CE thus requires rich information to be collected about the job seeker early in the employment process, and this is done using a qualitative approach known as Discovery. Then, flexible strategies such as adjustments to work schedules, creation of a set of job duties, and identification of a job location are used to negotiate a CIE position [2]. The formal introduction of CE began when the Federal Register first published a definition of CE in the early 2000 s [3]. Since then, it has been codified into law under the definition of supported employment in the Workforce Innovation and Opportunity Act of 2014 [2]. Despite incorporation of CE into policy and use by VR agencies across the United States, there has been considerable debate over the supporting research evidence of CE as a pathway to CIE [4–6].
Articles on CE have appeared in the peer reviewed literature since the early 2000 s, but these have been predominantly descriptive in nature and focused on implementation procedures [6]. Across two extensive reviews of the literature, Riesen and colleagues collectively examined the body of empirical evidence for CE published between 2001 and 2021 [4, 5]. Results from the initial review indicated that, of a total of 25 identified CE articles published between 2001 and 2015, only 10 included information on participant outcomes [7–16]. The remaining 15 were non-data based, reporting on intervention model details and aspects of CE, but not including information on employment outcomes linked to CE services [4]. In a follow-up review, Riesen and colleagues [5] added 8 more data-based articles to the 10 identified in the first review by conducting database searches on CE published between 2015 and 2021 [3, 17–23]. Among the resulting collection of 18 total data-based articles, Riesen and colleagues observed descriptive reports that indicated favorable employment outcomes among CE recipients, which support its status as a promising practice [5]. However, results also revealed that there were no experimental studies reporting outcome data associated with CE as an intervention and only one utilizing a correlational design [5].
In the last two years, the call to investigate CE with increased methodological rigor has been met by several studies reporting outcome data. For example, Kim et al. examined the employment outcomes of VR service recipients who participated in CE using the U.S. Rehabilitation Services Administration Case Service Report data (RSA-911). Findings indicated that CE was an effective intervention for VR service recipients with cognitive disabilities, as these individuals demonstrated a higher likelihood of CIE and higher wages in comparison to VR recipients who had other disabilities [24]. However, a second study by Kim et al. using RSA-911 data indicated that a low number of CE service recipients (only about 30.2%) were exiting into CIE [1]. While these studies offer increased causal comparative scrutiny over prior research designs, they are placed among a body of literature spanning more than two decades that does not include a single randomized controlled design investigating the impact of CE on employment outcomes.
To address this void in the literature, the purpose of this study was to conduct a randomized controlled trial (RCT) study investigating the impact of CE as an intervention for transition-age youth with intellectual and developmental disabilities (IDD). Given the absence of stringent research designs investigating CE within the extant literature, this study aimed to apply the type and rigor of research that are still needed to establish CE as an evidenced-based practice. The following research questions guided the study; What is the impact of participation in a CE intervention on obtaining CIE for youth with IDD? What is the impact of participation in a CE intervention on weekly hours worked for youth with IDD? What is the impact of participation in a CE intervention on hourly wages for youth with IDD?
Method
Participants
The criteria for participating included that the youth (a) had a diagnosis of IDD that substantially limited one or more major life activities including economic self-sufficiency, (b) were unemployed at intake, (c) consented to work at least 10 hours per week, and (d) consented to the intervention and data collection protocol. The Institutional Review Board at Virginia Commonwealth University reviewed and approved this study involving human subjects in March of 2017 (#HM20008986). Forty-six youth with disabilities between the ages of 18–24 participated in this RCT. A little more than half (n = 27, 58.7%) were male while the remaining participants identified as female (n = 19, 41.3%). Slightly less than half of the sample reported their race as white (n = 22, 47.8%), followed by Black/African American (n = 16, 34.8%), and Asian (n = 2, 4.4%). The remaining participants reported their race as other/mixed (n = 6, 13%).
The primary disability for the largest percentage of participants was autism spectrum disorder (n = 23, 50%), followed by intellectual disability (n = 12, 26.09%) and Down syndrome (n = 11, 23.91%). All participants had a significant disability with high support needs including functional impairments that were identified as barriers to employment. These barriers to employment were reported as limited communication skills (n = 33, 71.7%), poor social / interpersonal skills (n = 19, 41.3%), challenging behaviors (n = 17, 36.95%), physical limitations such as poor stamina and endurance (n = 12, 26.1%), and limited self-care skills (n = 5, 10.9%).
