Corps de l’article

Little research explores how diversity management efforts can foster competitive advantages through effective human resource management (Shen et al., 2009), especially for firms that intend “to grow and compete across borders” (Samuel and Odor, 2018, p. 44). Most research in international management focuses on the effects of specific diversity attributes, often classified by visibility. That is, surface-level diversity reflects visible attributes, observable at first sight, whereas deep-level diversity pertains to invisible attributes, such as experience or values, that may be communicated through complex verbal and nonverbal interactions (Shore et al., 2009). Research in turn focuses on how these various attributes (e.g., gender, race, experience) might facilitate entry into various foreign markets (Hagen and Zucchella, 2014; Mohr and Shoobridge, 2011) and expand firms’ internationalization scope (Singh and Point, 2004). In various studies of these outcomes, small and medium enterprises (SMEs) tend to be neglected (De Jong and Van Houten, 2014; Rivas, 2012), even though their unique characteristics likely lead to distinct effects of diversity on their internationalization. Moreover, the focus on specific diversity attributes leads to a lack of consensus regarding the overall effect of diversity; both positive and negative effects have been identified (Jehn et al., 1999; Mannix and Neale, 2005).

In response, we draw on previous research (Jehn and Bezrukova, 2004; Mannix and Neale, 2005) and argue that it is not diversity per se (i.e., attributes) that determines organizational performance but rather its management, especially by SMEs[1]. Because SMEs suffer limited resources, including human ones (Aldrich and Auster, 1986; Knight et al., 2004), they might be subject to more negative or neutral effects of diversity on internationalization, which rarely have been addressed in prior literature (Mannix and Neale, 2005; Williams and O’Reilly, 1998). In such contexts, good management is critical to SMEs (Tansky and Heneman, 2003) and even might compensate for a lack of (human) resources, if it involves the implementation of effective practices.

According to Samuel and Odor (2018), implementations of diversity management vary from firm to firm, reflecting the different perspectives identified in prior literature. Dass and Parker (1999) propose a typology of four perspectives on diversity management, from total ignorance to uses of diversity as a source of value creation. By leveraging resource-based theory (RBT) (Barney, 1991) and resource orchestration theory (Chadwick et al., 2015; Sirmon et al., 2011), we propose a novel view of these perspectives, with which we can identify associated management practices and their effects. From this perspective, diversity is a resource that must be managed (Cox and Blake, 1991; Richard, 2000), beyond its sole attributes. That is, we still account for diversity attributes, but as part of a global approach that also includes diversity management perspectives and the associated diversity management practices. With this theoretical foundation, we aim to identify the extent to which different perspectives (Dass and Parker, 1999) affect the scope of SMEs’ internationalization, according to their implementation of distinct practices. Thus, as our central research question, we ask, what effects do distinct perspectives on diversity management exert on SMEs’ internationalization?

To test our proposed framework empirically, we use data from a Luxemburgish survey of managerial practices, as well as official diversity data. The sample includes 1,348 Luxemburgish SMEs. According to a classification procedure, we identify practices associated with each diversity management perspective, then use an ordered probit model to measure the effect of these perspectives, based on various practices, on SMEs’ scope of internationalization.

Along with the theoretical contributions to both diversity management and internationalization literature, this study offers some useful recommendations for SME managers, public authorities, and international support services.

Theoretical Framework

The management of diversity and related perspectives

Previous research that has attempted to understand how diversity management is perceived, implemented, and defined among organizations identifies various perspectives (Dass and Parker, 1999; Singh and Point, 2004; Thomas and Ely, 1996) that likely have distinct effects on firms’ performance (Podsiadlowski et al., 2013). Among these research, only a few focus on internationalization. Dass and Parker’s (1999) typology includes four perspectives on diversity management, each of which can spark a different organizational response (Singh and Point, 2004). It spans all known perspectives on diversity management, including strategic and more pessimistic views that still tend to be neglected:

  1. With a resistance perspective, diversity is a non-issue or threat, so more homogeneity tends to result. This perspective fosters a reactive response to diversity questions, characterized by denial, avoidance, defiance, or manipulation (Singh and Point, 2004).

  2. A discrimination-and-fairness perspective acknowledges diversity as a cause of problems, notably for members of minority groups who are at risk of discrimination. To protect them, a defensive strategy is required, as provided by equal opportunity measures.

