Who’s in Charge?
An empirical study of executive staffing practices in foreign subsidiaries

 

Anne-Wil Harzing,
University of Bradford, Management Centre
Emm Lane,
Bradford,
West Yorkshire
D9 4GL,
United Kingdom

Tel: +44 (0)1274 234364
Fax: +44 (0)1274 546866

Email: a.harzing@bradford.ac.uk

 

 

Abstract

This study investigates executive staffing practices in foreign subsidiaries of MNCs. Firmly grounded in a literature review of the reasons for employing either PCNs (parent country nationals) or HCNs (host country nationals) in top management positions in foreign subsidiaries, a number of factors influencing the choice between these alternatives is identified. Using a combination of an archival and mail survey research method, the influence of each of these factors is empirically tested with a sample of nearly 3000 observations.

 

Introduction

In the international HRM literature, the discussion of staffing policies takes up a prominent place. The most important element in an international context is usually the executive nationality policy in foreign subsidiaries. Do companies have a policy to employ mainly Parent Country Nationals (PCNs) or Host Country Nationals (HCNs) as (top) managers in their subsidiaries and which circumstances would invoke a preference for one policy over another? Unfortunately, although many publications focus on the advantages and disadvantages of employing PCNs as opposed to HCNs and some give an overview of the factors that might influence executive staffing policies, there are surprisingly few studies that examine actual MNC staffing practices. If they give any empirical evidence about executive staffing practices, all of the above publications refer to Tung (1981, 1982, 1987).

Although Tung’s study was a primer in the field, it is now more than 15 years old. Recently, Kopp (1994) compared international human resource policies in Japanese, European and American multinationals. One of the issues studied was the nationality of top managers in overseas operations. Kopp’s results in this area, however, were limited to one table summarizing the percentage of HCNs and PCNs in overseas operations of MNCs headquartered in three different regions (Japan, US, Europe). Although Tung’s study included some information on the reasons for sending out expatriates to foreign subsidiaries, neither of the two studies systematically related reasons for expatriation to actual staffing practices. Boyacigiller (1990) provided an empirical analysis of factors influencing executive staffing practices. Her study, however, was based on a single American MNC, so that her results might not be universally applicable. In addition, although the study was published in 1990, data were actually gathered in 1983 so that the results of this study might be slightly dated as well.

A large-scale, international study that systematically investigates a number of factors that might influence executive staffing practices in foreign subsidiaries seems long overdue. As indicated above, previous studies were mostly conceptual. If they did contain any empirical work, this was based on a relatively small sample and/or investigated a very limited number of home and host countries. In addition, few studies offered a theoretical framework for their investigation. Unfortunately, large-scale international studies almost inevitably involve mail surveys, which are plagued by low response rates (Harzing, 1997). This article therefore provides a large-scale international study of MNC executive staffing practices in foreign subsidiaries that is mainly based on secondary data, gathered via desk research. Firmly grounded in a literature review of the reasons for employing either PCNs or HCNs in top management positions in subsidiaries, a number of factors influencing the choice between these alternatives is identified. After a description of the study’s methodology, the hypothesized effect of these factors is tested using both an innovative archival method and a smaller sample of data gathered by an international mail survey.

Theoretical Framework

At first sight, the study by Edström & Galbraith (1977) is the only one that theoretically explains why international transfer of managers occurs. Edström & Galbraith found three general company motives for making this type of transfers. The first was to fill positions, which mainly concerns the transfer of technical knowledge to developing countries, where qualified local nationals are not available. The second major motive is management development. The transfer gives the manager international experience and develops him or her for future important tasks in subsidiaries abroad or with the parent company. This kind of transfer would be carried out even if qualified host-country nationals were available. For the third reason for international transfers, the final goal is not individual development but organization development, that is transfers are used to change or maintain the structure and decision processes of the organization. In this case, international transfers are used as a coordination and control strategy. This strategy consists of two elements: socialization of both expatriate and local managers into the corporate culture and the creation of a verbal information network that provides links between subsidiaries and headquarters.

