Workforce diversity has the potential to create positive outcomes for employers and employees and thus provide competitive advantages. Employees who perceive equal access to opportunities and fair treatment in their organization are shown to have lower turnover (Chrobot-Mason and Aramovich, 2013). A climate that promotes inclusion of employees by taking into account their views and encouraging input is associated with higher productivity (Sabharwal, 2014). Research likewise shows that inclusion of people with disabilities is associated with positive outcomes such as lower absences and turnover, higher productivity and even higher stock prices (Fisher and Connelly, 2020; Gonzalez, 2021; Hernandez and McDonald, 2010; Kalargyrou, 2014; Lindsay et al., 2018).

Overall, employees with disabilities experience disparities and are largely excluded from the labor force despite the documented benefits of including and empowering them. These disparities are further heightened for individuals with disabilities from marginalized communities, highlighting the importance of using an intersectional lens. Intersectionality is a nuanced view that an individual’s lived experience is greater than the sum of their identities and that forms of inequality can operate together and exacerbate each other (Crenshaw, 2018).

Data from the Bureau of Labor Statistics show that, in 2021, only one in three working age people with disabilities (31.4%) were employed as compared to three-quarters (72.5%) of their nondisabled peers (US Bureau of Labor Statistics, 2022). In 2021, among working age individuals with disabilities, only 29% of women and 34% of men were employed compared with 68% of women and 79% of men without disabilities. Non-Hispanic whites with and without disabilities had the highest employment rates relative to other races and ethnicities (BLS, 2022). Research on the employment outcomes of other marginalized groups of people with disabilities, including lesbian, gay, bisexual, transgender, queer and individuals with other sexual or gender identities (“LGBTQ+”) demonstrates that these individuals also experience higher rates of unemployment relative to their counterparts (Charlton et al., 2018; Ciprikis et al., 2020; James et al., 2016; Leppel, 2021).

As mentioned, employer attitudes and organizational culture can hinder employment of people with disabilities (Blanck, 2020; Schur et al., 2014). Workplace culture that is unwelcoming discourages marginalized workers with disabilities from getting or keeping a job (Araten-Bergman, 2016; von Schrader et al., 2014). Research shows that employers discriminate in hiring against people with disabilities and other marginalized individuals (Ameri et al., 2018; Baert, 2018).

Once hired, access to workplace accommodations remains a barrier to many employees with disabilities (Schur et al., 2014). This barrier is more burdensome for those with less visible disabilities (Blanck et al., 2020; Hyseni et al., 2022), and those with multiple marginalized identities (Blanck et al., 2020; Hill et al., 2016). Marginalized employees are also more likely to experience and report discrimination in the workplace (Blanck et al., 2021; Dispenza et al., 2018; Medina et al., 2021; Shaw et al., 2012).

The COVID-19 pandemic has exacerbated employment and economic well-being disparities between people with disabilities from marginalized communities and their counterparts (Blanck, 2020, 2022). The pandemic hit people with disabilities especially hard because their jobs were more likely to be in occupations most affected, such as construction and food services (Schur et al., 2021). Consequently, the disability employment gap increased early in the pandemic. In particular, White and Black women with disabilities experienced relatively greater employment losses during the pandemic compared to White men without disabilities (Schur et al., 2021).

A substantial portion of the increased gap was explained by differential effects of the pandemic across occupations. The unexplained portion of the employment gap, however, may reflect the growing attitudinal and structural discrimination that people with disabilities experience in employment, especially those embodying multiple-marginalized identities (Schur et al., 2021). As such, the pandemic has highlighted existing inequalities and barriers to employment experienced by people with disabilities from marginalized communities and, in some cases, exacerbated them.

Adopting employer practices that target hiring of people with disabilities from marginalized communities, providing effective accommodations and increasing inclusion in organizations can reduce these employment barriers (Burke et al., 2013; Kaye et al., 2011; von Schrader et al., 2014). Nonetheless, even when employers emphasize diversity, equity and inclusion (“DEI”), they usually focus on one dimension of identity (e.g. gender) instead of considering intersecting experiences of people with multiple marginalized identities (Haymon and Wang, 2020).