  3. An access-and-legitimacy perspective indicates that diversity creates opportunities, so the focus is on acknowledging and celebrating all types of diversity. In firms with this perspective, managing a diverse workforce familiarizes them with the demographic characteristics of their different markets (Lorbiecki, 2011; Thomas and Ely, 1996), such that they favor individual development and respect for differences, and the strategic response is accommodative (Singh and Point, 2004).

  4. The learning perspective recognizes that diversity offers opportunities but also imposes costs. In comparison with the access-and-legitimacy perspective, the logic behind the learning perspective emphasizes organizational learning with a long-term orientation. The knowledge and competencies introduced by diversity can be shared among firms, with a real goal of learning from every employee in their pursuit of multiple, concrete objectives, such as efficiency, innovation, and customer satisfaction (Dass and Parker, 1999). The strategic response is proactive, so managing diversity allows the organization to “internalize differences among employees so that it learns and grows because of them” (Thomas and Ely, 1996, p. 86).

Managing or orchestrating diversity

Nkomo et al. (2019, p. 502) cite the RBT as one of the “dominant theoretical lenses for explaining the relationship between diversity and organizational performance.” This theory prioritizes resources, defined as “assets, capabilities, organizational processes, firm attributes, information, knowledge, etc. controlled by a firm that enables the firm to conceive of and implement strategies that improve its efficiency and effectiveness” (Barney, 1991, p. 101). It assigns the task of identifying and making the best use of the resources to managers of each firm, to sustain firm performance and create competitive advantage (Colbert, 2004). This argument clearly applies to diversity management (Jehn and Bezrukova, 2004; Mannix and Neale, 2005). Singh and Point (2004, p. 298) assert that Dass and Parker’s (1999) typology reflects the RBT, especially the learning perspective, in which “the value of people, human capital, can be enhanced by cultural and demographic diversity to provide sustainable and non-imitable advantage.” Similarly, Richard (2000) argues that diversity is a strategic resource that should help the firm access additional experiences, knowledge, and skills. The value-in-diversity (vs. inevitability) hypothesis similarly predicts that the value of human capital is enhanced by diversity (Cox and Blake, 1991).

Even if we accept the RBT as an appropriate theoretical framework to consider diversity as a resource that needs to be managed, the substance of this management, and the concrete practices underlying different perspectives, requires further consideration (Janssens and Steyaert, 2019; Shen et al., 2009). The conventional RBT fails to address concrete actions; an extension, the resource orchestration theory, explicitly focuses on “the role of managers’ actions in effectively structuring, bundling, and leveraging firm resources” (Barney et al., 2011, p. 1306). The concept of resource orchestration addresses resource-related processes and actions precisely, by highlighting how resources get managed or orchestrated across the firm, at various stages of maturity and levels (Barney et al., 2011; Sirmon et al., 2011). According to Sirmon et al. (2007), those processes and actions can be classified as structuring resources (acquiring, accumulating, and divesting), bundling resources (stabilizing, enriching, and pioneering), and leveraging resources (mobilizing, coordinating, and deploying). Using resource orchestration to exploit firm-specific resources, including diverse human resources (by structuring, bundling, and leveraging them), means that firms implement specific management practices (Chadwick et al., 2015).

Previous diversity management literature lists multiple diversity management practices (e.g., mandatory or voluntary diversity training, recruitment tests, quotas, mentoring, self-managed teams, diversity task forces, positive discrimination during recruitment; Dobbin and Kalev, 2016). In addition to management practices focused on diversity, Podsiadlowski et al. (2013, p. 161) suggest that “diversity measures may also be embedded within existing tools for training, personnel development, formalized recruitment, selection and assessment, mentoring, or coaching without being specifically identified.” However, these contributions do not allow us to link each practice identified in prior literature precisely with a specific perspective on diversity management (Shen et al., 2009).

From a resource orchestration lens, Dass and Parker’s (1999) four perspectives relate to specific bundles of practices (Sirmon et al., 2011). Because it denies diversity issues, a resistance perspective leads to the total absence of practices. A discrimination perspective instead would integrate positive discrimination practices, equal opportunities, and affirmative action programs, targeted at helping minority groups. The access-and-legitimacy perspective would integrate all types of diversity, whether legally protected or not, and aim to develop individual potential by addressing additional factors, such as the family situation or work–life balance. Finally, a learning perspective would prompt an organizational approach with embedded practices designed to foster organizational learning by members. We hypothesize:

Hypothesis 1. The four perspectives on diversity management prompt the implementation of distinct sets of managerial practices.

