‘Nationality diversity in Top Management
Teams: The effect of regional career models
and CEO non-nativity’
Floor Botden
10415777
Master Thesis
1 July 2014
MSc Business Studies - International Management
Supervisor: Dhr. dr. Niccolò Pisani
1
Abstract
While top management career models are highly embedded in national institutional contexts, the phenomenon of globalization forces multinational corporations to face different institutional contexts when internationalizing their organizations. Nevertheless, the vast majority of executives forming top management teams in multinationals are national (versus international). Aiming to explain the differences in nationality diversity in Top Management Teams (TMTs) across regions, this study discusses the institutional embeddedness of career models across North America, Latin America, Europe, and Asia and their effect on the extent of TMT nationality diversity. Moreover, this study aims at testing the moderating effect of CEO non-nativity on the association between regional career models and TMT nationality diversity. Four regional career models are proposed and tested on the TMTs of the Fortune Global 500 companies of 2013. The results obtained show clear differences in levels of TMT nationality diversity across regions and thus confirm the relevant role of regional institutional factors in shaping career models and affecting the extent of TMT nationality diversity.
Key words: Internationalization; Nationality diversity; Top management teams (TMTs);
2
Acknowledgements
I gratefully and sincerely acknowledge the excellent guidance and supervison by dhr. dr. Niccolò Pisani, Assistant Professor of International Management at the University of Amsterdam Business School. His always extensive feedback and great aspiration and dedication has pushed me to keep track and to continuously improve my writing and research skills.
3
Table of content
1. Introduction ... 4
2. Literature review ... 6
2.1 Institutional theory ... 6
2.2 Career patterns and models ... 8
2.3 TMTs nationality diversity ... 11
3. Theoretical framework ... 16
3.1 Home region institutional factors and career models ... 16
3.2 The Asian career model ... 17
3.3 The Latin American career model ... 19
3.4 The North American career model ... 22
3.5 The European career model ... 24
3.6 CEO non-nativity ... 27 4. Research methods ... 29 4.1 Data collection ... 29 4.2 Sample ... 29 4.3 Independent variable ... 30 4.4 Dependent variables ... 31 4.5 Moderating variable ... 32 4.6 Control variables... 33
5. Analysis and results ... 35
5.1 Descriptive statistics ... 38
5.2 Regional career models and TMT nationality diversity ... 39
5.3 Regional career models and TMT regional diversity ... 41
5.4 The moderating effect of CEO non-nativity ... 43
6. Discussion ... 46
6.1 Academic relevance ... 49
6.2 Managerial implications ... 50
6.3 Limitations and suggestions ... 51
7. Conclusion ... 52
4
1. Introduction
In recent years, various factors like corporate scandals and increasing feminism have called for more diversity in top management positions (Singh, Vinnicombe, & Johnson, 2001). Despite these calls, top management team members largely remain white, male, and middle aged (Singh et. al., 2001). However, with globalization intensifying and executive search crossing national borders, the number of non-natives on top management teams of multinational corporations is firmly increasing which allows for more nationality diverse TMTs (Staples, 2007). This is an important matter as with globalization pressures, it is the ability to anticipate and respond to internal and external pressures and opportunities that affects a firm’s competitiveness (Nielsen, 2009). In particular, this applies to multinational corporations (MNCs) that are regularly confronted with complex environments and thus rely highly on the ability of their Top Management Teams (TMTs) to adjust their strategic decisions to these opportunities and pressures for change. The composition of the TMT is therefore a subject of high importance in understanding a firm’s competitiveness in rapidly changing business environments. Although nationality diversity in TMTs is increasing, TMTs of large MNCs still show limited nationality diversity (Van Veen & Marsman, 2008). Globalization presses many companies all over the world towards internationalization and thus requires geographic mobility from executives (Elvira & Davila, 2005). Nevertheless, the vast majority of the executives appear to remain active in TMTs of firms based in their home country.
Having said that, when considering the composition of TMTs across regions, clear differences emerge. This sheds an interesting light on the varieties of composition of TMTs across regions and raises questions on the extent to which organizations are open to nationality diverse TMTs and on the institutional forces influencing this. In order to provide a better understanding of such issues, this study aims to answer the following research question: ‘How do regional career models impact the extent of TMT nationality diversity across regions?’ The
5
foundation of an answer to this question is found in Institutional theory. From an institutional perspective as proposed by North (1991), such difference in TMT compositions arise because of the different institutional environments in which MNCs are originated. Formal and informal rules of the game such as culture, customs, ethics, traditions, codes of conduct and legal regulations that are rooted in society are adopted by both individuals and corporations. Different beliefs, values, social structures, economic and labour force policies across regions define the traits of a legitimate management career and subsequently influence the way both managers shape their careers to a top management positions and organizations select their executives. The path to a top management position is therefore argued to be embedded in different contextual layers which depend on various contexts of origin, society and culture (Mayrhofer, Meyer & Steyrer, 2007). A clear understanding and comparison of different career patterns to top management positions across world regions could therefore lead to a better prediction and understanding of the openness of corporations around the world towards a nationality diverse TMT. By first investigating these dominant institutionally embedded career models across the North American, Latin American, European, and Asian region, five hypotheses are proposed and then tested on the TMTs of the Fortune Global 500 companies as listed in the 2013 ranking.
The following chapter will provide a discussion of previous research and findings relative to this topic. Subsequently, the institutional factors influencing regional career models are discussed in the chapter relative to the theoretical framework in order to provide an overview of the legitimized career models across North America, Latin America, Europe and Asia. In this chapter hypotheses will be proposed and discussed. The following chapter will discuss the research methods used in the analysis. Thereafter, the results from the statistical analysis will be presented in the analysis and results chapter. The interpretation, implications and academic relevance of the results will be discussed in the chapter relative to discussion and the last chapter will include a conclusion to this research.
