The Direct Support Workforce: An Examination of Direct Support Professionals and Frontline Supervisors During COVID-19

Abstract

Direct support professionals (DSPs) and frontline supervisors (FLSs) have critical roles in home and community-based services for people with intellectual and developmental disabilities. Low wages and high levels of responsibility created a long-term crisis in recruitment and retention and are exacerbated by the COVID-19 pandemic. A national sample of DSPs and FLSs were compared on demographics and work-related circumstances using data from the third Direct Support Workforce COVID-19 Survey. Significant differences were found in demographics, hours worked, wages, wage augmentations, and quality of work-life. Policy recommendations to address the worsening workforce crisis are provided.

Direct support professionals (DSPs) and frontline supervisors (FLSs) are instrumental in providing home and community-based services to persons with intellectual and developmental disabilities (IDD). DSPs provide various supports that include meeting individual needs related to health, social connections, employment, and other aspects of community living. FLSs often provide a significant amount of direct support to persons with IDD too, but their primary role is to guide and direct the work of DSPs. The work of DSPs and FLSs is the linchpin of state and national efforts to enact the full inclusion and participation of people with disabilities in their communities. However, this workforce is undervalued, as demonstrated by the low wages and lack of benefits noted in a report to the president about the direct support workforce crisis (President’s Committee for People with Intellectual Disabilities, PCPID, 2018). DSPs report that the supports and services they provide go unrecognized and that they have risked their lives during the pandemic to offer supports and services to individuals with disabilities during the pandemic (Kinder, 2020a).

Before the COVID-19 pandemic, there were 4.6 million people in the direct support workforce in 2019 (Campbell et al., 2021). The growth in the aging population from 47.8 million to 88 million by 2050 will increase the number of workers needed to provide these services (PCPID, 2018; Campbell et al., 2021). It is estimated that there will be an additional 1.3 million in-home care jobs created between 2016 and 2028, which will make this the largest-growing occupation in the United States economy (Campbell et al., 2021).

Description of the Workforce

Bogenschutz and colleagues (2014) described DSPs as “those workers who provide person-to-person assistance to people in need of daily support in activities of daily living, household tasks, personal health and safety, community access and integration, relationships, work, and a multitude of other activities.” The U.S. Bureau of Labor Statistics does not have an official classification for this essential workforce. This has likely contributed to many workforce issues, such as wage compression, because it is impossible to make direct comparisons of duties and wages with similar occupations. Because there is no occupational classification for DSPs, it’s possible they are put into the classifications of Personal Care Assistant (PCA) or Home Health Aide (HHA; Bureau of Labor Statistics, 2021). The lack of an occupation classification by the Bureau of Labor and Statistics makes it more difficult to clearly identify the DSP workforce in size, job responsibilities, and compensation, and to compare this workforce to other similar job classifications.

The largest source of data about DSPs is the National Core Indicators (NCI) Staff Stability Survey (National Core Indicators [NCI], 2020). The NCI Staff Stability Survey collects information from service providers about wages, benefits, turnover and other pertinent staff information. According to NCI, the average wage for DSPs in 2019 was $12.00 per hour (NCI, 2020). Forty-two percent of workers in this industry receive public assistance (Campbell et al., 2021). Low wages paired with a high level of responsibility for providing supports to people with significant support needs has likely contributed to the high turnover (42.8%) and vacancy rates (11.2%) of this workforce (NCI, 2020). Additionally, Pettingell and colleagues (2022) found that incentives (e.g., wage bonuses, paid time off, access to health insurance and/or retirement benefits, pay incentive or referral bonus programs) by themselves did not have a positive association with DSP retention. Rather, staff wages were the most prominent factor related to differences in DSP retention in addition to the state where the organization was located.

There is less information available about FLSs. Like the DSP role, FLS is not an identified occupation by the Bureau of Labor and Statistics, leading to the same challenges related to data describing the FLS workforce. However, a study of the direct support workforce that included FLSs found an average wage of $15.45 per hour and an annual turnover rate of 12.2% (Bogenschutz et al., 2014).

In 2017, the direct support workforce was 49% people of color and predominantly female (86%). Twenty-six percent of the workforce were immigrants. It was also an aging workforce, with an average age of 41 years and 24% of workers aged 55 and older (Campbell et al., 2021). In the National Core Indicator (NCI) Staff Stability 2020 Survey, agencies reported that DSPs were predominantly female (73.3%). Over 1/3 (38.0%) identified as White, 37.3% Black or African American, 5.6% Hispanic, 1.9% Asian, 1.4% more than one race/ethnicity, 1.0% Pacific Islander, and 0.8% for American Indian or Alaska Native and Other, respectively. Fifteen percent (15.4%) of DSPs had been employed less than 6 months, 14.3% between 6 to 12 months, 18.7% between 12 to 24 months, 12.5% between 24 and 36 months, and 39.0% 36 months or more. FLSs were also largely female (75.7%); however, they were more likely to be White (46.0%) compared to DSPs. Additionally, 33.6% of FLSs identified as Black or African American, 5.1% Hispanic, 2.1% Asian, 1.2% more than one race/ethnicity, 1.0% American Indian or Alaska Native, and 0.8% Pacific Islander and Other, respectively (NCI, 2022). There is less information available about FLSs then DSPs.