The classification of Edström & Galbraith is well accepted in the literature on international transfers. Virtually every publication that deals with international transfers refers to Edström & Galbraith’s now classic 1977 ASQ article. Borg perfectly describes this unanimous support in his dissertation: "The study of Edström & Galbraith (1977), which is very often quoted in both articles and textbooks, seems to be the only one which theoretically explains why international transfers of managers occur" (Borg, 1988:41). A further investigation, however, revealed a substantial number of German studies, both conceptual and empirical, on this subject. The fact that they appeared in the German language only seems to have blocked their way to the Anglo-Saxon research community. A summary of these studies and a comparison of their classifications to the one by Edström and Galbraith can be found in Table 1. There is a considerable consensus on the principal functions of international transfers, well represented by the classification of Edström & Galbraith. In many of the German studies, though, the focus is more on a direct type of expatriate control than on the informal type of control or coordination that Edström & Galbraith distinguish. However, the ultimate goal is similar in both occasions: making sure that the various organizational units strive towards common organizational goals.

 

One of the most important reasons to choose HCNs for top management positions in subsidiaries is their knowledge of the local market, business practices and cultural preferences (see e.g. Banai, 1992; Kobrin, 1988; Root, 1986; Schaffer & Rhee, 1995; Tung, 1982). Even if they have an international outlook, it will be very difficult and time-consuming for PCNs to become as intimately aware of local circumstances as HCNs are. Another, very pragmatic, reason to prefer HCNs is the fact that it is very expensive to send PCNs on an international assignment (see e.g. Banai, 1992; Kobrin, 1988; Root, 1986; Tung, 1982). In addition to compensation for removal and relocation costs, expatriates expect financial allowances for (international) education for their children, club memberships, return trips, loss of spousal income, "hardship", maintenance of property in their home country, etc. Finally, PCNs often have difficulty in adapting to local circumstances, causing them to "fail" on their international assignment (see e.g. Banai 1992; Kobrin, 1988). This might result in disturbed local business relationships or a premature return of the expatriate to the home country, although this is by no means as likely as is often claimed (Harzing, 1995). Especially if local circumstances are very different from the home country, HCNs might be the preferred option.

Having reviewed the most important reasons for choosing either HCNs or PCNs, we will now link these reasons with factors that are hypothesized to influence executive staffing practices (see Table 2 for a summary). The choice of factors was motivated by a combination of earlier (conceptual) studies and logical reasoning of factors that could be related to the three basic reasons for choosing either PCNs or HCNs. It was also limited, however, to factors that could be reliably measured. The first reason for sending out PCNs was position filling. Specific functions that have been subsumed under this reason are transfer of technical or managerial knowledge, lack of qualified local personnel and the training of local nationals. First, these functions are likely to be more important when the level of education in the host country is low (see also Boyacigiller, 1990; Scullion, 1991). Further, MNCs with a research intensive product are more likely to feel the need to transfer at least some of this knowledge to their subsidiaries and train local managers (see also Hamill, 1989; Ronen, 1986). Position filling is also likely to be more important for recently established subsidiaries, because MNCs will not have an intimate knowledge of the local labour market or might have difficulty in attracting high-caliber locals. When subsidiaries become more established, local recruitment might be easier and some transfer of knowledge and training of local managers will already have been effectuated (see also Boyacigiller, 1990; Franko, 1973; Hamill, 1989). The parent company’s lack of knowledge of the local labor market and a lack of recruitment potential will also be major reasons for greenfield establishments to attract PCNs for top management positions. In acquired subsidiaries, there will often be an established local managerial cadre, so that transfer for position filling will usually not be necessary.

Management development was the second major reason for sending out expatriates. Since the most important objective of transfers in this respect is to develop an international mindset among (future) top managers, this reason for international transfers is likely to be more important in highly internationalized MNCs. These MNCs will realize the importance of creating global awareness better than their less internationalized counterparts and will therefore send out more managers for this reason. In addition, larger MNCs are more likely to have a formal management development program in operation that involves the transfer of managers around the world. For their smaller counterparts, transfer for management development will be likely to take place at an ad hoc basis if at all.

