The effectiveness of employer practices to recruit,hire,and retain employees with disabilities: Supervisor perspectives

2.1. Findings of related studies

Over the past three decades in the United States, there have been large declines in employment and income among people with disabilities relative to those without disabilities (Houtenville, Brucker & Lauer, 2016) and sharp increases in the receipt of disability benefits (Livermore, Mamun, Schimmel, & Prenovitz, 2013). A great deal of the research on disability and employment has focused on the supply of labor by people with disabilities. Numerous studies investigate the work disincentives inherent to public benefit programs available to people with disabilities, most notably Social Security Disability Insurance (SSDI), Supplemental Security Income (SSI), Medicare, and Medicaid (Bound, 1989; Chen & van der Klaauw, 2008; Gruber & Kubik, 1997; Maestas, Mullen, & Strand, 2013; Von Wachter, Song, & Manchester, 2011). Many studies focus on the effectiveness of vocational supports aimed at boosting and supporting the labor supply (Sevak, O’Neill, Vandergoot, & Grossman, 2018).

The demand for labor by employers has been the focus of numerous studies. Over the last 20 years, several national surveys on employer perspectives and practices (listed in Table 1) were fielded to identify potential barriers to employing people with disabilities. The targeted respondents varied across these surveys (i.e., senior executives, human resource directors, and supervisors). These surveys used various sample designs (i.e., random, representative, and census) and drew from three sampling frames (i.e., Dun and Bradstreet [D&B] Registry members, Society of Human Resource Management [SHRM] members, and Federal agency staff members).

Many of these surveys ask respondents their perspectives about barriers from a predetermined list of potential barriers. These barriers typically fall into four categories: worker qualifications; attitudes; cost concerns; and the availability of helpful information. Depending on the wording of the survey, worker qualification barriers are the most frequently cited or considered the greatest barriers (Erickson et al., 2014; Taylor, Krane, & Orkis, 2010; Domzal, Houtenville, & Sharma, 2008).

Respondents are also asked about the potential usefulness of various practices from a predetermined list of practices. These practices typically fall into five categories: support for workers via accommodations and redefining tasks; support for workers to upgrade their skills, qualifications, and workplace network; cost-related practices; training other staff members; and organizational goal setting and planning. Perhaps not surprisingly, practices pertinent to all employees, such as flexible work schedules, are more frequently used than practices specific to workers with disabilities, such as disability affinity groups (Erickson, et al., 2014; Taylor, Krane, & Orkis, 2010; Domzal, Houtenville, & Sharma, 2008).

With regard to effectiveness, practices pertinent to all employees are more commonly viewed as potentially effective than practices specific to employees with disabilities (Taylor, Krane, & Orkis, 2010; Domzal, Houtenville, & Sharma, 2008). The 2011 Cornell/SHRM Survey asked about the perceived actual effectiveness of practices that were used. Very few employers perceive the practices they use as ineffective. Interestingly, some of the least frequently used practices were perceived to be among the most effective, such as centralized accommodation funds and disability affinity groups (Erickson et al., 2014).

Alternative research designs and sampling approaches have been used to investigate the effectiveness of practices, with a great deal of attention paid to the effectiveness of workplace accommodations. Follow-up surveys of employers that contact the Job Accommodation Network, a comprehensive resource for information on workplace accommodations and a source of technical assistance to employers and others to subsequently implement accommodations, suggest the provision of accommodations is positively related with increased productivity and retention for both the employee receiving the accommodation and for the organization as a whole (Harnett, et al., 2011; Schartz, Hendricks, & Blanck, 2006). Schur et al. (2014) found similar results based on a survey of 5,000 managers and non-managers at eight volunteer companies. Likewise, Solovieva, Dowler, and Walls (2011) found evidence of increased productivity and retention but also reduced training costs, based on a survey of 194 employers that had not contacted JAN. With regard to other practices, the results of case studies suggest that developing partnerships with workforce intermediaries creates networks and “pipelines” that improve the recruiting, hiring, and retention of people with disabilities (Katz, O’Connell, & Nicholas, 2012; Nicholas, Krauder, Krepcio, & Baker, 2011).