Demographic data on FLSs and DSPs from numerous fields (e.g., child mental health, individuals with IDD) vary consistently across several demographic factors. However, research comparing the two is scant. According to research conducted in 2014, senior managers tended to have higher educational attainment than frontline workers. Nearly half (46.2%) of senior managers had a master’s degree compared to 27.2% of frontline workers. Alternatively, 17% of frontline workers had a high school diploma (17.0%) compared to senior managers (7.7%) (Patterson Silver Wolf et al., 2014). In a 2022 study, Hall and colleagues found FLSs were more likely to have some college education (84% FLS vs 73% DSPs) and DSPs had a higher percent of a high school diploma or less (27%) compared to FLSs (Hall et al., 2022). Race differences were also consistent across fields. A higher percentage of senior managers (92.3%) and FLSs (81%) were White compared to 75.2% of frontline workers and 70% DSPs while fewer senior managers (3.8%) and FLSs (11%) were Black/African American compared to frontline workers (15.8%) and DSPs (21%; Hall et al., 2022; Patterson Silver Wolf et al., 2014). When looking at gender demographics, senior managers and FLSs tended to have a higher percentage of women (61.5% and 87%) than frontline workers and DSPs (58.8% and 82%; Hall et al., 2022; Patterson Silver Wolf et al., 2014).

COVID-19 Impact on the Workforce

In their report to the president, the National Council on Disability (NCD; 2021) notes that the shortage of direct support workers has been exacerbated by the COVID-19 pandemic. Prior to the pandemic, this workforce experienced difficult working circumstances, limited benefits, and low wages, which played a role in job turnover. During the pandemic, understaffing, increased work challenges, lack of hazard pay, lack of paid leave and childcare (with closed schools), and fear of catching or spreading COVID-19 led to additional turnover (National Council on Disability [NCD], 2021). Another study found similar factors related to turnover, with the additional difficulty experienced in keeping current staff and recruiting new staff with industries that had paid comparable wages in the past now paying more than they did and, in some cases, unemployment paying more than they did. Additionally, discontinuation of services, delays in launching new programs, and turning away new referrals impact the need for being able to keep current and attract new DSPs and FLSs (Dawson & Luechtefeld, 2021). The National Council on Disability (2021) also noted the difficulty in gauging the full effects of the pandemic on this workforce due to the lack of complete occupational data, which leaves some classes of workers undercounted or not counted.

A qualitative study conducted during the pandemic with home health care workers noted that these workers felt like they were invisible and not respected (Sterling et al., 2020). Little attention was paid to this workforce in the beginning of the pandemic. Many workers reported a lack of adequate training to prevent COVID-19 transmission and no access to PPE despite the close contact people providing direct support often have with the people they support (Allison et al., 2020; Kinder, 2020a; Sterling et al., 2020). In the spring of 2020, 46% of DSPs and FLSs in an online survey of 8,914 respondents reported having access to medical grade facemasks (Hewitt et al., 2020). In a follow-up survey of 8,846 DSPs and FLSs in the fall of 2020, 63% reported access to paper or disposable face masks, 36% medical grade face masks, and 36% fabric facemasks (purchased, not homemade; Hewitt, Pettingell, Kramme, et al., 2021). By summer of 2021, in a follow-up survey of 5,356 DSPs and FLSs, 91% reported they had sufficient PPE; however, one-fifth (20%) reported they had to pay out of pocket for their PPE (Hewitt, Pettingell, Bershadsky, et al., 2021).