The third reason for transferring PCNs to foreign subsidiaries, organizational development, will be discussed in two parts: one, the actual control and coordination function and two, transfer for the creation/improvement of communication channels. The control function will be likely to be more important for companies that come from a national culture that scores high on uncertainty avoidance (Hofstede, 1980, 1991). In these cultures, there is a strong preference for being "in control". There is suspicion towards foreigners as managers and a view that initiative of subordinates should be kept under control. Managers should be an expert in their field and should generally be selected based on seniority (Hofstede, 1980,1991). These characteristics would usually point to a trusted PCN as the preferred alternative for senior positions in subsidiaries. Direct control of subsidiary operations will also be more important if the level of cultural distance between home and host country is high. In this case, HQ managers might not trust the information they receive from local managers. In addition, HQ managers might fear that local managers are less committed to the company. A high level of political risk in the host country is also likely to make direct control through expatriates more important, because the risks of loss of income or assets might be substantial. In addition, several subsidiary characteristics might induce a higher importance of transfer for control or coordination reasons. When a subsidiary is very important to headquarters, keeping its operations under control will be felt to be more necessary (see also Boyacigiller, 1990). Large, majority-owned subsidiaries that report directly to headquarters are more important to headquarters than small, minority-owned subsidiaries that are much "lower" in the corporate reporting hierarchy. Finally, direct control will be more important when a subsidiary is recently established and when a subsidiary is under-performing and direct headquarters intervention is necessary.

Transfer of PCNs in order to create or improve information flow between headquarters and the subsidiary in question is likely to be more important if the level of cultural distance between headquarters and subsidiary is high. Communication between people from different cultural background can be very difficult even if they speak the same language and the opportunity for misunderstandings is usually high (see also Boyacigiller, 1990). Therefore, headquarters managers are likely to prefer to have at least some home country managers at important positions to facilitate the information flow. This is even more important in host countries with a high level of political risk, since speed and clarity of communication are vital in crucial circumstances (see also Boyacigiller, 1990). Further, transfer for communication reasons is likely to be more important for recently established subsidiaries, since the communication network with headquarters has to be build up from scratch.

 

The necessary familiarity with the local market and business leads us to expect a higher emphasis on local managers for some functions than for others. Because of the direct interaction with local employees and the necessary intimate knowledge of local laws, rules and practices, the function of personnel manager in a foreign subsidiary is expected to have the highest percentage of local managers. This function will be closely followed by the marketing function, which also requires a substantial knowledge of local tastes and preferences. Of the various functional disciplines, the finance function is expected to have the lowest percentage of local managers, since this function is less tailored to local circumstances and will elicit a higher level of headquarters control. Some evidence for this relationship can be found in a related research area: centralization of decision-making within MNCs. Many studies found that decisions in the field of marketing, HRM, industrial relations, production, R&D and finance are centralized to different degrees. Decisions in the first three fields are usually more decentralized than decisions in the latter three fields (see Van den Bulcke, 1984 for an overview). If decisions in a certain area are more decentralized, we would also expect a higher percentage of local managers.

The second reason for preferring HCNs in top management positions was a very pragmatic one: they are less expensive to employ than PCNs. This reason will be more important when the cost of living in the host country is higher than in the home country. In this case, an expatriate will expect to get more additional compensation to maintain his or her previous lifestyle. Local managers would probably have adjusted better to the high cost of living and would not require additional compensation.

A final reason to employ HCNs at top level management positions is to avoid adjustment problems that PCNs might have. The higher the cultural distance between home and host country, the larger the adjustment problems. Hence we would expect more HCNs in culturally distant countries.

All factors at the right-hand-side of Table 2 will influence executive staffing practices and can be interpreted as specific hypotheses to be tested in the empirical part of this paper. Figure 1 summarizes these hypotheses, except for the one concerning PCNs in different functions, in a more explicit form.

Industry can be seen as a proxy for two important firm-level variables that might have an impact on staffing practices: R& D intensity and corporate strategy. We were able to measure R&D intensity at firm level in the small-scale mail survey sample, but were unable to do the same for corporate strategy. In this respect industry might offer an approximation of differences in corporate strategies specifically in terms of global efficiency and local responsiveness (Bartlett & Ghoshal, 1989). Companies following a multidomestic strategy are likely to be found in industries where local responsiveness to the customer is more important than global efficiency, while the reverse is true for companies following a global strategy. Industry has therefore also been included as an independent variable in Figure 1.

