## Abstract

Direct support professionals (DSPs) and frontline supervisors (FLSs) play an integral role in the lives of people with intellectual and developmental disabilities (IDD) and are often the individuals directly responsible for assisting people with IDD to live and fully participate in their communities. These two groups of workers have typically been employed at lower wages with limited access to fringe benefits, contributing to high rates of turnover compared to a similarly skilled worker in the United States. This article summarizes findings and is the first investigation in several years to systematically examine the wages, fringe benefits, and stability of the DSP and FLS workforces supporting individuals with IDD. Findings suggest that a typical DSP may expect to earn about $11.25 per hour, while FLSs may expect wages of about$15.45 hourly. Of concern, however, is that fringe benefit provision was quite limited in this sample. Implications, including relation to past reports of DSP workforce development, are discussed.

## Review of Literature

### Defining Direct Support and Frontline Supervision

The profession of direct support encompasses a wide variety of workers in many different health and human service sectors. Generally, direct support professionals (DSPs) are 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. These workers may serve highly diverse populations, commonly finding work supporting people with intellectual and developmental disabilities (IDD), older adults, people with physical health care needs, people with mental illness, and people with substance misuse disorders in settings such as a participant's home, group homes, residential treatment facilities, hospitals, nursing homes, and job training and support programs. Taken together, DSPs form the backbone of the American health and social service system.

Direct support workers have many titles including direct care workers, personal care assistants, home health aides, and so on. In all service sectors, DSPs amount to roughly 3.38 million workers in the United States (PHI, 2013). The direct support workforce is also expanding at a rate faster than almost any other sector of the U.S. labor force, with an expectation of about five million people working in direct support by 2020, when projections suggest that direct support may be the largest job classification in the country (PHI, 2013). This rapid occupational expansion, which is driven largely by the greying of the population and the transition from institutional to community service provision, is accompanied by a contraction in the pool of workers who have traditionally worked in direct support, meaning service providers will likely need to engage a more diverse employee base in upcoming years (National Direct Service Resource Workforce Center, 2008; Scan Foundation, 2012; Ward, Smith, Bales, & Sandberg, 2009).

Related to direct support is frontline supervision. Frontline supervisors (FLS) are the first line of management in many human service organizations, typically providing supervision for DSPs and often engaging in direct support as a portion of their duties. There are two nationally validated competency sets that elaborate the depth of skill that is necessary to complete work as an FLS in the IDD field (Larson et al., 2007; Sedlezky, Reinke, Larson, & Hewitt, 2013). The competency conceptualization by Larson and colleagues (2007) included 14 broad competency areas, ranked by importance by a national sample of IDD service managers and FLSs. This work was expanded and revised by Sedlezky and colleagues (2013), who also built in methods for individualizing competencies based on work priorities of individual FLSs.

#### Fringe benefit provision

Provision of fringe benefits for DSPs generally occurs at low rates. Of particular importance is the availability of health care benefits. Among personal care aides, 31% are estimated to be without any form of health insurance, which compares unfavorably to the 18% of all American workers who are without health coverage (PHI, 2011). Only 47% of direct support workers (in all sectors) utilize health insurance coverage through their employer, while roughly 18% rely on public insurance programs to meet their health care needs (PHI, 2011). Complicating matters, many employers do offer health coverage benefits, but those benefits are not available to DSPs who work part time. Because many DSPs are hired to work part-time hours, many are excluded from eligibility for employer-sponsored health insurance plans.

As is the case with health coverage, availability of other fringe benefits is limited. Larson and colleagues (2005) compiled a comprehensive list detailing fringe benefit availability in 11 IDD-focused studies conducted between 1992 and 2003. Of note, the percentage of organizations providing access to retirement savings plans varied between 21–81%, and access to dental insurance ranged between 26–82% of organizations surveyed in various studies. Paid time off was provided by a majority of IDD services organizations surveyed in all studies but one, suggesting that this may be the most commonly provided form of fringe benefit for DSPs. In a statewide study of DSPs in North Carolina, Test, Flowers, Hewitt, and Solow (2003) found that more than half of organizations serving people with IDD offered health insurance, paid vacation, holiday pay, sick leave, retirement savings, dental insurance, and personal leave, though fewer organizations reported that these benefits were paid by the organization. In all of these studies, however, it is important to note that many DSPs may be excluded from eligibility for a particular fringe benefit based on being part-time employees (Larson et al., 2005).