Orchestrate diversity for internationalization

The effect of diversity on firms’ internationalization is a topic of great interest in international management literature. De Jong and van Houten (2014) argue for a positive effect of cultural diversity on the internationalization–performance relationship, and Rivas (2012) confirms a positive effect of functional diversity among boards and top management teams on internationalization. Most studies exclude SMEs though, despite their participation in economic and social globalization (Mohr and Shoobridge, 2011), suggesting the ongoing need to address diversity in SMEs (Rivas, 2012). Among the few available studies, Mohr and Shoobridge (2011) link ethnic diversity to increased internationalization by SMEs; Parrotta et al. (2016) also demonstrate empirically that ethnic diversity encourages internationalization. Khan and Lew (2018) indicate that human resource diversity may benefit internationalization and the firm’s chances of survival. However, they address diversity without accounting for how it is managed (Breuillot, 2021).

Diversity management should enable SMEs to orchestrate their limited resource base, even if they lack the means to recruit new team members with relevant experiences, knowledge, and skills (Kumar, 2012). Thus, they can overcome liabilities of newness (Stinchcombe and March, 1965), smallness (Aldrich and Auster, 1986), or foreignness (Zaheer, 1995). Previous empirical research (Zahra and George, 2002) indicates that internationalization can be analyzed according to three dimensions: speed, scope, and extent. We propose studying the scope of internationalization in relation to diversity management. First, previous research acknowledges a positive effect of some diversity attributes for reaching certain foreign markets (Mohr and Shoobridge, 2011; Singh and Point, 2004), such that they might expand firms’ internationalization scope (Hagen and Zucchella, 2014). Second, for SMEs, an international scope offers some critical benefits (Sapienza et al., 2006), because the “successful pursuit of international scope has the potential to produce … economies of scale, greater returns on investments, and an improved competitive stance” (Dai et al. 2014, p. 511). Therefore, we focus on the potential effect of diversity management on SMEs’ scope of internationalization.

As previously explained, we argue that resource orchestration theory (Sirmon et al., 2011) supports a consideration of the four perspectives of diversity management (Dass and Parker, 1999) as distinctive ways to orchestrate resource diversity, through the implementation of distinct sets of managerial practices. Following Podsiadlowski et al.’s (2013) argument that different perspectives on diversity management have distinct effects on organizational performance, we hypothesize:

Hypothesis 2. The four perspectives of diversity management have differentiated effects on the scope of SMEs’ internationalization.

Empirical methodology

Context and data

This study was conducted in Luxembourg, which offers interesting features for examining the relationship between diversity and SMEs’ internationalization. First, its economy is dominated by SMEs, which account for 66.6% of employment and nearly 68% of value added (17% above the EU average).[2] The small size of its domestic market also makes internationalization a pivotal determinant of SMEs’ growth (Vermeulen and Barkema, 2002). In 2013, Luxembourg was among the top six countries in terms of international SMEs; the proportion of manufacturing SMEs participating in international trade also was notably high (greater than 3 times the EU average for exporting SMEs). This position has remained steady from 2008 to 2019.[3]

Second, diversity is prominent in Luxembourg. Its population consisted of 44.5% foreigners in 2013,[4] a higher level of national origin diversity than any other European country. The labor market is powerfully dominated by foreigners (71.3%), 62.1% of whom are cross-border workers. Considering the age diversity, the “young-to-old” ratio shows that the share of young workers (15–34 years) was 1.6 times higher than older workers (50+ years) in 2013.[5] From 13.7% in 2003 to 23% in 2013, the employment rate among the older workers (60–64 years) has increased though, with similar patterns for workers 55–59 years (54.7%, +10%) and 50–54 years (79.5%, +11%) of age. Regarding gender diversity, since 1995, Luxembourg is the only country in Europe to have a Ministry for Gender Equality. It ranked above the European average (58.6%) in its female employment rate of 59.1% in 2013.

Third, considering Luxembourg’s general population and labor market demographics, diversity appears inevitable for Luxemburgish SMEs (Cox and Blake, 1991, p. 45), implying that their “competitiveness is a priori affected by the need (because of national and cross-national workforce demographic trends) to hire more women, minorities, and foreign nationals.” It creates an interesting setting for evaluating the effects of diversity management on SMEs’ internationalization.