6
2. Literature review
2.1 Institutional theory
In discussing differences in operations and organizations of firms across regions, institutional theory serves as one of the most fundamental theories. This theory suggests that such heterogeneities exist because of different institutions across the world, which can be defined as the formal and informal “rules of the game” (North, 1991: 98). Informal rules are best described as constraints such as sanctions, taboos, customs, ethics, culture, traditions, and codes of conduct and they are formed by values and believes. Formal rules are covered by the regulatory pillar and include legal regulations such as constitutions, laws and property rights. These formal and informal rules are not necessarily explicitly designed, they often simply emerge spontaneously from society and are adopted by individuals. Institutions define and shape the way a society lives, works and functions. However, the effect of institutions go beyond shaping human interaction. In North’s (1991) model, the rules of the game in every society are argued to be the humanly devised constraints and incentives that shape “the direction of economic change towards growth, stagnation, or decline” (North, 1991: 97). Stated otherwise, institutions matter not just for society but most certainly for the multinational corporations that work at the intersection of those various institutions. Even – or particularly – the world’s largest multinational corporations have to deal with the inevitable power of institutions in order to succeed in their cross-border operations. Corporations are argued to conform to the contextual rules and beliefs as this isomorphism earns them legitimacy in the various host environments where they operate (Deng, 2009). This isomorphism refers to the “generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed systems of norms, values, beliefs, and definitions” (Suchman, 1995: 574). Yet, different institutions exist not only in the environmental context of an organization but
7
exist throughout the entire organization. Even though there may be a consistent regulatory environment as a formal institutional setting in an organization, the informal institutions may differ between divisions, teams, groups, departments and levels throughout the entire organization. As institutions shape societies and individuals’ norms, believes and behaviours, MNEs are also shaped by institutions. As early as 1932, Berle and Means described the power of social institutions as follows:
“[…] a society in which production is governed by blind economic forces is being replaced by one in which production is carried on under the ultimate control of a handful of individuals. The organizations which they control have passed far beyond the realm of private enterprise – they have become more nearly social institutions” (1932: 46).
Thus, institutions are highly embedded in MNCs and vice versa. Given increased internationalization of activities, the question arises how these institutions differ across different countries and how these institutions affect MNCs not only in their operations but more interestingly in their organizational design. Considering the regulatory environment as the key institution of a society, the constitutions and laws enforced by the state regarding governmental policies, education, and economy structures have a tremendous impact on the organizational design of both home and host country MNCs. These institutions affect employees’ opportunities, paths and career patterns in a crucial way.
The influence of the institutional national context on career pathways of top managers has been researched by various studies since the late eighties. In discussing the different national heritages and policies concerning career development, Evans, Lank and Farquhar (1989) stressed the existence of national characteristics in career paths. Their study suggested that different national characteristics, policies and contexts have led to different approaches to
8
management development which in turn has resulted in the existence of different legitimate career models across nations. Evans et al. (1989) distinguished a “Latin” model, in which career paths depend on a political game where the political process is likely to outweigh achievement as it is mainly focused on elite educational qualifications, a “German” model which focusses primarily on functional career paths and expertise and an “Anglo-Dutch” model which is characterised by non-elitist recruitment policies focused on managed potential development (Evans et al., 1989). In line with this, Bauer and Bertin-Mourot (1996, in Davoine & Ravasi, 2012) also identified national models of top management careers in French, British, and German firms that are embedded in the institutional contexts, particularly in an educational context.
More recently, as international business is becoming increasingly globalized and different formal and informal institutions are forced to co-exist, several authors have studied how increased cross-border integration affects the internationalization of top management teams (Heijltjes, Olie, & Glunk, 2003; Ruigrok & Greve, 2008; Van Veen & Marsman, 2008). Evidently, institutional variables define what is considered to be a legitimized career path to a top-management position in a particular context and are therefore of great importance to understand differences among these institutionally embedded career patterns.
2.2 Career patterns and models
The term ‘career’ is often used rather broadly to refer to a sequence of jobs or a continuous commitment to a particular occupation. Several authors have used the term in a more refined way. Following Kerkhoff (1993: 13), in this study career patterns refers to “the pathway an individual follows between positions in the social structure occupied at different points in the life course”. Previous studies on career models often seem to assume that “career” is a universally-shared concept (Gunz, Mayrhofer, & Tolbert, 2011). However, when considering
9
management careers in multinational companies, to become a top manager of a large multinational is certainly not the result of a one-way-only career path. Careers are central to individuals, organizations and society and cannot be restricted to one single view of employees climbing the corporate ladder that is always valid, especially across geographical regions. An interesting question in the area of career models and different career patterns therefore is how career patterns across different countries and regions reflect national differences in formal and informal institutions. Different beliefs, values, social structures, economic and labour force policies across regions define what a “normal” career is. Top management career patterns are thus highly embedded in national institutional contexts as they are determined by cultural and institutional factors (Davoine & Ravasi, 2013). These institutions form fundamental bases for individual career paths as well as for career patterns of larger groups within a society (Mayrhofer et al., 2007). Previous career studies often cite Hughes (1937: 413) who defined the concept of careers as:
“The moving perspective in which persons orient themselves with reference to the social order, and of the typical sequences and concatenations of office – may be expected to reveal the nature and 'working constitution' of a society.”
Yet, the greater part of the studies on careers and career patterns often neglect this by focusing mostly on the careers of individuals and their immediate context such as personal background. Numerous studies have researched new concepts of careers over the past 20 years such as the “protean career” (Hall, 1996), and the “boundaryless career” (Arthur, 1994) and the study of career patterns is becoming increasingly pursued (Higgins & Dillon, 2007; Davoine & Ravasi, 2013). However, the broader context within which careers are shaped and lived are often ignored. Although embedded in the institutional environment, relatively little research has been conducted on the cross-cultural generalizability of career theories/models (Gunz, Mayrhofer, & Tolbert, 2011).
10
In spite of increasing internationalization, Davoine and Ravasi (2013) argued that little is known on the continuity of the national models of top management careers which are determined by cultural and institutional factors. In their research they investigated the extent to which certain features of national top management profiles are stable, bringing interesting empirical evidence on this topic. Their data showed a “certain erosion of national models due to the gradual arrival of non-national top managers, but this process occurs very slowly despite the strong international dimension” of the companies they studied from France, Germany, Great Britain and Switzerland(Davoine & Ravasi, 2013: 152). Davoine and Ravasi (2013) showed the increasing importance of international career steps and international MBAs in European top management characteristics and their study brought new evidence to the internationalization process of top management profiles. Moreover, their findings confirms the embeddedness of top management career patterns which result as highly influenced by national institutional contexts. In line with this assumption of institutional embedded career patterns, Sullivan and Baruch’s research on careers distinguish new and traditional career patterns and they discussed “the variety of career patterns that are being enacted in today's dynamic work environment” (2009: 1554). However, this study does not elaborate on exactly how these new managerial careers are shaped by the institutional environment and neither do they distinguish different patterns across the world. Research from Inkson (1995) elaborated on how the changing economic and organizational environment have affected managerial careers. However, Inkson (1995) also merely focussed on the individual careers rather than larger contextual career patterns.