More than half of the participants (n = 29, 63%) were still in school at intake, and the remaining were out of school (n = 17, 37%). Those who had left school were in segregated, non-work programs for people with disabilities or were at home with no day program. Almost all of the participants (n = 40, 86.96%) reported that they had participated in work experiences through their school programs, whereas a small percentage (n = 6, 13.04%) reported that they had never participated in school-initiated work experiences. None were participating in work experiences when they joined the study. When asked to describe these experiences, the youth or parents and guardians described them as activities in which all students with disabilities at the school participated rather than individualized experiences specific to each person’s interests and work preferences. Other activities were designed for students with disabilities in the school setting rather than in community businesses such as cleaning in the school cafeteria, selling cookies and coffee to other students and teachers, and other non-paid experiences.
A large majority of the participants (n = 38, 82.6%) reported that they were clients of the state VR agency at intake, whereas eight participants (17.4%) reported that they were not. When asked if they had ever held a CIE position, only two (4.3%) had previously been employed part-time. Forty-four participants (95.7%) had never held a CIE position. None of the participants were employed when they joined the study, as this was one of the study’s inclusion criteria.
Recruitment and randomization
Youth were recruited in three cohorts through community organizations that included local school districts, the Autism Society, Autism Speaks, the Down Syndrome Society, and the state VR agency. These organizations disseminated a flyer on the study and a form for parents to sign consenting to be contacted. Signed forms were sent to the study director who contacted the parents and potential participants for a home visit or a meeting at a private location of their choice. The purpose of the meeting was to describe the study and to obtain informed consent.
A block randomization procedure was used to assign participants to the control or treatment group. Once four youth consented to participate, they were randomized into the intervention or control group using an online random number generator. Youth who were randomized into the control group agreed to remain in that group assignment for at least one year. At the end of the year in the control condition, they were offered the opportunity to be re-randomized for the next year. As such, 24 participants were initially randomized into the control group; 22 were randomized into the treatment group. For the purpose of this analysis, of the 24 participants who were originally in the control group, seven requested to be re-randomized and later received the intervention for a total of 29 youth receiving the intervention. To account for this, the researchers used a statistical adjustment of standard errors.
Procedures
The CE intervention included three interconnected phases: discovery, customized job development, and ongoing supports. Inge and colleagues conducted a series of focus groups with nationally recognized experts in CE to identify the practices that are associated with CE [3]. Twelve core practices of CE were identified from the focus group analysis. These core practices formed the foundation of the intervention in this study and are presented in Table 1. The employment specialists who implemented the intervention received training in the CE intervention from the research team and a consultant who was a nationally recognized expert in CE services. Ongoing team meetings were held to discuss the intervention and to provide training and technical assistance (TA) to the employment specialists. These meetings were scheduled weekly as a group or individually with the project director. The three interconnected phases of CE services were implemented sequentially as follows; Discovery, Customized Job Development, and Ongoing Supports.
Core practices of customized employment
Core practices of customized employment
Discovery is a descriptive, capacity-based phase that does not include administering standardized tests or evaluating a job seeker’s skills and competencies. Discovery gathers information by direct observation, participating with the job seeker in familiar and novel settings, and in-depth interviewing with people who know the job seeker well [25, 26]. The outcome is a plan for customized job development including the identification of the job seeker’s vocational themes. Discovery is the foundation of CE and precedes job negotiation and customization.
Discovery began for the study participants in their homes and included family members of their choice. If the job seekers were not their own guardians, their Legally Authorized Representatives were included in the initial visit and subsequent activities. Some participants did not want project staff to come to their homes, and this request was honored by asking them to identify a private location of their choice. During the home visits and meetings at other preferred locations, the goal was to begin to learn about the job seeker’s interests and skills. This information was used to schedule activities that occurred in familiar and novel locations including businesses representative of the participants’ interests and vocational preferences. Family members were asked to provide contact information and introductions to other family members and acquaintances (known as social capital) who knew the participants well for in-depth interviews.
Customized job development
All of the information from discovery was summarized and reviewed with job seekers and their families leading to the identification of vocational themes, which were then explored for customized job development. Customized job development used informational interviews to learn more about businesses, their working conditions, and potential work tasks where the needs of the job seeker could be met [3]. Informational interviews in some cases lead to the identification of tasks that could be negotiated into a new job description for a participant. In other words, the goal was to get to know the needs of a business in a similar way that discovery focused on getting to know each youth in the study. Customized jobs were negotiated using an employment proposal based on each youth’s unique skills and interests.
