The resiliency of employees with intellectual and developmental disabilities during the COVID-19 pandemic and economic shutdown: A retrospective review of employment files

1.1Literature review

The employment rate of individuals with IDD remains discouragingly low despite federal legislation and state initiatives specifically designed to encourage CIE over sheltered employment and disability specific non-work day centers and sheltered workplaces have proven to be ineffective at increasing the rate of CIE (Wehman et al., 2018). To wit, the Workforce Innovation and Opportunity Act (WIOA) of 2014 which was enacted to increase access to CIE, high quality training, and job support services (U.S. Department of Labor, 2018; WIOA, 2014). Further, the Federal Medicaid Waiver Programs enable states to allocate funding toward employment intervention programs targeting individuals with IDD (Centers for Medicare and Medicaid Services, 2016). Despite these initiatives, individuals with IDD are still not gaining access to employment. Winsor et al., (2018) reported that the number of individuals with IDD accessing CIE has actually decreased from a high of 25% in 2001 to 18.8% in 2016. For individuals with IDD who attain CIE, careers remain elusive (Heyman et al., 2009). This is particularly true for BIPOC youth or other minorities with IDD who are also disadvantaged by minority and/or poverty status in addition to disability (Dutta et al., 2019).

There are evidenced based practices that have proven to be effective in supporting individuals with IDD to obtain and maintain CIE. One such evidence-based intervention is supported employment (SE). This practice seeks to assist individuals with the most significant disabilities to locate and maintain long-term, meaningful, competitive integrated employment (Schall et al., 2020). The supported employment model is a four-phase approach. Of particular relevance to this paper are the two phases of supports provided on the job site. They include job site training (JST) and long-term supports services (LTSS; Schall, et al., 2015). Another pathway to securing CIE is through customized employment (CE). CE supports individuals with IDD who may not have all the necessary skills required for an existing job description, so a new position is negotiated to fit the needs of the company and the individual job seeker (Brooke et al., 2018). The Workforce Innovation and Opportunity Act of 2014 included CE in the definition of Supported Employment, and federally defined CE as “based on individual determination of the strengths, needs and interests of the person with a disability, and is designed to meet the specific needs of the employer.” Both SE and CE continue to provide support to the new employee with a disability through JST and LTSS to ensure job satisfaction and career advancement. The goal with both SE and CE is establishing a successful and positive relationship between the employer and the new employee with IDD. The employment specialist (ES) maintains a relationship with both that is synergetic, collaborative, and benefits all parties involved.

State vocational rehabilitation agencies fund both SE and CE through the WIOA (WIOA, 2014). In order to qualify for either SE or CE, the job seeker with a disability must require intensive services as these services are designed for those with the most significant disabilities. While these practices are most helpful to individuals with IDD seeking CIE, their funding structure provides a view of the impact of the COVID-19 crisis on the employment of individuals with IDD. Namely, ES who provide services are responsible for reporting the time they provided supports, the phase of supports they are providing, and the workplace, hours worked, and wage of the individual they support for each month services are provided.

The purpose of this paper is to describe how CIE was impacted for employees with IDD during the COVID-19 pandemic and recession using data collected by ES who provided SE or CE. The overarching research question was; how did the COVID-19 pandemic and recession impact the employment of employees with IDD?

This study was guided by the following sub research questions:

2.1Design

In order to address the research questions posed by this study, the researchers conducted a retrospective review of records of individuals with IDD who were receiving on-the-job supports through JST or LTSS. All services were provided by Business Connections employment services organization (ESO) based at Virginia Commonwealth University that exclusively works with individuals with disabilities seeking CIE. While this is a university-based ESO, referrals for services come from the general population of individuals seeking VR services from the community. All services are funded through the state vocational rehabilitation agency. In order to track the impact of the COVID-19 pandemic and subsequent recession, the authors compared the month-by-month employment status of individuals with IDD employed in CIE and receiving services from the ESO from Feb. to July 2019 and Feb. to July 2020. Inclusion criteria for this records review was, 1) presence of an intellectual or developmental disability, 2) full or part time employment in CIE on Feb. 28, 2019 or 2020, 3) receiving either JST, or LTSS from the ESO. The researchers used an honest broker system and did not see the records or have access to any identifying information. The university IRB determined that this study was ‘not human research’ due to the lack of human subject interaction.