An online survey of 478 DSPs reported that 84% believed they were at risk for contracting COVID-19. However, 95% reported that they knew how to protect themselves and the people they supported (LoPorto & Spina, 2021). In the summer of 2021, 57% of DSPs and FLSs reported exposure to COVID-19, with 19% indicating a positive COVID-19 diagnosis (Hewitt, Pettingell, Bershadsky, et al., 2021). Pandemic-related challenges such as increased workload demands along with understaffing and the risk of COVID-19 transmission were also reported by workers in age-related support services (Cimarolli & Bryant, 2021). Nearly three in ten of these workers reported challenges such as financial hardships, separation from family members, and challenges with meeting the needs of their families (Cimarolli & Bryant, 2021). Workers in home-and community-based services were more likely to report challenges than those in facility-based settings such as assisted living or nursing homes (Cimarolli & Bryant, 2021). DSPs and FLSs also reported workplace challenges. In the spring of 2020, 26% of DSPs and FLSs responding to an online survey (8,914 respondents) reported being short-staffed (Hewitt et al., 2020). In November of 2020, 50% of DSPs and FLSs in a follow-up survey (8,846 respondents) reported that their workplace was short-staffed (Hewitt, Pettingell, Kramme, et al., 2021). By summer of 2021, more than half of 5,356 DSPs and FLSs (54%) reported that their work-life had worsened during the pandemic (Hewitt, Pettingell, Bershadsky, et al., 2021). An increase in hours and responsibilities can lead to exhaustion, stress, and detachment, all factors in the development of burnout (Hewitt & Larson 2007; Skirrow & Hatton, 2007). These factors are likely contributing to the current workforce crisis with high turnover and vacancy rates (NCI, 2022; NCD, 2021; Sterling et al., 2020).

As of August 2021, a third of the states had publicly available data about HCBS service sites and the impact of coronavirus on enrollees and vaccination rates (Watts et al., 2021). Staffing shortages since the start of the pandemic have been particularly notable on in-home and group home services. Adult day programs and supported employment programs were closed for extended periods of time in order to comply with social distancing measures. McCall and colleagues (2021) found that 4%, or 168,370 DSPs, were displaced from their jobs within the first 3 months of the pandemic. Nine percent, or 14,770 workers, of these displaced workers re-entered the workforce by March of 2021, however, none had returned to direct support work. The remaining 91%, or 153,610, direct care workers remained out of the workforce at the end of the first quarter of 2021.

During the pandemic, several bills were passed at the federal level that provided additional funds to states to address the workforce challenges caused by the COVID-19 pandemic for essential workforce sectors. However, DSPs were not always beneficiaries of these efforts. For example, the Families First Corona Virus Response Act (FFCRA) of 2020 provided emergency paid sick leave for essential workers. However, according to the National Council on Disability, certain employers, such as home care agencies, were allowed to exclude DSPs if they chose (NCD, 2021). Any of the provisions of FFCRA aimed at providing assistance to essential workers excluded independent contractors. This means that DSPs hired directly by individuals using self-directed programs could not access emergency paid sick leave or any of the other provisions of this act (NCD, 2021). A survey of DSPs supporting people with aging-related needs in HCBS settings identified financial hardships as one of their main challenges (Cimarolli & Bryant, 2021).

Some states used funds provided by the Coronavirus Aid, Relief, and Economic Security (CARES) act to temporarily increase pay to essential workers. The implementation varied across states. Some provided a one-time payment; others provided a temporary hourly pay increase (Kinder, 2020b). The hazard pay is no longer being paid in most cases despite the continuation of the COVID-19 pandemic. Due to the previously discussed difficulties in identifying the DSP and FLS workforce, there is a lack of comprehensive data enabling a complete understanding of how these programs have affected DSPs and FLSs.

Purpose of the Study

The purpose of this study was to explore the similarities and differences between DSPs and FLSs in the direct support workforce. Given the dearth of data comparing these groups and their work circumstances, our goal was to compare DSPs and FLSs on demographics and work issues during the COVID-19 pandemic. The research questions included:

1. Do DSPs and FLSs differ on demographic characteristics?

2. Do DSPs and FLSs differ with respect to their working hours before and during the COVID-19 pandemic?

3. Do DSPs and FLSs differ on their wages and wage augmentations during the COVID-19 pandemic?

4. Do DSPs and FLSs differ in how they view their work-life status during the COVID-19 pandemic?

Method

Instrument

The Direct Support Workforce 12-Month Survey was the third in a series of three online surveys. It was launched using the online survey platform Qualtrics on June 1, 2021, and closed on July 25, 2021. Information about the survey and how to access it was posted on our website (https://ici.umn.edu/covid19-survey) and circulated on social media. It was also promoted and distributed to DSPs and disability organizations across the country by The National Alliance for Direct Support Professionals (NADSP), The Arc, the American Network of Community Options and Resources (ANCOR), and the National Association of State Directors of Developmental Disabilities Services (NASDDDS). The survey contained 10 items about respondent characteristics, nine items about wages and work hours, five items related to staffing, three items addressing COVID-19 safety measures at their place of employment, seven items about the individuals whom the respondents supported, eight items on well-being and work-life, 11 items about vaccination experiences, and eight items on demographic information. Two additional optional items asked respondents for their name and email address.