It is challenging to study employer behavior and outcomes. Asking employers— senior executives, HR directors, managers, supervisors, and workers— to participate in research studies and report on sensitive subjects, such as employment practices related to disability, creates several demands. First, it is difficult to recruit participants and minimize “survey non-response bias,” which occurs when some participants do not respond to the request to take a survey and these individuals are systematically different from those that take the survey. Surveys that utilize D&B Registry members experience low and declining survey response rates (Matos, 2014; Domzal, Houtenville, & Sharma, 2008); thus, their results are potentially subject to greater survey non-response bias. Low response rates stem from out-of-date phone numbers, incorrect contact persons, and difficulty reaching the persons most knowledgeable about disability-related practices. Concerns about privacy and legal matters may also reduce response rates, particularly as respondents are being asked to represent their organizations and not themselves. This is one reason the 2008 ODEP Survey (Domzal, Houtenville, & Sharma, 2008) did not ask about direct actions and experiences but rather about concerns, knowledge of available resources, and potentially helpful strategies. Anecdotally, intervention and case studies also are associated with difficulty recruiting participants (i.e., research subjects) for similar reasons, in part because getting permission to participate must be run by legal counsel and upper management. It is possible to create and use sample weights to address survey non-response bias based on observable systematic differences between the sample and the population the sample is intended to represent, such as differences of the sample’s size and industry composition and the population’s size and industry composition (published by organizations like the Census Bureau). However, there may be unobservable systematic differences, such as the difference in organizational culture and future culture.

Second, once participants are recruited into a study, there is the potential for “social desirability response bias,” which is the tendency of survey respondents to answer questions in a manner that will be viewed favorably by others. This source of bias may be particularly acute when asking about issues that some may view as ethnical and social justice issues.

Third, employers and employment practices are complex and may be operationalized at multiple occupational levels, making it difficult to determine the appropriate unit of analysis. Some studies focus on a specific occupational level, such as supervisors (e.g., Bruyère, Erickson, & Horne, 2002), while other studies screen for the person most knowledgeable about disability-related practices (e.g., Domzal, Houtenville, & Sharma, 2008). Schur et al. (2014) use an interesting approach, surveying all employees regardless of disability status and occupational level, which allows the ex-post stratification of analyses of survey items asked of all participants.

Fourth, identifying the causal effect of employer practices on intended outcomes is perhaps “the” challenge of employer research. The ability to conduct randomized control trials is typically not feasible in an employment setting for many reasons. Longitudinal data (which track practices and outcomes over time for a given sample) may shed light on causal effects but are typically not available and expensive to collect. As a result, survey research has focused on self-reported experiences and opinions about the effectiveness of certain practices.

Furthermore, when inquiring about the effectiveness of employer practices related to disability, it is difficult to define the frame of reference. Many previous studies have focused exclusively on workers with disabilities, offering no frame of reference to determine whether practices may be differentially effective for workers with and without disabilities. As a result, if an employer reports being ineffective at recruiting people with disabilities, it may be that the employer is ineffective at recruiting in general. As one exception, Schur et al. (2014) compared employer provision of accommodations to employees with and without disabilities and found that employer practices regarding people with disabilities had implications for the general workplace culture for all employees.

Lastly, conducting research with employers can be cost prohibitive. As response rates have dwindled, costs to reach targeted populations have necessarily increased. For example, the 2015 Kessler Foundation National Employment and Disability Survey (KFNEDS) used random-digit dial selection to interview people with disabilities and cost over $390,000 for a sample of 3,000 participants (Sundar et al., 2018). However, the proliferation of web-based panels, such as those maintained by survey solutions companies like Qualtrics, offers an alternative method at a more affordable rate. A replication of the 2015 KFNEDS instrument using a Qualtrics panel to obtain an equivalently sized sample cut the cost by 95% to $19,500.