#### Workforce stability

The direct support workforce has traditionally been marked by high rates of employee turnover. The most recent studies have varied in the magnitude of turnover reported, but generally fall between 45% and 70% annually within a particular organization (Braddock & Mitchell, 1992; Larson et al., 1998; Larson et al., 2005). The most recent estimate of nationwide turnover puts the annual separation rate at 38% (ANCOR, 2010). Further complicating matters, many DSPs who leave their positions do so within the first 6 months after being hired (Larson et al., 1998), resulting in a revolving door of DSPs in many organizations.

The fiscal impacts of DSP turnover are staggering. Studies have placed the cost of replacing a single DSP in IDD services at $2,413 (Hewitt & Larson, 2007) to$4,872 (ANCOR, 2010). Conservatively, the cost of turnover nationwide is estimated to be a minimum of $784 million annually (Hewitt & Larson, 2007). ### Need for Study The current study was the first in several years to investigate issues related to wages, access to benefits, and stability of the direct support workforce serving people with IDD in community settings, including both residential and day training and habilitation that included vocational supports. In a rapidly changing landscape for the provision of supports for people with IDD, the current study provides an important summary of the current state of one of the nation's largest labor market segments. The main research questions were • What are the wage, turnover, and fringe benefit characteristics of the direct support workforce in community-based IDD services? • How do those workforce indicators differ based on geographic location and service type? ## Method This study was a cross-sectional examination of the direct support workforce serving people with IDD. Based on site-level data from 16 multisite organizations, the current investigation was a part of a larger longitudinal intervention project. Methods used in the study were reviewed and approved by a university-based institutional review board. ### Sampling and Participants A purposeful cross-sectional sampling procedure was utilized in the completion of this study. The authors began sampling by seeking cooperation from local developmental disability provider organization administrators who were informed about the objectives of the larger intervention project and were asked to pass information to other organizations that may be interested in participation. Interested organizations then contacted a member of the research team directly for screening and to ask pertinent questions about the study and the requirements of participant organizations. The researchers selected organizations with the objective of achieving a sample that was representative of community service organizations in Minnesota, the state involved in the study. Specifically, emphasis was placed on gaining an equal balance of residential provider organizations and day training and habilitation services providers. Additionally, geographic diffusion was sought, so the distribution of organizations in urban, suburban, and nonmetropolitan areas represented the population of the state. #### Inclusion criteria Organizations were screened for inclusion based on a number of mandatory inclusion criteria. All organizations included in this study were also a part of a larger piece of longitudinal research. Thus, sites enrolling in the research had to have multiple sites (such as group homes), access to appropriate Internet and computers needed for a component of a training intervention, as well as commitment from management to participate in all aspects of the project (e.g., roll out training, supervisor engagement, administration of surveys, consents processes). In cases in which DSPs were unionized (two organizations in total), approval for participation was also secured from union representatives. Participating organizations had to provide either day training and habilitation or residential services for people with IDD in a setting deemed community-based, as determined by state HCBS Medicaid waiver funding requirements. #### Participating sites In total, 147 sites participated and were included in the analysis for this study. These sites were distributed between those that provided residential supports (n = 122) and day training and habilitation supports (n = 25). Despite the lower number of day training and habilitation sites, these sites typically served much larger numbers of individuals with IDD, so the total person served census among the day supports programs (n = 1,925) was much larger than residential supports providers (n = 445). The data represented a total of 1,220 DSPs working across all sites including day services (n = 405) and residential services (n = 815). A summary of the site characteristics is found in Table 1. Table 1 Site Characteristics Of note, the sample contained sites that were representative of the state where the study was conducted. The distribution of residential and day training and habilitation sites, metropolitan and nonmetropolitan regions, and overall organization size were similar to the state's overall profile of multisite service providers. ### Instrument The instrument used in collection of the data reported in this study was a two-page survey that elicited a variety of site-level workforce indicators. Among the data collected were reports on the number of DSPs and supervisors working at each site (as well as more specific census figures for the number of DSPs who were full-time, part-time, and on-call), turnover rates in the past 12 months (including crude separation within 6 months and 1 year after hire), fringe benefit eligibility criteria, the number of critical incident reports filed at each site in the last 6 months, and average wage rates for DSPs and supervisors working in each site. Terms were defined, in writing, at the beginning of the survey to enhance standardization regarding the categorization of workers into appropriate categories. The survey had been thoroughly vetted with multiple stakeholder groups in previous studies (e.g., Lakin et al., 2003; Larson, Nord, Salmi, Doljanac, & Hewitt, 2008). Although validation of the instrument did not take place prior to the beginning of this study, the instrument had been used in other applications in which the survey's constructs, wording, and presentation had been utilized successfully. In the case of annual crude separation, which was used to report turnover, variables from the survey were used in a subsequent calculation. Crude separation rates were calculated for each site as an indicator of workforce stability. Crude separation was calculated by dividing the number of departing staff within a specified timespan by the sum of the total number of vacant positions and DSPs working at each site at the time of survey. Annual crude separation was calculated using the number of DSPs who had left the site for any reason (voluntary termination, involuntary termination, transfer, promotion, etc.) in the 12-month period immediately preceding survey completion. Additionally, turnover rates were calculated for DSPs who left their positions within the first 6 months after initial hire, using the same methods of calculation. ### Data Collection The survey was completed by an administrator or manager associated with each site, under the guidance of members of the research team. In most cases, the person completing the site information survey was a human resources manager or program director who had access to all of the information necessary to accurately furnish data. In the remaining cases, the person completing the survey was a site supervisor with access to the necessary information. In all cases, a member of the research team assisted the organization's personnel with definitions, employee classifications, and other technical aspects of data collection. This collaboration between site personnel and research team members ensured a good degree of internal consistency. Completed surveys were returned directly to an assigned member of the research team for data entry and analysis, either in person or via postal mail. Data were collected in a staggered manner, as organizations consented to participation in a larger intervention project. Data collection commenced in mid-2009 and was completed in mid-2011. ### Analysis Statistical analyses were completed using IBM SPSS Version 21. Summative descriptive statistics were initially utilized to gain a sense of the workforce in the participating sites in a general sense. In this analytic phase, general workforce status indicators were examined, including mean wage rates, turnover rates, early turnover, and fringe benefit availability. Additionally, comparisons were made based on service type to investigate whether there were significant differences in the workforce indicators based on whether the site provided day services or residential supports. This segment of the analysis involved a series of independent samples t-tests, which are reported in the results section. Likewise, because regional cost of living varied between the metropolitan and nonmetropolitan areas, comparisons were also made based on region for wage variables. ## Results Results obtained based on the analysis procedures previously outlined are presented in this section by topic. The DSP wage information is presented first, followed by findings about turnover and DSP stability and fringe benefit eligibility. Finally, general descriptive information about the status of frontline supervisors is presented. All results presented in this report were based on site-level data, which aggregated data for all DSPs or supervisors in a particular worksite. ### DSP Workforce Indicators The state of the DSP workforce was developed around three common types of indicators: wages, fringe benefit availability, and crude separation (annual turnover) rates. Each of these areas is expanded in the following section. #### DSP wages The DSP wages were reported as the mean hourly wage among all DSPs working in a particular site, as reported by a supervisor or human resource professional with access to site-level data within each organization. Thus, this report highlighted aggregated site averages, not individual wage rates. The overall mean hourly wage across all sites was$11.26 (SD  =  1.38). Mean site wages ranged from as low as $9.16 to a high of$18.00, representing a substantial differential. The mean wage at day training and habilitation sites (M  =  12.50, SD  =  1.47) was significantly higher than among residential sites (M  =  11.06, SD  =  1.27; t[109]  =  4.01, p < .001).