In sum, little research has been done on contextual career models and only few studies have researched and compared the different career models that exist across countries. Also, many studies compared different career models within one region, whereas a comparison of legitimized career patterns across regions is missing. This is rather unfortunate as a comparison
11
between regional career patterns would enable a better understanding of how individuals reach a top management position, how TMTs are composed and how this differs across regions. This is of great importance as many of the studies on career patterns are embedded in the ‘upper echelon perspective’ put forward by Hambrick and Mason (1984) who claimed that organizational choices, strategies and performance are strongly affected by the composition of the top management.
2.3 TMTs nationality diversity
Hambrick and Mason (1984: 193) argued that an organization’s strategic choices and performance levels are “partially predicted by the top management team’s background characteristics”. This stresses the importance of understanding how a TMT is composed as it is of great influence of the strategic decision-making and performance. TMTs are defined as “the executives who are full members of the executive committee as per the companies own definitions. […] This follows the commonly employed definition of the TMT as the most important group of decision-makers in a company” (Greve, Nielsen, & Ruigrok, 2009: 217). In the 1970s and 1980s, various authors tried to explain how organizational performance and strategy are influenced by a firm’s TMT and even long before that time, this relationship has been discussed widely (Certo, Lester, Dalton & Dalton, 2006; Finkelstein & Hambrick, 1990). Because the focus has been so much on the effects of the compositions of a TMT, the actual heterogeneities between TMTs across countries or regions have been studied to a lesser extent. Yet, recent studies are conducted on the characteristics of TMTs from mainly European and United States (US) firms and many of these studies focused on single measures, such as the years of organizational tenure or international experience (Biemann & Wolf, 2009; Carpenter, Geletkanycz & Sanders, 2004; Sambharya 1996).
12
Based on a sample of 54 multinationals from the US, Sambharya’s (1996: 739) research showed that “TMTs with a higher mean, greater heterogeneity, and a higher proportion of managers with foreign experience were significantly associated with the firm's international involvement”. In other studies in this field, the composition of the TMT in terms of its demographic features such as age, tenure, education, and functional background is at central focus (Chaganti and Sambharya, 1987; Finkelstein & Hambrick, 1990; Murray, 1989; Rivas, 2012; Wiersema and Bantel, 1992). Organizations are becoming more diverse in such areas and this heterogeneity has found its way up to the top management teams (Pitcher and Smith, 2001). Yet, nationality diversity is one dimension at the top management level which has been largely neglected in previous research (Heiltjes et al., 2003). Despite economic activities becoming increasingly international, there is relatively little known about how this internationalization is reflected at the level of nationality diversity in TMTs, why some organizations have more foreign top managers than others and what determines this division (Nielsen & Nielsen, 2010). Organizations are becoming increasingly more international in terms of for example their product markets, foreign sales, foreign assets and the foreign workforce they employ (Heiltjes et al., 2003). Subsequently, the question arises to what extent these organizations have also internationalized their top management teams.
Heijltjes et al. (2003) examined the extent of nationality diversity in the composition of top management teams in two highly internationalized European countries with the presence of many well-known multinationals, Sweden and The Netherlands. For the Swedish case, they studied the top 35 companies for the year 1999 and for Dutch case, they collected data for the top 45 companies for the period 1990–1999, which allowed Heijltjes et al. (2003) to investigate the internationalization of Dutch TMTs over time. Their results indicated that “the national diversity of top management teams in these countries has not progressed to the same level as the internationalization of the companies at large” (Heijltjes et al., 2003: 89). Moreover, their
13
research showed that the degree of internationalization of the company in terms of foreign workforce is an obvious factor which encourages the internationalization of top management teams. “The greater the number of foreign managers in a company at a particular time, the greater the chance that foreign managers will enter top management ranks in a future period” (Heiltjes et al., 2003: 93). Although their research showed meaningful results for the European business environment, Heijltjes and collegues (2003) based their conclusions only on the use of two sample European countries. Greve et al. (2009) conducted a more extensive research within the field and investigated how changes in a firm’s degree of internationalization are associated with the configuration of top management teams. Their dataset consisted of a population of 41 banking and insurance firms at the end of the year 2002 with headquarters in Austria, Belgium, Denmark, Finland, France, Germany, the Netherlands, Norway, Sweden, Switzerland, and the UK (Greve et al., 2009). The dataset also included detailed career profiles of the 264 executives that were working in the TMTs of these firms. Focusing on one particular industry allowed the researchers to avoid industry-effects and thus allowed for better comparisons. Greve et al. (2009) used a very refined research design in which they address the advantages of using indices accounting for both the duration of individual TMT member’s experience in a particular country and the geographic dispersion of the TMT member’s experiences abroad. Their method also captured the distinction between a single long international experience in one country and shorter yet more numerous experiences in different countries (Greve et al., 2009). Another interesting feature of their study is the fact that they did not use measures related to the financial dimension of internationalization, like foreign sales rates, as they rightly argued that production and sales of financial services is not inherently attached to physical locations. To incorporate country effects, they used Hall and Gingerich’s (2004) coordination index (CI) score of varieties of capitalism as a proxy. Their findings suggested that firms tend to match top executive profiles to their strategies. For example, the
14
entry into unfamiliar foreign markets and different cultures was found to be related to higher levels of internationality at the TMT. Another interesting contribution to the field comes from Rivas (2012) who used a mixed sample of US and European firms where both the board and TMT units were analysed at the same time. Rivas (2012) explored the role of differential importance that different characteristics like age, tenure and functional background at the board and at the TMT level simultaneously exercise on internationalization. This longitudinal research examined board members individual attributes in 2001 and firm’s internationalization in 2005–2007 and sampled the board members of 108 large European and United States firms (Rivas, 2012). The results showed a positive effect on internationalization for functional background diversity of both boards and TMTs.
Knowing that TMTs are increasingly formed by non-nationals (Staples, 2007), the question arises how this differs across regions and how this is affected by the institutional environment of the home regions of the multinational corporations. Focusing more on the US rather than Europe, Finkelstein and Hambrick (1996) argued that since the workforce of the US is increasingly diverse, diversity in nationality and race will be one of the most important diversity dimensions in the future. However, as mentioned before, in the majority of studies in this field, data originates from the US and Western European countries and various characteristics of TMT members have been studied such as age, company tenure, job tenure, education and team homogeneity (e.g. Heijltjes et al., 2003; Nielsen, 2010; Nielen & Nielsen, 2013; Rivas, 2012; Sambharya 1996; Van Veen & Marsman, 2008). TMT diversity in terms of nationality, however, has been researched poorly.
