Firm-level diversity management competencies: development and initial validation of a measure Jennifer Gytha Carstensa* and François Servaas De Kockb a
Department of Industrial Psychology, University of Stellenbosch, Stellenbosch, South Africa; b
Section for Organisational Psychology, School of Management Studies, University of Cape Town, Cape Town, South Africa
This work was supported financially by the Mandela Rhodes Foundation, as well as the Department of Industrial Psychology, University of Stellenbosch.
*Corresponding author. Email:
[email protected]
The International Journal of Human Resource Management
Firm-level diversity management competencies: development and initial validation of a measure The present study addresses the need for a conceptual model and empirical measure of diversity management at the firm level. Whereas earlier diversity management studies focused mainly on manager-employee interactions, we develop a firm-level framework and measuring instrument to diagnose firms’ diversity management competencies. We interviewed managers and content analysed interview transcripts to identify eleven broad firm-level diversity management competencies. To assess the reliability, construct- and criterion-related validity of our measure, we surveyed 157 respondents from 61 different companies. Results of this exploratory study showed that overall firm-level diversity management competencies predicted firm-level perceptions of both proximal and distal diversity management outcomes. In sum, the study develops a useful framework and measurement instrument to support diversity management efforts.
Keywords: diversity management; diversity management competencies; inclusive climate; business case
Introduction In the last few decades, few countries have been left untouched by the rapid diversification of the modern workplace. As a result of increasing workforce diversity, a fundamental concern is how to balance increasing levels of diversity with the need to maintain and increase firm competitiveness (Schuler and Walker 1990). The uptake of the ‘business case’ approach to diversity, which refers to whether a diverse workforce results in economic benefits for companies (Litvin 2006), has been slow (Jackson 1992) and stunted by inconsistent empirical evidence to support its utility (Barnett, Chadwick, Dwyer and Richard 2004; Bezrukova et al. 2
The International Journal of Human Resource Management 2002; Blake and Cox 1991; Ford, Ismail and Richard 2006; Joshi, Liao and Roh 2011; Martins and Milliken 1996). The business case for diversity is often easier justified through rhetoric, rather than with empirical evidence. We believe that one of the reasons for the dearth of research that investigates the business case for diversity has been a myopic focus on diversity-related inputs (e.g., diversity interventions) and outcomes (e.g., profitability), rather than diversity management as a process (Chanda, D’Netto, Monga and Shen 2009). To test the business case for diversity, earlier studies (e.g. Ammeter et al. 2003; Bezrukova et al. 2002) have generally adopted a state perspective to determine the value of diversity to the organisation. In a state view of diversity management, the relationship between levels of demographic diversity and corresponding outcome measures is typically determined. This ‘quota’-approach has various shortcomings, however. First, it lacks a strong theoretical argument why or how increased diversity may lead to increased firm performance. Although this link has been easier to justify at the team level (e.g. Martins and Milliken 1996), the inferential leap at the firm level is more tenuous. For example, it has been suggested that increased diversity would lead to greater access to diverse markets (Blake and Cox 1991; Cox 2001), but this assumption has only been tested anecdotally. Second, a diversity-state perspective on diversity management is flawed because firms are constrained in their ability to control their absolute levels of demographic diversity. For example, their demographic composition may merely reflect external socio-economic factors, such as shifts in the applicant population (Seymen 2006). In such cases, it makes little sense to focus diversity management strategies on employment equity targets or ‘headcounts’, but rather, company efforts should be directed at managing diversity better within the firm, through the optimal use of structures, policies, procedures, and interventions. That is, firms generally have
3
The International Journal of Human Resource Management more control over the way they deal with diversity on a day-to-day basis, than how they look from an employment equity (EE)/staff representation perspective. For these reasons, we believe that it is time to recast the diversity debate away from a state perspective, to a process-perspective. In this view, the focus falls on creating and sustaining the conditions for effective diversity management or, stated otherwise, setting the scene within which a diverse workforce can thrive by means of deliberate, focused diversity-related practices. In contrast to a state perspective on diversity, a process perspective emphasises ‘good practices’ directed at managing diversity in pursuit of optimal performance outcomes. Stated otherwise, a process view acknowledges that success lies more in ‘what you do’ than ‘what you’ve got’. Despite their appeal, process-approaches to diversity management remain largely untested in the published literature. Conceptual models that indicate how workforce diversity could be managed operationally (e.g. Chanda et al. 2009; Cox 1994; Gilbert, Ivancevich and Stead 1999; Kochan, McMillan-Capehart and Richard 2002; Martins and Milliken 1996; Seymen 2006) to enhance firm-level outcomes are rarely empirically tested. We believe that one of the reasons for limited empirical support for process-models of diversity management is because of a lack of widely accepted measures of workforce diversity management. Existing measures have generally focused on the individual or dyadic level (i.e. where managerial behaviours or manager-employee interactions are assessed), rather than the sphere of management practices. In sum, a need exists for a firm-level measure of diversity management that addresses ‘good practices’, structures, policies, procedures, and interventions that may support effective diversity management. Developing such a measure could hold several potential benefits. First, it could allow the diagnosis of firms’ diversity management strengths and weaknesses. Such diagnostic information
4
The International Journal of Human Resource Management may help users to target specific areas that may benefit more from investments of time, money, and energy. Second, the ability to measure diversity management practices may also help further diversity research at the organization level of analysis, because diversity-related inputs can be operationalized better as independent variables in ‘business-case’ research studies. The research objectives of our investigation were twofold. First, to develop a conceptual framework and measurement tool for managing diversity by drawing on literature and integrating feedback from diversity management experts and other practitioners. Second, we wanted to determine if the resulting diversity management competencies are related to important proximal and distal firm outcomes. To these ends, we conducted our study in two stages. In the first stage, we used qualitative semi-structured interviews and derived a firm-level diversity management competency model and corresponding measurement instrument, the Diversity Management Competencies Questionnaire (DMCQ) 1. We also conducted content validation of the measure and report on these results. In the second stage of our research study, we administered the measure to a field sample of respondents working in different firms. The resulting data were analysed at respondent level in order to determine the measurement properties of the DMCQ. More specifically, we studied its internal consistency, internal factor structure, and discriminant validity. Finally, we aggregated responses for participants within each firm in order to ‘step up’ the unit of analysis to the firm level. This allowed us to conduct a small-scale criterion-related validity analysis of the measure against two firm-level criteria, namely proximal diversity-related outcomes and distal firm performance variables. In sum, practitioners and researchers may potentially benefit from the firm-level diversity management competency model and measure developed here. 1
The DMCQ measure is available to researchers as supplemental material to this article; it is also available for download on our research site (https://supplemental material to article).
5
The International Journal of Human Resource Management Development of the Diversity Management Competency Framework Prior Conceptual Models Diversity management is a planned and systematic managerial process aimed at creating an organisational environment where all employees contribute to organisational effectiveness (Thomas, 1996). Diversity, in turn, refers to any difference in which people categorize themselves or others and which has a significant effect on group interaction and outcomes (DiTomaso, Parks-Yancy and Post 2007). A number of conceptual models on diversity management suggest how diversity management contributes to organisational effectiveness. For example, the Interactional Model of Cultural Diversity (IMCD) (Cox 1994) proposes that the diversity climate of an organisation is determined by multi-level influences ranging from individual factors such as stereotypes to organisational factors like structural integration of diverse people. In turn, the diversity climate of an organisation influences individual employee outcomes, which collectively affects organisational effectiveness. This interactional model implies that a multi-level approach to managing diversity is required for it to have a positive effect on organisational outcomes. Another conceptual model, proposed by Gilbert et al. (1999), considers diversity management as a complete cultural change, rather than an isolated component of policy design. According to this view, to drive the culture change process, company CEOs highlight diversity management as a strategic imperative and transform the human resource function, because they are convinced that diversity management makes sense from both a justice as well as a business perspective. The resulting benefits of diversity managed well will, in turn, influence individuals’ attitudes and behaviours, thereby cascading down to affect important organisational outcomes (Gilbert et al. 1999).