Sample

There were 7,366 surveys submitted in Qualtrics. Of those, 13% opened the link without answering any items, 11% reported they were DSPs or FLSs but only answered the first three questions or left the survey blank, 3% were not FLSs or DSPs, < 1% were duplicate testers (those who provided the optional name and/or email address items and could be verified to have taken it a second time), and <1% resided outside the United States. This left a usable sample of 5,356 respondents who were located in nearly all 50 states, the District of Columbia, Guam, and Puerto Rico. There were 4 states with no respondents (7%), 33 states or territories that had 1-100 respondents (61%), 9 states that had 101-250 respondents (17%), 5 states that had 251-400 respondents (9%), and 3 states that had more than 400 respondents (6%). Only DSPs and FLSs were included in analyses, therefore, the final analytic sample had 5,242 respondents. Of those 4,295 (82%) were DSPs and 947 (18%) were FLSs.

Variables

Demographic Variables

• Age was a continuous measure.

• Gender Identity was a single item with four categories: woman including transgender woman, man including transgender man, nonbinary, and prefer to self-describe.

• Race was a single item with six categories: American Indian or Native American, Asian, Black or African American, White, Some Other Race, or Two or More Races. Race groups were collapsed into Black or African American, White, and Other to explore the relationship with work role (DSP vs. FLS). Due to the small number of respondents in each category, the “Other” group included Asian, American Indian/Native American, Some Other Race, and Two or More Races.

• Ethnicity was a single item with two categories: No, I am not of Hispanic, Latino, or Spanish origin, and Yes.

• Education Level was a single item with six categories: postgraduate education, a 4-year degree, some college, a 2-year degree, a high school diploma or GED, and less than a high school diploma.

• Household Income was a single item with five options: over $100,000, $40,000 to $99,999, $22,000 to $39,999, $15,000 to $21, 999, and $14,999 or less.

• Setting Worked In was a single item asking where the participant provided the majority of their services to people with four response categories: agency or facility, family or individual home, community employment or job site, and another site not included in the options (e.g., community nonemployment [recreation, fun], school setting, telehealth/virtual).

• Primary Wage Earner in the Household was a single item with two categories: yes and no.

Hours, Wages, and Work-Life Variables

• Number of Hours Worked Per Week Before the Pandemic was a single item with five categories: less than 15 hours, 16-30 hours, 31-40 hours, 41-50 hours, and 51+ hours.

• Number of Additional Hours Worked Per Week Due to the Pandemic was a single item with five categories: none, 1-15 hours, 16-30 hours, 31-40 hours, 40+ hours.

• Hourly Wage Pre-pandemic was a continuous measure.

• Current Hourly Wage was a continuous measure.

• Receiving a COVID-19 Wage Augmentation or Bonus was a single item with two categories: yes and no.

• Amount of COVID-19 Wage Augmentation or Bonus was a single item with six categories: $0.01 to $1.00 per hour, $1.01 to $2.00 per hour, $2.01 to $3.00 per hour, $3.01 to $4.00 per hour, more than $4.01 per hour, and a lump sum bonus.

• Since the Beginning of the Pandemic, Work-life Status was a single item with five categories: much better, better, the same, worse, and much worse.

Analysis

All analyses were conducted in SPSS version 27 (IBM Corporation, 2020). Frequency distributions provided descriptive statistics. Crosstabulation tables with Chi-square tests (χ²) and t tests were run to examine differences between DSPs and FLSs. Analyses were evaluated at alpha level (α = 0.003) adjusting for the number of comparisons.

Results

Descriptive Results

Demographics

There were 5,242 respondents who were either DSPs (82%) or FLSs (18%) in the analytic sample. The average age was 45 years (SD = 13 years). Over four-fifths (83%) identified as women, including transgender women; 15% as men, including transgender men; and 1% nonbinary and preferred to self-describe, respectively. Nearly three-fourths (73%) identified as White, 19% as Black or African American, 2% as American Indian or Native American, 1% as Asian, 2% as another race not listed as an option, and 4% as two or more races. Additionally, 6% came from a Hispanic, Latino, or Spanish heritage. Fewer than 2% did not have a high school diploma, 25% had a high school diploma or GED, 15% had a 2-year degree, 30% had some college, 20% had a 4-year degree, and 8% had postgraduate education. Nearly two-thirds (63%) of respondents provided the majority of services in agency or facility sites, 28% in family or individual homes, 7% in community employment or job sites, and 2% in other settings. Nearly three-fourths (71%) are the primary wage earner in their households. Four percent of respondents had an annual household income of $14,999 or less, 10% $15,000 to $21, 999, 35% $22,000 to $39,999, 43% $40,000 to $99,999, 8% over $100,000. Lastly, two-thirds (66%) worked for their primary employer for more than 36 months, 10% between 24 to 36 months, 11% between 12 to 24 months, 8% between 6 to 12 months, and 5% less than 6 months.