Moreover, the difference in levels of internationalization of TMTs have not been studied across world regions even though it has been argued that career patterns to top management positions are institutionally embedded and thus vary across regions. So far, there has not been much empirical evidence and conceptualization on various career models that are embedded in
15
institutional environments and most of these studies have focused merely on US based or European career models. A clear comparison of different institutionally embedded career models across world regions and the related investigation of the institutional forces shaping such differences at the level of TMT nationality diversity are missing. Accordingly, this study aims to answer the following research question: How do regional career models impact the extent of TMT nationality diversity across regions?
16
3. Theoretical framework
3.1 Home region institutional factors and career models
Previous studies in the field of top management have suggested that the composition of the TMT is highly dependent on various factors such as age, tenure and experience. Moreover, it has become clear that the process of reaching a top management position is the result of what is considered to be a legitimized career pattern. Various authors have suggested that a career pattern is a process that is embedded in different contextual layers that depend on various contexts of origin, society, culture and the global context (e.g. Mayrhofer, Meyer & Steyrer, 2007; Rivas, 2012; Davoine & Ravasi, 2013; Qian, Cao, and Takeuchi, 2013). Mayrhofer et al. (2007) distinguished four contexts on which career paths depend. They suggested that career models depend on the work context, which entails the external labour market, the context of origin such as education and work history, the context of society and culture and the global context, which comprises the internationalization of businesses and organizations (Mayrhofer et al., 2007). Various studies have brought empirical evidence to this. Biemann and Wolf (2009) for example found that each country in their sample, i.e. Denmark, Germany, Japan, the United Kingdom, and the United States, showed one to four predominant career patterns. Norburn (1987) also found significant differences between TMTs in North American and the United Kingdom in terms of corporate experiences, education, and self-concept in relation to aspiration and executive succession rates. In comparing the backgrounds of top managers in the United Kingdom, Germany and France, Mayer and Whittington (1999) also found notable differences.
Furthermore, studying the effects of national culture, Geletkanycz’s (1997) research showed that values of individualism, uncertainty avoidance, power distance, and long-term orientation are significantly related to executives’ adherence to existing strategy and leadership profiles. As Biemann and Wolf (2009) argued, there are many varieties of country differences
17
that affect the development of career patterns. In line with this, Hofstede’s (2001) research to national culture suggested that cultural factors can have a strong influence on the implementation of various career patterns as culture affects both individuals and organizations. For example, in a country that scores high on uncertainty avoidance, both individuals and organizations will prefer a stable employer–employee relationship (Hofstede 2001) which in turn will be reflected in the way those individuals establish their careers and in the way in which those organizations will select and recruit their top managers. Therefore, it is expected that such cultural factors alongside other institutional factors will have a strong effect on the composition of the TMT and particularly on the level of nationality diversity of TMTs.
Hypothesis 1. Regional career models impact the level of TMT nationality diversity.
3.2 The Asian career model
Being the world’s largest and most populous continent, Asia is a region known for its contradictions at various facets of society. Lowe (1998) thus suggested that ‘modern’ Asian societies like Hong Kong and Singapore have not developed their institutional environment in the same way as other Asian countries. Although modernised, these Asian societies have not wholly adopted Western values, traditions, norms and institutions (Lowe, 1998). Hong Kong for example “may be described as a modem society with a traditional culture, which, in terms of business values, means a complex combination of convergence and divergence” (Lowe, 1998: 344). In line with this, Ralston, Gustafson, Cheung and Terpstra (1993) brought evidence of convergence, divergence and 'crossvergence' between Hong Kong, China and the North America. As a British colony, Hong Kong's formal institutions such as the political, economic, educational and legal systems are rooted in English norms and are thus more comparable to those of the US rather than those in China (Ralston et al., 1993). However, approximately 98 per cent of the ‘modern Asia’ population are Cantonese-speaking natives who follow traditional
18
Chinese cultural patterns (Ralston et al., 1993). Stated otherwise, there are certain informal institutions which predominate, even in countries with a different formal institutional environment.
Paik and Jang (2009) argued that the economic and business environment in prominent Asian countries such as China and Korea are dominated by long-lived social norms and traditions and networks of personal relationships. These networks strongly influence the career patterns in Asia. In line with this, Chowdhury and Mahmood (2012: 1811) argued that
“[…] although there are formal policy guidelines for the recruitment and selection of non-managerial employees in individual departments in the public sector, in practice, organizations recruit employees without having necessary vacancies, and when vacancies do exist, they are filled on an ad-hoc basis without any external advertisement. It is, therefore, almost impossible to secure employment in the public sector without strong backing of key people within the organization or in the political party in power.”
As a result, Asian career patterns to TMT positions are often associated with seniority promotion, lifetime employment, relations and political connections (Biemann and Wolf, 2009; Lowe, 1998; Wilkinson, 1996; Qian et al., 2013). According to the findings from Chowdhury and Mahmood (2012: 1811), managers in the private sector consider the recruitment and selection of employees “as personal matters and maintain informal channels to fulfil the purpose. Besides friendship and kinship, pressure from politicians from the ruling party takes precedence over qualifications and skills of the candidates”. Without these interpersonal relations and connections, it seems rather hard, if not impossible, to climb the corporate ladder in Asian organizations.
19
Moreover, Asian managers value a strong and stable work and culture relationship over their whole careers more than others (Biemann & Wolf, 2009). The Chinese particularly value strong and long-lasting interpersonal relationships, known as ‘guanxi’ (Qian et al., 2013). They are raised and trained in maintaining collective harmony, and avoiding conflicts (Tjosvold, Poon, & Yu, 2005). As TMT functional diversity may trigger cognitive and affective conflict (Qian et al., 2013), it is expected that this strong and unique culture does not incentivize highly diverse TMTs, especially in relation to nationality and cultural traits. Because of the distinct culture, traditions, social norms, complex formal institutions and language barriers, it is therefore expected that Asian companies face the strongest institutional forces against increased nationality diversity in TMTs.
3.3 The Latin American career model
Next to Asia, Latin America (including South- and Central-America) is also a significant emerging economy. Yet, while Asian management research has been reviewed extensively, research on Latin America management remains limited as the top management journals have given Latin America only little attention (Bruton & Lau, 2008; Nicholls-Nixon, Davila Castilla, Sanchez Garcia, & Rivera Pesquera, 2011). Also in research on managerial career patterns, there is a lack of empirical evidence from Latin American countries. However, research on the formal and informal institutional environment provides a meaningful insight in the way they may shape career patterns. Similar to Asia, the Latin American region exists of various countries that have not developed their business environment equally. However, Latin American countries do share at least parts of a common history. They have experienced various political and social developments which have gradually established a certain ‘Latin American way of life’ and a religious faith (Elvira & Davila, 2005). Their common history such as the Spanish and Portuguese domination, their paths towards independence, the inheritance of
20
traditions and more recently their growing importance in the global market has resulted in a strong feeling of regional coherence.