2.2Data collection and verification

The researchers used existing records of employment services to document disability, industry where the individual was employed, average weekly hours worked, hourly wage, gender, age, race, and employment status at the end of each month during the two periods reviewed (February–July 2019 and February–July 2020). To further clarify employment status and input missing data, the honest brokers contacted individual employment specialists to confirm and clarify findings. Finally, as the data is collected prospectively for the purposes of billing for services, and the data set was reviewed and updated through individual interviews between the honest brokers and employment specialists, the researchers determined the data set to be complete and accurate. The brokers, who are responsible for collecting such data for billing purposes, transferred data in de-identified aggregate form for analysis.

3.1Demographic description

Using the three inclusion criteria for the two time periods selected, resulted in the identification of 156 individuals with IDD who were employed in CIE while receiving JST or LTSS on February 28, 2019 and, again, 156 with IDD who were employed in CIE while receiving JST or LTSS on February 28, 2020. Across the two time periods reviewed, 133 individuals who received JST or LTSS services in 2019 were still receiving services in 2020, while 23 individuals were unique to the 2019 and 2020 records review. All participants represented in these records received either SE or CE. This is an indication that all of the records reviewed represented individuals with significant impact from their disability. Table 1 presents the demographic descriptions of these samples by age, gender, race, and disability for each set of records reviewed.

Given that 133 of the 156 individuals in each time period were the same, it is not surprising to find that there were no significant differences between the 2019 and 2020 cohort with respect to age, gender, race, disability, prior education and prior employment experience. For both time periods, these individuals were in their mid-twenties, overwhelmingly male, with just less than half (45.5% in 2019 and 46.2% in 2020) being BIPOC, and just over half (54.5% in 2019 and 53.8% in 2020) being white. Autism was the most common primary disability reported following by intellectual disability. These two disabilities comprised 84.7% of 2019 and 87.2% of 2020 participants. The majority of these individuals completed high school, but the type of diploma earned was not reported by the ESO. Finally, the overwhelming majority had no prior employment history. Nevertheless, these individuals had to present with severe disabilities in order to qualify for SE or CE services.

3.2Baseline employment description

As of February 28, 2019, the individuals with IDD earned a mean wage of $9.88 per hour and worked a mean of 82.29 hours a month or approximately 20.57 hours weekly. Prior to the pandemic, on February 28, 2020 the individuals with IDD receiving services in that group earned a mean wage of $10.21 per hour and worked a mean of 85.79 hours a month or approximately 21.45 hours per week. The majority of these individuals at both time periods (2019 = 61.56%; 2020 = 64.73%) received disability benefits including Supplemental Security Income (40 [25.64%] in 2019, 44 [28.2%] in 2020), Social Security Disability Insurance (14 [9%] in 2019 and 13 [8.33%] in 2020), or two or more disability specific benefits (42 [26.92%] in 2019 and 44 28.2%] in 2020).

The majority of both of these groups worked in the following industries:

3.3Research question 1: Employment status by month Feb. to July in 2019 and 2020

In order to assess the impact of the COVID-19 pandemic and subsequent recession, we reviewed the hourly wage, hours worked per month and employment status for the 156 individuals with IDD employed in CIE on February 28, 2020 and again on March 31, April 30, May 31, June, 30, and July 31, 2020. To confirm changes in employment were due to the COVID crisis, we compared that data to the same time period in 2019. We used the following terms to identify an employee’s employment status:

Working on-site: In order to complete their job, the employee with IDD went to work at the employer’s workplace

COVID furlough: The employee was not working but would return to their job once the shutdown was lifted.