 

Methodology

Research method and sample

It would be very difficult to gather a substantial amount of information on executive staffing practices by means of either interviews or surveys. Interviews are unlikely to give information for more than a couple of companies and international mail surveys are plagued by very low response rates. We therefore used an archival method and gathered most of our data from secondary sources (e.g. the Directory of Corporate Affiliations and annual reports). Complete data were gathered for 2689 instances, i.e. the choice for either a PCN or a HCN in a specific subsidiary. These 2689 instances represent more than 200 MNCs from 11 different home countries, subsidiaries located in 48 different countries and operating in 23 different industries. For a more detailed description of the archival sample see Table 3.

For variables that were not available from secondary sources, we used a second sample of data gathered in a smaller-scale mail survey. This was conducted as part of the author’s earlier study on control mechanisms in MNCs. Complete data were gathered for 239 subsidiaries that represented 104 different MNCs, headquartered in Japan, the United States and seven European countries. These subsidiaries were located in 22 different countries and operated in eight different industries. For a more detailed description of the mail survey sample see Table 4.

Since the way in which we measured our dependent variable, the choice for either a PCN or a HCN at top management levels in foreign subsidiaries, is rather unconventional, we will discuss it as part of the research method. All studies on staffing policies or practices in MNC subsidiaries focus on top level managers. At lower levels in the organization, most employees will be host country nationals. In this study, we limited ourselves to the managing director or general manager, although some information is available for marketing, financial and personnel managers as well. To determine whether the manager was a PCN or a HCN (or a third country national (TCN)), we considered the name of the managing director, personnel manager, marketing manager or finance manager as included in the Directory of Corporate Affiliations. The research method for this study thus consisted of inferring the nationality of the manager's from his or her name, usually both first and last names were available, and subsequently classifying the manager as PCN, HCN or TCN. Coding was done independently by three different coders and any unresolved differences (less than 1% of the cases) were excluded from the sample.

At first sight this research method might seem fraught with difficulties. However, an independent check of nationality in the smaller-scale mail survey showed that only 3% of the cases were classified incorrectly. In the mail survey, the responding managing directors were asked to state their nationality (PCN, HCN or TCN), so that their answer could be compared to the result of the "name guessing". All misclassifications involved PCNs or HCNs that turned out to be TCNs, so in none of the cases was a HCN mistaken for a PCN or visa-versa (see Harzing, 1999 for a more detailed description).

However, the research method might be problematic in the United States, a country in which immigration has been underway for decades and in which names say little about nationality. US MNCs and US subsidiaries were therefore not included in the archival sample. Large-scale immigration in other countries occurred centuries ago (e.g. the Vikings) and names will be likely to have indigenised since then. Although at present many European countries have a large influx of immigrants, it is highly unlikely that any of them will have risen to the position of managing director of a MNC subsidiary. Most of them even struggle to get a job at all.

The choice was usually as simple as: is Carlos Gonzales de Castejon Spanish or Swedish, is Mikko Tanhuanpää Finnish or German or is Diarmuid O’Colmain Irish or French? Of course some combinations of headquarters and subsidiary countries proved to be more difficult. What about the difference between Swedish/Finnish/Danish/Norwegian names and between German/Austrian/Swiss names? For these groups, totaling some 50 cases, assistance was sought from local academics with a language background and/or extensive international experience. They were able to classify more than 70% of the names without having any doubts. The remaining 30% of these cases remained unclassified. Even more difficult was the combination of a British headquarters with subsidiaries in Canada, Australia and New Zealand. However, the sample included only 5 Canadian subsidiaries of a British headquarters, of which 2 were excluded because of missing data on one or more of the independent variables. The remaining three managing directors had a French name and were therefore concluded to be Canadian. Subsidiaries in Australia and New Zealand with a British headquarters (about 15 in total) were not classified.

The two previous major studies on MNC staffing practices asked HQ-managers to either indicate whether a particular region was "primarily staffed" by certain types of employees (Tung) or to indicate the overall percentage of PCN, HCN and TCN managers (Kopp). A specific problem with the first approach is that the answer is unlikely to be accurate. Our data show that most companies use PCNs as well as HCNs in many regions, so though easier for the respondent, this type of question looses many - possibly important - details. In addition, not all companies are likely to have detailed data on the nationality of managers of their subsidiaries readily available. More importantly, however, even if they have, respondents are unlikely to bother to consult them, especially if nationality policies are only one of the many issues included in the questionnaire, as was the case in both Tung’s and Kopp’s study. Therefore, instead of describing the actual situation the respondent might be responding with any of the following: his most recent experience, a particularly striking experience, a socially desirable answer or stated company policy, while the reality gradually emerged to be different. (Negandhi and Welge’s (1984) study showed large differences between stated policy and actual practice). Of course, the same problem applies for all mail questionnaires, since the researcher has no opportunity to verify or explore answers. Although we do not want to question the overall validity of Tung’s and Kopp’s studies, we do feel that in the case of staffing practices our research method provides data that are at least as accurate as questionnaire data. We verified what actually happens in subsidiaries, measuring emergent rather than espoused strategy.