Elvira and Davila (2005) argue that these combined historical and global developments have resulted in unique characteristics for business management. They argue that “Latin America’s work culture has developed through a unique historical process shaping its economic, political and societal structure. Thus, Latin American organizations are embedded in their historical tradition even though globalization forces have deeply changed the region” (Elvira & Davila, 2005: 2166). According to Coatsworth (2008) the institutions and policies that Spain and Portugal imposed and adapted during the colonial era form the base for the difficulties many Latin American countries faced, or are facing, in terms of inefficient legal systems and policies (Coatsworth, 2008; Sánchez-Martín, De Arce, & Escribano, 2014). Even though the prior influence of authoritative and leftist regimes still affects the political environment, Latin American countries today are characterized primarily as liberal/electoral democracies (Freston, 2004). Research by Nicholls-Nixon et al. (2011) showed that the majority of the Latin American countries have adopted US-style federalist forms of government, organized around the presidency, executive branch, legislative branch and judiciary. However, the stability of some Latin American democracies today remains insecure because of concerns about high levels of corruption and political distrust.
Furthermore, exploring the human resource management from a Latin American perspective, Elvira and Davila (2005) argued that Latin America is a region where people, social relationships and their management are supreme and where strong common values are shared. Social relationships and respect for authority are cultural traits that characterize the Latin American management environment (Elvira & Davila, 2005). As a result, selection processes for organizations are often subordinated to personal relationship ties. In Chile for example, in order to promote within organization it is more important to have social relationships with
21
primary social groups (e.g. ties with wealthy families, university colleagues, country club members or birthplace geographical bonds) than having talent (Abarca, Majluf, & Rodriguez, 1998). The importance of social relationships in order to get a promotion is also confirmed by Davila and Elvira (2005:12) who argued that “whereas glass ceilings for women or minorities appear in developed countries due to gender or racial discrimination, in Latin America they relate to physical appearance and social contacts”. In line with this, several authors have argued that family traditions and ties are so important that many organizations retain a policy of hiring family members or close relatives (e.g. Abarca et al., 1998; Elvira & Davila, 2005; Nicholls-Nixon et al., 2011). The assumption here is that trust, loyalty and responsibility will be guaranteed by the strong family ties and familiar traditions. As a result, many large and smaller organizations in Latin America have been family-run for several generations (Husted & Serrano, 2002).
While some Latin Americans have made it through top management positions at foreign multinational corporations, “a strong cultural value for closeness to the nuclear and extended family reduces executives’ willingness to move in support of a firm’s geographic expansion strategy” (Davila & Elvira, 2005: 19). Although Latin American top executives may not be very open to transfer to a foreign TMT, their own TMTs may be more open towards other nationalities compared to Asian TMTs. Opposed to Asia, Latin America’s informal institutions such as culture, traditions and language and formal institutions such as governmental policies and politics are relatively more similar to those in the United States or (parts of) Europe than the Asian ones.
Hypothesis 2. Latin American-based companies are more likely to have a higher level of
22
3.4 The North American career model
The North American region – including the US and Canada – is often associated with a typical standard image of an “American”: a person of white, middle-class status who aims at living the ‘American dream’. However, the North American history characterized by European settlement and colonial domination, has generated waves of immigrations which made the United States the nation with the highest immigrant population in the world (Patterson, 2000). It is therefore hard to speak of one typical American citizen. This heterogeneity is, however, not that strongly reflected in the way they develop their careers. Research on career paths to TMT positions over the past decades in the North American region shows a rather general career path that is nationally distinct, rather than ethically or racially. Regardless of ethical or racial background, the path to a top management position in North America has been characterized for a long time by a long organizational tenure, a business or finance degree at an Ivy League institutions and preferably an MBA degree (e.g. Cappelli & Hamori, 2005; Martelli & Abels, 2010; Sussman, 1979; Swinyaid & Bond, 1980). The following composite picture emerged from a study of 1700 executives in 750 of the largest American corporations:
“He was born in 1926 and spent most of his childhood in the urban Midwest. Eldest of three children, he attended public schools, helped put himself through college (a large public institution) by working part-time, and then earned his MBA at a top private university. He started with his employer 19 years ago as a functional specialist. He is married with three children, and his wife has no outside career; his family has had to relocate during his career, but not often. He believes concern for results, integrity, and desire for responsibility increase one’s chances for success- more so than creativity, ambition, and aggressiveness.” (Sussman, 1979: 15)
23
Yet, this profile is the result of a rather dated research and although similarities exist and it remains an image of North American top managers that many people share, it is not the contemporary picture of a top manager in North America. The labour market in the United States has experienced critical changes in the past three decades resulting in a decreasing tenure of employees and vanishing employee loyalty (Hamori & Kakarika, 2009). In this new labour environment various authors suggested that the alternation of TMT members and more importantly the career patterns to a TMT position have also changed (e.g. Cappelli & Hamori, 2005; Frycell & Lerner, 1989; Hansen, 1993; Hamori & Kakarika, 2009; Hayes & Abernathy, 1980). There has been a concern with corporate social performance in responding to changes in underlying social and political trends (Frycell & Lerner, 1989). In line with this, Hansen (1993) argued that the growth of ethnic minority populations and commitment to equal opportunity produced widespread initiatives in managing workplace diversity which had a significant impact on American career development.
Besides the social concerns, Fryxell and Lerner (1989) argued that the composition of top management teams in the North American region has received much attention because of the economic concern with arranging the best composition of top managers, which is mainly measured by educational background in the form of MBA’s and international business experience. In examining age, gender, promotions, years of education, nature and number of educational institutions attended, tenure within companies, and the time it took executives to achieve their positions, Cappelli and Hamori (2005) investigated how Fortune 100 executives and their career tracks changed between 1980 and 2001. They found that today's top managers of Fortune 100 companies are fundamentally different: “They are younger, more of them are female, and fewer of them were educated at elite institutions. They're making it to the top faster and taking fewer jobs along the way. And they are increasingly moving from one company to another as their careers unfold” (Cappelli & Hamori, 2005: 25). Even though degrees from elite
24
Ivy League institutions are becoming less prominent, various studies confirmed that graduate degrees in business, especially an MBA degree, has become one of the most important instruments to reach a top management position in corporate America (Cappelli & Hamori, 2005; Hamori & Kakarika, 2009; Hansen, 1993).