3.1. Subject recruitment

The 2017 KFNEDS-SP was a survey of 6,530 supervisors from U.S. organizations employing at least 25 workers, the minimum size of organizations responsible for complying with guidelines set forth by the Americans with Disabilities Act (ADA). These supervisors were recruited in the summer of 2017 from the Qualtrics, Inc. Business-to-Business (B2B) Research Panel, a group of pre-screened and willing respondents maintained by Qualtrics, Inc. Panel members were informed of the availability of a survey and subject matter of a survey, and the survey remained available until the target number of respondents completing the survey was achieved. It is not possible to calculate a response rate in the traditional sense. It took 12 days to achieve the first set of 5,410 completed surveys and 24 days to achieve the second set of 6,061. (Of the 11,471 completed surveys, 6,530 were retained for analysis. See Section 3.4, sample characteristics, for more details.)

With regard to survey non-response bias, respondents completed the survey on their own time, not company time, and potentially when at home. They did not have to seek upper management approval to complete the survey and were not recruited as representatives of their company or organization; thus, it is expected that survey non-response bias was reduced relative to attempting to recruit them explicitly at their place of work. This approach is based on the style of the National Survey on Drug Use and Health ([NSDUH]; Center for Behavioral Health and Statistics Quality, 2017), which uses a household sample frame to inquire about workplace substance use and employer policies.

Similarly, surveying supervisors on their own time increase anonymity and is expected to reduce pressures to safeguard company interests around sensitive issues and reduce social desirability response bias. The NSDUH uses a household sample frame. In addition, using supervisors as the respondents rather than human resources (HR) professionals may also reduce social desirability response bias. HR professionals seldom perceive the practices they use as being ineffective (Erickson, von Schrader, Bruyère, & VanLooy, 2014).

3.2. Questionnaire design and content

The KFNEDS-SP questionnaire was developed by researchers at the UNH Institute on Disability in consultation with Kessler Foundation and an advisory board. The overall goal was to collect information about processes and practices used by their organizations to employ people with disabilities, and the effectiveness of those practices. The survey protocol and procedures were approved through exempt-level review by the UNH Institutional Review Board.

When inquiring about effectiveness and frame of reference, the KFNEDS-SP questionnaire asks supervisors whether their organization had a process (e.g., “Do you have a process to recruit qualified employees?”). Supervisors could respond “yes,” “no,” or “I don’t know.” When indicating that a process was in place, supervisors were next asked if the process was effective, in general. If they indicated on the 4-point Likert-type scale that the process was effective (i.e., “somewhat effective” or “very effective” as opposed to “not at all effective” or “not very effective”), they were asked whether it was “as effective” for people with disabilities. In this way, information specific to employees with disabilities may be contextualized within the larger picture of the organizations’ overall effectiveness in relation to its employees. A process or practice was “as effective” if they selected “equally effective” or “more effective,” as opposed to “less effective,” or “I don’t know.”

In addition, supervisors were asked to characterize upper management’s commitment to hiring, training, and accommodating and retaining employees with disabilities. These items used 4-point Likert-type response choices ranging from “not at all committed” to “very committed.” To contrast with their perspective on upper management’s commitment, similar items also asked how important each of these areas was to the supervisors themselves. This not only provided a context through which to contrast the relative importance or commitment to each outcome but also served as a mechanism to reduce pressure on supervisors to respond in socially desirable ways by allowing them to distinguish their own beliefs from those attributed to upper management.

Furthermore, the survey asked about specific practices (e.g., “reviewing hiring practices to determine their accessibility for people with disabilities,” “allowing job sharing,” “flexible work schedules”). Supervisors were provided response options to indicate whether each practice automatically was used for employees in general, was used in the past or upon employee request (or at supervisors’ discretion), or not typically used. As with the process questions, follow up items regarding effectiveness (both in general and specific to people with disabilities) were asked whenever a practice was reported. When a practice was not used by an organization, or if supervisors were not certain whether a practice was used, they reported whether or not they thought the practice would be “feasible” in order to improve employment of people with disabilities (responses were selected from a 4-point Likert-type scale ranging from “not at all feasible” to “very feasible”). In this way, the survey helps to identify promising practices as those that are perceived to be most effective by employers who have used them and most feasible by employers who have not.

