The case for inclusion

Increasingly, organizations are searching for ways to recruit, hire, and retain a diverse workforce. As diversity in the workforce increases, research on the relationship between diversity and performance in organizations has served to illuminate the boundary conditions by which increased diversity leads to positive, rather than negative, workplace outcomes. Research has shown that focusing on inclusion in addition to diversity is more likely to contribute to positive workplace outcomes (Ferdman, Reference Ferdman, Ferdman and Deane2014; Shore et al., Reference Shore, Cleveland and Sanchez2018). In this way, inclusion can be conceptualized as moderating the relationship between demographic diversity and performance.

Brewer’s (Reference Brewer1991) theory of optimal distinctiveness allows a deeper understanding of the importance of inclusion being coupled with diversity in organizations. The basis of the theory is that individuals simultaneously need to feel uniqueness and belongingness in their environments. These needs motivate individuals to prefer a balance between assimilating and differentiating themselves within a social group. Optimal distinctiveness can be used to understand the importance of inclusion in the workplace. In essence, employees want to feel a sense of belonging while also feeling their uniquenesses are valued. Shore et al. (Reference Shore, Randel, Chung, Dean, Holcombe Ehrhart and Singh2011) presented a model of workgroup inclusion aligned with the theory of optimal distinctiveness by emphasizing both valuing uniqueness and facilitating belongingness. In their model, the outcomes of workgroup inclusion included increased job satisfaction, well-being, organizational citizenship behavior, job performance, creativity, and intention to stay, among others. In total, the model theoretically explained the benefits of coupling workgroup inclusion with diversity.

The role of leaders in creating inclusion

Leaders are uniquely positioned to encourage workgroup inclusion because of the power and influence they hold (Veli Korkmaz et al., Reference Veli Korkmaz, van Engen, Knappert and Schalk2022). As such, research has begun to integrate the diversity and leadership literatures to understand the mechanisms through which leaders can promote inclusion (Roberson & Perry, Reference Roberson and Perry2021). To create a multi-level model of inclusive leadership, Veli Korkmaz et al. (Reference Veli Korkmaz, van Engen, Knappert and Schalk2022) expanded upon Shore et al.’s (Reference Shore, Randel, Chung, Dean, Holcombe Ehrhart and Singh2011) model of workgroup inclusion. This expanded model added to Shore et al.’s dimensions of uniqueness and belongingness the dimensions of demonstrating appreciation (i.e., commending the contributions of both individual employees and the team as a unit) and aiding in organizational efforts (i.e., supporting diversity by prioritizing the organization’s inclusion goals and allowing changes to occur to achieve those goals). These dimensions are illustrated in Table 1. Veli Korkmaz et al.’s model provides a basis for understanding the components of inclusive leadership at employee, team, and organizational levels of analysis. Importantly, Nishii and Leroy (Reference Nishii and Leroy2022) extended the conceptualization of inclusive leadership by emphasizing that each employee’s historically marginalized identity(ies) must be explicitly considered by leaders for inclusive leadership behaviors to be enacted.

Table 1. Inclusive Leadership Construct as Described by Veli Korkmaz et al. (Reference Veli Korkmaz, van Engen, Knappert and Schalk2022)

Inclusive leadership is distinct from, yet similar to other leadership constructs. For example, transformational leadership is associated with a climate for inclusion. Further, both inclusive leadership and transformational leadership aim to help employees perform at their full potential (Brimhall & Palinkas, Reference Brimhall and Palinkas2020; Randel et al., Reference Randel, Galvin, Shore, Ehrhart, Chung, Dean and Kedharnath2018). Nonetheless, inclusive leadership is a distinct construct and diverges from transformational leadership due to its focus on valuing each employee so they can feel comfortable bringing their ideas and perspectives to the table; in contrast to transformational leadership, which prioritizes supporting employees so they can fulfill organizational needs (Randel et al., Reference Randel, Galvin, Shore, Ehrhart, Chung, Dean and Kedharnath2018).

In addition to differing from transformational leadership, inclusive leadership is also distinct from other related interpersonal constructs. In considering Arthur et al.’s (Reference Arthur, Day, Mcnelly and Edens2003) established framework of AC dimensions, consideration/awareness of others comes the closest. Even though both consideration/awareness of others and inclusive leadership are characterized by behaviors consistent with making group members feel as though they belong; inclusive leadership (but not consideration/awareness of others) additionally requires behaviors that show followers their value to the group.