To further investigate wage differentials, a one-way analysis of variance (ANOVA) was performed to bring additional insight to wage disparities between DSPs in day and residential sites located in metropolitan and nonmetropolitan areas. The DSP wages differed significantly based on type and region of service (F[3, 117]  =  33.56, p < .001).

Post-hoc tests using Tukey's HSD method were performed to investigate the wage differences indicated by ANOVA. Tukey post-hoc comparisons indicated statistically significant differences between wages of DSPs working in day training and habilitation sites in the metropolitan region (M  =  $13.62, SD = .69) and their counterparts in nonmetropolitan day programs (M =$11.71, SD  =  1.24), and residential services in all regions (metropolitan M  =  $11.94, SD = 1.41; nonmetropolitan M =$10.49, SD  =  .77). Only when comparing nonmetropolitan day services DSPs and residential DSPs in the metropolitan region were the differences in wages not statistically meaningful. Taken together, these findings suggest notable wage disparities among DSPs in different service settings and regions, which will be discussed in greater depth in the discussion section of this article. DSP and FLS wages and turnover rates by service type and location are summarized in Table 2.

Table 2

DSP and FLS Wages and Turnover Rates by Service Type and Location

#### DSP stability and composition

Beginning with annual crude separation rates, calculated using the method described in the instrumentation section, the mean turnover rate reported by all sites participating in this study was 25.76% (SD  =  19.88), indicating that the typical site replaced over a quarter of its workers in the previous year. Turnover rates reported by the sites were quite variable, with the range spanning from 0% to 100% annual crude separation. Among residential sites, the mean annual crude separation rate was 25.38% (SD  =  19.18), while it was slightly higher in day training and habilitation locations at 27.49% (SD  =  23.17). The difference between residential sites and day program sites was not statistically significant.