Finally, the survey included several follow-up, open-ended questions to delve more deeply into supervisors’ perspectives and experiences working with people with disabilities. For example, several of these items asked supervisors to share what made certain practices more (or less) effective for people with disabilities, when these options were selected, or what made practices unfeasible to implement if indicated as such. Qualitative data from these questions, especially when paired with the quantitative findings, will help to generate specific, actionable information and reveal emerging practices not yet fully recognized in the field of employment and disability.

3.3. Data collection

The 2017 KFNEDS-SP was conducted using standard, replicable survey practices. Respondents of the survey included adults ages 18 and over who worked as supervisors in U.S. businesses and organizations. Starting in the summer of 2017 a sample of supervisors from employers around the country were invited to take the survey as part of Qualtrics and partners’ business-to-business (B2B) panel. Respondents were recruited by Qualtrics and its partner organizations using a variety of methods, including web intercept, targeted email lists, panel member referral, and social media. Incentives for respondents included cash payments, free downloads, and/or membership points; all incentives were decided and allocated by Qualtrics and its partners. Informed consent to participate was obtained in accordance with requirements of the UNH Institutional Review Board. The median time to complete the survey was 15 minutes. Descriptive statistics for all responses were analyzed using Stata 15 (StataCorp, 2017).

3.4. Sample characteristics

More than 11,000 respondents (n = 11,471) ages 18 and over consented to participate in the survey. Of those, 1,805 were dropped because their organization employed fewer than 25 people (and thus was not required to adhere to standards of the ADA). Another 788 respondents were removed for not supervising employees in their current position, and 55 respondents were dropped for indicating that they either did not understand the definition of disability as presented or did not wish to continue the survey. An additional 1,993 respondents were dropped for inattentive or rushed responding. Complete and valid surveys were available from 6,580 respondents. Of those, 50 respondents were removed by the researchers for incomplete responses on key demographic variables. The final analytic sample was 6,530 supervisors, ranging in age from 18 to 78 and representing all 50 states. Table 2 shows sample demographics and socioeconomic information.

Of the 6,530 supervisors, 60 percent of supervisors were female, and 48 percent were between the ages of 35 and 54. Most (72 percent) had a college degree or higher, and 69 percent earned at least $60,000 annually. White non-Hispanic respondents comprised 77.5 percent of the sample; 6.6 percent were Black non-Hispanic, and 10 percent were Hispanic. Respondents were presented with the federal definition of disability as stated in the ADA (U.S. Department of Justice, 2009). According to this definition, many of the supervisors surveyed had some personal experience with or exposure to disability: 19 percent experienced a disability themselves, and 39 percent reported that someone close to them (e.g., family member, friend, colleague) had a disability.

Employers of various sizes were represented by supervisors in the analytic sample of 6,530 supervisors. Organizations with 500 or more employees comprised 44 percent of the sample. Compared to recent data from the Census Bureau’s Statistics of U.S. Businesses (SUSB, 2015), the sample slightly overrepresented midsized organizations: 30 percent of respondents came from organizations with 100 to 499 employees, compared to 15.6 percent nationally. The remaining 26 percent represented smaller organizations with 25 to 99 employees. Details appear in Table 3.

Most (72 percent) respondents worked for private for-profit organizations. Another 15 percent worked for non-profit organizations, and 13 percent were employed in the government sector. As shown in Table 3, supervisors who answered the survey came from a variety of industries. Besides the “other” category (23 percent), the most represented industries were the service industry (14 percent), health (12.9 percent), professional, scientific and technical (12.8 percent), and education (10 percent).

4.1. Recruiting qualified applicants

Table 4 shows the percent of organizations with a process to recruit new employees and supervisors’ estimation of the effectiveness of these processes. Most organizations (83 percent) had a process for recruiting new employees, and in 88 percent of cases, the supervisors felt these processes were effective at attracting qualified applicants. Far fewer (53 percent) felt the processes were as effective for recruiting applicants with disabilities. This may be due to the fact that a minority of organizations expend “moderate” to “a lot” of effort on recruiting people with disabilities. Only 45 percent reported expending this amount of effort recruiting people with disabilities, compared to 68 percent who expended moderate to a lot of effort recruiting for diversity, and 72 percent who exerted this amount of effort on recruiting, in general. These findings are illustrated in Fig. 1.