Measures and statistical methods

The independent variables for the large scale archival sample were all measured using secondary data. The level of uncertainty avoidance was measured by Hofstede’s (1980) Uncertainty Avoidance Index score for each country. The higher this score, the higher the level of uncertainty avoidance. Kogut and Singh’s (1988) composite index, which summarizes the difference between two countries on each of Hofstede’s dimensions, was used to measure the overall cultural distance between home and host country. Size of the company was measured by the level of total sales. Data for this variable were gathered from the Directory of Corporate Affiliations. Political risk indicators provided by Political Risk Services (1996) were used to measure the level of political risk for all of the host countries included in the sample. Since higher indicators reflected a lower the political risk in a country, the variable was transformed so that a higher indicator reflected a higher political risk. The level of education in the host country was measured as the number of students (education at third level) per 100 inhabitants. Data for this variable were gathered from UNESCO’s statistical yearbook (Unesco, 1997). Cost of living data were gathered from the yearly comparison of costs of living in 56 cities as published by the Union Bank of Switzerland (UBS, 1997). Although the cost of living in major cities might not be representative for the country as a whole, it forms a good basis for international comparison, especially since many expatriates will live in these centers of economic activity. Cost of living for home and host country was compared. A dummy variable was created that took the value of 1 if the cost of living in the host country was higher than in the home country and took the value of 0 otherwise. A dummy variable of ownership was created that took the value of 1 if the subsidiary was majority-owned or wholly-owned by the parent and 0 if this was not the case. Data on the percentage of ownership were taken from the Directory of Corporate Affiliations. This source was also used for the reporting distance of the subsidiary to headquarters. This variable took the values of 1, 2 or 3, the higher the number the greater the reporting distance from headquarters. Hence value 1 meant that the subsidiary reported directly to headquarters, while value 2 and 3 meant a subsidiary reported through one or two intermediate levels respectively.

The independent variables for the smaller-scale mail survey sample were measured using a combination of primary and secondary data. The level of R&D intensity was measured by dividing a company’s R&D expenses by its total level of sales. The level of foreign sales was measured by dividing the MNCs foreign sales by its total level of sales. Data for both variables were gathered from annual reports. The entry mode (greenfield or acquisition) was measured by asking respondents whether the company had been acquired at any point in its existence and the variable took the value 1 if the subsidiary had been acquired and 0 if it was a greenfield. Subsidiary size was measured by asking respondents for the number of employees and subsidiary age was measured by asking for the year of foundation and subtracting this from the year in which the survey was conducted. Since we expected that many respondents might be hesitant to supply exact performance data (see also Roth, Schweiger & Morrison, 1992), we asked them to evaluate a subsidiary’s profitability in relation to other subsidiaries in the group. Subsidiary performance was measured using a five point Likert scale that asked respondents how their subsidiary’s profitability compared to other subsidiaries (far below, below, about the same, above, far above).

To explore the influence of the variables described above on the likelihood of a PCN as managing director, a binomial logistic regression analysis was conducted. Logistic regression is similar to linear regression but is suited to models where the dependent variable is dichotomous. The dependent variable staffing policy is captured by a dummy variable that takes the value of one if the managing director is a PCN and zero if the managing director is a HCN. The probability of a PCN as managing director is explained by the independent variables described above. As for linear regression, the regression coefficients estimate the impact of the independent variable on the probability that the managing director is a PCN. A positive sign for the coefficient means that a variable increases the probability of a PCN. A negative sign indicates the reverse.