6
The International Journal of Human Resource Management More recently, Chanda and colleagues (2009) proposed a framework that focuses on the contribution of Human Resource Management (HRM) to diversity management. Within this model, a firm’s diversity management philosophy is key. That is, the diversity management philosophy rests upon a commitment to equal employment opportunity (EEO), affirmative action (AA), and appreciating and making use of diversity at the strategic level. So, the broader philosophy determines how and to what extent diversity management is practiced through HR practices at the strategic, tactical and operational level. This model also emphasizes the role of line managers: they should drive the diversity management initiative at all organisational levels (Chanda et al. 2009). While these conceptual models suggest how effective diversity management may influence firm performance, they do not outline how effective diversity management would manifest at the level of organisational practices. As such, a framework is needed which comprises essential elements of managing diversity that may guide firms’ diversity initiatives. Although ‘best practices’-lists for diversity management exist (e.g. Beale and Cox 1997; Human 2005; Reichenberg 2001), they often lack coherent conceptual structure and logical organisation—existing approaches do not seem to provide a useful theory describing the domains of effective diversity management practices, nor how these relate to diversity-outcomes and other firm-level criteria. To address the limitations of existing diversity management frameworks, we develop a firm-level model for effective diversity management using an approach which is based competency modelling. Competency models have become important to describe individual employee behaviours in industrial/organisational psychology, as they effectively distinguish top performers from average performers (Campion et al. 2011). A competency can be defined as a
7
The International Journal of Human Resource Management set of behaviours that are instrumental in the attainment of preferred results (Bartram 2011). In response to individual-level competency modelling, organisational-level competency modelling has come to the fore in organisational practice. For example, the ‘core competencies’ (Prahalad and Hamel 1990) of a firm may indicate what technologies, skills, and/or processes provide companies with a distinctive competitive advantage. In our study, we apply this approach to develop the Diversity Management Competency Framework—a framework that describes effective diversity practices clustered around eleven core diversity management competencies (see Table 1). On the basis of prior literature and subject matter expert (SME) interviews that were content analysed, we distilled the conceptual framework and constitutive definitions for each competency. Next, we developed a measuring instrument, the Diversity Management Competencies Questionnaire (DMCQ), to tap into the respective competencies, and assessed its content validity, construct, and criterion-related validity in a sample of respondents drawn from multiple firms.
Diversity Management Competency and Firm Outcomes Diversity management competency and firm performance The notion that diversity-related investments result in positive firm performance outcomes is termed ‘the business case for diversity’ (Barnett, Chadwick, Dwyer and Richard 2004; Bezrukova et al. 2002; Blake and Cox 1991; Ford, Ismail and Richard 2006; Litvin 2006; Martins and Milliken 1996). The business case for diversity proposes that effective diversity management may improve firm performance through (1) better problem-solving, (2) increasing creativity and innovation, (3) increasing organisational flexibility, (4) improving the quality of personnel through better recruitment and retention, and (5) improving marketing strategies (Blake and Cox, 1991). 8
The International Journal of Human Resource Management The theoretical origins of the business case perspective lies in the resource-based view (RBV) (Dunford, Snell and Wright 2006) of the firm. This view emphasises a firm’s internal resources as sources of competitive advantage (Dunford et al. 2006). Competitive advantage is defined as “a company’s ability to maintain and gain market share in its industry” (Gerhart, Hollenbeck, Noe and Wright 2008, p. 4). Firms draw their competitive prowess from internal resources which may include tangible and intangible assets such as management skills, the knowledge and information that an organisation controls, as well as organisational processes and ‘routines’ (Barney, Ketchen and Wright 2001). Within a competitive environment, some business routines may be more efficient and effective than others (Barney 2001). Whereas ineffective routines are detrimental to a company's survival in the long run, effective routines create competitive advantages for a company (Barney 2001). Diversity-related ‘routines’, or the way a diverse workforce is managed, may also enable firms to create competitive advantage (see Pfeffer 1998). Diversity-related routines would manifest as practices that collectively constitute diversity management competency. In our framework, diversity management competency can be defined as the firm’s ability to effectively respond to opportunities and challenges that derive from a diverse workforce. We consider diversity management competency as the sum total of a firms’ performance on the full set of eleven diversity management competencies. Drawing on the resource-based view, firms that can more effectively respond to diversity challenges and opportunities will create competitive advantage and, therefore, are also likely to perform better. The first hypothesis (see Figure 1) of the current study is therefore as follows: Hypothesis 1: Firm-level diversity management competency is positively related to firm performance.
9
The International Journal of Human Resource Management Diversity management competency and intermediary outcomes Diversity management may also affect firm performance indirectly, rather than directly (Boon, Boselie and Dietz 2005). What is the process through which diversity management may influence firm performance or, stated otherwise, what is in the ‘black box’ of diversity (Lawrence 1997)? In one study (Kirby and Richard 2000), the major employee affective outcomes that result from diversity management were investigated in different companies that have implemented diversity management. Employees (N = 269) were asked to list the major diversity-related issues they experienced. Results showed that three general issues were salient, including perceptions of (a) inclusive climate, (b) perceived advancement opportunities and (c) perceived power distribution (Kirby and Richard 2000). First, inclusion referred to respect and acceptance of differences, integration within the workplace, harassment and unsuccessful recruitment of certain identity groups. Inclusion competencies that are imbedded into an integrated management system create an inclusive climate (Maak and Pless 2004). The actions, processes and structures that organisations put in place (which may constitute elements of diversity management) may play a fundamental role in developing a climate of inclusion (Herdman and McMillan-Capehart 2010; Nishii, 2013). Second, perceived advancement opportunity involved recognition of contribution, the representativeness of a qualified applicant pool, unsuccessful training of unrepresented groups and reverse discrimination to previously advantaged groups. An organisation's compensation, promotion, and hiring systems can, for example, engender perceptions amongst employees about whether there are equal opportunities for advancement for all employees (Bartle, Hayes and Major 2002), as "organization-wide policies, practices, and procedures send implicit and explicit messages to employees about the organization’s stance toward the allocation of opportunities" (p. 452). 10
The International Journal of Human Resource Management Last, perceived power distribution referred to the demographic distribution of identity groups in upper management levels, as well as the resulting difference in influential power of various identity groups. When employees believe that all diversity groups have equal influence on making decisions and exercising authority, a balanced power structure exists (see Ely and Thomas 2001; Gonzalez 2010; Konrad and Linnehan 1999). When firms successfully employ diversity management competencies, they are also likely to ensure (ethnically) fair distribution of power in the organisation. Taken together, employees’ perceptions of inclusion, opportunities for advancement, and ethnic power distribution may be important consequences of diversity management competency (see Chanda et al. 2009; Cox 1994; Ely and Thomas 2001; Gilbert et al. 1999; Kirby and Richard 2000; Kochan et al. 2002; Martins and Milliken 1996). Therefore, we posit: Hypotheses 2-4: Diversity management competency is positively related to inclusive climate (H2), perceived advancement opportunities (H3), and perceived power distribution (H4).
Intermediary outcomes and firm performance These intermediary (diversity-related) outcomes may, in turn, affect firm performance. For example, in inclusive climates, employees feel that they fit into the organisation and, therefore, are more likely to perform better than employees that do not perceive fit (Roberson 2006). The notion of supplemental person-organisation fit (Sacco and Schmitt 2005) suggests that similarity between an employee and the immediate work environment will result in positive work outcomes. This argument is also supported by an empirical study showing that perceived level of inclusion strongly predicts job performance (Cho and Mor Barak 2008). A recent study also showed the benefits of climate for inclusion for gender-diverse groups: within inclusive climates,
11
The International Journal of Human Resource Management interpersonal bias is reduced, in turn reducing the effect of group conflict on unit-level satisfaction and turnover (Nishii, 2013). In addition to inclusion, perceived advancement opportunities may also lead to firm performance. The expectancy theory (Vroom 1964) of motivation suggests that employee perceptions of opportunities for advancement – essentially beliefs about instrumentality – may affect individual performance. That is, if employees believe there is a future for them in the organisation (or possibilities for advancement) employees tend to work harder (Hammer and Landau 1986). Collectively, firm performance should be positively affected when beliefs about equal advancement opportunities encourage increased individual work efforts. Finally, ethnic power distribution in a firm may also impact firm performance. For example, Anderson, Gruenfeld and Keltner (2003) suggested that the distribution of power has certain effects upon individuals’ affect, cognition and behaviour. They argue that sources of power, such as the ability to control rewards and opportunities, trigger approach-related processes such as positive affect, perceptual attunement to rewards, and goal-directed behaviour. To our knowledge, the notion that perceived power distribution may be related to firm performance has not been tested empirically. Therefore, in sum, we propose: Hypotheses 5-7: Inclusive climate (H5), perceived advancement opportunities (H6), and perceived power distribution (H7) are positively related to firm performance.
(Insert Figure 1 about here) Method: Development of Diversity Management Competency Framework and Questionnaire Expert participants for interviews In the first stage of our study, the first author interviewed subject matter experts in the field of diversity management to develop the Diversity Management Competency Framework. To this 12
The International Journal of Human Resource Management end, we contacted diversity management consultants and requested them to provide the contact details of practitioners involved in diversity management roles in medium-to-large South African firms. We used additional snowball sampling to identify experts involved in diversity management. To be included in the expert panel, we required that participants had at least two years of experience in diversity-related consulting or direct responsibility for diversity management. As a result, the sample of 12 subjective matter experts included both senior line managers (83.3%) and diversity management consultants (16.7%). These were predominantly male (75%). In terms of ethnicity, they were mostly White (75%), whereas some were Black (16.7%) or Asian (8.3%).