To justify this research focus on the Luxemburgish context, we also provide a synthetic introduction to its legal framework pertaining to diversity.[6] In 2006, following European Commission directives, it enacted a law promoting equal treatment and condemning all forms of discrimination, but most notably in workplaces. Then a new title, related to equal treatment in matters of employment and work, was introduced into the Labor Code. The law also allowed for positive discrimination measures, though they remain optional and at the discretion of the organization, as does adoption of the “diversity charter” created in 2012.[7] Since we ran our survey in 2013, two provisions have been introduced. The first, a modification of the labor code in 2016 to guarantee equal pay between men and women, falls outside the scope of our study, because we do not consider salaries. The second provision dates from 2017 and modifies the 2006 law by adding a prohibition of discrimination based on nationality.

The survey that informs this study was carried out by the Luxembourg Institute of Socio-Economic Research (LISER) in 2013, designed to gather information on corporate social responsibility (CSR) strategies, managerial practices, and the characteristics and outcomes of Luxemburgish companies, including their internationalization. This survey has informed previous studies of CSR strategies, human resource diversity, and their effects on firm innovation and performance (e.g., Bocquet et al., 2017, 2019). But the data have never been used to address the specific issue of diversity management as it pertains to SME internationalization. This survey included 2,819 firms, according to a stratified sampling technique based on firm size and sector. The sample is restricted to SMEs with 10–250 employees, in line with the European definition,[8] representing 1,348 firms. For analysis purposes, we merged the survey data with administrative data from the social security administration to obtain detailed information about the diversity of the employees (age, gender, and nationality) of each SME and control for diversity attributes. Finally, we used a weighting procedure, based on the inverse of the response rate per stratum, to ensure representative results for the target SME population.

Variable definitions

Dependent variable. According to a definition by Lin (2012), internationalization scope (Inter_scope) indicates the concentration and geographical diversification of SMEs’ activities. For Inter_scope, we consider three outcomes: Firms (1) remain in their domestic market (Luxembourg), (2) penetrate other European markets, or (3) transcend European borders. According to gradualist models and stages theories of internationalization, traditional SMEs (Dominguez and Mayrhofer, 2017) often choose neighboring countries with short psychic distances (Johanson and Vahlne, 2009). In contrast, early internationalizing firms (EIFs) internationalize soon after their inception and “seek to derive significant competitive advantage from the use of resources and the sale of outputs in multiple countries” (Oviatt and McDougall, 1994, p. 49). Thus, the international scope of EIFs tends to be wider and more dispersed than that of traditional SMEs, often transcending European borders (Trudgen and Freeman, 2014).

Independent variables. To capture the four perspectives (resistance, discrimination, access-and-legitimacy, learning perspectives), and in accordance with our theoretical framework, we identify 12 managerial practices that reflect these four perspectives and that were implemented by SMEs in 2012. To account for a wide span of practices associated with orchestrating resource diversity, we include both diversity management practices (Dobbin and Kalev, 2016) and general management practices (Podsiadlowski et al., 2013) that have been identified in prior diversity management literature but never concretely associated with a specific diversity management perspective.

The two dedicated diversity management practices we address are a mentoring system and positive discrimination during recruitment (Dobbin and Kalev, 2016). We go beyond traditional frameworks of diversity management programs and include practices implemented through other management tools to manage or orchestrate diversity. Following Podsiadlowski et al. (2013), we include worker training practices, such as the implementation of various training, providing a choice of which training to follow to employees, and implementation of skills development incentives. Dass and Parker (1999) insist on the importance of family relationships and work–life balance, so we also consider two related practices: scheduling flexibility and extra-legal family benefits. The prior practices mainly focus on individual potential, but other options involve organizational learning and communication. Diversity opponents assert that it can hinder social integration and increase conflict (Williams and O’Reilly, 1998). To overcome such concerns, practices such as brainstorming and creating discussion spaces are helpful (Mannix and Neale, 2005). Finally, performance evaluation is an important component of diversity management (Williams and Mavin, 2012); we investigate two individual-level practices, individual recognition system and individual performance bonuses, and one organizational-level practice, team performance bonuses (Yang and Konrad, 2011).

By applying k-means clustering[9] to these 12 managerial practices (Table 1), we identify four clusters of SMEs; we present the characterizations in the Appendix. As expected, the different groups of SMEs are characterized by different managerial practices, reflecting the four perspectives on diversity management. Hypothesis 1 thus receives support.