OPEN IN VIEWER

Fig. 1

Employer Recruitment Efforts (In General, For Diversity and For People with Disabilities) and Hiring Goals (For Diversity, For People with Disabilities, and Disability Included as a Type of Diversity).

Past research indicates that organizations sometimes partner with disability organizations to help them recruit people with disabilities (Hernandez et al., 2008; Schur et al., 2014). Supervisors were asked whether their organizations used this practice, and only 28.5 percent answered affirmatively. These results are also shown in Table 4. Of those who do partner with disability organizations to recruit people with disabilities, almost all (92 percent) felt it was effective. Of supervisors who were not sure whether their organizations do this already or whose companies do not partner with disability organizations to recruit, 75 percent agreed it would be feasible to do so in order to attract more qualified applicants with disabilities.

Supervisors reported that, to some degree, upper management in their organizations was committed to hiring people with disabilities. As shown in Fig. 2, one-fifth (20 percent) of upper level managers were seen as “very committed” while another 45 percent were “somewhat committed.” Supervisors’ own importance attributed to hiring people with disabilities mirrored the commitment they saw from their management: 22 percent of supervisors reported it was “very important,” and 46 percent said it was “somewhat important.”

OPEN IN VIEWER

Fig. 2

Supervisor-Rated Upper Management Commitment (and Supervisor’s Own Rating of Importance) to Hiring People with Disabilities (Percent of Total).

The generally low levels of strong commitment or importance given to hiring people with disabilities were reflected in the organizational hiring goals, as reported by supervisors (see Fig. 1). Only 30 percent had company goals to hire people with disabilities, although a majority (57 percent) reported diversity hiring goals. Of the 57 percent with diversity hiring goals, only 12 percent reported that people with disabilities were included as a diversity population.

More than half (60 percent) of the responding supervisors reported they had ever knowingly hired a person with a disability. Supervisors were next asked about specific employer practices related to hiring people with disabilities, and they provided information about the perceived effectiveness and feasibility of those practices. A summary of findings is provided in Table 4.

Fewer than half of organizations represented were reported to review or audit hiring practices to ensure they were accessible to people with disabilities. Of the 43 percent whose organizations did this, 87 percent felt it was effective for hiring people with disabilities. Of those who did not report that their organizations regularly reviewed the accessibility of their hiring practices, 79 percent felt it would be a feasible practice to implement.

Supervisors were also asked whether their organizations provided training for supervisors in accessible application and interview techniques. Among the 41 percent who used the practice, most (87 percent) felt it was effective for hiring people with disabilities. Most (76 percent) supervisors whose organizations did not offer this training indicated that it would be a feasible strategy to make it easier for their organizations to hire qualified people with disabilities.

Most supervisors (87 percent) reported that their organizations had a process for supporting new hires to learn their jobs, and most (93 percent) who had a process felt it was effective. Three-quarters (76 percent) also indicated that the process for supporting new employees to learn their jobs was as effective for people with disabilities as it was for employees, in general. Figure 3 shows the perceived commitment of upper management and supervisors’ own attribution of importance to supporting employees with disabilities to learn their jobs. Although 46 percent of supervisors saw upper management at their organizations as being “very committed” to supporting employees with disabilities to learn their jobs, this is slightly less than the 53 percent who were very committed to supporting employees in general to learn their jobs. It is also substantially lower than the importance supervisors themselves attributed to employees with disabilities learning their jobs; 80 percent of respondents said it was “very important” to them as supervisors to support employees with disabilities to learn their jobs. Still, this figure is slightly lower than the 82 percent who indicated it was very important to support employees in general to learn their jobs.

OPEN IN VIEWER

Fig. 3

Supervisor-Rated Upper Management Commitment (and Supervisor’s Own Rating of Importance) to Supporting People with Disabilities to Learn their Jobs and Providing Employees with Requested Accommodations (Percent of Total).