Ongoing supports
All of the participants received on-the-job support and training to learn the job duties in their individualized job descriptions. Although job duties were negotiated based on the abilities and interests of each youth, they could not perform their negotiated job duties with 100% independence. The job supports included training on the negotiated tasks and skills such as social skills as well as facilitating support from supervisors and coworkers at the businesses. As the youth demonstrated independence and the businesses took over their supervision and support, the employment specialists faded from the workplaces. Ongoing workplace visits were made based on the needs of each youth thereafter to ensure that they, as well as the businesses, were receiving needed support.
Data analysis
To estimate the treatment effect of participation in the CE intervention, three sets of analyses were conducted. The first examined the impact of participation on CIE, a binary outcome (1 = Yes, 0 = No). Unadjusted risk ratios (RRs) were calculated using the “riskratio.boot” function found within the “epitools” package [27, 28]. This function uses unconditional maximum likelihood estimation paired with bootstrap resampling to estimate confidence intervals. The researchers compared unadjusted RR estimates with adjusted estimates obtained from risk ratio regression. These models were fit using the “glm()” function with a quasi-Poisson link in base R and included a binary variable indicating treatment assignment (1 = Treatment, 0 = Control), along with participant age, gender identity, race, and ethnicity as covariates.
The second analysis examined the impact of participation on hourly wages. Unadjusted treatment effect estimates were obtained using a linear regression model via the “lm()” function in R. Hourly wage was treated as the outcome, and a binary variable indicating treatment variable was included as a predictor. These results then were compared with adjusted estimates, which also included age, gender identity, race, and ethnicity as covariates. The third analysis examined the impact of participation on participants’ number of hours worked per week. The researchers again used a linear regression model, comparing both unadjusted and adjusted estimates in the same approach as used above for hourly wages.
Results
Competitive integrated employment outcomes
In the treatment group, 18 of the 29 participants (62.07%) who received the CE intervention achieved a CIE outcome. In comparison, only four of the 24 participants while assigned to the control group attained a CIE outcome (n = 4, 16.67%). The difference was statistically significant (RR = 3.87, 95% CI [1.83, 17.24], p < .001). This result was robust to the choice of statistical model and the inclusion of age, race, and gender identity as potential controls. The researchers also fit a relative risk regression using the “glm” function with a quasi-Poisson link in base R, which allowed us to include age, race, and gender identity to control for potential confounders of the treatment effect. After the inclusion of these covariates, the adjusted risk ratio was still positive, significant, and in the same magnitude as the unadjusted estimate (Adjusted RR = 4.11, t = 2.93, p = .005). Table 2 includes full model estimates for the relative risk regression, including coefficients, standard errors, and 95% confidence intervals.
Treatment Effect Estimates for the Effect of Participation in CE Intervention on Obtaining CIE
Treatment Effect Estimates for the Effect of Participation in CE Intervention on Obtaining CIE
n = 53 for all models. CE = Customized Employment. CIE = Competitive Integrated Employment. (Cluster-robust) standard errors were calculated using the Satterthwaite degrees of freedom approximation. Results from a relative risk regression with quasipoisson link. Additional control variables included participant age, sex, race/ethnicity, and school enrollment status.
Looking only at participants who obtained CIE by the end of the study period, the treatment and control participants earned very similar wages. The median wage was $9.13/hr. for the treatment group and $10.00/hr. for the control group. The salary range for the treatment group was $7.25 (minimum wage in the state where the study was conducted) to $17.50 per hour. For the participants in the control group who became employed, the salary range was $7.55 to $10.55. The mean hourly wage for the treatment group was $9.44 compared to a mean hourly wage for the control group of $9.35.
Hours worked per week
Participants in the control group who obtained CIE worked more hours compared to participants in the treatment group who achieved CIE. The hours worked per week for the treatment group participants ranged from as little as four hours per week to 20 hours. While the range of hours worked for participants in the control group was 12 hours to 27 hours per week. The median hours per week was 20 hours for the control group and 7.5 for the treatment group. The average hours per week for participants in the control group was 19.7 hours compared to 8.88 hours per week for those in the treatment group.
Discussion
The findings of this study support the use of CE services to facilitate CIE. CE is designed to meet the specific needs and interests of individuals with significant disabilities as well as the needs of the employer by using flexible strategies at every stage of employment [25]. It is a service for supporting individuals who have a history of being unemployed or underemployed and those who have previously been excluded from receiving support in finding employment. The sample for this study met this profile because they were unemployed at intake and, for the majority of participants, the job they obtained during the study was their first time becoming employed. Forty-four (95.7%) had never held a CIE position and only two individuals (4.3%) had previously been employed part-time.