Interviews with subject-matter experts The interviews were semi-structured and consisted of a standard set of open-ended questions. Semi-structured interviews have the objective of attaining the perspectives of the interviewees, but also to confirm the information the researcher already gathered on the topic (Blumberg, Cooper and Schindler 2008). The interview questions were formulated using the Critical Incident Technique, which is a useful approach to gather information about effective and ineffective individual behaviours (Barrick, Field and Gatewood 2008). In our study, however, we gathered critical incidents of firm-level diversity management practices, both effective and ineffective. To this end, participants were asked to recall effective and ineffective diversity management practices they have identified in any organisation. Example questions were “Please describe organisational policies, practices, structures, processes, or actions (formal and informal) which were very helpful in managing diverse employees” and “Please describe policies, practices, structures, processes, or actions (formal and informal) that had damaging effects with regards to managing diverse employees”. 13
The International Journal of Human Resource Management The interview protocol included an introduction of the interviewer and aim of the study, participant rights (confidentiality, consent and right to withdraw), and a list of standard questions and prompts. The interview protocol was pilot tested prior to administering it within the sample of experts. The researcher asked the same questions to all subject matter experts, in the same order, and introduced standard prompts where needed, as probes. Most of the interviews were completed within two hours. After the interviews, interviewees were thanked for their participation. The subject matter expert interviews were recorded on an audio recorder, with permission, and fully transcribed by three professional transcribers. The full set of transcripts was provided to a separate panel of experts for subsequent content analysis.
Thematic content analysis The transcribed interview data were analysed through content analysis, which involves “interpreting, theorizing, or making sense of data by first breaking it down into segments that can be categorized and coded, and then establishing a pattern for the entire data set by relating the categories to one another” (Camara, Deummond and Jackson 2007, p. 24). A panel of experts (three professors in industrial-organisational psychology) extracted themes from the transcribed interviews in a facilitated discussion. To distil common themes from the transcripts, we used an open analysis coding procedure to extract general themes (Blumberg et al. 2008). Each expert listed the themes they identified. Next, the facilitator asked the panel to group similar themes, identify overlaps, and reach consensus about items of disagreement. The panel reformulated each of the theme groupings until an appropriate definition (or code) was formulated for each. Explicit definitions for each code were formulated to represent a final list of eleven diversity management competencies (see Table 1). Example competencies include ‘Capitalizing on diversity’, defined as the ability of a firm to ‘Use(s) diversity by capitalizing on the diversity of knowledge, skills, 14
The International Journal of Human Resource Management abilities and perspectives of employees to create competitive advantage’, and ‘Diversity measurement’, which is considered to occur when a firm ‘Measures, evaluates and monitors diversity management progress and performance through metrics and other measuring or monitoring instruments to ensure continuous progress on diversity management goals.’
(Insert Table 1 about here)
Together, these competencies form part of a broader conceptual framework, namely the Diversity Management Competency Framework (see Figure 1). This framework, developed from diversity management literature, also proposes proximal (intermediate) and distal performance outcomes for diversity management competencies.
Questionnaire development The items for the Diversity Management Competencies Questionnaire (DMCQ) were developed using the procedures outlined in Hinkin (1998). That is, the constitutive definitions for each diversity management competency were used to generate items that represented effective and ineffective diversity management practices (or critical incidents). Example practices were drawn from the interview transcripts, and these were supplemented with examples provided by a diversity management consultant, as well as those identified in earlier literature. Next, items were filtered by identifying overlapping practices and those considered to be generic human resources practices not directly relevant to diversity management. This resulted in a provisional item pool of 280 items.
15
The International Journal of Human Resource Management Content validation Content validity of the measure was ensured in two ways. First, the content validity of each item was assessed by calculating its Content Validity Ratio (CVR) (Lawshe, 1975). That is, the set of preliminary items for each competency, along with their constitutive definitions, were administered to a panel of 12 judges, consisting of three Industrial-Organizational (I-O) psychologists and nine postgraduate students in I-O psychology. Respondents were asked to rate each item as ‘essential’, ‘useful but not essential’, or ‘not essential’ in describing the defined competency. The content validity ratio was calculated for each item, using the following formula: 𝐶𝐶𝐶𝐶𝐶𝐶 =
𝑛𝑛𝑒𝑒 − 𝑁𝑁 2
𝑁𝑁 2
(1)
, where ne is the number of judges who indicated an item as ‘essential’ and N is the total number of judges in the panel. As such, a negative content validity ratio would indicate that fewer than half of the judges indicated an item as essential to measure the construct (Lawshe 1975). Finally, to control for the possibility of chance agreement, the content validity ratio for each item was evaluated against critical cut-off values (Pan, Schumsky and Wilson 2012). This resulted in the elimination of 101 items. The second content validation procedure required a separate independent panel of four judges (a mixture of HR managers and I-O psychologists) to sort the remaining 179 items into their respective competency categories. The researchers presented a randomly sorted list of items to judges, along with the names and constitutive definitions for each of the eleven competencies. Judges had to assign the full list of items to a single competency. Cohen's (1960) kappa was calculated for each pair of judges and items that decreased agreement between judges were not retained, resulting in a final set of 123 try-out items. 16
The International Journal of Human Resource Management Pilot study to Shorten the Questionnaire In order to identify items for the final survey measure, the 123 try-out items were administered to a small convenience sample of 25 managers and professional staff from three large companies (defined as >250 employees) from diverse industries (construction, financial services and alcoholic beverages). The participants were mostly White (75%), but also included Mixed Race (21.9%) and Black African (3.1%) individuals. Managers were from middle (50%), senior (25%), and junior management/supervisors (21.9%); some were individuals with professional roles (for example, chartered accountants) (3.1%). The sample was relatively balanced in terms of sex (59.4% male; 40.6% female) and represented various age categories. We conducted analyses of basic item characteristics 2 of the pilot study data by inspecting the central tendency (means), variability (standard deviations), as well as the item-total correlation (between the item and competency score summed across items in the respective scale) of each item. Poor items (for example, those with limited variability, extremely low/high item means relatively to other items in the scale, or low item-total correlations) were identified and eliminated. In addition to these analyses, we improved the clarity of, and reduced possible ambiguity within, a few items, as suggested by some pilot study respondents. The final measure, the Diversity Management Competencies Questionnaire (DMCQ) 3, contains 98 items designed to assess 11 diversity management competencies (see Table 1). The procedures we employed ensured that the instrument was content valid as a measure of conceptually distinguishable firm-level diversity management competencies. We consider the measure useful in both private and public sector organisations; can be applied in any country where English is spoken; and may be completed by various types of employees. However, 2
For the sake of brevity, the descriptive statistics of the pilot study are not reported here. The complete measure is available as supplemental material to this article; it may also be accessed at https://sites.google.com/site/dmcquestionnaire/home for research or practice use.
3
17
The International Journal of Human Resource Management respondents should be limited to those individuals who would be familiar with diversity management practices employed within a particular firm.
Method: Reliability and Validity of the Diversity Management Questionnaire (DMCQ) Participants and procedure In the second stage of our study, we assessed the measurement properties of the measure, more specifically, its reliability of scores, internal factor structure, and criterion-related validity. To assess internal consistency and construct validity, we relied on individual-level respondents’ perceptions of the respective study variables. Companies were invited to participate by means of an electronic newsletter of the South African Board for People Practices (SABPP), a professional society for human resource practitioners and executives. The newsletter email contained a study description and a link to the online survey. In addition to this approach to recruit participants, the authors also contacted firms through their personal and social networks, in a ‘snowballing’ fashion. The survey cover page informed respondents of their rights to confidentiality, early withdrawal, and provided other research ethics information. In addition, informed consent was requested. In order to categorise respondents from each firm, the survey asked the name of the respondent’s employing firm, although participants were informed that this information would only be used to group the responses by company. In our survey, 157 respondents from 61 firms participated. Most respondents described themselves as ‘professional’ (32.5%), whereas others were line managers (17.5%), HR managers (13%), CEOs/COOs/MDs (11%), or other employees (26%). The majority of the sample was White (67.5%), with other indigenous ethnic groups forming the remainder of participants. They spoke mainly Afrikaans (58%) or English (30.6%) as their first language. Gender was approximately equally distributed (male: 49%; female: 48.4%; and gender not reported: 2.5%). The participants were evenly distributed across the
18
The International Journal of Human Resource Management adult working age, namely in their twenties (18-29 age group; 32.5%), thirties (30-39 age group; 31.8%), 4 or older (40-69 age group; 34.4%). Firms were sampled in a ‘snowball’ fashion from a single country
(South Africa) and represented industries across a broad spectrum.