Finally, cross-cultural competence (Deardorff, Reference Deardorff and Kim2017) and cultural intelligence (Earley & Ang, Reference Earley and Ang2003) have been suggested as similar constructs to inclusive leadership in that all three involve the consideration of group members’ individual identities and the uniqueness that comes with each identity as affecting an individual’s perspective. However, unlike inclusive leadership, cross-cultural competence and cultural intelligence focus on the agility with which the leader can navigate different cultural contexts; versus (as is the case with inclusive leadership) ensuring the belongingness and psychological safety of followers holding a range of identities and degrees of power in the workplace. Table 2 provides the definitions and discriminant validity evidence for inclusive leadership and each of these similar, yet distinct, constructs.

Table 2. Constructs Related to, Yet Distinct From, Inclusive Leadership

Defining the domain of inclusive leader behaviors

In this section, we review the current definitions and conceptualizations of inclusive leadership and incorporate information from the allyship literature in order to propose a measurable AC dimension. With the heightened focus on inclusive leadership in the workplace, recent research has sought to define inclusive leadership behavior as a construct. For example, Brimhall and Palinkas (Reference Brimhall and Palinkas2020) conducted a qualitative study to understand inclusive leadership behaviors, uncovering themes related to equitable consideration, shared power, collective motivation, universal belonging, and authentic transparency. Exemplary behaviors were provided for each theme (see Table 3). For example, the equitable consideration theme included behaviors such as “recognizes that everyone has different needs and abilities,” “values the unique talents of others,” and “gives the same amount of consideration to everyone’s unique needs” (p. 364). These themes are consistent with Veli Korkmaz et al.’s (Reference Veli Korkmaz, van Engen, Knappert and Schalk2022) model and are helpful in beginning to conceptualize the performance domain of inclusive leadership. However, the behaviors comprising this framework lack the necessary specificity to be uniquely representative of inclusive leadership, and easily identifiable within an AC context.

Table 3. Inclusive Leadership Themes and Codes From Brimhall and Palinkas (Reference Brimhall and Palinkas2020)

Thus, to develop these behaviors further, we turned to the allyship literature. Selvanathan et al., (Reference Selvanathan, Lickel and Dasgupta2020), defined allies as “those who provide support for the disadvantaged group and engage in informed actions to challenge inequality” (p. 1344; see also Brown & Ostrove, Reference Brown and Ostrove2013; Ostrove & Brown, Reference Ostrove and Brown2018). In this conceptualization, the relative power of the advantaged and disadvantaged groups is emphasized, and members of each group have different needs, with advantaged group members pursuing moral acceptance, and disadvantaged group members pursuing empowerment and respect (Selvanathan et al., Reference Selvanathan, Lickel and Dasgupta2020). Collier-Spruel and Ryan (Reference Collier-Spruel and Ryan2022) argue that because the intent and impact of allyship may differ, it is important to consider the impact of allyship behaviors from the perspective of the disadvantaged group member. Equally important is considering allyship not as an identity, but as an action that must be repeatedly enacted (Carlson et al., Reference Carlson, Leek, Casey, Tolman and Allen2020). Therefore, Collier-Spruel and Ryan (Reference Collier-Spruel and Ryan2022) categorized enacted allyship behaviors as either effective or ineffective as seen from the perspective of the disadvantaged group member.

Examples of effective allyship behaviors include amplifying the voice of a disadvantaged group member when they are being ignored and ensuring that the needs of a disadvantaged group member are taken into consideration (Collier-Spruel & Ryan, Reference Collier-Spruel and Ryan2022). Enacted effective allyship behaviors fall under the umbrella of inclusive leadership behaviors, but inclusive leadership behaviors also include a broader range of behaviors. That is, whereas allyship behaviors are more concerned with using one’s power to challenge inequity, inclusive leadership behaviors not only challenge inequity but also ensure that each employee feels like they belong, and their uniqueness is valued even when inequity is not salient.

Understanding that allyship behaviors can exemplify aspects of inclusive leadership provides further insight into the specific behaviors that fall within the inclusive leadership behavioral domain. For instance, informed by the allyship literature (Carlson et al., Reference Carlson, Leek, Casey, Tolman and Allen2020; Collier-Spruel & Ryan, Reference Collier-Spruel and Ryan2022; Selvanathan et al., Reference Selvanathan, Lickel and Dasgupta2020), oftentimes the historically marginalized identity of the individual impacted by the behavior is important in identifying the leader’s behavior as inclusive. Importantly though, inclusive leadership behaviors can also be demonstrated towards those who do not hold historically marginalized identities or towards those whose historically marginalized identities are not visible or salient in a certain context.