Turnover of DSPs within their first 6 months of employment was also calculated. This early turnover was found to be at a rate of 7.45% (SD  =  12.19) in the total sample, and at 8.19% (SD  =  13.68) in day services sites and 7.29% (SD  =  11.91) in residential support locations. Differences between day and residential sites were not significant.

Of additional interest was the distribution of full-time, part-time, and on-call DSPs. Full-time workers in the average day program site comprised the vast majority of DSPs (87.6%, SD  =  21.1), while the composition of the direct support workforce in residential settings is comprised of mainly part-time workers, who made up 58.4% (SD  =  25.4) of the DSPs in a typical residential site. As illustrated in Table 3, statistically significant differences existed in the mean utilization rates of full-time, part-time, and on-call DSPs, compared across day training and habilitation services and residential services sites.

Table 3

DSP Site Workforce by Service Type and Full-Time, Part-Time, or On-Call Status

#### DSP fringe benefit eligibility

All of the sites in this study reported that they officially offered both paid leave and health care coverage to some employees, although the proportion of employees that actually took advantage of these fringe benefits was not known in the context of this study. Although these benefits were available in all respondent sites, the sites varied considerably in the hours DSPs had to work in order to qualify.

Overall, health insurance eligibility for DSPs was available to workers who were employed for at least 27.49 hours per week (SD  =  9.06). Although some sites made health insurance available to employees who worked as little as 1 hour per week, others required up to 35 hours of hourly scheduled work to qualify.

Likewise, the range of work hours required for eligibility for paid leave time was quite variable, ranging from only 1 hour to 32 hours per week. The mean requirement was 12.22 (SD  =  10.84) scheduled hours per week for paid leave eligibility. The differences between the hours required for paid time off eligibility among residential sites (M  =  11.11, SD  =  11.19) and day supports (M  =  17.16, SD  =  7.47) were significant (t[134]  =  2.57, p  =  .011).

### Frontline Supervisor Workforce Indicators

This study also investigated basic workforce status of FLSs. The overall annual salary reported by this study's sites was $32,130.56/$15.45 hourly (SD  =  6,245.24) and ranged from $7,696 to$52,900. The difference between the FLS salaries in day services sites and residential services sites was statistically significant (t[105]  =  3.99, p < .001). At a mean of $37,955.50/$18.25 hourly (SD  =  8,969.55), day training and habilitation service FLSs made significantly higher annual salaries than their counterparts in residential services (M  =  31,253.69/$15.02 hourly, SD = 5,256.37). Wages for FLSs in the metropolitan region (M = 33,532.88/$16.12 hourly, SD  =  7,734.51) were marginally higher than their counterparts in nonmetropolitan parts of the state (M  =  31,293.36/$15.04 hourly, SD = 5,038.66), but the difference was not statistically significant. #### FLS turnover Annual crude separation rates for frontline supervisors were calculated using the same formula as described for DSPs in the section above. Six-month departures were not gathered for this group, so only annual turnover rates were reported. Annual turnover for FLSs in the sites surveyed occurred at a rate of 12.32% (SD = 28.74) in the total sample. As was the case in the sites' reports of DSP turnover rates, a large range existed among annual crude separation rates for FLSs, with several sites reporting no FLS turnover and some sites reporting rates up to 100% in the previous year. Although differences between residential and day program sites were not statistically significant, qualitative differences did exist, with residential programs reporting 13.04% (SD = 29.68) turnover in the previous year and day programs experiencing only a 9.23% (SD = 24.61) change in frontline supervisors annually. ## Discussion The purpose of this study was to update current knowledge of the characteristics of the wages, turnover rates, and fringe benefits of DSPs and FLSs in community services for people with IDD, with particular attention to the differences in workforce indicators that may exist based on location and service type. Building on the results, the discussion investigates issues of particular importance for policy makers, human service managers, and frontline workers. ### DSP Wages, Benefits, and Turnover At$11.26 per hour, the mean wage for DSPs in Minnesota's community services for people with IDD reported in this study was somewhat higher than might be expected based on cross-sector wage trends recently released by the U.S. Bureau of Labor Statistics (2012), ANCOR (2010), and PHI (2013) for the broad categorical group of direct support workers. It is also somewhat higher than most wage studies conducted on DSPs in community services for people with IDD in the past decade. It is highly likely that the reason for this higher-than-average wage distribution is the strong representation of vocational services in the day training and habilitation sites, which had wages that significantly outpaced residential sites. When looking solely at the residential wage in this study the findings are more consistent with previous studies. This finding is also consistent with previous studies that have revealed higher wages among DSPs in vocational services. Some day training and habilitation sites in the current study had mean wages at $18.00 per hour, which is substantially higher than the typical wage for direct support professionals and likely contributed to bringing the overall mean wage upwards. Although it is a positive development that wages were found to average$11.26 across all sites, some cautions are still warranted. Based on a 40-hour per week work schedule, this wage equates to only about $23,400 annually, which would still place many DSPs in a family of four just below the federal poverty guideline, which was set at$23,850 in 2014 (U.S. Department of Health and Human Services, 2014). Because it is known that many DSPs are not offered a full 40 hours per week work schedule, it is likely that many workers make far less from their work as a DSP, meaning that many must rely on multiple forms of income in order to adequately provide for their families.