Finally, as a small-scale criterion-related validation of the measure, we also investigated the relationships between diversity management competencies and a number of organisational criteria. As it would be inappropriate to study relationships among the respective variables by using individual-level data—this would ignore the ‘nested’ structure of employee data within firms (Hitt, Beamish, Jackson, and Mathieu 2007), and also be prone to common source bias (Podsakoff, MacKenzie, Lee and Podsakoff 2003)—our study treats the organisation as the level of analysis, by using respondents’ aggregated perceptions (that is, within the same firm) of the study variables to test our hypotheses. It is important to note that the latter analyses relied on an assumption that our proposed constructs exist at the firm-level, and that these constructs may be operationalised by collapsing individual respondents’ scores within each company to yield an overall estimate of the study variables at the organisation-level (Kozlowski and Klein 2000). 5 We address this assumption further in our statistical analyses. We tried to recruit respondents that would represent a broad range of perspectives, that is, both managerial and non-managerial staff, in each company. In doing so, it helped to ameliorate the limitations of ‘single-source’ perspective-bias—when respondents from a single organisation level, for example, the executive committee, have a different view of diversity-related matters than those shared by employees lower in the organisation, because it is obscured by their
4
It must be noted that a considerable proportion of these firms were local operations of well-known multinational firms, or South African listed firms with operations and markets in countries outside of Africa. 5 In doing so, we acknowledge the important issues associated with individual vs. organisation-level constructs (see Kozlowski and Klein, 2000), as raised by an anonymous reviewer.
19
The International Journal of Human Resource Management organizational vantage point, interests, and opportunities to observe 6. So, by including the perspectives of multiple respondents within the same organisation into an overall firm-level score for the each study variable, we tried to limit the threat of common source bias (Podsakoff et al. 2003) that is prevalent in analyses relying only on individual-level data only.
Measures Diversity
Management
Competencies
Questionnaire
(DMCQ).
The
Diversity
Management Competencies Questionnaire was used to measure overall diversity management competency in each firm. Its 98 items are divided amongst 11 sub-scales, each tapping a specific diversity management competency (see Table 1). Example items are “the organisation influences perceptions of diversity in broader society”; “the organisation assembles work teams to capitalise on strengths of diversity”; and “the organisation measures, evaluates and monitors diversity management progress and performance”. The items are answered on a 5-point Likert scale ranging from strongly disagree (1) to strongly agree (5). We included a sixth response option (i.e., don’t know) to allow for the possibility that an employee might have insufficient exposure to, or knowledge of, diversity management in an organisation. In addition to the content validity evidence in the instrument development phase of our study, the data from our subsequent field survey showed that the subscales were internally consistent (.80 < α < .97; see Table 2). We report on these analyses, as well as the factor structure, in our Results section. A firm’s overall diversity management competency score was calculated as the unweighted linear composite (that is, the mean) of scores on the 11 competency measures. (Insert Table 2 about here)
6
We thank an anonymous reviewer for highlighting this important issue.
20
The International Journal of Human Resource Management Inclusive Climate. The 9-item Inclusive Climate scale was developed from literature and earlier measures (Avery et al. 2007; Herdman and McMillan-Capehart 2010). An example item is: “The company has a culture of openness and listening to diverse views of employees”. Perceived
Advancement
Opportunities.
The
6-item
Perceived
Advancement
Opportunities scale was based upon an earlier measure (Riordan and Shore 1997). An example item is: “There is equal access to advancement opportunities for all employees”. Perceived Power Distribution. The 5-item Perceived Power Distribution measure was a modified version of an earlier measure (Ledford, Mohrman and Pelled 1999). An example item is: “Diversity groups have equal authority or power in the organisation”. Performance Index. The 56-item Performance Index (Spangenberg and Theron 2004) was used to measure subjective perceptions of firm performance. The Performance Index was developed from a comprehensive model of work unit performance effectiveness. The questionnaire measures eight independent dimensions. We used subscales tapping Productivity and Efficiency, Adaptability, Capacity, Market Standing, and Projected Future Growth. The measure uses a five-point Likert response scale. The Performance Index scales we used have shown very good internal consistency reliability (.80 < α < .89) and factorial validity (Spangenberg and Theron 2004). An overall firm performance index was calculated for each firm by averaging scores on the various performance dimensions.
Results Preliminary Analyses The measurement properties of the respective measures were assessed at the individual (respondent) level of analysis. Prior to analyses, we tested the data for adherence to statistical assumptions, including normality, linearity, homoscedasticity and extreme values: no serious 21
The International Journal of Human Resource Management violations were found, except for skewness for some dimensions and lack of multivariate normality. Measurement Properties Length Considerations. A design specification of the DMCQ was that it should be both diagnostic, as well as informative. Diagnostic, because it is useful as a broad measure of diversity management competencies, but also informative, that is, the questionnaire should suggest ways to respondents in which they may enhance diagnosed deficiencies. As a result, each competency measure had to represent a comprehensive sample of organizational practices. By using longer lists of individual practices in each subscale (as opposed to only the minimum required to achieve internally consistent measures) we were not only able to (a) sample the content domain of each competency dimension in a ‘saturated’ fashion, but also, (b) suggest the good practices in the questionnaire items. However, the drawback of its comprehensive nature is its length, when all the subscales are used. When all test items are used for each dimension, large sample sizes are needed to assess its measurement properties, given the need for a relatively high participant-to-item ratio (Tabachnick and Fidell 2013). As we foresee different ways in which our measure may be used—either as an abbreviated broad, ‘shallow’ measure of all 11 dimensions, or as an exhaustive, ‘deep’ list of practices for only selected dimensions—we report the reliability and construct validity analyses in ways that may support either use of the measure (see ‘validity’, Nunnally and Bernstein 1994). Reliability. First, item analysis was used to assess the internal consistency of scores from the respective measures. The reliability estimates (Cronbach’s α; see Table 2) were satisfactory for the Diversity Management Competency Questionnaire (DMCQ) scales (.80 < α < .97), the intermediate diversity outcomes (.93 < α < .96) measures, as well as the Performance Index
22
The International Journal of Human Resource Management scales (.83 < α < .88) measuring firm performance. DMCQ Factor Structure. Next, we assessed the internal factor structure of the DMCQmeasure with confirmatory factor analysis 7. Robust Maximum Likelihood (ML) estimation was employed to estimate all models. We used Lisrel 9.1 (SSI, 2012) to conduct all analyses. As a result of a high item-to-participant ratio (98:157) in our study, a test of the complete measurement model of the DMCQ at the item level was not recommended, because it would result in insufficient statistical power (Hair et al. 2013). Consequently, we tested the complete measurement model—one that included all 11 subscales—using item-parcels as indicator variables. We created item parcels, three for each competency dimension, using procedures outlined by Little, Cunningham, Shahar, and Widaman (2002). Support was found as the model showed good fit, χ2 (440, N = 157) = 924.440, p < .01, Satorra-Bentler χ2 (440, N = 157) = 711.270, p < .01, Robust CFI = .99, RMSEA = .08, 90% CI: [0.076; 0.091]. The large chi-square values are largely due to a number of modification indices that suggested freeing error-covariances—this could not be theoretically justified in our study— as well as some cross-loadings. All 33 item parcels were reasonable indicators of their respective latent factors (range of lambda factor loadings =.44 to .99; Mlambda = .87, SDlambda = .12). Consistent with a competency-model perspective all factors were significantly interrelated (range of latent factor intercorrelations 8 = .53 – .83). It is common for dimensions of performance to show ‘positive manifold’ (e.g., Viswesvaran, Schmidt, and Ones, 2005) and therefore, realistic to expect some cross-loading between constructs and indicators of constructs. DMCQ Discriminant Validity. To determine the discriminant validity (Campbell and 7
Prior to fitting the hypothesized models within each subscale, we used exploratory factor analysis to test the unidimensionality assumption (Hair et al. 2013)—uni-dimensionality was evident for all rotated solutions (using Principal Axis Factoring and Direct Oblimin rotation). 8 These correlation estimates have been corrected for measurement error and, therefore, are somewhat higher than would be observed with standard correlation analyses.