Demographic Comparisons Between DSPs and FLSs

There were significant differences between DSPs and FLSs on demographic characteristics. As seen in Table 1, There were statistically significant differences between DSPs and FLSs with respect to race, χ²(2) = 34.264, p < 0.001. DSPs had a significantly higher percentage indicate Black or African American compared to FLSs (20% vs. 11%); FLSs had a significantly higher percentage indicate White compared to DSPs (80% vs. 71%).

Table 1.

Demographic Comparisons Between Direct Support Professionals (DSPs) and Frontline Supervisors (FLSs)

Variable	DSPs		FLSs
Gender Identity	N	%	N	%	p-value
Man (including transgender man)	556	16.0 a	104	13.0b	0.085
Woman (including transgender woman)	291	83.0 a	701	86.0 b
Nonbinary	38	1.0 a	6	1.0 a
Prefer to self-describe	24	<1.0 a	3	<1.0 a
Total	3,534	100.0	814	100.0
Race	N	%	N	%	p-value
Black or African American	693	20.0 a	91	11.0 b	<0.001
White	2,418	71.0 a	633	80.0 b
Other	308	9.0 a	68	9.0 a
Total	3,419	100.0	792	100.0
Hispanic, Latino, or Spanish Background	N	%	N	%	p-value
Yes	197	6.0 a	53	7.0 a	0.348
No	3,209	94.0 a	743	93.0 a
Total	3,406	100.0	796	100.0
Education Level	N	%	N	%
Less than high school	71	2.0 a	8	1.0 a	<0.001
High school diploma or GED	983	27.0 a	138	17.0 b
Some college	540	15.0 a	126	15.0 a
2-year degree	1,102	31.0 a	212	26.0 b
4-year degree	661	18.0 a	241	29.0 b
Postgraduate education	269	7.0 a	98	12.0 b
Total	3,626	100.0	823	100.0
Annual Household Income	N	%	N	%
$14,999 or less	156	5.0 a	4	1.0 b	<0.001
$15,000 to $21,999	375	12.0 a	15	2.0 b
$22,000 to $39,999	1,220	37.0 a	172	23.0 b
$40,000 to $99,999	1,239	39.0 a	474	62.0 b
Over $100,000	218	7.0 a	95	12.0 b
Total	3,208	100.0	760	100.0
Type of Work Setting	N	%	N	%
Agency or facility	2,560	59.0 a	717	76.0 b	<0.001
Family or individual home	1,321	31.0 a	156	17.0 b
Community employment or job site	330	8.0 a	51	5.0 b
Other site	83	2.0 a	23	2.0 a
Total	4,294	100.0	947	100.0
Primary Wage Earner in Household	N	%	N	%
Yes	3,066	72.0 a	646	68.0 b	0.036
No	1,206	28.0 a	299	32.0 b
Total	4,272	100.0 a	945	100.0
Age (average)	45 years		44 years		0.002

Note.

Subscript letters ^a and ^a in a row indicate column proportions do not differ significantly at the 0.05 level. Subscript letters ^a and ^b in a row indicate column proportions differ significantly at the 0.05 level.

p-values in bold represent relationships that are significant at the 0.003 level.

There were statistically significant differences between DSPs and FLSs on education level, χ²(5) = 93.905, p < 0.001. DSPs had a significantly higher percentage with a high school diploma or GED (27% vs. 17%), a significantly higher percentage with a 2-year degree (31% vs. 26%), and a significantly lower percentage of 4-year degrees (18% vs. 29%). Statistically significant differences were also present between DSPs and FLSs for annual household income, χ²(4) = 234.802, p < 0.001. DSPs had significantly higher percentages of annual household incomes of $14,999 or less (5% vs. 1%), $15,999 to $21,999 (12% vs. 2%), and $22,000 to $39,999 (37% to 23%). FLSs had significantly higher percentages making $40,000 to $99,999 (39% vs. 62%) and over $100,000 (7% vs. 12%). DSPs were significantly older (M = 45 years; SD = 14 years), on average, than FLSs (M = 44 years; SD = 12 years; see Table 1), t(1,420) = 3.500, p = 0.002. There was a significantly lower percentage of DSPs working in agency or facility settings (59% vs. 76%) and significantly higher percentages in family or individual homes (31% vs. 17%) and community employment or job sites (8% vs. 5%) compared to FLSs. These differences were statistically significant, χ²(3) = 94.959, p < 0.001 (see Table 1).