The models were estimated with SPSS 7.51 using the maximum-likelihood method. The null hypothesis that none of the independent variables has a significant impact can be tested with the model c 2. When the model c 2 is significant, this null hypothesis can be rejected. A test whether specific independent variables have a significant impact can be based on the Wald statistic. Significance levels of separate independent variables based on the Wald statistic are indicated in the models in Table 5. The partial correlation of each predictor variable (its relative impact) with the dependent variable is indicated by R. R can range in value from –1 to +1. A positive value indicates that as the independent variable increases in value, so does the likelihood of a PCN as managing director. If R is negative, the opposite is true.

Results

Descriptive data

Of the 2689 managing directors included in our archival sample, 40.8% were parent country nationals. Table 3 shows the percentage of PCNs for MNCs originating from different countries and the percentage of PCNs in various subsidiary country clusters and industries. When we look at home countries, we see that subsidiaries of Japanese MNCs are much more likely to have a PCN as managing director than subsidiaries of European MNCs. The exact percentage of PCNs in subsidiaries of European MNCs does differ considerably across the various countries, however, from a low of 18.1% for Denmark to a high of 47.1% for Italy. At subsidiary level, the highest percentage of PCNs can be found in Latin America, Africa and the Far and Middle East, while expatriate presence is much lower in Canada and Western Europe and is particularly low in Scandinavia. In general, MNCs operating in the financial sector and the automobile industry show the highest percentage of PCNs as managing director. A low expatriate presence is found in some service industries and in "multidomestic" industries such as food.

 

Table 4 that describes the mail survey sample shows a similar overall result with 37.7% PCNs. The country and industry patterns are broadly similar to Table 3, although individual countries/industries might have a percentage of PCNs that differs from the archival sample. Given the small size of most of the individual country/industry samples this is to be expected, however.

 

The results above describe the percentage of PCNs in the managing director functions only. Although much less information was available for the other functions, we found the level of expatriate presence to be generally lower in other functions. Only 17.2% of the subsidiary finance directors (N=358) was a PCN, while this was the case for 10.1% of the marketing directors (N=218). The lowest percentage of PCNs, however, was found in the personnel director’s function (2.2%, N=92). In general, MNCs tend to have more PCNs for the managing director function than for any of the other functions.

Logistic regression results

The descriptive results above confirmed our hypothesis with regard to the PCNs in different functional areas. In this section, we will investigate the other hypotheses as summarized in Figure 1 by means of logistic regression. Table 5 shows these regression results. Model 1 presents the results for the large-scale archival sample, while model 2 presents the results of the smaller-scale mail survey sample that provided information on six independent variables that were not available for the large-scale archival sample. Both models have a high explanatory power, with a high and highly significant c 2. Another way to assess the performance of the maximum likelihood models is to measure the percentage of correct observations and compare it to the classification rate that would be obtained by chance (the baseline rate, which is equal to a2 + (1-a)2, where a is the proportion of PCNs in the sample). As Table 5 shows, our models predict the likelihood of a PCN as managing director better than a random model would, with classification improvements of 46.3% and 49.9%. These improvements lie well above the minimum improvement of 25% as suggested by Hair et al. (1995). In both models the specificity (its capacity to correctly predict HCNs) is very good (84.45% to 86.58%), while sensitivity (its capacity to correctly predict PCNs) is good (61.87% to 67.78%). Pseudo R-square measures confirm that both models have a good explanatory power.

As model 1 shows, the results for all of the independent variables included in the large-scale archival sample are in the expected direction and all coefficients are (highly) significant. As predicted, the likelihood of a PCN as managing director for subsidiaries is higher in MNCs that are headquartered in a country with a high level of uncertainty avoidance, in large MNCs and in a situation where there is a high level of cultural distance between home and host country. At the host country level, the level of political risk and the level of education are, as expected, positively and negatively related to the likelihood of a PCN as managing director. A HCN is more likely if the cost of living in the host country is higher than in the home country. Finally, majority-owned subsidiaries are more likely to have a PCN as managing director and the larger the reporting distance from headquarters the lower the chance of a PCN as managing director.

With regard to industry only the industries with a significantly higher or lower level of PCNs as managing director were included in the model. We find a high percentage of PCNs as managing director in the financial services, printing & publishing, telecommunications equipment and paper industry, while a low percentage of expatriates is found in the advertising and business & management services, the computers & office equipment, the electronic & electric equipment and the food industry.