23
The International Journal of Human Resource Management Fiske 1959) of the DMCQ scales, we compared (a) a baseline model in which the correlations between the competency dimensions are freely estimated; and (b) a nested comparison model in which the correlations are constrained to be unity. We used the raw data as input for the analysis and found poor fit of the nested model, χ2 (495, N = 157) = 2519.294, p < .01, RMSEA = .16 (90% CI: .155; .168). A chi-square difference test indicated that the nested model (specifying the relationship between competencies as perfectly correlated) showed significantly poorer fit, compared to the baseline model, χ2diff(55, N = 157) = 1594.854, p < .001. We also conducted a more stringent test of discriminant validity. We examined the variance extracted estimates and the squared correlation estimates between the constructs (Hair, Black, Babin, Anderson, and Tatham, 2006). Results showed that average variance extracted (AVE) between constructs was good (MAVE = .76, SDAVE = .14) and the average squared correlation between constructs was relatively lower (Mr2 = .52, SDr2 = .12). Only four (7.2%) of 55 possible intercorrelations between competency dimension showed poor evidence of discriminant validity; that is, in the majority of cases, latent constructs explained more variance in their measures than they shared with one another. This provides good evidence of discriminant validity and suggests that the constructs are distinct (Hair et al. 2006). Therefore, the evidence suggests that the competency dimensions are empirically distinct from one another. In sum, the DMCQ showed acceptable internal construct validity. Item-level CFA of DMCQ. In acknowledgement of potential concerns with item parceling as an approach to create observed variables in CFA (see Little et al. 2002), we executed a number of confirmatory factor analyses using randomly drawn subsets of three items per subscale. A CFA including all the 98 questionnaire items was not expected to converge due to the high item-to-respondent ratio. This approach is akin to using a ‘short-form’ of the DMCQ.
24
The International Journal of Human Resource Management Resulting model fit estimates were generally consistent to those observed with the analysis using item parcels 9. Overall, our factor analyses showed support for the hypothesized factor structure, although the competency dimensions showed sizeable intercorrelations. We recommend that the construct validity of the DMCQ is further assessed and cross-validated in larger, independent samples. Tests of Hypotheses: Criterion-related Validity Data Aggregation. In order to test our study hypotheses, we required scores for variables aggregated at the company level. As such, it was necessary to demonstrate that the data collected at a lower level of analysis (e.g., individual-level perceptions of diversity management and firm outcomes) were similar enough to one another to aggregate as an indicator of a higher-level construct (e.g., shared perceptions of firm-level variables)(Kozlowski and Klein 2000; Lebreton and Senter 2008). We assessed the inter-rater agreement (rwg; James, Demaree, and Wolf, 1993) between multiple respondents within the same organization and found it was generally strong (Mean rwg across all study variables, across firms = .75; SDrwg = .11) and therefore supported aggregation (Lebreton and Senter 2008). However, our default criterion for inclusion of a firm (i.e, that it would require at least three or more respondents) limited the data available for analyses (k = 21). We repeated the agreement analyses, but this time, using a liberal cut-off of at least two respondents per firm in order to qualify for inclusion in the firm-level analyses. As this relaxation in the inclusion constraint resulted in similar inter-rater agreement (Mean rwg across all study variables, across firms = .75; SDrwg = .10) and enhanced the firm-level observations available for analysis (k = 35), we decided to conduct the firm-level analyses accordingly. Therefore, we calculated a linear composite (that is, the mean) of employees’ dimensions scores 9
The results of these analyses are available on request from the second author.
25
The International Journal of Human Resource Management to arrive at an aggregate score for a particular firm. Table 3 presents the intercorrelations between the study variables, determined at the organizational unit of analysis. For the sake of brevity, we use overall diversity management competency as the independent variable in the hypothesis tests. However, correlations between the complete list of competencies and the outcome variables are reported in the table. (Insert Table 3 about here) Hypothesis 1 stated that overall firm-level diversity management competency would be positively related to firm performance. The correlation was highly statistically significant (r = .56, p < .01) with a large effect size (Cohen, 1988), supporting Hypothesis 1. Across firms in our study, their overall perceived diversity management competency was related to employees’ general perceptions of firm performance. Hypotheses 2-4 stated that firm-level diversity management competency would be positively related to proximal diversity-related outcomes, including inclusive climate (H2), perceived advancement opportunities (H3), and perceived power distribution (H4). These hypotheses were also supported as correlations were significant between diversity management competency and inclusive climate (r = .67, p < .01), advancement opportunities (r = .50, p < .01) and distribution of power (r = .62, p < .01). In sum, firms’ overall diversity management competency was related to proximal diversity-related outcomes. Hypotheses 5-7 proposed that these proximal outcomes would, in turn, be associated with a distal firm performance criterion. That is, we expected that inclusive climate (H5), perceived advancement opportunities (H6), and perceived power distribution (H7) would be positively related to perceived firm performance. Results showed that correlations were statistically significant between inclusive climate (r = .42, p < .05), distribution of power
26
The International Journal of Human Resource Management (r = .46, p < .01), and firm performance, but advancement opportunities (r = .30, p > .05) did not predict perceived firm performance. As a result, Hypotheses 5 and 7 were supported, but not Hypothesis 6. Overall, diversity-related outcomes showed medium effects with perceived firm performance in our sample of organizations.
Discussion In the research reported here, we developed a framework and measure of diversity management competencies at the firm level. Our approach draws the focus towards a process mind-set of diversity management which assumes that the way in which a company manages diversity may ultimately have an effect on proximal or distal firm-level outcomes. In a concurrent validation study conducted in a sample of 61 firms, we tested the psychometric properties of a newly developed measure of diversity management competencies and evaluated the relationships between firms’ overall diversity management competency and perceptions of diversity-related outcomes and firm performance. Main findings and implications for theory Our diversity management competency framework represents a novel approach to understanding how firms manage diversity from a practical point of view, that is, it proposes 11 clusters of practices that should affect proximal diversity-related outcomes and distal firm performance— the so-called ‘business case for diversity’, explained. By enhancing employees’ perceptions of an inclusive climate, advanced opportunities for all, and equal power distribution, diversity management competencies may ultimately affect ‘bottom-line’ firm performance. By applying a competency-modelling approach to organisational diversity management practices, we have essentially identified the ‘critical incidents’ of diversity management and logically ordered them
27
The International Journal of Human Resource Management into a sequential conceptual framework. In doing so, our study extends the literature on conceptual diversity management frameworks (Chanda et al. 2009). The results of our instrument development stage showed that diversity management practices at the firm level can be parsimoniously clustered into eleven diversity management competencies. We showed it is possible to measure these competencies reliably and validly using the measure developed here, the Diversity Management Competency Questionnaire (DMCQ). In our data, the DMCQ exhibited acceptable content-, construct-, and criterion-related validity. So, our study provides a practical tool that allows the measurement of firms’ diversity management efforts that is easily understandable and ‘face-valid’. In the final part of our study, we tested relationships suggested in our conceptual model in a sample of firms representative of a broad range of industries. Results suggest that firms with better developed diversity management competencies generally also enjoy higher firm performance. The findings build on previous evidence that indicate that diversity management is related to firm performance (e.g. Bezrukova et al. 2003; Choi and Rainey 2010; Ferris, Hiller, Kroll and Wright 1995; Johnson and Richard 2001). Despite our relatively small sample of firms (k = 61), our study is one of a few to provide some empirical evidence of the ‘business-case’ for diversity. We further develop this literature by also showing how effective diversity management may affect firm performance. Our framework proposed that intermediary diversity management outcomes may play an intermediary role. Results showed that firms’ overall diversity management competency was associated with proximal diversity-related perceptions (such as a climate of inclusion, perceptions of equal power distribution, and perceived advancement opportunities) which, in turn, also correlated with perceptions of firm performance. In doing so,
28
The International Journal of Human Resource Management our results mirror earlier findings that diversity efforts may successfully impact diversity climate (Herdman and McMillan-Capenhart 2010). In their words, “diversity climate is emerging as an important intermediary outcome in the prediction of key organizational and individual outcomes” (p. 49). We add to this emerging line of work some evidence that employees’ perceptions of power distribution and career advancement opportunities may be extra pieces of the puzzle in the ‘black box’ of diversity management. Taken together, our study suggests plausible mechanisms that may explain how diversity management ultimately leads to firm performance. We urge more studies that test these potential mediating mechanisms (for an example of tests of human resource management as mediating mechanisms, see Jiang, Lepak, Hu, and Baer, 2012).
Implications for practice The present study also provides practical guidance for diversity management at the organisational level. First, the measure 10 that we developed can be used as a diagnostic tool to evaluate firms’ diversity management practices. Not only does the measure allow for diagnosis, but it is also suggestive of diversity management ‘good practice’. We encourage practitioners to use it to extract ‘critical incidents’ that may enhance their diversity management programmes, systems, and procedures. Another way to use our measure is in a normative fashion. For example, we have prepared diagnostic reports that summarise and normatively compare firms’ performance on the set of competencies against those of other firms that we surveyed. This diversity management competency profile highlights areas of diversity management strength and potential development areas.