Cappelli and Hamori (2005) furthermore expect that MBAs will be even more prominent in top management teams in the next generation. Because MBAs are such highly international programs which attract many aspiring top managers from all over the world, the increased importance of this program has facilitated the access of foreigners to TMTs in the North American region. Based on these findings it is clear that career patterns to a top management position in the North American region are more related to formal institutional factors such as education and functional background rather than informal factors such as cultural background, traditions and relationships as for Latin American and, in turn, Asian corporations.
Hypothesis 3. North American-based companies are more likely to have a higher level of
TMT nationality diversity than Latin American companies.
3.5 The European career model
Today, moving to a neighbour country in Europe is similar to moving to a neighbour state in the United States. The shorter distances involved, coupled with educational efforts and common business practices that encourage international experience at early career stages result in high levels of exposure to and awareness of languages, cultures, and laws among Europeans (Palmer & Varner, 2007). Although as proven by previous studies it remains rather hard to speak of one single integrated European career model, there are important common elements across European countries vis-à-vis other world regions. As the great majority of the largest European
25
multinational corporations are headquartered in the Western European countries, this study refers to Western Europe in particular.
Various authors have suggested that regardless of the strong ties within the European Union, a top management ‘elite’ presides that varies across European borders (Mayer & Whittington, 1999; Davoine & Ravasi, 2013; Heiltjes et al., 2003; Mathews, Ueno, Kekäle, Repka, Lopes Pereira, & Silva, 2001). In studying the integration of managerial elites in major economies in Europe, Mayer and Whittington (1999) found that top manager profiles in France, Germany and the United Kingdom have not changed much even though these economies faced great internal and external pressures over the past two decades. French top managers still rely substantially on family ties and state connections, German managers remain highly engineering-oriented and the number of British top managers with a background in finance is steadily increasing (Mayer and Whittington, 1999). This is also in line with a more recent study by Davoine and Ravasi (2013) who discussed the stability of certain characteristics of national top management profiles in the above mentioned countries and Switzerland. They underlined the increasing importance of knowledge in economics and management that is necessary to reach top management positions in European firms.
Despite the above mentioned differences, all countries showed a significant importance of educational background in top managers’ profiles. The British managerial career pattern is characterized by a significantly smaller percentage of university graduates, which ranges from 90 per cent in Germany to 95 per cent in France and less than 72 per cent in the United Kingdom (Davoine & Ravasi, 2013). However, this gap is compensated with a large participation in MBA and other executive education programs in the United Kingdom. Notable is the importance of a leading and prestigious – often international –educational background. In fact, 83 per cent of the university-educated top managers in France graduated from at least one of the three most prestigious universities (Davoine & Ravasi, 2013). Likewise, in the United Kingdom almost a
26
third of the British top managers has a degree from the prestigious universities of Oxford or Cambridge (Davoine & Ravasi, 2013). German top managers on the other hand, remain rather independent of the exclusive universities so significant in France and the United Kingdom. In Germany, no institution seems to be clearly associated with a top management function. However, as their educational system does not arrange university or educational institutional hierarchy, here a doctorate is seen as a sign for high potential. In their sample of German firms, nearly half of the managers had a doctoral degree (Davoine & Ravasi, 2013). In the German academic system, a doctorial title seems to play a similar role as the role of the Grandes écoles in France. The importance of an extraordinary educational background is also reflected in the high numbers of top managers who completed an MBA found by Davoine and Ravasi (2013). Interestingly, the highest percentage of MBA graduates was found in Switzerland and the United Kingdom, which are also the countries with the highest number of non-native top managers. This confirms that the internationally focused MBAs facilitates the access of foreigners to TMTs. Moreover, their result also showed that international experience and professional experience at different firms, which is highly valued, is also more common and concerns more than half of the top managers (Davoine & Ravasi, 2013).
In line of this, a study by Freye (2009) showed that the model of the ‘mountain climber’ working for a single firm is gradually being replaced by top managers who have worked at various functions and positions at different companies. In Europe, emphasis is placed on a prestigious or exclusive – educational – background in order to achieve a TMT position. Moreover, the increasing number of foreign managers indicates that European companies value education and professional experience over cultural background. Considering European countries’ increasing integration, political collaboration and seemingly fading borders, this may not seem surprising. Additionally, as Heijltjes et al. (2003) argued, smaller European countries such as Switzerland, Sweden, and The Netherlands house a relatively large number of
27
multinationals because their home market is relatively small. In line with this, Hill (1998) found that eighty per cent of the top twenty Nordic firms operates internationally. Because many European companies have internationalized at a relatively early stage, highly value a prestigious university degrees and MBA degrees, and are familiar with different cultures, languages and traditions, a significantly high proportion of foreign managers is expected to be allowed to climb the last steps of the corporate ladder towards a TMT position (Heijltjes et al., 2003). Accordingly, the following hypothesis is proposed:
Hypothesis 4. European-based companies are more likely to have a higher level of TMT
nationality diversity than North American-based companies.
3.6 CEO non-nativity
In the majority of studies in the field of top management teams, the Chief Executive Officer (CEO) is treated as any other member of the TMT, averaging their characteristics to assess overall TMT demographic composition (Buyl, Boone, Hendriks, & Matthyssens, 2011). However, various studies have suggested that CEOs may play a unique and decisive role in TMT composition and its effects (e.g. Arendt et al., 2005; Buyl, Boone, Hendriks, & Matthyssens, 2011; Jaw and Lin, 2009; Kisfalvi & Pitcher, 2003). Moreover, various authors suggested that their impact should be considered separately from that of the other TMT members (Haleblian and Finkelstein, 1993; Jaw and Lin, 2009; Papadakis and Barwise, 2002). Considering the CEO not as just as a regular TMT member but as the leader who has a strong effect on TMT dynamics (Kisfalvi & Pitcher, 2003) and whose job it is “to conciliate and soothe the aftermath of intense exchanges among team members […] for avoiding group fragmentation and destructive rivalries” (Hambrick, 1995: 123), it seems plausible that CEO’s familiarity with different institutional environments will positively affect the level of openness to a nationality
28
diverse TMT. Moreover, it seems plausible that if the highest position within an organization, a CEO position, is open to a non-native, subsequently this organization will also be more open to non-native managers in its TMT. Furthermore, a recent systematic analysis by Ghemawat and Vantrappen (2014) on the mid-2013 data from the Fortune Global 500 list, showed that the appointment of a non-native CEO of any origin is still an unusual occurrence. To illustrate, only three non-Indian firms were led by Indian CEOs, only three non-Brazilian firms were run by Brazilian CEOs, and only five non-South African firms were led by South African CEOs (Ghemawat & Vantrappen, 2014). It is therefore expected that a non-native CEO will affect the association between regional career models and TMT nationality diversity.