The correlation matrix of the independent variables showed some degree of multicollinearity. In particular, the level of political risk in the host country was significantly correlated with the level of education in the host country (-.360), and with the cost of living differential (-.243) and the cultural distance (-.186) between headquarters and subsidiary. Further the level of education in the host country was significantly correlated with the level of cultural distance between headquarters and subsidiary (-.248). We therefore estimated two additional models that excluded the level of political risk and the level of education respectively. This did not change the signs or significance of the other variables.

 

In model 2 all variables that are significant have the expected influence on the choice between PCN and HCN. Three of the variables that were included in both the archival and the mail-survey sample (level of UAI in home country, size of the company and host country educational level) are significant in both samples, while another three variables (cultural distance, political risk in host country and cost of living differential) are insignificant in the mail survey sample. Of the six variables that were available in the mail-survey sample only, R&D intensity and internationalization are insignificant. All subsidiary characteristics were significant at the 0.05 level or better. As predicted, a PCN as a managing director is more likely in greenfields than in acquisitions, in larger subsidiaries, in younger subsidiaries and in under-performing subsidiaries. The result with regard to industry confirm the results in the archival sample with a higher percentage of expatriates as a managing director in the telecommunications equipment and paper industry and a lower percentage in the computers & office equipment and the electronic & electric equipment industry.

Again some degree of multicollinearity was present between the level of political risk in the host country and the level of education in the host country (-.416), and the cost of living differential (-.340) and the cultural distance (-.205) between headquarters and subsidiary, but a model excluding political risk did not result in changes to the signs or significance of other variables. Further, exclusion of the level of foreign sales, which was significantly correlated with the level of UAI in the home country (-.369), the size of the company (-.239) and R&D intensity (.252) did not show an impact on the signs or significance of other variables nor did the exclusion of the size of headquarters, which was significantly correlated with the level of UAI in the home country (-.218), R&D intensity (-.158), foreign sales (-.239) and the size of the subsidiary (.228).

Since Japanese MNCs are known to differ considerably from European MNCs on a number of characteristics apart from staffing practices, some of the effects might simply be a "Japan" effect. We therefore included a dummy variable in the analysis, indicating whether headquarters was in Europe or in Japan. Even when this variable was included, however, all variables kept their expected sign, although some became slightly less significant. Except for cost of living (p=.1569, one-tailed) and size of the company (p =.075, one-tailed), all variables were significant. Moreover, except for subsidiary education at 0.029 and reporting level at 0.0042 all variables were still significant at a 0.001 level or better.

Discussion and conclusion

With regard to home country, the descriptive results for Japan with 76.5% PCNs confirm both Tung’s and Kopp’s study in which the percentage of home country senior managers in Japanese subsidiaries was 70% and 74% respectively. In Tung’s study, around 57% of the senior managers in subsidiaries of European MNCs were parent country nationals, in Kopp’s study this was 48%, while in our study it is only 31%. This difference may be party due to differences in sample composition (e.g. Tung’s sample contained more non-Western subsidiaries) and research method (e.g. Tung asked respondents whether a particular region was "primarily staffed" by certain types of employees). We may also conclude, however, that while Japanese MNCs continue to rely strongly on expatriates, European MNCs have localized their subsidiary management in the past two decades. In contradiction to this, Mayrhofer & Brewster (1996) argue that European multinationals still organize their staffing policies along ethnocentric lines and put forward a spirited analysis of why they are right to do so. However, the empirical evidence the authors cite to support their claim for ethnocentric staffing policies is very scattered and rather contentious. Our results show very clearly, however, that there are large differences between MNCs headquartered in different European countries. In previous studies, these differences were not recognized as all European MNCs were considered as one single group.

At subsidiary level, our results generally confirm Tung’s (Kopp did not differentiate her findings according to subsidiary country) findings in that the highest percentage of PCNs can be found in Latin America, Africa and the Far and Middle East, while expatriate presence is much lower in Canada and Western Europe. Tung’s study did not include a separate category for Scandinavian subsidiaries and had only two (European) companies operating in Eastern Europe, so that a comparison in these categories is not possible.