Cluster 1 aligns with the discrimination perspective and contains SMEs that have exclusively implemented positive discrimination practices in hiring, assigning priority to candidates from underrepresented groups, such as older workers. Cluster 2 corresponds to the access-and-legitimacy perspective. These firms offer training to encourage employees to develop their skills and careers. Due to their focus on individual potential, they give employees a choice of training. They also rely on discriminatory hiring practices and extra-legal benefits, including days off for family reasons, in line with the access-and-legitimacy emphasis on inclusion. In Cluster 3, the SMEs intensively adopt all noted managerial practices except positive discrimination. This intensive, broad adoption implies a learning perspective, which embraces an organizational view (vs. individual in the access-and-legitimacy perspective) of diversity management. Finally, SMEs in Cluster 4 are poor adopters, always below the average of each managerial practice, reflecting a resistance perspective.

We dichotomize the classification variable to produce 4 dummies: discri_persp (= 1 if the SME belongs to Cluster 1, 0 otherwise), acces_persp (= 1 if the SME belongs to Cluster 2, 0 otherwise), learning_pers (= 1 if the SME belongs to Cluster 3, 0 otherwise), and resist_persp (= 1 if the SME belongs to Cluster 4, 0 otherwise). The last variable provides the reference category for the empirical model.

Table 1

Variable definitions, clustering

Variable definitions, clustering

-> Voir la liste des tableaux

Control variables. We include traditional antecedents of SMEs’ internationalization as control variables. Noting the potentially nuanced effects of various diversity attributes (Stahl et al., 2010), we control for them and take a multidimensional approach toward three attributes. Specifically, we include gender, age, and nationality diversity, due to their ability to represent both surface- and deep-level diversity (Eagly and Chin, 2010). Similar to previous studies (Bocquet et al., 2019; Richard, 2000); we use the Blau (1977) index[10] to measure distributions of gender, age, and nationality among each SME’s workforce in 2012. We thus introduce three measures, Hgender, Hage, and Hnationality. The highest possible value of the Blau index depends on the number of groups in the population, so the maximum value for gender diversity is .5 (i.e., equal proportions of women and men). For age diversity, we consider nine groups (24 years and younger, 25–29, 30–34, 35–39, 40–44, 45–49, 50–54, 55–59, and 60 years or older), and the highest value is .89. For nationality, the maximum value of the Blau index is .86, because our sample features seven nationalities (Luxembourgish, German, French, Belgium, Portuguese, Italian, and other[11]).

We control for firm age, because older firms have had more opportunities to internationalize and grasp the process (Johanson and Vahlne, 2009); they typically have more resources too (Fernhaber and Li, 2013). However, younger firms might possess the flexibility and learning capabilities needed to adapt to distant foreign markets (Dai et al., 2014). Moreover, the firm’s age at its initial internationalization might influence its performance and degree of internationalization (Knight et al., 2004). We therefore differentiate SMEs according to three dummies: Age_4 captures very young SMEs (including early internationalizing firms, created less than four years ago; Age_5-9 are SMEs created between five and nine years ago; and SMEs older than 10 years (reference category) are denoted Age_10P. Similar to prior studies of SMEs internationalization, we use Indus as a dummy variable, controlling for the manufacturing sector (vs. service sector) (Dai et al., 2014; Fernhaber and Li, 2013). Ownership also informs SMEs’ internationalization process (Morais and Ferreira, 2020), so we determine whether SMEs belong to a group and include a group dummy. Belonging to a group enhances the probability of internationalization, especially to more distant markets, because these SMEs benefit from additional resources (Caldera, 2010). Technological innovation is an important explanatory factor that allows SMEs to adapt to the requirements of international markets (Ramos et al., 2011). The R&D dummy equals 1 if the SME made R&D expenditures between 2010 and 2012 to develop new products or technologies. Finally, we account for the difficulties SMEs face in recruiting qualified employees (recruit_barriers), because a lack of new skills can severely constrain their growth abroad (Onkelinx et al., 2016). Table 2 summarizes the variables in the ordered probit model.

Econometric procedure

To test the effects of diversity management perspectives on SMEs’ internationalization scope, we use an ordered probit model, for two main reasons. First, a binary logit or probit model, in which Y = 1 for international scope and Y = 0 for domestic scope, would be too rudimentary to evaluate SMEs’ internationalization fully. Many European SMEs engage in international business, but few transcend European borders (Knight et al., 2004). The proximity of European countries, psychologically (Johanson and Vahlne, 2009) and geographically (Fernhaber and Li, 2013), as well as the important size of the European market (Knight et al., 2004), motivates many European firms to stay in Europe. In addition to psychic distance, which is an important determinant of firms’ international behavior (Dominguez and Mayrhofer, 2017), the resources needed to transcend European borders differ from those required to export within Europe (Kumar, 2012). In this sense, Europe represent a sort of intermediate threshold that should be taken into account when studying European firms’ internationalization scope (Knight et al., 2004). Second, multinomial logit or probit models allow for more than two categories, but they suffer from the well-known independence of irrelevant alternatives assumption (Greene, 2003), in that the errors are assumed to be independent for each category. To overcome this problem, an ordered probit model[12] accounts for the ordinal nature of the dependent variable (Greene, 2003). Recall that our dependent variable, the scope of SMEs’ internationalization (Inter_scope), has three outcomes, ranging from 1 (firms stay in the domestic market) to 3 (firms transcend European borders).