Meaningful participation of marginalized groups in the organization is a general indicator of the effectiveness of DEI practices in the workplace. As suggested, organizational citizenship behavior (OCB) is one aspect of employee participation in the workplace. OCB includes discretionary individual actions that extend beyond formal job requirements (Organ, 1988). OCB has been shown to be positively associated with overall employee engagement and job satisfaction (Babcock-Roberson and Strickland, 2010; Kataria et al., 2012; Rurkkhum and Bartlett, 2012), and negatively associated with turnover intent (Lambert, 2010).

Not all employees engage effectively and equally in OCB due to a variety of complex individual and organizational factors. Marginalized employees are less likely to engage in OCB (Mamman et al., 2012). In particular, people of color (“POC”) are less likely to engage in OCB (Aquino and Bommer, 2003; Jones and Schaubroeck, 2004).

The relationship between OCB and gender is complex (Moon, 2016). Social norms and gender roles determine who engages in discretionary work outside of formal job requirements. As such, women are more likely to engage in OCB (Cameron and Nadler, 2013; Thompson et al., 2020). For other groups with less visible identities such as gay and lesbian employees, engagement in OCB is often determined by their outness in the workplace and organizational climate (Brenner et al., 2010).

The importance of climate in predicting OCB has been shown in multiple studies. A conducive climate of inclusion and leadership commitment to DEI have been shown to increase levels of OCB (Babcock-Roberson and Strickland, 2010; Carter et al., 2014; Dang et al., 2022; Moon, 2016; Mousa et al., 2020; Panicker et al., 2018). Likewise, experiencing discrimination is associated with lower levels of OCB (Ensher et al., 2001). Other factors such as type of venue (Grego-Planer, 2019) and tenure (Kegans et al., 2012; Moon, 2016) determine engagement in OCB.

This study focuses on perceptions of the extent to which employees in general engage in OCB in the organization. Considering the overarching goal of the study to explore the extent to which employees engage in behaviors that support their organization’s goals rather than those of particular co-workers and supervisors, we adopt the approach of Lee and Allen (2002) and use items from the pool developed by them to measure OCBO. In addition, OCBO can be conceptualized in terms of the intended target, which in this case is the work group (Lee and Allen, 2002; Ehrhart and Naumann, 2004). This group-level approach to measuring OCBO differs from the conventional approaches to measuring OCB as it “aggregates the individual evaluations of the collective OCB atmosphere in the organization, regardless of the behavior of a specific employee” rather than aggregating individuals’ evaluations of co-workers or supervisors (Vigoda-Gadot et al., 2007, p. 469). As such, the focus is on perceptions of the overall OCB performance within the work group or the organization (Ehrhart and Naumann, 2004).

Perceptions of OCB are closely connected to perceptions of how employees are treated by the organization, as well as to individual attitudes of organizational commitment, job satisfaction and turnover intention (Babcock-Roberson and Strickland, 2010; Brenner et al., 2010; Kataria et al., 2012; Kennedy and Daim, 2010; Pearce and Randel, 2004; Podsakoff et al., 2009; Randhawa and Kaur, 2015; Rurkkhum and Bartlett, 2012; Singh et al., 2013). Past research has found that employees with disabilities tend to have worse perceptions of employee–management relations and management treatment, and lower job satisfaction but similar levels of organizational commitment compared to employees without disabilities (Schur et al., 2017). The lower job satisfaction among employees with disabilities is largely explained by perceptions of worse management treatment and lower job security (Schur et al., 2017).

While we know from prior research that OCB has been shown to be positively associated with overall employee engagement and job satisfaction (Babcock-Roberson and Strickland, 2010; Kataria et al., 2012; Rurkkhum and Bartlett, 2012), there have been no studies relating disability directly perceptions of OCB in the organization. Furthermore, the COVID-19 pandemic has likely had important implications for OCB. Initial studies show that increased work stress has reduced overall OCB (Yu et al., 2021). Nonetheless, we know little about whether workers with disabilities perceive OCB at different rates and if such engagement is shaped by other individual and organizational-level factors.