Using ACs to identify inclusive leadership behaviors

ACs are a commonly utilized method for assessing employees, especially leaders (Kleinmann & Ingold, Reference Kleinmann and Ingold2019). ACs comprise multiple assessment components, at least one of which is a behavioral simulation exercise. An AC may consist solely of simulation exercises, or combine them with other methods, such as interviews, personality inventories, and/or ability tests. The result is a comprehensive, partially or fully behavioral evaluation of an assessee’s proficiency on a set of job-relevant, behaviorally defined performance dimensions. ACs can be used as a method of selecting leaders, as a diagnostic method to identify dimensions on which a leader can improve, and as a method to help leaders develop specific KSAOs (Thornton et al., Reference Thornton, Rupp and Hoffman2015). Many AC exercises are explicitly interpersonal in nature, which provides an opportunity to rate (potential) leaders on a variety of interpersonal dimensions, such as communication, consideration of others, and influencing others (Thornton & Rupp, Reference Thornton, Rupp, Jackson, Lance and Hoffman2012). Common types of AC simulation exercises of an interpersonal nature include role plays, leaderless group discussions, interviews, and inbox simulations (Povah & Povah, Reference Povah, Povah, Jackson, Lance and Hoffman2012).

Even though ACs can be costly and time consuming to administer compared to other selection methods, they offer advantages such as strong predictive validity, decreased adverse impact as compared to cognitive ability tests, and the ability to measure and combine ratings on a wide range of criteria—especially those that are interpersonal in nature (Eurich et al., Reference Eurich, Krause, Cigularov and Thornton2009; Thornton et al., Reference Thornton, Rupp and Hoffman2015). ACs measure dimensions multiple times through multiple exercises to increase the reliability of the measured dimensions. Meta-analyses have estimated the criterion-related validity of AC overall ratings to range from .27 to .47, demonstrating the method’s strength as a predictor (Arthur et al., Reference Arthur, Day, Mcnelly and Edens2003; Gaugler et al., Reference Gaugler, Rosenthal, Thornton and Bentson1987; Hermelin et al., Reference Hermelin, Lievens and Robertson2007; Monahan et al., Reference Monahan, Hoffman, Lance, Jackson and Foster2013).

Contemporary research has led to the acknowledgment that the AC method offers flexibility, which is another benefit of using ACs. Specifically, research supports the use of the AC method to accurately measure an assessee’s proficiency on a set of behavioral dimensionsFootnote ² (Thornton et al., Reference Thornton, Rupp and Hoffman2015), as well as general performance within a behavioral simulation exercise (Hoffman et al., Reference Hoffman, Melchers, Blair, Kleinmann and Ladd2011; Monahan et al., Reference Monahan, Hoffman, Lance, Jackson and Foster2013; Putka & Hoffman, Reference Putka and Hoffman2013). Specifically, exercise scores could be used to predict performance when the exercise is contextually very similar to and has comparable situational cues to what is experienced in the focal job.

We argue that ACs present a ripe opportunity to assess inclusive leadership for multiple reasons. First, many organizations already utilize ACs for leader assessment and development. Additionally, inclusive leadership is distinct from other commonly assessed AC dimensions. Furthermore, the types of behaviors that define inclusive leadership (at various levels of proficiency) can be induced via interpersonal behavioral simulation exercises. Finally, because inclusive leadership is not typically assessed within ACs, adding an exercise (or adding additional behavioral prompts into existing exercises; see Lievens et al., Reference Lievens, Chasteen, Day and Christiansen2006) to measure inclusive leadership has the potential to provide incremental validity above and beyond current methods. Despite this potential, in reviewing the literature on AC dimensions (Arthur et al., Reference Arthur, Day, Mcnelly and Edens2003; Meriac et al., Reference Monahan, Hoffman, Lance, Jackson and Foster2008, Reference Meriac, Hoffman, Woehr and Fleisher2014), inclusive leadership has yet to be included within AC dimension taxonomies.

Summary and operationalization of leader inclusion as an AC dimension

We illustrate the theoretical propositions posed in this paper within Figure 1. As is shown, we expect inclusive leadership, assessed as an AC dimension, to have utility in predicting a wide range of both leader and follower outcomes. Further, we expect inclusive leadership to complement other, traditionally assessed AC dimensions in predicting not only diversity-related outcomes, but also more general outcomes such as individual and group performance.