Although wage rates were on the high end of what may be expected, turnover rates were slightly lower than expected. With mean turnover across all sites being 25.76%, the turnover rate in the present study is about 11 percentage points lower than the most recent general estimate (ANCOR, 2010) and substantially lower than previous turnover estimates in community services for people with IDD (Braddock & Mitchell, 1992; Larson et al., 2005).

Although the precise reason for this reduced turnover is unknown, at least two possibilities may exist. First, the higher wage rates reported in the overall sample may have contributed to organizations' ability to retain DSPs, who may have seen benefit in remaining in their positions rather than leaving for jobs in other sectors (such as retail) that typically offer wages lower than $11.26 per hour. Second, it is likely that the scarcity of secure work opportunities due to the nationwide recession prompted some DSPs to remain in their positions because other opportunities may not have been as readily available as would have been expected prior to the economic downturn. Additionally, this study had strong representation from work sites in nonmetropolitan areas, where alternate employment opportunities are more limited. Regardless of the reasons for the decline in turnover, it remains comparable to human service professions. For instance, the U.S. Bureau of Labor Statistics estimated that educational services had an industry-wide crude separation rate of 26.1% in 2013, while the health care and social assistance sector had a total turnover rate of 29.4% (U.S. Bureau of Labor Statistics, 2014). This turnover continues to produce high costs to organizations, both fiscally and in terms of service quality. The finding that all sites represented in this study offered fringe benefits to at least some DSPs is promising, though it does leave questions. This study did not seek to understand how many DSPs are actually eligible for benefits (many part-time workers are excluded, for instance), nor how many eligible DSPs eventually enroll in benefit plans. These are significant factors that may mean many DSPs do not have access to employer-sponsored health care, despite the fact that employers technically do offer benefits to some employees. This may be the case particularly in residential services, where the majority of DSPs are either part-time or on-call, based on this study's findings. Additionally, DSPs who are not considered to be full-time are often excluded from benefits eligibility or may be offered fringe benefits at an out-of-pocket cost. Further study is necessary on this matter, particularly in light of new regulations of the Patient Protection and Affordable Care Act (PPACA) that places regulations on both employers and individuals regarding the purchase of health insurance. The finding of strong reliance on full-time DSPs in day training and habilitation services and part-time and on-call DSPs in residential services is also worthy of discussion. Particularly considering that DSPs in day services receive significantly higher wages than their counterparts working residentially, the availability of full-time hours for DSPs in day training and habilitation services may mean that such workers may be able to consider direct support as a stand-alone career, with less need for seeking additional employment. Additionally, because many employers only offer fringe benefits such as health insurance to full-time employees, there may be built-in incentives to seek work in day programs versus residential sites, especially considering provisions of the PPACA, which mandates that larger employers offer health insurance benefits to full-time workers. Although these findings are positive for day service providers, the converse may be true for residential services, where lower wages, unavailability of full-time hours, and more constricted access to important fringe benefits may result in challenges to attracting highly qualified DSPs who will remain employed over time. ### FLS Status This study's findings related to the status of frontline supervisors in community services for people with IDD are quite important, as they are the first snapshot of this population to emerge in over 7 years (Larson et al., 2007). Calculated hourly, FLSs in the current study made a wage of$15.45, which compares favorably to the most recent prior estimate of \$11.98 (Larson et al., 2007). Based on the U.S. Consumer Price Index (U.S. Department of Labor, 2014) rate of inflation from 2007 to 2013 (12.4%), observed wage rates for FLSs in this study outpaced the inflation-adjusted wage reported by Larson and colleagues in 2007. Of note, however, is that FLS wages are significantly higher in vocational supports, represented in day training and habilitation sites, than they are in residential supports, as was found with DSP wages. With such wage discrepancies by service type for both DSPs and FLSs, there is some concern that there may be a hierarchy in attractiveness of positions in vocational and residential service sites. It is also interesting that residential sites in this study provide more integrated, community-focused services and supports, yet these workers are paid less than those in day training and habilitation programs, which are more segregated with regard to service model. Consideration by policy makers may be important in setting more equitable reimbursement rates for services, enabling more consistency in wages. Comparable wages in residential services may be able to attract higher-skilled workers and will ensure that workers who are providing more inclusive services that are focused on community integration and participation are rewarded for this outcome.