Table 5 summarizes organizations’ utilization of three specific training practices, as well as supervisors’ perspectives about their effectiveness (when used) or feasibility (when not used). Supervisors were asked whether their organizations offered certain practices to help new employees to learn their jobs, and they could select one of three responses. They could indicate that the practice was (a) “automatically offered for all new hires,” (b) “offered at supervisor discretion or upon new hire request,” or (c) “not typically offered.”

Ninety-five (95) percent of organizations offered onsite training by a supervisor or coworker as a way to help new employees to learn their jobs; it was automatically offered for all new hires by 75 percent. An overwhelming majority (97 percent) found this to be an effective practice. Somewhat fewer, although still a large proportion (81 percent) of supervisors from these organizations reported onsite training by a supervisor or coworker to be as effective for helping employees with disabilities to learn their jobs. Nearly all (80 percent) whose organizations did not offer onsite training by a supervisor or coworker indicated it would be feasible to implement.

Job shadowing was reported by 89 percent of organizations; 64 percent of supervisors said it was automatically offered to all new hires to help them learn their jobs. Overall, 97 percent of supervisors found job shadowing to be an effective way to help new employees learn their jobs, and 79 percent felt it was as effective for employees with disabilities as it was for all employees, generally. Of those who did not implement job shadowing, 73 percent felt it would be a feasible way to help new employees learn their jobs.

Supervisors next reported whether their organizations offered short-term outside assistance, such as job coaching to help new employees learn their jobs. This practice was the least used of the three practices (47.5 percent of organizations), but where it was used, supervisors usually found it effective for employees in general (88 percent), and as effective for employees with disabilities (85 percent). Short-term outside assistance was automatically offered for all new hires by 22 percent of supervisors and some of the time by 26 percent. Of the majority of supervisors reporting that their organization did not offer short-term outside assistance to help employees learn their jobs, more than half (54 percent) felt it would be feasible to implement.

Questions about the provision of accommodations were disability-specific, and supervisors were not asked to contrast these processes with all employees, generally. About two-thirds (65 percent) of supervisors reported that their organizations had a process for employees with disabilities to request accommodations such as specialized equipment, job reassignment, a modified work schedule, or extra assistance (see Fig. 4). When a process was in place, nearly all supervisors (97 percent) felt it was effective for employees with disabilities. It was less clear how employees were made aware of the process; only 63 percent reported that the process was discussed at a new employee orientation. As shown in Fig. 4, only 18 percent of supervisors said that their organizations had a centralized accommodation fund, a general account specifically to pay for applicant and employee accommodations. When organizations had such a fund, it was almost always (94 percent) viewed as an effective at improving the ability to provide accommodations to employees with disabilities when requested.

OPEN IN VIEWER

Fig. 4

Effectiveness and areas for improvement regarding employers’ process for employees to request needed accommodations and centralized accommodation funds.

Providing requested accommodations was important to supervisors (see Fig. 3). Most (68 percent) indicated that it was “very important” and another 26 percent felt it was “somewhat important.” In comparison, supervisors perceived that upper management were a bit less committed to providing employees with requested accommodations. Supervisors indicated that 48 percent of upper-level managers were “very committed” and 38 percent were “somewhat committed” to fulfilling accommodation requests. Only 6 percent of supervisors felt accommodations were “not very” or “not at all” important, compared to 14 percent of upper management who were seen as not committed to providing the same.

Supervisors were asked to report on three specific practices relating to flexible working arrangements, as these kinds of accommodations are widely available to employees with and without disabilities. As with training practices, accommodation practices were described as being automatically offered, offered at supervisor discretion or upon employee request, or not typically offered. When practices were not typically used, supervisors were asked whether those practices would be a feasible strategy to improve the organizations’ ability to employ people with disabilities. Table 6 shows organizations’ flexible working arrangements and their respective effectiveness for employees, generally, and employees with disabilities, specifically.