Those in the treatment group were almost four times (and significantly) more likely to achieve CIE than those in the control group. Indeed, approximately 62.07% who received the CE intervention achieved CIE outcomes. However, the number of hours worked per week by those who received CE was low. Hours ranged from as little as four hours a week to 20 hours, with a median of 7.5. It is important to note that this sample consisted of youth (ages 18–24), and some were still enrolled in school during the course of the study, which limited the number of hours they were able to work.
Including only the participants who obtained CIE by the end of the study period, the youth who received the CE intervention earned a median wage of $9.13 per hour working a median number of hours per week of 7.5. A median hourly wage combined with the median hours worked per week results in an annual salary clearly below the poverty level ($9.13 x 7.5 = $68.48 per week or 52 × $68.48 = $3,560.70 per year). While hourly wage is more than the federal minimum wage, the total wages earned is an issue for a CIE outcome. However, this was a first job for many participants, and it is acknowledged that, for youth with disabilities, a first job may be a building block to future employment and should lead to other opportunities (just as it should for their peers without disabilities).
In an analysis of the RSA 911 Case Service Report data from program years (PY) 2017 to PY 2020, Kim et al. [1] also found that VR recipients exiting VR services who achieved a CIE outcome after receiving CE worked a limited number of hours. The median hours worked by these individuals was 13 hours per week, and the median hourly wage was $11.5 (mean = $11.25, SD = 5.81 [1]. These findings from Kim and colleagues were more than the median hours worked and median wages earned by the youth in this study, but they are still problematic as employment outcomes. It is important to note that the mean age for the Kim et al. [1] study was 31.61 years old (SD = 12.39, ranging from 18 to 84), whereas this RCT’s participants ranged from 18–24 years of age.
During this study, the researchers identified several factors that limited the number of hours worked per week by youth who achieved a CIE outcome. First, over half of the participants (63%) were still in school at intake, which limited the time that they could engage in study activities. As an example, one youth who was still in school routinely went with her family out of town during the weekends and could not participate in the intervention. Her in-school activities also limited her availability, because she and her family did not want her to miss specific school activities. And, this limited the days of the week that she was willing to work. The result was a limited window of days a week that the employment specialist could engage her in the intervention. Limiting the number of days that she was available to work impacted job negotiations, because a number of employers who were open to negotiating a position for her did not match her availability. Future research needs to consider the impact of negotiating positions on the number of hours worked and the weekly wages earned.
Second, most of the jobs involved negotiating one or two job duties to develop a position for the youth with IDD who achieved a CIE outcome. These were often job duties that were not getting done at all or not completed regularly at the business but were identified as matching the job seeker’s skills and interests. The study team did not intentionally negotiate only one or two job duties but was willing to coordinate with the businesses to develop a position that met the needs of the business.
When only one or two job duties were negotiated, it obviously impacted the number of hours per week the transition-age youth could work. Although this was not an issue for participants in this study because they only wanted to work a few hours a week, this could be an issue for adults who do not have other daytime activities. The number of tasks negotiated is of crucial importance because it can not only impact the number of hours and days the person works in an integrated position, but also their wages. Practitioners need to consider if working more than one customized job per week could increase the number of days and hours worked while also increasing the weekly wages of people with significant disabilities who receive this service. Research needs to explore this critical issue in order to improve the CIE outcomes for individuals who receive CE services.
Not having opportunities to engage in early work experiences can negatively impact a transition-age youth’s ability to develop work skills that are essential in finding and maintaining employment [31]. No one in this RCT study was participating in work experiences at the time of intake. The work activities that they had experienced were typical entry level tasks reflecting the low work expectations for people with disabilities. These activities had been developed for all students with disabilities and not individualized for the youths’ specific interests. A head start on Discovery during school may be beneficial when transitioning to the State VR agency for services post-graduation. Because many of the experiences that this sample had prior to the study were in segregated settings and activities, there was limited information to facilitate Discovery activities during the intervention. This may have increased the amount of time needed to learn about the youth’s skills and interests in integrated settings. Given that a lack of individualized work experiences was a barrier for all participants in both the treatment and control groups, the researchers do not know if providing other individualized, integrated opportunities during school could have impacted the employment outcomes achieved by this sample. Future research needs to explore the relationship between Discovery activities during high school and the job outcomes of youth with disabilities.