As was the case with DSPs, turnover rates for FLSs have reduced notably since the last systematic study. This study found FLS turnover across all sites to be about 12%, compared with about 24% in the last study by Larson and colleagues (2007). Although the reason for this reduction is not clear, it seems likely that it may be related, at least in part, to difficulties in postrecession job mobility, wage rates for FLSs that appear to outpace inflation, and variation in the organizations participating in the study.

### Implications

The present study provides an important update to what we know about the direct support and frontline supervision workforce that supports people with IDD. Although wages are marginally higher among respondents in this study than has been observed in similar studies in the past decade, low wages prevail, making it difficult to recruit and retain DSPs who may best exhibit the broad competencies required of their positions. In the absence of higher wages and better benefits, employers must consider creative strategies for building a competent, sustainable workforce. Likewise, more study is needed to determine which workforce development strategies provide the greatest returns in community support settings. Studies that focus on specific outcomes of carefully implemented, high-fidelity workforce development interventions are in particularly high need.

The provision of fringe benefits, particularly health insurance, remains somewhat of an unknown. Although the results of this study suggest that most employers offer health insurance coverage to some employees, it is unclear whether most DSPs and FLSs are eligible and able to buy into employer-sponsored insurance programs. The implementation of the PPACA also raises questions about how health insurance will be provided to frontline personnel supporting people with IDD. The PPACA's individual insurance mandate will mean that DSPs and FLSs will be covered in coming years, but it is uncertain the extent to which human service providers will be able to provide employer-sponsored coverage, or the extent that DSPs working in IDD services will be able to afford insurance provided through state or federal marketplaces, even when subsidized. It is also likely many DSPs will benefit from Medicaid expansion in states where this has occurred, as many continue to live near the federal poverty line. However, because not all states have expanded Medicaid, many DSPs may be in danger of falling into the Medicaid gap, where they make too much money to qualify for Medicaid, but not enough to comfortably afford marketplace-based insurance premiums. Future research may wish to focus on the patterns of health care insurance provision that DSPs pursue.

### Limitations

This study was a cross-sectional analysis of the direct support and frontline supervision workforce in one state. Because states differ in their IDD service systems, caution must be exercised in extrapolating results to other IDD systems. Data were collected over a 2-year period in which many organizations serving individuals with disabilities were facing an uncertain fiscal climate, so it is possible that wage and turnover trends may have shifted during the data collection period due to extraneous influences. In addition, this study only looked at DSPs and FLSs working in agency-based community services. Findings may not be extended beyond this sampling frame to be applied to self-directed supports or institutional settings, which are both known to differ from agency-based community services on some workforce status indicators (Bogenschutz et al., 2010; Braddock & Hemp, 2004; Duffy, 2004; Hewitt, Larson & Lakin, 2000; Lakin et al., 2003). Additionally, it is possible that the wage rates may be higher than expected due, in part, to the study's inclusion criteria. It is feasible that organizations that self-selected to participate in a study of workforce development are already aware of the importance of employee compensation and pay their DSPs a higher wage as a result. Additionally, the requirement for all participating organizations to have multiple sites (thereby excluding some small, single-site organizations) may have resulted in higher overall wage rates.

Finally, the use of site-level data must be viewed within its limitations. All figures in this study's findings are based on site-level means. Although this has relatively little effect on measures of turnover and wages, workforce status indicators about fringe benefit provision and eligibility are largely made at the organizational level. This may lead to some skewing of results in this study, because there are many more residential sites, which often abide by the same organization-level policies regarding fringe benefit eligibility. Despite these limitations, this study provided the first comprehensive look into the frontline workforce supporting people with IDD in several years, as well as being one of the first examinations of the state of frontline supervisors in the IDD field.

## Acknowledgments

This article was supported in part by Grant H133B080005 from the National Institute on Disability and Rehabilitation Research, U.S. Department of Education to the Research and Training Center on Community Living at the University of Minnesota.