A flexible work schedule was reported as an accommodation practice by 78 percent of supervisors, including 28 percent who said it was automatically available to all employees. Another 50 percent indicated that flexible work schedules could be offered at supervisor discretion or upon employee request. Overall, 93 percent found the practice to be effective, and 83 percent said it was as effective for employees with disabilities. Among the 22 percent of supervisors whose organizations did not typically offer flexible work schedules, just over half (51 percent) felt it would be a feasible strategy to improve the organizations ability to employee people with disabilities.

Working from home was the second flexible working arrangement posed to supervisors as an accommodation practice. Table 6 summarizes the results. Working from home was an option frequently offered (54 percent) by organizations; 19 percent of supervisors said their organizations automatically allowed employees to work from home at least some of the time, and another 35 percent said they allowed occasional or frequent working from home upon employee request or at the supervisors’ discretion. In total, 79 percent of supervisors whose organization offered working from home as an accommodation felt this practice was effective and 76 percent believed it was as effective for employees with disabilities. A large proportion of respondents (46 percent) reported that working from home was not typically allowed. Of these, less than one-fourth (22 percent) felt allowing employees to work from home at least some of the time would be a feasible strategy to improve their ability to employ people with disabilities.

Job sharing was the third flexible working arrangement accommodation practice about which supervisors responded. Overall, 46 percent of organizations offered job sharing. Most (90 percent) found it effective, and the vast majority (88 percent) felt it was as effective for employees with disabilities. Among the 55 percent of organizations that reported not offering job sharing, just over half (51.5 percent) of supervisors felt it would be a feasible strategy to improve their organizations’ ability to employ people with disabilities.

Only 17 percent of supervisors reported that their organizations had a process that allowed employees to anonymously disclose disabilities or health conditions. Another 29 percent indicated they were not certain whether there was a disclosure process in place, and the remaining majority (54 percent) said there was no process to anonymously disclose disabilities or health conditions.

As such, supervisors were asked whether their organizations offered training to employees about disability issues and cultural competence in order to improve knowledge and attitudes about colleagues with disabilities. Results are summarized in Table 6. In total, more than two-thirds (70 percent) of organizations offered cultural competence training related to disability. This proportion includes 44 percent who reported that their organizations do this regularly and 26 percent who indicated their organizations had done so in the past. Overall, 92 percent of those supervisors believed it was effective at improving the organization’s ability to employ or accommodate people with disabilities. Only about one-third (30 percent) of supervisors’ organizations did not offer cultural competence training regarding disability issues, but 81 percent of those supervisors said it would be a feasible practice to implement.

Because retention and accommodation practices often are linked when it comes to employees with disabilities, supervisors were also asked whether they consult guidance from federal, state, or local resources regarding the provision of accommodations. This practice was reported by 73 percent of supervisors; 40 percent of supervisors’ organizations do this regularly, and 33 percent had done so in the past. Overall, 91 percent of supervisors whose organizations used or had used this practice reported it was effective. Among the 27 percent of supervisors from organizations that did not use this practice, 78 percent felt it would be a feasible strategy to improve their ability to employ or accommodate people with disabilities.

5.1. Underused but highly effective processes and practices

Most employers have processes in place for recruitment and training, and these processes are typically seen as effective for employees, generally, and for employees with disabilities, specifically. One exception is the recruiting process; just over half of supervisors who believed their organizations effectively recruited new employees felt that they did as well recruiting people with disabilities. Supervisor perspectives about their companies’ recruitment efforts and hiring goals tell part of the story. Fewer than half of the supervisors reported that their companies put forth moderate to a lot of effort to recruit people with disabilities, whereas nearly three-quarters expend moderate to a lot of effort recruiting, in general. Organizations are also much less likely to have goals for hiring people with disabilities than they are to have goals for hiring other types of diversity. Only 12 percent consider disability as a type of diversity. Employers could improve their recruitment practices by partnering with disability organizations to help them recruit qualified applicants with disabilities. Only about a quarter of the supervisors surveyed reported that their employers work with disability organizations to recruit, but of those who did, the overwhelming majority (92 percent) found this practice to be effective.