Table 2

Variable definitions, ordered probit model

Variable definitions, ordered probit model

-> Voir la liste des tableaux

Results

Table 3 presents the results of the ordered probit model. The model achieves good fit and correctly predicts approximately 71% of the cases. The mean variance inflation factor (VIF) score is 1.14, and the highest VIF, for the category Learning_persp, is 1.53,[13] well below the conservative threshold of 4 (Hair et al., 2010).

We start with the effect of diversity management—that is, the effects of different ways to orchestrate resource diversity. We hypothesize that SMEs that adopt distinct diversity management perspectives (Dass and Parker, 1999) do not achieve the same international scope. Our results confirm this prediction when we use the resistance perspective, which represents the lowest level of engagement, as a reference. First, the discrimination perspective coefficient is non-significant, so implementing affirmative action or positive discrimination practices does not affect the scope of SMEs’ internationalization. This result makes sense; this perspective does not entail a search for competitive advantage using diversity (Dass and Parker, 1999). Second, SMEs with an access-and-legitimacy perspective implement certain managerial practices (skills development incentives, extra-legal family benefits, individual performance bonuses) to help individual members feel good at work and perform better; they also exhibit a greater likelihood of being internationalized (p <.10), though mostly in Europe. Third, the coefficient for the learning perspective variable is significant and positive at 1% (p <.01), indicating a crucial effect of adopting various diversity management practices, especially general management ones (training employees, creation of discussion spaces, brainstorming, team performance bonuses), to foster the scope of internationalization. The coefficient is significant for SMEs that internationalize both within Europe and beyond European borders. Hypothesis 2 thus receives support.

Among the control variables and antecedents of internationalization, we expected nuanced effects of various attributes (Horwitz and Horwitz, 2007). Age diversity (p <.01) has a significant positive effect on the scope of SMEs’ internationalization. In our theoretical framework, age diversity can be classified as a surface-level attribute, but it also imposes psychological effects at deeper levels (Eagly and Chin, 2010), such that it aligns with experiential or knowledge diversity, which also have positive effects on firms’ internationalization (Kumar, 2012). Our study identifies a positive effect of gender diversity on SMEs’ scope of internationalization (p <.05). However, SMEs that feature diverse nationalities exhibit a lower probability of reaching distant markets (p <.05). This result contradicts some previous findings (Parrotta et al., 2016) but is consistent with international management research that identifies a negative effect of cultural origin diversity on team outcomes (Stahl et al., 2010). Borrowing from previous research (Richard, 2000), we posit that the negative effect reflects the national context: Luxemburg’s general population and labor market encompass many nationalities, which may be too extensive, relative to the size of SMEs, to create a positive effect. Moreover, every firm is similarly equipped with diverse nationalities.

When we control for firm age, we find a positive effect for very young firms (p <.05), including early internationalizing firms, with regard to internationalizing in Europe but not transcending European barriers. This result seems logical, considering the size of the European market (Knight et al., 2004) and the time usually needed to obtain sufficient market resources (Kumar, 2012) to overcome the barriers of psychic distance (Johanson and Vahlne, 2009). Firms created between five and nine years ago similarly exhibit a positive effect for the scope of internationalization; this finding refers to both Europe (p <.05) and beyond (p <.10). We find positive, significant effects of operating in the manufacturing sector (p <.01), belonging to a group (p <.01), and undertaking R&D activity (p <.01) on SMEs’ internationalization, in Europe and beyond. In a rather obvious link, firms that struggle to recruit qualified workers are less likely internationalize (p <.01).

Discussion

Distinctive effects of various diversity management perspectives

We investigate the effect of diversity management on the internationalization of SMEs. Noting the variations among firms in terms of managing diversity (Samuel and Odor, 2018), we combine the RBT (Barney, 1991) and resource orchestration theory (Sirmon et al., 2011) to consider four perspectives on diversity management as distinctive ways to orchestrate resource diversity. The four perspectives comprise different bundles of practices, which have different effects on SMEs’ scope of internationalization, as we show. This evidence extends Podsiadlowski et al.’s (2013) proposition that different perspectives on diversity management have distinct effects on organizational performance, to include internationalization.