Given the need to increase the employment rates of people with disabilities and their meaningful engagement in the workplace, this study examines the relationship between individual- and job-level characteristics, and OCB involving employees with disabilities. In what we believe is a first-of-its-kind study, we examine how perceptions of group-level OCB vary with disability status and with other intersectional identities such as race, gender and age. Specifically, we address four research questions.

RQ1. Are employees with disabilities less likely to perceive group-level OCB?

RQ2. How are factors such as size of venue, type of venue (private or otherwise) and tenure associated with perceptions of group-level OCB?

RQ3. Does a supportive company culture have a positive relationship with perceptions of group-level OCB?

RQ4. Does embodying multiple marginalized identities have a negative relationship with perceptions of group-level OCB?

To address these research questions, we employ data from the first phase of a longitudinal survey project of lawyers in the USA. In collaboration with our recruiting partners, we invited American Bar Association (“ABA”) and non-ABA members, a total of nearly 200,000 people, to participate in an online survey about their experiences in the legal profession. Because the focus of this study is on the experiences of marginalized attorneys, we targeted our recruitment efforts through entities associated with these communities. Specifically, we worked closely with ABA leadership, the LGBTQ + community and a representative work group in developing the survey. In addition, our survey recruitment efforts were supported by the ABA Commission on Disability Rights and the ABA Commission on Sexual Orientation and Gender Identity to increase participation rates of their members. While these efforts may have resulted in some overrepresentation of these marginalized groups relative to their prevalence in the overall legal profession, this has the advantage that it allows us to more precisely identify the effects of marginalized group membership.

The survey used both quantitative and qualitative questions, with fixed-choice and openended response opportunities, and is discussed in detail elsewhere (Blanck et al., 2020). In accordance with our strategy, we sent email requests that included the survey link to members of national and state legal organizations. We were unable to track how many people saw and opened the email. Nonetheless, our efforts led 5,543 people to open the link included in the email to learn more about the survey. Of these people, 4,532 opened the survey. In the end, 3,590 people completed and submitted the survey. While the full sample of lawyers responded to the survey, not all of them necessarily completed all the survey questions. The overall analytical sample used to answer the research questions discussed above is 1,865. The analytical sample is derived by including those participants who answered all the questions included in our models, such as questions about demographics, OCB and other job-related characteristics.

Table 1 provides an overview of the respondents’ characteristics, indicating the proportion or mean for each variable included in our models as well as the range for the analytical sample.

Disability is coded as 1: “has a disability” and 0: “no disability.” This variable was created as a combination of seven questions. First, we used the six disability measures from the American Community Survey (“ACS”). In addition, respondents were asked “Do you have a disability or health condition not reflected in the previous question?” Those who answered yes to one or more of these seven questions are coded as 1 or “Yes,” while those who responded no to all the questions are coded as 0 or “No.” Overall, around one-quarter (23%) of our respondents reported at least one disability.

Gender is coded as a binary variable: 1 for “Women,” 0 for “Men.” Women make up 56% of respondents while men make up 44%.

Race and ethnicity are coded as one binary variable reflecting whether the respondent is a person of color (“POC”). This variable is coded as 1 when the respondent identifies as Black, Hispanic/Latino, Asian, American Indian/Alaska Native/Native Hawaiian/PacificIslander, or Multiracial, and 0 for “white, non-Hispanic.” About one in seven (16%) of respondents identify as POC.

Age is coded as a categorical variable with five categories for different age groups. Attorneys below the age of 35 years make up the largest proportion of our analytical sample (24%) followed by those 46–55 years old (23%).

Sexual orientation and gender identity is a binary variable, coded as 1 when the respondent identified as lesbian, gay, bisexual, transgender, queer, or other sexual orientations and gender identities (“LGBTQ +”), and as 0 for heterosexual and cisgender attorneys (“non-LGBTQ+).” Around 16% of respondents identified as LGBTQ +.

Job tenure reflects the number of years the respondent has worked at the current organization at the time of the survey. Responses range from less than 1 year (coded as 0) to 70 years, with the average tenure slightly longer than 10 years.