10
This measure may be accessed as supplemental material to this article.
29
The International Journal of Human Resource Management In addition to the measurement tool, our broader Diversity Management Competency Framework (see Figure 1) can be used as a simple and effective guiding framework to manage diversity. Our content validation procedures ensured that the competencies were coherent, parsimonious and that they ‘make sense’ to practitioners conceptually. In addition, the broader framework gives an explanatory model of how diversity management may lead to outcomes that are salient to firm stakeholders. Our experience is that practitioners are more likely to adopt competency models when they understand how it leads to things that are important to them.
Limitations The first limitation of the study relates to sampling issues. To identify participants for our study, non-probability sampling was used and our sample of companies was limited. As such, the sample may not be representative of the population and sampling error cannot be determined (Blumberg et al. 2008). We took this approach because of the practical difficulty of finding respondents from multiple organisations. Our sampling relied heavily on voluntary participation from practitioners on the national distribution list of a professional HR society and further snowball sampling. Future studies should replicate our work in larger samples of firms, ideally drawn from populations using probability sampling techniques. Larger samples would enable more rigorous testing of the proposed direct and mediating relationships (as path coefficients) in our conceptual model. Second, we averaged the responses of multiple employees within the same firm to operationalize the firm-level scores on the study variables. While we tried to ensure that a minimum of five respondents from each organisation participated, we had limited success. The relatively small number of respondents per firm makes it likely that the correlation results still contain some common rater effects, which refer to “any artifactual covariance between the 30
The International Journal of Human Resource Management predictor and criterion variable produced by the fact that the respondent providing the measure of these variables is the same” (Podsakoff, MacKenzie, Lee and Podsakoff 2003, p. 881). We suggest that future studies include at least five respondents from the same firm, as averaging across multiple raters reduces individual rater effects (Conway 2002). Not only would this provide a more reliable estimate of firm-level construct levels, but also, it would open up the opportunity to assess direct and cross-level effects at individual and firm-levels of analysis in a multilevel fashion (Aguinis, Chen, Culpepper and Mathieu 2012). Last, in addition to the subjective measures of firm performance we employed, future studies may benefit from including objective measures of firm performance.
Further research The present study creates an agenda for further research. First, the measurement properties of the Diversity Management Competencies Questionnaire (DMCQ) should be further explored. In addition to the analyses we conducted, it is also important to cross-validate its measurement properties in other, larger samples. We did not consider our sample size (N = 157) large enough to ensure statistical power for confirmatory factor analysis at the item-level, primarily due to a very high item-to-participant ratio when the full questionnaire is used (Tabachnick and Fidell 2013). As a result, we had to rely on item parcelling to create observed variables for our CFA. Future studies should also assess the factor structure and discriminant validity of the DMCQ using item-level data. Such studies could also determine whether, or not, a higher-order factor of diversity management competency influences scores at the first-order competency level. Secondly, the Diversity Management Competency Framework (Figure 1) needs further empirically evaluation. Our early empirical tests reported here were rather elementary, because of the limited sample size at the firm-level, and therefore represent only a first step towards 31
The International Journal of Human Resource Management evaluating the broader framework. And by using the measure we developed in our study, it becomes possible to test the suggested ‘causal’ paths in diversity-firm performance models using more sophisticated techniques, such as structural equation modelling (SEM) and hierarchical linear modelling (HLM). Such studies may produce useful evidence-based models that speak directly to the so-called ‘business-case for diversity’.
Conclusion Overall, our research has joined the debate on the business-case for diversity by developing a framework and measure of firm-level diversity management competencies. We also provided initial tests of links between these competencies and other firm-level outcomes, those that are proximal, and others that are distal. Our results represent early evidence that managing diversity better, through a set of eleven diversity management competencies, may hold potential benefits for companies that are diverse. As such, the present study gives evidence-based guidance to firms on effective diversity management. In doing so, we hope to promote sustainable workforce diversity management to the benefit of the employee, employer, and society.
Acknowledgements We thank the Mandela Rhodes Foundation and the Department of Industrial Psychology, University of Stellenbosch, for their financial support, and the South African Board for People Practices (SABPP) for help with distributing our survey. We also thank Proff C. C. Theron and A. S. Engelbrecht for their helpful comments.
32
The International Journal of Human Resource Management References Aguinis, H., Chen, G., Culpepper, S. A., and Mathieu, J. E. (2012), ‘Understanding and Estimating the Power to Detect Cross-level Interaction Effects in Multilevel Modeling,’ Journal of Applied Psychology, 97, 5, 951-966. Anderson, C., Gruenfeld D. H., and Keltner, D. (2003), ‘Power, Approach, and Inhibition,’ Psychology Review, 110, 265–84. Ammeter, A. P., Arthur, M. M., Ferris, G. R., Frink, D. D., Kaplan, G. R., Morrisette, H. S., Reithel, B., and Robinson, R. K. (2003), ‘Gender Demography and Organization Performance,’ Group & Organization Management, 28, 1, 127-147. Avery, D. R., Hebl, M. R., Hernandez, M., Mckay, P. F., Morris, M. A., and Tonidandel, S. (2007), ‘Racial Differences in Employee Retention: Are Diversity Climate Perceptions the Key?’ Personnel Psychology, 60, 35–62. Barnett, T., Chadwick, K., Dwyer, S., and Richard, O. C. (2004), ‘Cultural Diversity in Management, Firm Performance, and the Moderating Role of Entrepreneurial Orientation Dimensions,’ Academy of Management Journal, 47, 2, 255–266. Bartram, D. (2011), ‘The SHL Universal Competency Framework,’ White Paper, 1-10. www.shl.com. Barney, J. B. (2001). ‘Resource-based Theories of Competitive Advantage: A Ten-year Retrospective on the Resource-based View,’ Journal of Management, 27, 643–650. doi: 10.1177/014920630102700602 Barney, J., Ketchen, D. J., and Wright, M. (2001), ‘The Resource-based View of the Firm: Ten Years After 1991,’ Journal of Management, 27, 625-641.
33
The International Journal of Human Resource Management Barrick, M., Field, H. S., and Gatewood, R. D. (2008), Human Resource Selection (6th ed.), Mason: Thomson South-Western. Bartle, S. A., Hayes, B. C., and Major, D. A. (2002), ‘Climate for Opportunity: A Conceptual Model,’ Human Resource Management Review, 12, 445–468. Beale, R. L., and Cox, T. (1997), Developing Competency to Manage Diversity: Reading, Cases & Activities, San Francisco: Berrett-Koehler Publishers. Bezrukova, K., Ely, R., Jackson, S., Jehn, K., Joshi, A., Kochan, T., Levine, D., Leonard, J., and Thomas, D. (2003), ‘The Effects of Diversity on Business Performance: Report of the Diversity Research Network,’ Human Resource Management, 42, 1, 3-21. Blake, S., and Cox, T. H. (1991), ‘Managing Cultural Diversity: Implications for Organizational Competitiveness,’ The Executive, 5, 3, 45-56. Blumberg, B., Cooper, D. R., and Schindler, P. S. (2008), Business Research Methods (2nd ed.), London: McGraw-Hill Higher Education. Boon, C., Boselie, J. P., and Dietz, G. (2005), ‘Commonalities and Contradictions in Research on Human Resource Management and Performance,’ Human Resource Management, 15, 3, 67-94. Camara, S., Drummond, D. K., and Jackson, R. L. (2007), ‘What is Qualitative Research?’ Qualitative Research Reports in Communication, 8, 1, 21-28. Campion, M. A., Carr, L., Fink, A. A., Odman, R. B., Phillips, G. M., and Ruggeberg, B. J. (2011), ‘Doing Competencies Well: Best Practices in Competency Modeling,’ Personnel Psychology, 64, 225–262.