There were no statistically significant differences between DSPs and FLSs on gender identity, χ²(3) = 6.619, p = 0.085, ethnicity, χ²(1) = 0.882, p = 0.348, and primary wage earner in their household, χ²(1) = 4.383, p = 0.036.

Comparisons Between DSPs and FLSs on Hours Worked

Hours worked before the beginning of the COVID-19 pandemic and additional hours worked due to the COVID-19 pandemic were examined between DSPs and FLSs. As seen in Table 2, there were statistically significant differences between DSPs and FLSs in the number of hours worked weekly before the COVID-19 pandemic, χ²(4) = 293.617, p < 0.001, and additional hours worked weekly due to the pandemic, χ²(4) = 71.692, p < 0.001. A significantly higher percentage of FLSs worked 16 or more hours pre-pandemic (43% versus 21%). DSPs were significantly more likely to report not working any additional hours due to COVID (41% versus 27%) and a significantly higher percentage of FLSs reported working an additional 1 to 15 hours weekly due to the pandemic.

Table 2.

Weekly Hours Worked Comparisons Between Direct Support Professionals (DSPs) and Frontline Supervisors (FLSs)

Weekly Hours Worked Pre-pandemic	DSPs		FLSs		p-value
	N	%	N	%
Less than 15 hours	269	6.0a	3	<1b	<0.001
16 to 30 hours	568	13.0 a	19	2.0 b
31 to 40 hours	2,147	51.0 a	419	45.0 b
41 to 50 hours	894	21.0 a	402	43.0 b
51+ hours	378	9.0 a	96	10.0 a
Total	4,256	100.0	939	100.0
Additional Weekly Hours Due to COVID-19	N	%	N	%	p-value
None	1,667	41.0 a	244	27.0 b	<0.001
1 to 15 hours	980	24.0 a	307	34.0 b
16 to 30 hours	493	12.0 a	128	14.0 a
31 to 40 hours	326	8.0 a	64	7.0 a
40+ hours	619	15.0 a	159	18.0 a
Total	4,085	100.0	902	100.0

Note.

Subscript letters ^a and ^a in a row indicate column proportions do not differ significantly at the 0.05 level. Subscript letters ^a and ^b in a row indicate column proportions differ significantly at the 0.05 level.

p-values in bold represent relationships that are significant at the 0.003 level.

Comparisons Between DSPs and FLSs on Wages and Wage Augmentations

Hourly wages, both pre-pandemic and current, and wage augmentations due to the COVID-19 pandemic were examined between DSPs and FLSs. As seen in Table 3, DSPs (M = $14.18; SD = $3.37) on average were making significantly less pre-pandemic per hour than FLSs (M = $18.10; SD = $5.48), t(1,016) = −20.284, p < 0.001. The same trend was seen with respect to current wages. DSPs (M = $14.60; SD = $3.21) were currently making significantly less per hour, on average, than FLSs (M = $18.86; SD = $5.51), t(986) = −21.936, p < 0.001. Of note, both groups had experienced increases in average wages during the pandemic.

Table 3.

Hourly Wage and Wage Augmentation Comparisons Between Direct Support Professionals (DSPs) and Frontline Supervisors (FLSs)

Continuous Variables
Variable	DSPs		FLSs		p-value
Hourly Wage Pre-pandemic (average)	$14.18		$18.10		<0.001
Hourly Wage Currently (average)	$14.60		$18.86		<0.001
Categorical Variables
Receiving a Wage Augmentation	N	%	N	%	p-value
Yes	1,064	27.0a	234	26.0 a	0.515
No	2,867	73.0 a	666	74.0 a
Total	3,931	100.0	900	100.0
Amount of COVID-19 Wage Augmentation	N	%	N	%	p-value
$0.01 to $1.00 per hour	175	18.0 a	20	9.0b	0.001
$1.01 to $2.00 per hour	262	27.0 a	59	27.0 a
$2.01 to $3.00 per hour	221	22.0 a	74	33.0 b
$3.01 to $4.00 per hour	50	5.0 a	6	3.0 a
More than $4.01 per hour	41	4.0 a	10	4.0 a
Received a lump sum bonus	234	24.0 a	54	24.0 a
Total	983	100.0	223	100.0

Note.

Subscript letters ^a and ^a in a row indicate column proportions do not differ significantly at the 0.05 level. Subscript letters ^a and ^b in a row indicate column proportions differ significantly at the 0.05 level.

p-values in bold represent relationships that are significant at the 0.003 level.