Organization size is coded as a binary variable, with two categories to indicate small size venues (less than 24 employees) and larger venues (25 or more employees). The cut-off point between small and larger firms was established based on industry standards, discussions with experts, and existing and widely accepted definitions of small businesses. The cut-off point was selected from the 19–29 range used in small-business research (Hurst and Pugsley, 2011) and research on law firms (Thomson Reuters Institute, 2022). Attorneys from small venues make up 40% of our analytical sample and those in larger venues 60%.

Organization type is a binary variable coded 1 for “private” and 0 “other” (e.g. in-house legal department, public sector, non-profit, education). Around 58% of organizations were private law firm practices.

Co-worker support is a continuous variable showing the degree of co-worker support perceived by respondents. It ranges from 1 to 10 and has a mean of 8. 1 indicates unsupportive co-workers and 10 indicates supportive co-workers.

Presence of DEI policy is coded as a binary variable. It is coded 1 to indicate presence of effective DEI policies and practices and 0 to indicate when such policies are lacking. More than one-third (38%) of respondents reported to have experienced such effective policies.

To present associational statistics for our sample, we calculate correlation coefficients and their statistical significance levels for the variables used in our models to address our core research questions. To answer the primary research questions, we use multivariate Ordinary Least Squares (“OLS”) regression analysis to determine the relationship between disability status, intersectional individual and job characteristics, and OCB.

Our first model tests the association between OCB perceptions and individual demographics. To assess the contribution of other covariates to the variation of OCB, we progressively add to this model job tenure, organization type and organization size in model 2, co-workers support and effective inclusion strategies in model 3 and interaction terms between demographic variables to conduct intersectional analyses in model 4. We also run these models separately for attorneys with disabilities and those without disabilities to assess the unique relationship between perceptions of OCB and culture for these two subgroups. Finally, to test the robustness of the model, we analyze each of the three components of the OCB using ordered logit. Since not all models meet the proportional odds assumption, we replicated the results using generalized ordered logit models.

Table 2 shows the pairwise correlation coefficients and their statistical significance levels for the variables used in our models to address the core research questions. The results indicate that disability, gender, race/ethnicity, age, tenure, size of venue, and presence of supportive co-workers and effective DEI strategies are significantly associated with OCB perceptions. Being a person with a disability, a woman, a person of color and younger is negatively associated with OCB perceptions. Being older, having longer tenure, working in a small organization, having supportive co-workers and having experienced effective DEI strategies is positively associated with OCB perceptions.

Table 3 shows the results from the OLS multivariate regressions. Model 1 predicts the OCB score from individual demographics. Results show that the OCB score is expected to decrease by −0.63 points for people with disabilities compared to those without disabilities and −0.43 points for POC compared to non-Hispanic white attorneys, controlling for other demographics. Being older on the other hand is expected to increase OCB scores.

In Model 2, we add specific job and organizational factors to predict the OCB score. Coefficients for disability, race and ethnicity, and age remain unchanged. Working for a small organization is shown to be associated with higher OCB scores compared to a larger organization, controlling for other covariates. Factors such as tenure and type of venue are not shown to play a role in determining OCB scores.

In model 3, we include measures of organizational culture such as supportive co-workers and effective DEI strategies. Model 3 beta coefficients for individual demographics were compared against Model 2 beta coefficients. Race and age coefficients were reduced in size and lost some significance once we include proxies for organizational culture, in Table 3. This suggests that company culture plays a mediating role in the relationship between OCB perceptions and age and race. For example, around 70% of the association between OCB and race/ethnicity is due to the fact that POC are less likely to experience a supportive organizational culture. Overall, the OCB score is predicted to increase by 0.76 for every increase in the level of co-worker supportiveness and by 1.03 when effective DEI policies are present.