Seasonal furlough: The employee worked in a public school or public school supplier where there was a known regular furlough between June and August.

Laid off: The employee no longer had a job with the employer.

Ill/quarantined: The employee took sick leave more than one week, or was required to stay home in quarantine because of exposure to COVID-19.

The frequency and percentage for each condition month-by-month for 2019 and 2020 are presented in Table 2. The month-by-month comparison of percentage working on site, mean hourly wage, and mean hours worked are presented in Fig. 1a, b, and c.

Fig.1

a. Month-by-month comparison of percentage of individuals working on site in 2019 and 2020. b. Month-by-month comparison of hourly wage for employed individuals with IDD in 2019 and 2020. c. Month-by-month comparison of hours worked monthly in 2019 and 2020.

3.4Illness

There was very little illness noted in this group of employees with only four of the 156 taking extended time for illness or quarantine. Two of the four individuals were out for medical procedures unrelated to COVID. Both returned once cleared by their doctor. One individual was sick with COVID symptoms but did not have access to a COVID test at the time of the illness. They returned to work once the symptoms resolved and their doctor cleared them. The fourth individual was required to quarantine due to exposure to a COVID-19 positive co-worker, but never developed symptoms and returned to work after 14 days of quarantine and a negative COVID test.

3.5Work status

Unlike the 35.2% of the United States workforce that worked entirely from home, none of these employees with IDD were afforded the opportunity to work from home (Bick et al., 2020). Instead, most employees who were unable to work due to business closure were furloughed. The month of highest impact was April 2020 where only 41.6% of employees worked on site in essential jobs. Each month following April showed an increase in the percentage of individuals working on site. By July, however, there were still substantially more employees furloughed, laid off, or quarantined due to COVID-19 (34.5%), versus those who were furloughed due to regular seasonal employment in 2020 (5.1%) or compared to the employment statuses for the same time period in 2019. Some of those who were furloughed in 2020 due to COVID-19 elected that option due to their own health concerns. While only one employee with IDD elected to take leave in March due to the pandemic (0.6%), this number remained steady, but relatively low for April through July with a range of 8 (5.13%) to 11 (7.05%) employees with IDD electing to refrain from working on-site during those months of the pandemic. A Pearson Chi-Square was completed to compare working on site versus not working between the records reviewed from the 2019 and 2020 periods. Overall, the differences between the 2019 and 2020 groups were significantly different based upon their work status for all months reviewed. Specifically, we found significant differences in March (χ² = 4.271, df = 1, p = 0.034), April (χ² = 100.953, df = 1, p = 0.000), May (χ² = 73.688, df = 1, p = 0.000), June (χ² = 48.767, df = 1, p = 0.000), and July (χ² = 30.422, df = 1, p = 0.000) indicating that COVID-19 had a very significant impact on these individual’s employment status for all months studied.

3.6Wages

The COVID-19 crisis did not appear to have an effect on the wages of those who retained their positions. In fact, the mean hourly wage of these workers rose steadily from $10.21 per hour to $10.91 per hour from February to July 2020. This is similar to the pattern of wage growth observed in 2019 when the mean hourly wage increased from $9.88 per hour in February 2019 to $10.06 per hour in July 2019.

3.7Hours worked monthly

Nevertheless, it appears that the income for these individuals was impacted by the number of hours available to work. There was a significant impact on hours worked per month. Specifically, the drop in hours worked per month is evident with a 19.2% drop in March with the onset of mandated shutdowns of non-essential businesses. This loss of hours continued into April when the hours worked monthly was 49.84% of the hours worked in February. Beginning in May, employees increased the hours worked monthly through July with modest increases each month (3.69% from April to May, 19.64% from May to June and 11.36% increase from June to July). Even with these modest increases in hours per month, the mean hours worked per month in July remained 30.87% below the February mean hours worked per month. Independent t-tests indicated these differences were significant for all months compared as follows: March (t = 2.558, df = 309, p = 0.11), April (t = 6.567, df = 298, p = 0.000), May (t = 6.429, df = 300, p = 0.000), June (t = 2.936, df = 290, p = 0.004), and July (t = 2.507, df = 299, p = 0.013).