Figure 1. Proposed Theoretical Model.

To test our propositions, the question becomes, how might inclusive leadership be best operationalized in an assessment center context? The first step would be to provide a behavioral definition of inclusive leadership, as well as example behaviors at varying levels of proficiency (Thornton et al., Reference Thornton, Rupp and Hoffman2015). Consistent with Veli Korkmaz et al. (Reference Veli Korkmaz, van Engen, Knappert and Schalk2022) we behaviorally define inclusive leadership as using the resources a leader has access to, including status and power, to foster employees’ uniqueness, show appreciation for individual employees and for the team as a whole, increase the belongingness team members experience, and support organizational inclusion efforts. Furthermore, for a leader’s behaviors to be conceptualized as inclusive, they must be indicative of fostering uniqueness, showing appreciation, and increasing belongingness for employees both with and without historically marginalized identities. For example, we might expect leaders highly proficient in inclusive leadership to encourage employees with and without historically marginalized identities to share and expand upon their perspectives that do not align with the group’s consensus. In contrast, we might expect leaders low in inclusive leadership proficiency to discourage an employee with a historically marginalized identity from sharing their perspective that does not align with the group’s consensus. Table 4 provides our operational definition of inclusive leadership as an AC dimension, with behavioral examples of high and low effectiveness for each component.Footnote ⁴

Table 4. Behavioral Examples of Inclusive Leadership Components

Figure 2 takes our operationalization one step further by providing a sample behaviorally anchored rating scale (BARS) for this dimension. Both this and our behavioral examples in Table 4 are general in nature, and don’t contain exercise-specific behaviors that might be expected in the context of an actual AC. As has been advocated for in the literature, our operationalization considers how the features of the situation may or may not elicit a behavioral expression of an individual’s underlying trait (Lievens et al., Reference Lievens, De Koster, Schollaert, Cartwright and Cooper2008). That is, trait activation theory can explain when behavioral manifestations of traits are more likely to occur (Tett & Guterman, Reference Tett and Guterman2000). In the context of ACs, trait activation theory posits that the relevance of a particular behavior to the exercise increases the likelihood that it can be observed. Second, trait-relevant cues within each exercise can increase the likelihood that an underlying trait is manifested in observable behavior within the AC.

Figure 2. Example of a Behaviorally Anchored Rating Scale for the Dimension of Inclusive Leadership.

As an example, consider a hypothetical simulation exercise in which inclusive leadership is one of the dimensions assessed, where specific behavioral cues have been built in to provide behavioral opportunities to display inclusive leadership behaviors. This could be an intra-group simulation with an assigned leader. In the simulation, the group has been tasked with reviewing brief grant proposals and allocating funds. Four distinct dimensions, one of which is inclusive leadership, will be assessed to increase the accuracy of assessor judgments (Gaugler & Thornton, Reference Gaugler and Thornton1989). The group is composed of the assigned leader and five other role players who are trained in their individual roles. The group will be given twenty minutes to review background information, task instructions, and brief grant proposals. The leader will be instructed to use the next thirty minutes to guide the group through a discussion to reach a consensus concerning the allocation of funds. To elicit behaviors of inclusive leadership proficiency, situational cues will be provided through group member actions (Lievens et al., Reference Lievens, Chasteen, Day and Christiansen2006). For example, a confederate group member with a historically marginalized identity cued through either a salient visible identity or shared personal experience, such as race, disability status, pregnancy status, or gender identity, will share their perspective that differs from the group consensus. The leader’s behavioral response to this group member will then be recorded by assessors using the BARS provided in Figure 2.

Importantly, those constructing ACs used to measured inclusive leadership should make power differences among individuals salient to elicit inclusive leadership behaviors. An assessee demonstrating proficiency in inclusive leadership should be able to engender belongingness and uniqueness among employees with historically marginalized identities, who have less power in the organization, as well as individuals with more power in a certain context. To make less powerful identities salient, situational cues must be utilized to evoke relevant behaviors (Lievens et al., Reference Lievens, Chasteen, Day and Christiansen2006; Schollaert & Lievens, Reference Schollaert and Lievens2012). Careful thought must be devoted to ensuring that these cues feel realistic to assessees. For instance, AC participants may interpret a confederate wearing a gay pride t-shirt as contrived and manufactured, thereby potentially decreasing the relevant behaviors elicited and decreasing the face validity of the exercise.