As the results in Table 3 show, the disability effect decreases substantially in magnitude (−0.7216 to −0.4000), but people with disabilities remain significantly lower in their OCB perceptions compared to those without disabilities. The substantial decrease in the effect when controlling for co-worker support and DEI policies indicates that the lower OCB perceptions of workers with disabilities are at least partially mediated by their lower perceptions of co-worker support and the presence of DEI policies compared to people without disabilities. This suggests that other measures of organizational culture could also play a mediating role. The lower perceptions of co-worker support among people with disabilities are consistent with prior evidence that disability affects co-worker judgments (Beatty et al., 2019).

In Table 4, we rerun model 3 for two subgroups—attorneys with and without disabilities—to further explore the association between organizational culture and perceptions of OCB. Our results show that the presence of effective DEI policies is important in improving perceptions of OCB for workers with and without disabilities. Nonetheless, there are small differences in the magnitude of the coefficients between these two groups suggesting that presence of DEI policies is more important for attorneys with disabilities in increasing their OCB compared to those without disabilities.

In our final model in Table 3, we include interaction terms between disability and other demographics such as gender, race, and LGBTQ + status. The aim is to explore the compounding role of multiple marginalized identities on OCB. Our results show that all three interactions are not statistically significant at p < 0.1, indicating that disability and other demographic variables likely have a simple main effect on OCB. However, when dealing with a purely linear model, the estimated coefficient is equal to the marginal effects. Such is not the case for more complicated models such as those with interactions terms. Therefore, calculated adjusted predictions at the means and average marginal effects shown in Figures 1 and 2 are presented to ease interpretation of Model 4 that includes three interaction terms.

Results show that for a person with average characteristics, having a disability decreases the OCB score by 0.42 points (p < 0.1). On the other hand, for an average person, being in the 36–45-year age group as opposed to the under-35 group increases the OCB score by 0.04 points; being 46–55 years old increases the score by 0.43 points, being 56–65 years old increased the score by 0.40 points; and being over 65 years old increases the score by 1.03 points (p < 0.1). Overall, keeping other characteristics at their average, individuals 66 years and older report the highest average OCB score (15.84). Marginal effects for other demographic characteristics are not statistically significant.

We check the robustness of our results by running three separate ordered logit regressions for each item in our dependent variable. Table 5 shows that some results change depending on which dimension of OCB is considered. For example, results from Model 4 show that the log odds of keeping up with developments in the organization (item 1) and protecting the organization (item 2) decrease with each year of tenure. Working for a private venue is associated with higher log odds of engaging in activities to protect the organization (item 2). Disability seems to interact with sexual orientation and gender identity in determining one’s likelihood to engage in activities that improve the functioning of the organization (item 3). Similarly, the importance of DEI policies is particularly heightened for this last item. We replicated these models using generalized ordered logit regression. Our findings did not deviate from the results of the OLS and ordered logit models.

This study represents a first step in exploring in-depth labor force outcomes of workers with disabilities by focusing on meaningful workplace participation, DEI and OCB. As such, due to this nature, it has certain limitations. Using an online survey involves a chance of introducing selection biases in our findings as it limits our reach only to those who have access to the technology to complete the survey. In addition, our sample of law professionals limits the generalizability of the findings. Our survey is focused on attorneys, an educated white-collar group of employees, and as such our findings should be interpreted with caution when considering other groups of employees. Nonetheless, our findings show that people with disabilities in the legal profession are less likely to perceive OCB compared to their nondisabled counterparts. Taking into account the fact that the legal profession as an industry is generally more aware of discriminatory practices in the workplace, these results might be magnified in other industries.

Furthermore, to ensure representation of marginalized workers in our sample, we have used strategies to oversample certain groups such that the sample is not representative of the overall legal profession in the United States. As noted earlier, this has the advantage that we can more precisely estimate the effects of marginalized group membership. In addition, we see these findings as exploratory and laying the groundwork for future studies to investigate more systematically and with fully representative samples opportunities for meaningful workplace participation for employees who are usually placed on the margins of organizations.

The next phase of the survey will explore post-COVID-19 workplace experiences of legal professionals. In addition, it will provide a longitudinal component to analyzing workplace participation and inclusion. Future work will thus explore other dimensions of workplace inclusion such as trust and belonging, decision-making and fair treatment.