Table 3

Relationship of diversity and SMEs’ internationalization (ordered probit regression)

Relationship of diversity and SMEs’ internationalization (ordered probit regression)

*Statistically significant at the .10 level; **.05 level; ***.01 level.

Source: Survey on Organizational and Managerial Practices in Luxembourg Enterprises (2013), LISER

-> Voir la liste des tableaux

As we have outlined, a resistance perspective consists of refusing to orchestrate resource diversity, without any attempts to structure, bundle, or leverage it. These firms simply are not convinced by value-in-diversity arguments and only respect the minimal legal requirements. The discrimination perspective reflects a fearful orchestration, such that firms just want to make sure that minority groups are protected (Dass and Parker, 1999). Positive discrimination practices are implemented, but this perspective does not extend to bundling or leveraging resource diversity. At most, human resource diversity is structured, and only in terms of equal opportunity measures beyond the legal requirements, leading to a non-significant effect on SMEs’ internationalization.

The latter perspective focuses on practices designed by diversity management programs (Dobbin and Kalev, 2016; Yang and Konrad, 2011). The access-and-legitimacy perspective instead spans general management practices (Podsiadlowski et al., 2013). Firms structure and bundle their human resource diversity by implementing practices to enrich them (e.g., skills development incentives, training choices), though only at the individual level (e.g., extra-legal family benefits, individual performance bonuses). Therefore, they stop short of the needed coordination of resource diversity at the organizational level. The result is a limited effect on SMEs’ internationalization, circumscribed to Europe.

Finally, in addition to structuring and bundling, firms adopting a learning perspective leverage their resource diversity by adding general management practices at the organizational level (e.g., creating discussion spaces, brainstorming, mentoring). These firms can enjoy the positive effects of diversity through their organizational learning (Thomas and Ely, 1996) and foster their internationalization scope more effectively than ones that adopt the access-and-legitimacy perspective. They are expanding internationally beyond European borders.

A learning perspective to orchestrate diversity

The results indicate significant positive effects of a learning perspective for SMEs’ internationalization. Following resource orchestration theory (Sirmon et al., 2007), this positive effect arises because the learning perspective is the only one to adopt structuring, deploying, and leveraging practices. The wish to create value for customers and leverage competitive advantages from specific resources is fundamental to firms (Sirmon et al., 2007). In that regard, a learning perspective reflects the value-in-diversity hypothesis (Cox and Blake, 1991), namely, the belief that diversity can create value and competitive advantages (Dass and Parker, 1999) and entails a strategic response (Singh and Point, 2004).

This perspective mainly involves general management practices located both at the individual and organizational levels. In line with Richard and Johnson’s (2001) recommendation, it can move beyond the framework of diversity programs and include general management notions (Podsiadlowski et al., 2013). After structuring the resource (acquiring it), SMEs implement practices to support every individual member’s potential to thrive, and then complement them with organizational practices to mobilize, coordinate, and deploy diversity in a way that fosters organizational learning (Thomas and Ely, 1996).

Attributes or management of diversity? A global approach

Previous studies mainly focus on the attributes of diversity per se, which lead to a lack of consensus regarding the effect of diversity on firms’ performance (Jehn et al., 1999; Mannix and Neale, 2005). To overcome this impasse, we argue for a global approach when researching diversity effects on firms’ performance. Attributes of diversity, diversity management, and related practices can all affect organizational performance, such as internationalization, and they need to be considered together. Gilbert and colleagues (1999) refer to diversity management as a new organizational paradigm that embraces the value of diversity. We believe this organizational paradigm should consider diversity as a whole, spanning all types of attributes and the way they are managed and orchestrated, using concrete practices (Dobbin and Kalev, 2016; Podsiadlowski et al., 2013; Shen et al., 2009). This recommendation does not mean that we deny the effect of diversity attributes; they must be considered in their variety. However, we argue that they cannot intrinsically be a source of performance improvement or deterioration by themselves and that management is key (Jehn et al., 1999; Mannix and Neale, 2005). Such an approach also requires the consideration of diversity at different levels (Nkomo et al., 2019). Most studies prioritize the individual or managerial level, such as by examining diversity in the top management team (Rivas, 2012), or else investigating team diversity outcomes (Tekleab and Quigley, 2014). With this study, we observe management practices at the organizational level, among all workers.