Findings from the 2017 KFNEDS-SP suggest other opportunities to further investigate and engage employers regarding specific practices they undertake. Some practices are currently used by relatively few employers but are seen as very effective by those who use them and feasible by those who do not. Examples include utilizing short-term outside assistance (e.g., a job coach) to help employees with disabilities learn their jobs, and offering job sharing to make it easier to employ or accommodate people with disabilities.

About two-thirds of employers have a process to provide requested accommodations to employees, and nearly all supervisors believe these processes are effective. This suggests an opportunity to support the 35 percent of supervisors’ organizations that do not yet have such a process in place. Similarly, the vast majority of supervisors were not aware of a process for employees to disclose their disabilities. This is significant because disability disclosure often is a precursor to an inclusive workplace environment and an employer’s ability to provide useful accommodations for employees who need them (Charmaz, 2010; von Schrader, Malzer, & Bruyère, 2014). Moreover, fewer than one in five supervisors were aware of a centralized accommodation fund to help them provide requested accommodations to employees. The fund was found to be an effective tool by nearly all who had it, suggesting that the broader adoption of a centralized accommodation fund would be an effective way for more employers to accommodate and retain workers with disabilities.

5.2. otential influence of upper management commitment

The contrast between supervisors’ perception of upper management’s attitudes and their own attitudes toward hiring, training, and accommodating employees with disabilities revealed areas of opportunity to improve employment outcomes. For example, the importance that supervisors ascribe to hiring people with disabilities mirrors their perceived commitment of upper management to hire people with disabilities. This raises the possibility that upper managers who show increasing commitment to their organizations’ employment of people with disabilities may influence supervisors to likewise ascribe more importance. Moreover, because directionality is not implied in our findings, the reverse might also be true. More vocal support from supervisors may help to sway the attitudes of upper management toward greater commitment to employees with disabilities.

On the other hand, supervisors’ attitudes sometimes differed from those they perceived in their upper managers. Compared to the perceived commitment of upper management, supervisors attach much more importance to helping employees learn their jobs and to providing employees with requested accommodations. This is not surprising; although upper managers may be more focused on overall organizational hiring and/or diversity goals, supervisors carry out the day-to-day operations, responding to employees’ direct needs and overseeing their development and productivity. Findings from the 2017 KFNEDS-SP suggest that both supervisors and workers would benefit from more commitment on the part of upper management to ensuring supervisors can dedicate the time required and provide the tools that employees with disabilities sometimes need to learn their jobs and fully succeed in their positions.

5.3. Limitations

The research presented does have some limitations. First, the sample drawn for this study is national in scope but neither representative nor acquired using random selection. As such, it may contain inherent biases, and caution must be used when attempting to generalize to the population of U.S. supervisors or employers. Second, although Qualtrics uses a double verification process to ensure the identity of survey respondents, it is impossible for researchers to know that participants are telling the truth. As with all self-report studies, the possibility exists that some supervisors shared perspectives that were either intentionally or unwittingly inaccurate. To minimize this risk, both attention and speeding checks were utilized.

5.4. Conclusion

This article presents findings from the 2017 KFNEDS-SP, the first national survey to look at the effectiveness of the practices that employers use to recruit, hire, train, accommodate, and retain people with disabilities in their organizations, from the unique perspective of supervisors of employees with and without disabilities. By focusing on practices that are used and deemed effective, the 2017 KFNEDS-SP highlights feasible approaches to improve employers’ ability to successfully recruit, hire, accommodate, and retain qualified and talented employees with disabilities. In addition to reinforcing the idea that employer practices most effective for employees with disabilities are also highly effective for employees, generally, the results offer important new information for employers, policymakers, legislators, and the disability community. Information about the utilization and effectiveness of workplace practices will support new priorities, policies, and programs to educate and assist employers and service providers as they strive to expand practices with the greatest potential for improving employment outcomes of people with disabilities. Ongoing, inferential analyses of the quantitative findings, as well as qualitative analyses of data from the survey’s open-ended questions, will further contribute to the creation of specific, actionable information to support effective employer practices.