When considering that exercises and situational cues need to feel realistic to assesses to cue relevant behaviors, the issue of assessee and assessor wellness becomes salient. A well-constructed AC exercise that measures the inclusive leadership dimension will produce variance in behavior so that some assessees demonstrate high proficiency in inclusive leadership whereas others demonstrate low proficiency, which could offend others participating in the exercises (including role players and assessors). As industrial-organizational psychologists, it is our responsibility to always support the wellness of AC participants, including when they experience discriminatory behaviors (Lefkowitz & Lowman, Reference Lefkowitz, Lowman, Farr and Tippins2010). Therefore, we propose suggestions that simultaneously address aims to create realistic situational cues of historically marginalized identities and also promote the wellness of those involved.

First, participants in the AC should have the option to participate in a debriefing session following the conclusion of the AC to discuss any low-inclusion behaviors witnessed. Second, AC exercises can consist of one-on-one conversations between two managers about selection, promotion, termination, etc. of an applicant/employee who holds a marginalized identity but is not present for the conversation. In this example, an individual with a marginalized identity does not have to witness any potential low-inclusion behaviors, thereby mitigating wellness concerns. Alternatively, an assessee in an AC exercise could have a selection, promotion, termination conversation with an employee with a historically marginalized identity who is visually represented by animation technology and is not physically present, thereby decreasing well-being concerns on behalf of the confederate. In summary, consideration of AC participants’ well-being should be a priority throughout all exercises and stages of the process.

Future research directions

With an operational definition, BARSs, and exercise design parameters, research into the validity (and incremental validity) of inclusive leadership as an AC dimension is within reach, and collaboration with organizational partners would best facilitate the testing of our theoretical model. Future research should first examine the interrater reliability of assessors’ ratings of inclusive leadership behaviors, the internal consistency validity of the subdimensions (i.e., items) of inclusive leadership, the convergent validity of AC-measured inclusive leadership with other ratings of inclusive leadership (such as other-reported leader inclusion), and the discriminant validity of inclusive leadership as an AC dimension with similar dimensions and constructs (i.e., those listed in Table 1). Then, research should empirically investigate the criterion-related validity of inclusive leadership in predicting follower performance, leader performance, and follower diversity; and to include the incremental validity of leader inclusion above and beyond other common AC dimensions.

AC exercises, such as a single intra-group simulations, commonly measure maximum (rather than typical) performance (Thornton et al., Reference Thornton, Mueller-Hanson and Rupp2017). Because inclusive leadership should be consistently enacted to influence important outcomes, future research should assess inclusive leadership through multiple simulations over time to ensure an accurate estimate of leaders’ motivation to perform typical-level inclusive behaviors.

Future research can also examine whether a low, moderate, or high level of fidelity in a simulation maximizes the predictive validity of AC-measured inclusive leadership. Although high-fidelity simulations are often recommended for training individuals, low to moderate levels of fidelity can be suitable when using a simulation to select employees, because all participants may not have the same familiarity with the organizational and job context (Thornton et al., Reference Thornton, Mueller-Hanson and Rupp2017). Furthermore, research could explore how utilizing different components of the simulation, such as the stimuli, response, content, and difficulty at varying levels of fidelity can affect the simulation’s validity (Fluckinger et al., Reference Fluckinger, Dudley and Seeds2014).

In this paper, we mainly discuss adding the dimension of inclusive leadership to an AC to increase the predictive validity of the OAR. However, ACs can also be used for purely diagnostic or developmental purposes (Thornton et al., Reference Thornton, Rupp and Hoffman2015). Therefore, we recommend future research evaluate adding an inclusive leadership dimension to diagnostic and developmental assessment centers. In these contexts, inclusive leadership ratings would not be combined with other dimension ratings, but rather used together to create a profile of strengths and developmental opportunities, which can be further developed through subsequent training/coaching or feedback and further practice facilitated by additional AC exercises. In this situation, validity of the inclusive leadership dimension would not be demonstrated by cross-sectional correlations to performance indicators, but rather further explorations into the factor structure of the AC dimensions (as having utility for diagnostic purposes) as well as training evaluations to demonstrate learning and development over time.

Finally, we suggest that future research incorporate refining exercises with situational cues that occur in the focal job to best elicit inclusive leadership behavior relevant to a certain context. Ratings of exercise performance (rather than ratings of dimension performance across several exercises) could then be used as a situationally specific measure of inclusive leadership behavior and could be used to predict performance.