## References

References
American Association on Intellectual and Developmental Disabilities and The Arc of the United States
. (
2008
).
Employment: Joint position statement
.
American Network of Community Options and Resources
. (
2010
).
2009 direct support professionals wage study
.
Alexandria, VA
:
Author
.
Americans With Disabilities Act of 1990, Pub. L. No. 101-336, 104 Stat. 328 (1991).
Bogenschutz
,
M
.,
Hewitt
,
A
.,
Hall-Lande
,
J
.,
&
LaLiberte
,
T
.
(
2010
).
Status and trends in the direct support workforce in self-directed supports
.
Intellectual and Developmental Disabilities
,
48
,
345
360
.
,
D
.,
&
Hemp
,
R
.
(
2004
).
Developmental disabilities in North Dakota: A study of the structure, financing and quality of residential and community services
.
Boulder, CO
:
University of Colorado, Department of Psychiatry
.
,
D
.,
&
Mitchell
,
D
.
(
1992
).
Residential services and developmental disabilities in the United States: A national survey of staff compensation, turnover and related issues
.
Washington, DC
:
American Association on Mental Retardation
.
Butterworth
,
J
.,
Migliore
,
A
.,
Nord
,
D
.,
&
Gelb
,
A
.
(
2012
).
Training of employment consultants leads to better employment outcomes for job seekers with intellectual and developmental disabilities
.
Journal of Rehabilitation
,
78
,
20
29
.
Duffy
,
N
.
(
2004
).
Keeping workers covered: Employer-provided health insurance benefits in the developmental disabilities field
.
New York, NY
:
John F. Kennedy Jr. Institute for Worker Education, City University of New York
.
Erickson
,
W
.,
Lee
,
C
.,
&
,
S
.
(
2013
).
Disability statistics from the 2011 American Community Survey (ACS)
.
Ithaca, NY
:
Cornell University Rehabilitation Research and Training Center on Disability Demographics and Statistics (StatsRRTC)
.
Heller
,
T
.,
Miller
,
A. B
.,
&
Factor
,
A
.
(
1999
).
Autonomy in residential facilities and community functioning of adults with mental retardation
.
Mental Retardation
,
37
,
449
457
.
Hewitt
,
A
.,
&
Larson
,
S
.
(
2007
).
The direct support workforce in community supports to individuals with developmental disabilities: Issues, implications, and promising practices
.
Mental Retardation and Developmental Disabilities Research Reviews
,
13
(
2
),
178
187
.
Hewitt
,
A
.,
Larson
,
S
.,
&
Lakin
,
K. C
.
(
2000
).
An independent evaluation of the quality of services and system performance of Minnesota's Medicaid Home and Community Based Services for persons with mental retardation and related conditions
.
Minneapolis, MN
:
University of Minnesota, Research and Training Center on Community Living
.
IBM SPSS Statistics for Windows (Version 21.0) [Computer software]
.
Armonk, NY
:
Author
.
Lakin
,
K
.,
&
Bruininks
,
R
.
(
1981
).
Occupational stability of direct-care staff of residential facilities for mentally retarded people
.
Minneapolis, MN
:
University of Minnesota: Center on Residential and Community Services
.
Lakin
,
K
.,
Hill
,
B
.,
Bruininks
,
R
.,
Hauber
,
F
.,
&
Krantz
,
G
.
(
1983
).
Factors related to job stability of direct-care staff of residential facilities for mentally retarded people
.
Journal of Community Psychology
,
228
235
.
Lakin
,
K. C
.,
Polister
,
B
.,
&
Prouty
,
R. W
.
(
2003
).
Wages of non-state direct-support professionals lag far behind those of public direct-support professionals and the general workforce
.
Mental Retardation
,
41
,
141
146
.
Larson
,
S. A
.,
Doljanac
,
R
.,
Nord
,
D. K
.,
Salmi
,
P
.,
Hewitt
,
A. S
.,
&
O'Nell
,
S
.
(
2007
).
National Validation Study of Competencies for Frontline Supervisors and Direct Support Professionals: Final report
.
Minneapolis, MN
:
University of Minnesota, Research and Training Center on Community Integration
.
Larson
,
S
.,
Hewitt
,
A
.,
&
Knoblauch
,
B
.
(
2005
).
Recruitment, retention, and training challenges in community human services: A review of the literature
.
In
S
.
Larson
,
&
A
.
Hewitt
,
(
Eds.
),
Staff recruitment, retention, & training strategies for community human services organizations
(pp.
1
18
).
Baltimore, MD
:
Paul H. Brookes
.
Larson
,
S
.,
Lakin
,
K
.,
&
Bruiniks
,
R
.
(
1998
).
Staff recruitment and retention: Study results and intervention strategies
.
Washington, DC
:
American Association on Mental Retardation
.
Larson
,
S
.,
Nord
,
D
.,
Salmi
,
P
.,
Doljanac
,
R
.,
&
Hewitt
,
A
.
(
2008
).
Critical competencies and priority training needs for frontline supervisors
.
Journal of Intellectual Disability Research
,
52
,
763
.
Larson
,
S. A
.,
Salmi
,
P
.,
Smith
,
D
.,
Anderson
,
L
.,
&
Hewitt
,
A. S
.
(
2013
).
Residential services for persons with intellectual or developmental disabilities: Status and trends through 2011
.
Minneapolis, MN
:
University of Minnesota, Research and Training Center on Community Living, Institute on Community Integration
.
The Lewin Group
. (
2008
).
A synthesis of direct service workforce demographics and challenges across intellectual/developmental disabilities, aging, physical disabilities and behavioral health
.
Washington, DC
:
Author
.
Migliore
,
A
.,
Mank
,
D
.,
Grossi
,
T
.,
&
Rogan
,
P
.
(
2007
).
Integrated employment of sheltered workshops: Preferences of adults with intellectual disabilities, their families, and staff
.
Journal of Vocational Rehabilitation
,
26
,
5
19
.
National Direct Service Workforce Resource Center
. (
2008
).
A synthesis of direct service workforce demographics and challenges across intellectual/developmental disabilities, aging, physical disabilities, and behavioral health
.
Olmstead vs. LC and EW. 527 U.S. 581. (1999).
Omnibus Budget Reconciliation Act of 1981, Pub. L. No. 97-35, § 2176 (1981).
PHI
. (
2011
).
Health care coverage for direct-care workers: 2009 data update
.
New York, NY
:
Author
.
PHI
. (
2013
).
Occupational projections for direct-care workers: 2010–2020
.
New York, NY
:
Author
.
The Scan Foundation
. (
2012
).
Who provides long term care in the United States
.
Long Beach, CA
:
Author
.
Sedlezky
,
L
.,
Reinke
,
J
.,
Larson
,
S
.,
&
Hewitt
,
A
.
(
2013
).
National frontline supervisor competencies: Frontline supervisor assessment
.
Minneapolis, MN
:
University of Minnesota, Research and Training Center on Community Living, Institute on Community Integration
.
Stancliffe
,
R. J
.,
Lakin
,
K. C
.,
Larson
,
S. A
.,
Engler
,
J
.,
Taub
,
S
.,
&
Fortune
,
J
.
(
2011
).
Choice of living arrangements
.
Journal of Intellectual Disability Research
,
55
(
8
),
746
762
.
Test
,
D. W
.,
Flowers
,
C
.,
Hewitt
,
A
.,
&
Solow
,
J
.
(
2003
).
A statewide survey of the direct support professional workforce
.
Mental Retardation
,
41
,
276
285
.
U.S. Bureau of Labor Statistics
. (
2012
).
Occupational employment and wages, May 2012: 39-9021 personal care aides
.
U.S. Bureau of Labor Statistics
. (
2014
).
Economic news release: Table 16. Annual total separations rates by industry and region, not seasonally adjusted
.
U.S. Department of Health and Human Services
. (
2014
).
2014 poverty guidelines
.
U.S. Department of Labor
. (
2014
).
Bureau of labor statistics: Consumer price index
.
Ward
,
K
.,
Smith
,
C
.,
Bales
,
S
.,
&
Sandberg
,
K
.
(
2009
).
Survey of direct care workers in Alaska
.
Anchorage, AK
:
University of Alaska Anchorage Center for Human Development.
.
Wheeler
,
B
.,
Kurtz
,
D
.,
&
Smith
,
T
.
(
2002
).
Evaluation of the impact of WIC section 4681.4 (rate increase) on staff turnover for direct support workers in licensed community care facilities for people with developmental disabilities 1998–2000
.
Los Angeles, CA
:
University of Southern California
.

## Author notes

Matthew Bogenschutz, Virginia Commonwealth University, School of Social Work; Amy Hewitt, University of Minnesota, Research and Training Center on Community Living, Derek Nord, University of Minnesota, Research and Training Center on Community Living; and Renee Hepperlen, University of Minnesota, Research and Training Center on Community Living.

Address correspondence concerning this article to Matthew Bogenschutz, Virginia Commonwealth University, School of Social Work, PO Box 842072, 1000 Floyd Ave., Richmond, VA 23284 (e-mail: mdbogenschut@vcu.edu).