3.8Displaced worker benefits

There were two sources of paycheck support for workers during the period explored. They were extended unemployment insurance and the Paycheck Protection Program (PPP) provided through the CARES Act (Coronavirus Aid, Relief, and Economic Security [CARES] Act, 2020). Table 3 presents the number of furloughed and laid off employees who accessed compensation while unemployed via either program. While over half had access to some compensation while furloughed or laid off in March and April of 2020, access to these programs drops dramatically through May, June, and July. This provides a cause for concern for these workers as the CARES Act unemployment extension and PPP loans have closed.

3.9Research question 2: Variability of impact of the COVID-19 pandemic by individual characteristics

The researchers analyzed four individual independent variables to determine their impact on the two dependent work variables most affected by COVID; work status and hours worked per month. The dependent variables analyzed were age, gender, primary disability, and race. Because of the low number of Hispanic and Asian American employees, researchers recoded the race variable as a dichotomous variable including white and BIPOC. The research team used the independent t-test analysis to identify the degree to which work status and hours worked was impacted by age, gender, primary disability, and race. The repeated independent t-tests were performed for the months affected by COVID-19, specifically March–July, 2020 for both work status and hours worked.

None of the independent t-tests were significant when comparing the dependent variables to working status at any time point. Thus, age, gender, race, nor primary disability appeared to have an impact on working status. The same was true of hours worked with one exception. There was a significant impact on hours worked by race in February, March, April, and a trend toward significance in May. This interesting finding prompted us to consider if this difference also existed in 2019. Thus, we completed the same analysis on the hours worked in 2019 and found similar significant differences with BIPOC individuals with IDD working significantly more hours per month than their white peers for Feb. – July 2019. Table 4 presents the means and p-values for these comparisons for 2019 and 2020 when evaluating the effect of race on hours worked using an independent t-test. Figure 2a and b presents the mean hours worked per month in 2020 and 2019 by BIPOC and White individuals.

Fig.2

a. Month-by-month comparison of hours worked monthly by BIPOC and white individuals in 2020. b. Month-by-month comparison of hours worked monthly by BIPOC and white individuals in 2019.

The final variable we considered was the differential impact of industry type on hours worked and work status. We analyzed this data with a Pearson Chi-Square procedure. In this case, there were significant differences between industry and work status in March (χ² = 26.257, df = 12, p = 0.01), April (χ² = 36.283, df = 12, p = 0.000), May (χ² = 44.86, df = 12, p = 0.000), June (χ² = 59.44, df = 12, p = 0.000), and July (χ² = 43.61, df = 12, p = 0.000). It appears that individuals who worked in education, entertainment, food service, hospitality, and retail were more likely to be furloughed while individuals working in health care and distribution and supplies were less likely to be furloughed. This same difference was observed when analyzing the monthly hours worked by industry where individuals who worked in health care and distribution and supplies had higher mean hours worked over all other industries. This data was analyzed through the use of an analysis of variance and all p-values for each month of analysis were p = 0.000. The range of mean hours worked in distribution and supplies ranged from a high of 181.2 hours per month in April to a low of 160.8 hours per month in July. Likewise, the mean hours worked in health care ranged from a high mean of 87.97 hours per month in March to a low of 69.31 hours per month in April. By comparison, entertainment ranged from a high mean of 62 hours per month in March to a low of 0 hours per month in April while food service ranged from a high mean of 40 hours per month in March to a low of 12 hours per month in April. Likewise, retail saw similar losses in hours worked per month.