34
The International Journal of Human Resource Management Chanda, A., D’Netto, B., Monga, M., and Shen, J. (2009), ‘Managing Diversity Through Human Resource Management: An International Perspective and Conceptual Framework,’ The International Journal of Human Resource Management, 20, 2, 235-251. Cho, S., and Mor Barak, M. E. (2008), ‘Understanding of Diversity and Inclusion in a Perceived Homogeneous Culture: A Study of Organizational Commitment and Job Performance Among Korean Employees,’ Administration in Social Work, 32, 4, 100-126. Choi, S., and Rainey, H. G. (2010), ‘Managing Diversity in U.S. Federal Agencies: Effects of Diversity and Diversity Management on Employee Perceptions of Organizational Performance,’ Public Administration Review, 109-121. Cohen, J. (1960). ‘A Coefficient of Agreement for Nominal Scales,’ Educational and Psychological Measurement, 20, 1, 37-46. Cohen, J. (1988). Statistical Power Analysis for the Behavioral Sciences, (2nd ed.), Hillsdale, NJ: Lawrence Erlbaum Associates. Conway, J. M. (2002), ‘Method Variance and Method Bias in I/O Psychology,’ in Handbook of Research Methods in Industrial and Organizational Psychology (2002 ed.), ed. S. G. Rogelberg, Oxford: Blackwell Publishers, pp. 344–365. Cox, T. (1994), Cultural Diversity in Organizations: Theory, Research & Practice, San Francisco: Berrett-Koehler Publishers, Inc. Cox, T. (2001), Creating the Multicultural Organization: A Strategy for Capturing the Power of Diversity, San Francisco, CA: Jossey-Bass. DiTomaso, N., Parks-Yancy, R., and Post, C. (2007), ‘Workforce Diversity and Inequality: Power, Status, and Numbers,’ Annual Review of Sociology, 33, 473-501.
35
The International Journal of Human Resource Management Dunford, B. B., Snell, S.A., and Wright, P. M. (2006), ‘Human Resources and the Resource Based View of the Firm,’ in Strategic Human Resource Management: Theory and Practice (2nd ed.), eds. J. Billsberry, G. Salaman, and J. Storey, London: Sage Publications, pp. 17-39. Ely, R. J., and Thomas, D. A. (2001), ‘Cultural Diversity at Work: The Effects of Diversity Perspectives on Work Group Processes and Outcomes,’ Administrative Science Quarterly, 46, 229-273. Ferris, S., Hiller, J., Kroll, M., and Wright, P. (1995), ‘Competitiveness Through Management of Diversity: Effects on Stock Price Valuation,’ Academy of Management Journal, 38, 272– 87. Tabachnick, B. G., and Fidell, L. S. (2013), Using Multivariate Statistics (6th ed.), Boston: Pearson Education. Ford, D., Ismail, K., and Richard, O. C. (2006), ‘Exploring the Performance Effects of Visible Attribute Diversity: The Moderating Role of Span of Control and Organizational Life Cycle,’ International Journal of Human Resource Management, 17, 12, 2091–2109. Gerhart, B., Hollenbeck, J. R., Noe, R. A., and Wright, P. M. (2008), Human Resource Management: Gaining a Competitive Advantage (6th ed.), New York: McGrawHill/Irwin. Gilbert, J. A., Ivancevich, J. M., and Stead, B. A. (1999), ‘Diversity Management: A New Organizational Paradigm,’ Journal of Business Ethics, 21, 1, 61-76. Gonzalez, J. A. (2010), ‘Diversity Change in Organizations: A Systemic, Multilevel, and Nonlinear Process,’ The Journal of Applied Behavioral Science, 46, 2, 197–219.
36
The International Journal of Human Resource Management Hair, J., Black, W., Babin, B., Anderson, R., & Tatham, R. (2006). Multivariate Data Analysis. (7th ed.), Upper Saddle River, NJ: Pearson Education Inc. Hammer, T. H., and Landau, J. (1986), ‘Clerical Employees’ Perceptions of Intraorganizational Career Opportunities,’ Academy of Management Journal, 29, 2, 385-404. Herdman, A. O., and McMillan-Capehart, A. (2010), ‘Establishing a Diversity Program is Not Enough: Exploring the Determinants of Diversity Climate,’ Journal of Business Psychology, 25, 39–53. Hinkin, T. R. (1998), ‘A Brief Tutorial on the Development of Measures for Use in Survey Questionnaires,’ Organizational Research Methods, 1, 1, 104-121. Human, L. (2005), Diversity Management for Business Success, Pretoria: Van Schaik Publishers. Hitt, M. A., Beamish, P. W., Jackson, S. E., and Mathieu, J. E. (2007), ‘Building Theoretical and Empirical Bridges Across Levels: Multilevel Research in Management,’ Academy of Management Journal, 50, 6, 1385-1399. Jackson, S. E. (1992), ‘Preview of the Road to be Travelled,’ in Diversity in the Workplace: Human Resource Initiatives (1992 ed.), ed. S. E. Jackson, New York: Guilford Press, pp. 3-12. James, L. R., Demaree, R. G., and Wolf, G. (1993), ‘rwg: An Assessment of Within-group Interrater Agreement,’ Journal of Applied Psychology, 69, 85-98. Jiang, K., Lepak, D. P., Hu, J., and Baer, J. C. (2012), ‘How Does Human Resource Management Influence Organizational Outcomes? A Meta-Analytic Investigation of Mediating Mechanisms,’ Academy of Management Journal, 55, 6, 1264-1294. Johnson, N. B., and Richard, O. C. (2001), ‘The Impact of Human Resource Diversity Practices on Firm Performance,’ Journal of Managerial Issues, 13, 2, 177-195.
37
The International Journal of Human Resource Management Joshi, A., Liao, H. and Roh, H. (2011), ‘Bridging Domains in Workplace Demography Research: A Review and Reconceptualization,’ Journal of Management, 37, 521-552. Kirby, S. L., and Richard, O. C. (2000), ‘Impact of Marketing Work-Place Diversity on Employee Job Involvement and Organizational Commitment,’ The Journal of Social Psychology, 140, 3, 367-377. Kochan, T. A., McMillan-Capehart, A., and Richard, O. C. (2002), ‘The Impact of Visible Diversity on Organizational Effectiveness: Disclosing the Contents in Pandora’s Black Box,’ Journal of Business and Management, 8, 3, 265-291. Konrad, A. M., and Linnehan, F. (2003), ‘Affirmative Action as A Means Of Increasing Workforce Diversity’. In Individual Diversity and Psychology in Organizations (2003 ed.), eds. M. J. Davidson, and S. L. Fielden, Chichester: John Wiley & Sons, pp. 95-112. Kozlowski, S. W. J., and Klein, K. J. (2000), ‘A Multilevel Approach to Theory and Research in Organizations: Contextual, Temporal, and Emergent Processes’. In Multilevel Theory, Research, and Methods in Organizations (2000 ed.), eds. K. J. Klein, and S. W. Kozlowski, San Francisco: Jossey-Bass, pp. 349-381. Little, T. D., Cunningham, W. A., Shahar, G., and Widaman, K. F. (2002), ‘To Parcel or Not to Parcel: Exploring the Question, Weighing the Merits,’ Structural Equation Modeling: A Multidisciplinary Journal, 9, 151-173. Lawrence, B. S. (1997), ‘The Black Box of Organizational Demography’, Organization Science, 8, 1-22. Lawshe, C. H. (1975), ‘A Quantitative Approach to Content Validity,’ Personnel Psychology, 28, 563-575.
38
The International Journal of Human Resource Management Lebreton, J. M., and Senter, J. L. (2008), ‘Answers to 20 Questions About Interrater Reliability and Interrater Agreement,’ Organizational Research Methods, 11, 4, 815-852. Ledford, G. E., Mohrman, S. A., and Pelled, L. H. (1999), ‘Demographic Dissimilarity and Workplace Inclusion,’ Journal of Management Studies, 36, 7, 1013–1031. Litvin, D. R. (2006), ‘Diversity: Making Space for a Better Case,’ in Handbook of Workplace Diversity (2006 ed.), eds. A. M. Konrad, P. Prasad and J. K. Pringle, London: Sage, pp. 75–94. Maak, T., and Pless, N. M. (2004), ‘Building an Inclusive Diversity Culture: Principles, Processes and Practice,’ Journal of Business Ethics, 54, 2, 129–147. Martins, L. L., and Milliken, F. J. (1996), ‘Searching for Common Threads: Understanding the Multiple Effects of Diversity in Organizational Groups,’ Academy of Management Review, 21, 2, 402-433. Nunnally, J. C., and Bernstein, I. H. (1994), Psychometric Theory (3rd ed.), New York: McGrawHill. Nishii, L. H. (2013), ‘The Benefits of Climate for Inclusion for Gender-Diverse Groups,’ Academy of Management Journal, 56, 6, 1754-1774. Pan, W., Schumsky, D. A., and Wilson, F. R. (2012), ‘Recalculation of the Critical Values for Lawshe's Content Validity Ratio,’ Measurement and Evaluation in Counseling and Development, 45, 197-210. Pfeffer, J. (1998), The Human Equation: Building Profits by Putting People First, Boston: Harvard Business School Press.