All of the treatment group members who achieved a CIE outcome received on-the-job support after they were hired and subsequent ongoing support after they had adjusted to work. The “place and then train” philosophy has been the foundation of supported employment and is the service needed by many people with significant disabilities to maintain employment [6]. Kim and associates [1] in their analysis of the RSA 911 data examined the frequency of VR services provided with CE services in the ten states with the highest use of CE services. Supported employment was one of these services, with 17.5% (n = 311) of the sample receiving supported employment in addition to CE. The authors further analyzed the data by grouping the VR recipients who exited the program in CIE after receiving CE services. These were identified as Group A and Group B, with Group B more likely to use referral for SE in addition to CE. In the employment outcome, consumers served in Group B were more likely to exit into CIE (59.4%) than individuals served in Group A (15.1%; (χ2(1, N = 1,779) = 371.620, p < 0.001).
This is an important finding because some sources such as The Essential Elements of CE for Universal Application do not include guidelines for on-the-job training and support as part of CE [26]. Instead, it is suggested that people receiving CE services should be transitioned to SE services once they become employed. Further research is needed regarding the service needs of VR recipients with the most significant disabilities such as merging customized and supported employment services for those individuals who have been traditionally unemployed or underemployed.
Though the findings from this study indicate some concern regarding wages and hours for CE recipients, the findings that CE resulted in superior CIE outcomes is important. Research is increasingly demonstrating that participation in CIE impacts more than just wages and the number of hours that workers with disabilities spend in integrated community settings. Indeed, there are therapeutic benefits of working in terms of personal growth and life skills development for youth and young adults with IDD [29–31]. To illustrate, research findings from another analysis of the outcomes of the participants involved in this study found that transition-age youth in the CE treatment group also benefited in areas of personal growth and independence [32]. Specifically, the transition-age youth who participated in CE experienced significant increases in independence in three domains when compared to the control group on the Supports Intensity Scale -Adult Version, including Home Living, Employment, and Protection and Advocacy. Findings suggest a therapeutic effect of work activities on personal growth and skill development in important life domains. These effects may be as important as the hours and wages earned at this stage of a youth’s career development.
Study limitations
This study was limited by the COVID-19 Pandemic, which suspended the research for several reasons. First, the university Institutional Review Board (IRB) put on hold all research involving human subjects for the protection of those subjects. Second, many businesses were closed during the pandemic or had reduced staffing. All of the participants were laid off from work, and most were not re-hired when business restarted. Third, a number of the parents and the participants themselves did not want to meet the requirements to be vaccinated and/or wear a mask in order to return to work. Due to these limitations, there is no ongoing retention data available for the youth who were employed during this study. Future research is needed regarding the long-term job retention of people with disabilities who receive CE services.
Conclusion
The present study was the first to examine the impact of CE on employment for transition-age youth using a RCT design. The differential outcomes achieved by the sample in this study demonstrate that CE can result in youth with significant disabilities achieving CIE. These findings are consistent with previous research suggesting that rehabilitation professionals generally view CE as a relevant strategy for individuals with significant disabilities [25]. Inge and colleagues conducted a survey of rehabilitation providers in the U.S. and found that participants agreed regarding the unique characteristics of CE but reported that there was a gap between knowledge of CE and implementation [25]. As states continue to integrate CE into service delivery, they will need guidelines on best practices and how to ensure the fidelity of these practices [18]. Providing CE to assist people with the most significant disabilities is critical, because they remain chronically unemployed and many could benefit from the service. This RCT study provides data on the effectiveness of CE to assist transition-age youth with IDD in achieving CIE while raising many questions for future research.
Footnotes
Acknowledgments
The authors have no acknowledgements.
Conflict of interest
The authors declare no conflicts of interest.
Ethics statement
This study was approved and monitored by the Institutional Review Board at Virginia Commonwealth University (#HM20008986).
Informed consent
Informed consent was obtained from all participants prior to study activities in accord with procedures approved by Virginia Commonwealth University’s Institutional Review Board (#HM20008986).
Funding
The contents of this paper were developed under a grant from the National Institute on Disability, Independent Living, and Rehabilitation Research (NIDILRR, Grant Number: 90DP0085). NIDILRR is a Center within the Administration for Community Living (ACL), Department of Health and Human Services (HHS). However, the contents do not necessarily represent the policy of the U.S. Department of Health and Human Services, and you should not assume endorsement by the Federal government.