We defend the idea that diversity management matters (Jehn et al., 1999), beyond the effects of diversity attributes. With this approach, it is possible to go beyond the distinction of surface- versus deep-level diversity (Nkomo et al., 2019) and capture the reality of individual identities. Dennissen et al. (2020) argue that to capture multiple attributes of diversity, or intersectionality, diversity management is paramount. As Bleijenbergh et al. (2010) explain, diversity management practices can be the heart of personnel management, and Shen et al. (2009) defend the integration of diversity management into human resource management systems. We concur that diversity management and its associated practices can overcome simple frameworks of diversity management programs (Podsiadlowski et al., 2013).

Conclusion

To determine effective diversity management for SMEs’ internationalization, we take a more detailed view on the four ways to manage diversity (Dass and Parker, 1999; Singh and Point, 2004) and thereby reveal the substance underlying these four perspectives. In turn, we make three main theoretical contributions to diversity and international management literature. First, this study confirms quantitatively that diversity management matters; different perspectives on diversity lead to distinct organizational outcomes (Podsiadlowski et al., 2013), such as with regard to SMEs’ internationalization. Managerial practices can be linked to these four perspectives and analyzed according to resource orchestration theory, which supports a better understanding of how to mobilize diversity as a resource (Barney et al., 2011; Chadwick et al., 2015) and deeper analyses of the differentiated effects of managerial practices (Sirmon et al., 2011) on SMEs’ scope of internationalization. Second, we identify the learning perspective of diversity management as the most effective strategy for managing diversity to foster internationalization across European borders. In more detail, we establish a list of concrete practices that can structure, bundle, and leverage resource diversity. We confirm the importance of general management practices at both the organizational and individual levels to simultaneously foster organizational learning and develop the potential of each employee. Third, this study contributes to discussions of diversity as an organizational paradigm (Gilbert et al., 1999) and recommends a global approach to study the effect of diversity on organizational performance, considering diversity management various perspectives, their distinct management practices, and various diversity attributes (Nkomo et al., 2019).

Several recommendations for SME managers and international support services also emerge from this research. Our results indicate the need to prioritize the learning perspective of diversity management to foster SMEs’ internationalization and, more specifically, its scope. A learning perspective logic entails organizational learning. It leads to a proactive strategic response (Singh and Point, 2004), which results in the implementation of the various managerial practices we have identified herein. In detail, adopting a learning perspective leads to the implementation of 11 managerial practices, out of the 12 practices we consider (positive discrimination during recruitment being the one excluded). These 11 practices refer to employees’ training and skills development (i.e., skills development incentives, fostering employee training), means to communicate within the firm (i.e., mentoring programs, internal discussion spaces, brainstorming sessions), recognition systems (i.e., individual recognition system, along with individual and team performance bonuses), and employee empowerment (i.e., employees choose their training, scheduling flexibility, extra-legal family benefits).

This insight leads us to note several limitations that suggest the need for further research, the first of which relates to the context. Our results are specific to the Luxemburgish context, which represents an “extreme” case of diversity in terms of nationalities. Although it offers a stimulating study context, further studies are needed to clarify the effects in other national and institutional contexts. Second, the sample presents some limitations. We have no information about the age of SMEs, which is not included in Luxembourgish databases, due to changes in the legal status of companies over time. Nor is the date of the first international entry available. Both these pieces of information could support the identification of a subsample of early internationalizing firms (Rialp et al., 2005) and thus a more nuanced exploration of the effect of diversity management among distinct types of internationalizing SMEs. Third, our measure of international scope is simplistic yet efficient, delineating a European threshold for European SMEs. Further research might attempt to add the two other dimensions of internationalization, speed and scale. Fourth, our study does not depict the evolution of the four perspectives or interactions among them. An interesting research avenue could be to develop a processual approach to diversity management, to observe the potential transition from one perspective to another over time. Such an approach might consider the perspectives as a continuum instead of independent elements in a typology. Fifth, by conducting this study at the organizational level, we cannot control for the characteristics of the owner, which likely affect adopted diversity management practices. As Bebbington and Özbilgin (2013) explain, diversity at the leadership level influences diversity interventions; leaders usually must be diverse to support diversity at the organizational level. Sixth, it seemed relevant to study the effect of the four diversity management perspectives at the organizational level, but studying their effects on a sample of employees, accounting for their diversity attributes, also could provide greater precision with regard to the types of practices to implement, in accordance with the diversity attributes exhibited within the firm.