Hypothesis 5. All else equal, the appointment of a non-native CEO positively moderates
the association between a particular regional career model and TMT nationality
29
4. Research Methods
4.1 Data collection
This study will have an exploratory and explanatory foundation. In order to understand the effect that a particular regional career model has on the extent and patterns of globalization of TMTs, a quantitative study has been conducted with a data collection of the Fortune Global 500 companies as listed in the 2013 ranking. The data collected on the TMT members of these companies concerned the company’s nationality, the total number of TMT members, the number of different nationalities represented, the number of international and national members, the division of nationalities categorized in six world regions (North America & Canada, Europe, Asia, Latin America, Oceania and Other regions), number of males and females and biographical information on the CEO. All data was retrieved exclusively from official corporate sources. The following paragraphs will elaborate on the data collection by discussing the sample, the independent and dependent variables and the moderating and control variables.
4.2 Sample
The population in this study includes the TMTs of multinational corporations which have organizational activities in more than two countries. The hypotheses are tested with a dataset using a sample of the TMTs of the 500 companies that are listed in the Fortune Global 500 of 2013. This sample is selected by including all companies listed by Fortune Magazine (2013). By the use of this sample, the research question is answered based on the results of a rather large sample which enhances the reliability and generalizability of the results. Moreover, because most of these listed companies have provided an extensive annual report of 2012, it has limited the amount of missing values. However, some companies had to be excluded from the sample because of the inability of determining their TMT members’ nationality with
30
accuracy. Moreover, some companies were excluded of the sample because of the inability to collect information about the composition of the TMT in general. For this matter, a total of 98 companies were removed from the sample. This resulted in a total of 402 companies that are used in the statistical analysis. To control for the accurate information and to enhance the information access for specific details such as tenure and background, the online database of Orbis was used next to the companies’ annual reports of 2012. To make sure both the annual report and the reports from Orbis correspond with the same listed company, both reports were checked on corresponding turnover and CEO as noted in Fortune Global 500 list. This minimized the chances of non-corresponding information and an invalid database, which in turn enhances this study’s reliability and validity. Because this is a quantitative study, the results are based on numerical data. In order to draw conclusions based on numerical dataset regarding the different paths and patterns to top management positions across world regions, a set of variables is used. These variables include both an independent variable and a dependent variable, as well as a moderating variable and control variables.
4.3 Independent variable
The independent variables in this study are related to career models relative to a particular geographic regions. As discussed in the previous chapter, this study examines four regional institutionally embedded career models from North America, Europe, Latin America and Asia. Thus the particular regional career model is identified with the construction of dummy variables that take on the value of 1 if the company is based in the corresponding region and 0 otherwise. This creates four independent variables: European-, North-American-, Asian-, and Latin
American career model, to test which regional career models are more likely to promote a high
level of TMT nationality diversity. Furthermore, in order to test the overall impact of these regional career models on TMT nationality diversity, a ranked variable is constructed named
31 regional career models. This will explain whether being in a specific region – and thus
following that particular career model – increases the chances of a more nationality diverse TMT. This variable is operationalized with a ranked order from 1 to 4 in which 1 represents the regional career model that is least open to TMT nationality diversity and 4 represents the regional career model that is most open to TMT nationality diversity. Based on the hypotheses formulated, the regional career models variable takes on the following values: 1-Asia, 2-Latin America, 3-North America and 4-Europe.
4.4 Dependent variables
The dependent variable in this research is TMT nationality diversity. As the TMT knows various translations across countries, it is narrowed down to the definition of the TMT as the core executive management occupied by the top-tier of executives only (Certo et al., 2006). In most cases, the TMT corresponds with the notion of ‘Executive committee’, ‘Senior management’ or ‘Top management’. In this study the TMT members’ nationality was determined based on how they are reported by companies in their annual reports or in other secondary sources. This variable is operationalized by the Blau index (Blau, 1977). This index measures group heterogeneity and it is often used in top management team research (Carpenter, 2002; Finkelstein & Hambrick, 1996; Greve et al., 2009; Nielsen & Nielsen, 2010) to aggregate data on nationality from the individual to the team level. The index is calculated by the following formula:
Here, p is the proportion of group members in a given category and i is the number of different categories of the feature across all groups (Blau, 1977). A high value of B implies a high level of heterogeneity on the particular variable, which would imply a high level of TMT nationality diversity in this study. Not only is the Blau index the most frequently used approach for
32
measuring diversity (Harrison & Klein, 2007), it is also an eligible index for measuring diversity where the variable is categorical and members can only be part of one category, as in this case. As this study focusses on the differences of diversity between world regions, the Blau index is the appropriate measurement for TMT nationality diversity as a dependent variable. In this sample, TMT members’ nationality is divided into six different world regions which function as the six categories used to calculated the Blau index: Noth-America, West-Europe, Latin America, Asia, Oceania and Other regions. Yet, as the Blau index is based on data that only contains regional demographics instead of country demographics, this index provides information about how the nationalities of TMT members are divided across regions as opposed to countries. Therefore, it is rather a measurement of regional diversity. Accordingly, this variable is referred to as TMT regional diversity. In order to also capture TMT diversity on country demographics, a different variable is used. TMT nationality diversity is operationalized by dividing the number of nationalities represented in the TMT by the total number of TMT
members. The analytical tests are repeated for each operationalization in order to enable a
comparison between regional diversity and nationality diversity thus provide an insight on how TMT nationality is divided and how this differs across world regions.