39
The International Journal of Human Resource Management Podsakoff, P. M., MacKenzie, S. B., Lee, J-Y, and Podsakoff, N. P. (2003), ‘Common Method Biases in Behavioral Research: A Critical Review of the Literature and Recommended Remedies,’ Journal of Applied Psychology, 88, 5, 879-903. Prahalad C., and Hamel, G. (1990), ‘The Core Competency of the Corporation,’ Harvard Business Review, 79-91. Reichenberg, N. E. (2001), ‘Best Practices in Diversity Management,’ GLADNET Collection, Paper 309, retrieved from http://digitalcommons.ilr.cornell.edu/gladnetcollect/309 Riordan, C. M., and Shore, L.M. (1997), ‘Demographic Diversity and Employee Attitudes: An Empirical Examination of Relational Demography Within Work Units,’ Journal of Applied Psychology, 82, 3, 342–358. Roberson, Q. M. (2006), ‘Disentangling the Meaning of Diversity and Inclusion in Organizations,’ Group & Organization Management, 31, 212-236. Sacco, J. M., and Schmitt, N. A. (2005), ‘A Dynamic Multilevel Model of Demographic Diversity and Misfit Effects,’ Journal of Applied Psychology, 90, 203-231. Scientific Software International (SSI) Inc. (2015), Lisrel 9.1 for Windows, Skokie: Author. Schuler, R., and Walker, J. (1990), ‘Human Resource Strategy: Focusing on Issues and Actions,’ Organization Dynamics, 5-19. Seymen, O. A. (2006), ‘The Cultural Diversity Phenomenon in Organisations and Different Approaches for Effective Cultural Diversity Management: A Literary Review,’ Cross Cultural Management: An International Journal, 13, 4, 296-315. Spangenberg, H. H., and Theron, C. (2004), ‘Development of a Questionnaire for Assessing Work Unit Performance,’ South African Journal of Industrial Psychology, 30, 1, 19-28.
40
The International Journal of Human Resource Management Thomas, A. (1996), Beyond Affirmative Action: Managing Diversity for Competitive Advantage In South Africa, Randburg: Knowledge Resources. Viswesvaran, C., Schmidt, F. L., and Ones, D. S. (2005), ‘Is There a General Factor in Ratings of Job Performance? A Meta-Analytic Framework for Disentangling Substantive and Error Influences,’ Journal of Applied Psychology, 90, 1, 108-131. Vroom, V. H. (1964), Work and motivation, New York: Wiley.
41
The International Journal of Human Resource Management Table 1. Definitions of diversity management competencies, number of items and internal consistency reliabilities (Cronbach’s α) of scales in the Diversity Management Competencies Questionnaire (DMCQ). Competency 1. Leading
diversity 2. Capitalizing on
diversity 3. Sustainable
workforce transformation
Definition Showing diversity leadership in the market and society by voicing support for diversity and investing in initiatives that establish greater equal employment opportunity and redress societal inequality. Uses workforce diversity by capitalizing on the range of knowledge, skills, abilities and perspectives of employees to create competitive advantage. Balances workforce transformation needs with operational business realities to ensure sustainable workforce transformation and bottom-line financial performance.
4. Fair practices
Promotes fair employment practices to enhance employment equity, equal opportunity and fair treatment. The firm identifies and removes barriers to fair employment practices and implements best practice guidelines for enhancing equal opportunity and preventing unfair discrimination.
5. Systemizing
Plans and organizes how diversity management can be integrated into the organizational system by transforming HR functions, establishing additional policies, and implementing systems and procedures to integrate diversity management into the formal functioning of the organization.
diversity management 6. Diversity
measurement 7. Diversity
change management 8. Diversity
management credibility 9. Affirmative
development
Measures, evaluates and monitors diversity management progress and performance through metrics and other measuring or monitoring instruments to ensure continuous progress on diversity management goals. Manages diversity through a change management process and OD interventions which include constant communication and feedback of the importance and progress of diversity management and driving the process through champions for change. Shows authenticity and transparency in managing diversity through congruent actions so that trust is built amongst stakeholders. Supports aggressive accelerated functional skills development of individuals from protected/designated groups by proactively identifying, recruiting, nurturing and developing promising individuals.
10. Diversity skills
Empowers employees and managers to demonstrate effective interpersonal behaviour within a diverse workforce, including competency in diversity-related knowledge, skills, abilities and other characteristics.
11. Promoting
Actively promotes awareness, respect and valuing of differences (of any kind) within the organization through symbolic actions that reveal that diversity is celebrated.
inclusiveness
42
The International Journal of Human Resource Management
Table 2. Descriptive statistics and Cronbach’s α internal consistency of subscales. Descriptives
Reliability
M
SD
# Items
α
Leading diversity
3.74
0.67
9
0.92
Capitalising on diversity
3.49
0.68
8
0.92
Sustainable workforce transformation
3.74
0.59
7
0.80
Fair practices
3.72
0.43
11
0.88
Systemizing diversity management
3.83
0.53
7
0.82
Diversity measurement
3.21
0.79
11
0.96
Diversity change management
3.06
0.88
10
0.97
Diversity management credibility
3.25
0.73
10
0.96
Affirmative development
3.68
0.62
6
0.89
Diversity skills
3.01
0.8
10
0.97
Promoting inclusiveness
3.25
0.73
9
0.93
Perceived advancement opportunities
3.88
0.71
6
0.93
Perceived power distribution
3.48
0.79
5
0.95
Inclusive climate
3.86
0.67
9
0.96
Production and efficiency
3.66
0.59
5
0.83
Adaptability
3.63
0.55
7
0.88
Capacity/Sufficient resources
3.5
0.61
7
0.85
Market standing
3.72
0.58
7
0.87
Projected future growth
3.56
0.57
5
0.88
Measure Diversity Management Competencies
Diversity Management Outcomes
Performance Index (PI)
Notes: N = 157. Unit of analysis for the reliability coefficients was at the individual level.
43
Table 3. Intercorrelations of study variablesa. Variable 1. Leading diversity
1. -
2. Capitalizing on diversity .80**
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
-
3. Sustainability
.25
.47**
-
4. Fairness of practices
.59** .64** .68**
5. Systemizing diversity
.77** .57** .29
.64**
6. Measuring diversity
.79** .74** .42*
.67** .77**
7. Change management
.80** .84** .34*
.60** .57** .76**
8. Credibility
.71** .81** .53** .63** .52** .65** .90**
9. Development
.77** .77** .52** .82** .68** .71** .70** .68**
10. Diversity competence
.79** .80** .24
.56** .63** .72** .90** .81** .68**
11. Promoting inclusion
.79** .88** .42*
.61** .57** .71** .93** .91** .72** .90**
12. DMC (Total)
.89** .91** .52** .77** .75** .86** .93** .90** .86** .89** .93**
13. Advancement
.27
14. Power distribution
-
.44** .79** .73** .22
-
.38*
.45** .54** .49** .57** .34*
.43*
.60** .70** .45** .59** .58** .62** .57**
15. Inclusive climate
.42*
.40*
.57** .72** .70** .51** .63** .67** .72** .64**
16. DM Outcomes (Total)
.44** .65** .74** .74** .34*
.45** .60** .74** .64** .51** .63** .68** .87** .86** .89**
17. Production/efficiency
.43** .48** .37*
.25
.29
.45** .57** .23
18. Adaptability
.57** .68** .47** .41*
.41*
.42*
.53** .67** .45** .52** .61** .62** .33
.48** .54** .52** .72**
19. Capacity
.47** .53** .30
.34*
.26
.35*
.45** .56** .34*
.51** .56** .51** .31
.41*
.38*
.42*
.79** .65**
20. Market standing
.42*
.39*
.18
.19
.35*
.27
.31
.40*
.17
.34*
.36*
.37*
.10
.28
.21
.23
.81** .56** .79**
21. Projected growth
.27
.40*
.32
.32
.36*
.26
.32
.45** .19
.41*
.45** .40*
.29
.35*
.33
.37*
.59** .45** .74** .76**
22. Performance (Total)
.51** .59** .39*
.36*
.36*
.38*
.49** .62** .31
.33*
.43*
-
.36*
.73** .68** .65** .32
.51** .54** .23
.50**
-
.47** .47** .47** .26
.52** .57** .56** .30
-
.48** .36*
.46** .42*
Notes: k = 35. a
Unit of analysis is at the organizational level. Variable scores were created by aggregating scores for all respondents within each company.
** Level of significance = p < .01. *Level of significance = p < .05.
*Corresponding author. Email:
[email protected]
.42*
-
.45** .91** .79** .92** .90** .80**
The International Journal of Human Resource Management
H1
Inclusive climate H2
Diversity management competency
H3
H5
Advancement opportunities
H4
H6
Firm performance
H7
Power distribution
Figure 1. A conceptual model of the relationship between diversity management competency, diversity-related outcomes, and firm performance.
45