DSPs and FLSs were asked about receiving a wage augmentation or bonus because of the COVID-19 pandemic. There were no statistically significant differences in percentage of DSPs (26%) and FLSs (27%) receiving a COVID-19 wage augmentation or bonus, χ²(1) = 0.424, p = 0.515. However, for those DSPs and FLSs who did receive a wage augmentation or bonus due to COVID-19, there were significant differences in the amount received, χ²(5) = 19.588, p = 0.001. A significantly higher percentage of DSPs received $0.01 to $1.00 per hour (18% vs. 9%); FLSs had a significantly higher percentage who received $2.01 to $3.00 per hour (33% vs. 22%). About a quarter (24%) of both groups received a lump sum bonus.

Comparisons Between DSPs and FLSs on Quality of Work Life Since the Beginning of the COVID-19 Pandemic

Finally, DSPs and FLSs differed significantly regarding their perspective of their work-life status compared to the beginning of the COVID-19 pandemic, χ²(4) = 43.012, p < 0.001. DSPs had significantly higher percentages of feeling their work-life was better (19% vs. 16%) whereas FLSs were significantly more likely to report their work-life was worse (31% vs. 24%) or much worse (13% vs. 8%) than DSPs (see Table 4).

Table 4.

Quality of Work Life Since the Beginning of the Pandemic Comparison Between Direct Support Professionals (DSPs) and Frontline Supervisors (FLSs)

Quality of work life since beginning of the COVID-19 pandemic	DSPs		FLSs		p-value
	N	%	N	%
Much better	257	7.0a	45	5.0 a	<0.001
Better	732	19.0 a	137	16.0b
The same	1,603	42.0 a	297	35.0 b
Worse	938	24.0 a	264	31.0 b
Much worse	302	8.0 a	107	13.0 b
Total	3,832	100.0	850	100.0

Note.

Subscript letters ^a and ^a in a row indicate column proportions do not differ significantly at the 0.05 level. Subscript letters ^a and ^b in a row indicate column proportions differ significantly at the 0.05 level.

p-values in bold represent relationships that are significant at the 0.003 level.

Discussion

The respondents to this survey were predominantly female (DSPs, 83%; FLS, 86%), which is consistent with other data (Campbell et al., 2021; Kinder, 2020a; NCI, 2022). They were also largely white (DSPs, 81%; FLS, 86%) which is higher than other studies. Campbell et al. (2021) reported 49% of the direct care workforce were people of color in 2017. In 2020, NCI data showed only 38.0% of DSPs and 46% of FLSs identified as White (NCI, 2022). The majority of DSPs (72%) and FLSs (68%) were the primary wage earners in their household. However, the DSPs in this sample were more likely to report an income of less than $22,000 per year (17%) than were FLSs (3%). FLSs were more likely to report making $40,000 per year or more (75% vs 45%). The average hourly wage increased slightly for both DSPs and FLSs during the pandemic ($0.42 for DSPs and $0.76 for FLSs). The increase in pay may be related to wage enhancements provided from COVID-19 relief packages; however, nearly 75% of DSPs and FLSs reported that they did not receive a wage augmentation. Given the high-risk nature of their jobs during a pandemic, identifying ways to increase their wages as essential workers during pandemics is important.

As noted previously, a survey of DSPs conducted by this research team 6 months into the pandemic showed that the staffing shortage had worsened during the pandemic, with an increase from 26% at the beginning of the pandemic (Hewitt et al., 2020) to 50% 6 months later (Hewitt, Pettingell, Kramme, et al., 2021). Now, 12 months into the pandemic, 59% of DSPs and 73% of FLSs reported working more hours due to COVID-19. For DSPs, 24% reported working 1-15 additional hours per week while 15% reported working an additional 40 or more hours per week. One-third (34%) of FLSs worked an additional 1-15 hours per week and 18% reported working an additional 40 hours per week. The additional hours worked by FLS may reflect that the FLS position is often a salaried position and the expectation in many agencies is that FLS will cover open shifts in the settings that they supervise. Providers must recognize and reward the sacrifices made by FLSs to ensure the provision of services to individuals needing support. Considering the important role that FLSs play in guiding, directing and supporting DSPs, the failure to do so will only add to staffing shortage.

The toll of working additional hours (and not receiving wage augmentation) was evident in the views of work life quality reported by DSPs and FLSs. Nearly half (44%) of FLSs reported that their work life was worse or much worse. About 1/3 of DSPs (32%) also reported a decline in work-life quality. The decline in work-life quality is likely deepening the workforce crisis that existed before the pandemic. McCall and colleagues (2021) reported that an estimated 91% of the direct care workers displaced from the workforce in 2020 had not returned to their same occupation in 2021 which is one indication of the need to urgently address the workforce crisis before the system collapses. Studies in several states reported of group homes closing and the cessation of other kinds of supports due to the lack of available staff (for example, in Florida, Minnesota, and New York; McGivern, 2021; Moore, 2021; Steiner, 2022). Efforts to address these compounded workforce issues must be implemented on national, state, and local levels to ensure that community living remains a viable option for individuals with intellectual and developmental disabilities.