4.5 Moderating variable
Baron and Kenny (1986: 1174) defined a moderator as ‘‘a variable that affects the direction and/or strength of the relation between an independent, or predictor, variable and a dependent, or criterion, variable’’. In this case, it is expected that the relationship between home region institutional factors and TMT nationality diversity is positively moderated by CEO non-nativity. Therefore CEO non-nativity serves as the moderating variable in this study. This variable is operationalized as a dummy variable that takes on the value of 1 if the company has a non-native CEO and 0 if the company has a native CEO. To test the moderator effect, an
33
interaction variable is created: regional career models (standardized) x CEO non-nativity. Information on CEO’s nativity was obtained from corporate websites, annual reports and other corporate online sources such as Orbis, Investment Week and World of CEOs. From the 402 companies in the sample, there is only one company, CPC, which did not provide any information on the CEO’s characteristics. Seven companies from the sample (i.e. Renault, Bayer, Coca-Cola, Sanofi, Sberbank, Westpac Banking and Mitsubishi Electric) have a CEO with two or more nationalities. In five of these cases, one of these multiple nationalities corresponds with the company’s nationality. Therefore, for these CEOs the nationality of their company (Renault, Coca-Cola, Sberbank, Westpac Banking and Mitsubishi Electric) was filed. In determining the nationality from the other two CEO’s (Bayer and Sanofi), the first nationality mentioned in his/hers biography was chosen.
4.6 Control variables
Previous research has suggested that more factors may affect the level of nationality diversity in TMTs. Therefore, this research includes control variables in order to test the relative impact of regional career models on TMT nationality diversity. The first control variable is degree of
internationalization (DOI) as it has been argued that TMT internationalization, or nationality
diversity, is simply a dimension of a firm’s level of internationalization (Caliguri, Lazarova & Zehetbauer, 2004; Sullivan, 1994). It is expected that firms with a high DOI are more likely to have a nationality diverse TMT than firms with a low DOI. Therefore, DOI serves as an important control variable. Sullivan (1994) argued that the degree of firm internationalization is a multi-dimensional construct that consists of three dimensions: performance (foreign sales), structural (foreign assets) and attitudinal (international orientation). However, the latter dimension is hard to measure with secondary data as it entails psychological features. Therefore, in this study DOI is operationalized by the first two attributes and is calculated as
34
the average of (1) foreign sales to total sales and (2) foreign assets to total assets. The second variable is controlled for is TMT size as this has been an important variable in previous research which has proved to affect both TMT diversity and its effects (Carpenter et al., 2004). Larger TMTs have a higher chance on foreigners being admitted to the executive levels, which would subsequently result in a higher level of nationality diversity (Kaczmarek & Ruigrok, 2013). TMT size is determined by the total number of members of the TMT of each company as reported by the companies in their annual reports. Last, firm size was previously found to correlate with both firm’s and TMT’s level of internationalization and nationality diversity (Ruigrok & Greve, 2008; Sambharya, 1996) and is therefore included as a control variable. Larger firms are more likely to be able to overcome structural and financial issues that may result from the internationalization process, which may refer to a higher nationality diversity, thanks to the scale of their operations (Kaczmarek & Ruigrok, 2013). In this study the logarithm of total number of employees is used as a proxy for firm size.
35
5. Analysis and results
As discussed in the previous chapter, this analysis works with two different operationalizations of the same dependent variable, TMT nationality diversity. Therefore two sets of hierarchical multiple regressions are conducted, each with a different operationalization of the dependent variable. In the first set, TMT nationality diversity is operationalized by dividing the total number of nationalities represented by the total number of TMT members. In the second set, TMT regional diversity is operationalized by the Blau index. Before running any regression analysis, multicollinearity was tested by assessing the bivariate correlations between all four independent dummy variables regarding the particular regional career model. According to Pallant (2005) only correlations above 0.7 are critical. As shown in Table 1 all bivariate correlations in this study show values below 0.7 and are therefore not problematic. Also a collinearity diagnostic test was performed to pick up on problems with multicollinearity that may not be evident in the bivariate correlation. In this test, two values are calculated for all four independent dummy variables and the ranked regional career models variable: Tolerance and VIF. The collinearity diagnostic test did not show any tolerance levels below 0.10 and no VIF values higher than 10.0, which indicates no existing issues with multicollinearity (Pallant, 2005). Together with the bivariate correlation matrix, this indicates that multicollinearity is not problematic in this study. Therefore, all variables were retained.
36
38
5.1 Descriptive statistics
The descriptive statistics of the dependent, independent, control, and moderator variables used in this study are presented in Table 1. There is 1 missing value for CEO non-nativity and 161 missing values for DOI. It was impossible to obtain these from the available data. However, DOI is an important control variable and therefore it is retained. This results in a total of 241 companies to be analysed. The mean value of DOI is 0.43. This indicates that the number of companies with a high DOI is relatively close to the number of companies with a low DOI. This relatively even distribution makes the sample suitable for controlling for DOI. Also, a frequency analysis shows that only 22 of these firms score 0.0 on DOI. Moreover, the descriptive statistics show that the mean number of TMT members is 10. When looking at the descriptive values regarding the levels of regional and nationality diversity in TMTs, on average the sample companies score lower on regional diversity relative to nationality diversity. Frequency analysis shows that in the sample both fully diverse (maximum = 1.0) and non-diverse TMTs (minimum = 0.0) exist when measuring nationality diversity on both a national and a regional level. Furthermore, it shows that only 2 companies score 0.0 on nationality diversity whereas 250 companies score 0.0 on regional diversity. However, the number of companies that show the highest scores possible (e.g. 1.0) on nationality and regional TMT diversity are more similar; 2 companies have a fully diverse TMT measured on a nationality level and 3 companies have a fully diverse TMT measured on a regional level. Moreover, frequency analysis on CEO’s non-nativity shows that from the 401 companies that have provided information on CEO’s nationality, 48 CEOs are non-native, a mere 12 per cent.
As this study focuses on the differences of TMT nationality diversity among different regions, it is important to know how the companies within this sample are divided across regions. Figure 1 shows the frequency statistics on the home regions of 397 companies. Five companies from the sample (i.e. National Australia Bank, Commonwealth Bank of Australia,
39
Westpac Banking, Australia & New Zealand Banking Group and Telstra) are headquartered in a region other than the ones used in this study. This figure shows that apart from Latin America, the companies in the sample are relatively evenly distributed. This enables a good comparison across regions. The limited number of Latin American companies will be taken into consideration when discussing the results.
Figure 1. Frequency statistics on home regions.
5.2 Regional career models and TMT nationality diversity
The hypotheses are tested using a hierarchical multiple regression analysis to test the effects of regional career models independently from the influence of the control variables. To make sure that firm size, TMT size and DOI do not distort the association between regional career models and TMT nationality diversity, they are inserted into the model first. This ensures that they will get ‘credit’ for any shared variability that they may have with regional career models. Any observed effect of these factors can then be argued to be independent of the effects of the
variables that already have been controlled for. In order to determine the relationship between regional career models and the level of