Limitations

This study has several limitations. Although statistical significance was found in many of the relationships, there were a few cell sizes that were small (e.g., annual household income, weekly hours worked before the COVID-19 pandemic, and wage augmentation amounts). The sample was large, yet it is important to recognize that the survey methods used a convenience sampling approach and thus generalization should be avoided. Another limitation of this study is that participation by people of color was lower than expected compared to other studies. Additionally, the years of service of this sample, with 66% having been at their primary employer for 36 months or longer, may have contributed to a higher wage than has been reported in other studies. Because wages often rise with tenure, this may be particularly true given the high levels of turnover reported in this field (e.g., NCI, 2020)

Conclusion and Policy Recommendations

Whereas our sample was not as diverse as other national samples (e.g., NCI, 2020; NCI, 2022), the experiences of these respondents’ mirrors that of other studies and is likely an accurate reflection of the state of the direct support workforce. The challenges facing this workforce existed before the pandemic, as did the lack of attention to the crisis by policy makers. There are practices that providers can implement that have been shown to be effective in recruiting and retaining DSPs and FLSs. These practices include:

1. Marketing campaigns to promote direct support work (e.g., McCall et al., 2021).

2. Increasing base wages to make the positions more competitive (e.g., McCall et al, 2021).

3. Implementing evidence-based retention strategies such as realistic job previews, competency-based orientation and training, career paths, and mentoring (e.g., Hewitt & Larson, 2007).

4. Improved support for FLSs from organization leadership (e.g., Hewitt & Larson, 2007).

Although there are things that providers can do to address the crisis, the issue is largely systemic and requires systemic solutions on federal and state levels. Important policy recommendations for addressing the DSP and FLS workforce crisis include:

1. The U.S. Department of Labor needs to establish a standard occupational classification (SOC) code for DSPs (Hewitt, Pettingell, Kramme, et al., 2021) to identify this specific workforce and ensure that federal and local policies specifically include DSPs and FLSs. Having a SOC code for DSPs would allow DSPs to be categorized based on the skill requirements for their work rather than being inaccurately lumped into classifications with PCAs or HHAs (U. S., Bureau of Labor Statistics, 2021), would provide the mechanism for appropriately setting reimbursement rates for services provided by DSPs and would create the capacity to consistently identify staffing needs and gaps in services (National Alliance for Direct Support Professionals, 2018).

2. DSPs are primary wage earners and often single parents (Hewitt et al., 2019; PHI, 2019). McCall et al. (2021) found that 16% of men and 10% of women with children were less likely to re-enter the workforce then men without children at home. Access to affordable childcare and strategies that connect DSPs to childcare and other supports is essential for this workforce’s continued participation in providing supports (McCall et al., 2021). The pandemic only exacerbated the challenges workers have in finding affordable, reliable childcare.

3. Policy makers need to address the underlying causes for the workforce crisis including reimbursement rates for long term services and supports so that it is possible to provide a living wage for the DSPs and FLSs who do this essential work. Low wages have been shown to be a predictor for high turnover (Houseworth et al., 2020), therefore, increasing wages and Medicaid funding would provide agencies the opportunity to offer living wages and benefits. This may in turn ameliorate some of the factors contributing to high turnover.

The Coronavirus Aid, Relief and Economic Security (CARES) Act and the American Rescue Plan Act (ARPA) provided important assistance for providers to address workforce issues during the pandemic. The CARES Act, for example, included a Provider Relief Fund for provision of health care services, including community-based organizations to compensate for pandemic related expenses and lost revenue (ANCOR Foundation and United Cerebral Palsy, 2022; Kaiser Family Foundation [KFF], 2020). The CARES Act ended in 2021. ARPA, enacted in 2021, specifically targeted funding for Medicaid-funded home- and community-based services (HCBS) by allowing states to apply for a 10 percentage-point increase to the federal matching rate (known as “FMAP,” or the Federal Medical Assistance Percentage). The intention of this funding was to strengthen states’ HCBS programs and services (ANCOR Foundation and United Cerebral Palsy, 2022; KFF, 2021). Among the allowed expenditures include programs aimed at workforce recruitment and retention (KFF, 2021). Forty states used the ARPA funds to strengthen home and community-based services in areas such as recruitment and retention, support and training, and reimbursement (Manz, 2022). Although the CARES Act helped stabilize community-based providers during the pandemic, ARPA has the potential for providing a foundation for improving working conditions for DSPs and addressing recruitment and retention challenges. However, states need to be creating policies and practices that sustain and programs developed during ARPA to ensure a more stable and competent workforce into the future.