Neither Tung’s nor Kopp’s study differentiated their findings according to industry, so that our study is the first to give information on expatriate presence in different industries. Some of the industry effects are easily explained. The control aspect will lead companies to employ a large percentage of PCNs in financial services, while the importance of knowledge of the local market will lead companies to employ a large percentage of HCNs in advertising, business & management services and the food industry. The results for the other industries are less straightforward and would merit further investigation in a more controlled sample.

As expected the percentage of PCNs in more locally-oriented functions such as personnel and marketing was lower than for the finance function, although the highest percentage of PCNs was found for the managing director function. For both German and Italian MNCs, however, the percentage of PCNs for the financial director function comes close to the percentage of PCNs for the managing director function. Even more interesting, British MNCs are the only ones in which the financial director in subsidiaries is more likely to be a PCN than the managing director is. This conforms perfectly to the well-known emphasis of British companies on financial results (Lane, 1995; Prowse, 1994). These results confirm our hypothesis that employing HCNs because of their knowledge of the local market and business practices, will be more important for functional areas where this local knowledge is particularly relevant. In general, the personnel function is the most localized function in MNCs, since it has to comply with local employment laws and regulations and has to adapt its practices to the different cultural background of employees. The marketing function is also locally oriented, although some MNCs are aiming at a more globalized approach. Of the three functions, the finance function would clearly be the one where intimate knowledge and feel for local circumstances is least important. In addition, many MNCs will feel the control function of expatriation is as important for the finance function as it is for the managing director function.

Except for R&D intensity and internationalization, all factors that were hypothesized to impact on the choice between HCN and PCN were statistically significant in the expected direction, often at high levels of significance. Since our hypotheses were firmly grounded in the reasons to choose for either PCNs or HCNs as identified in previous International HRM literature, our results give considerable support for the validity of these reasons. However, for one variable, cultural distance, the proposed direction of the relationship was ambiguous. A higher cultural distance could lead to more PCNs for control and communication reasons and more HCNs for reasons of local knowledge and PCN adjustment problems. Our data show that the first reasons seems to be more important for companies: a larger cultural distance leads to a higher felt need for control and communication and hence more PCNs. To investigate whether the effect of cultural distance might be different in different circumstances, analyses were performed at the level of individual industries and headquarters countries. Although the strength of the relationship differed, it was significantly positive for every industry. Concerning headquarters country, most countries showed a strong and positive relationship. For Finland, France and Italy, however, the relationship was very weak and non-significant, which might be the result of two equally strong opposite effects. A more strictly controlled design will probably be necessary to further explore the role of cultural distance in executive staffing practices. In the smaller-scale mail survey sample, cultural distance, the level of political risk and the cost of living differential were no longer significant. This might be caused by the fact that variability for these variables in the small sample was smaller than in the larger, archival sample. The latter sample in particular included a larger number of non-Western companies.

Our study shows that a differentiated approach to subsidiary management, as advocated by many scholars in the field of international management (see e.g. Bartlett & Ghoshal, 1989; Ghoshal & Nohria, 1992; Martinez & Jarillo), is important for staffing practices as well. Less than 10% of the companies in our sample had a uniform staffing policy (only HCNs or only PCNs). These companies were mostly Japanese MNCs in the financial sector, a category that on average had a very high percentage of PCNs as a managing director. Other companies differentiated their approach according to host country and subsidiary characteristics.

This study identified several key variables that influenced the choice for either a PCN or a HCN as a managing director for foreign subsidiaries. We have not been able to relate these particular choices to performance, and cannot therefore offer any firm conclusions concerning their effectiveness. In spite of this, our results could be helpful for newly internationalizing companies that would like to learn from their more experienced colleagues concerning subsidiary staffing policies. Our results can easily be linked with some recent attempts to model IHRM. Schuler, Dowling and De Cieri’s (1993) model links various exogenous and endogenous factors to SIHRM issues, functions and policies/practices. Since staffing is one of the SIRHM policies/practices and many of our independent variables resemble their exogenous and endogenous factors, our model can be seen as a further specification of their general model. Two other recent models (Rosenzweig and Nohria, 1994 and Taylor, Beechler and Napier, 1996) mainly focus on the similarity between the subsidiary HRM practices and headquarters HRM practices. Nearly all of the independent variables included in these models can also be found in Figure 1. This is not surprising since the factors that will lead decision-makers to choose for a HCN as managing director will also favor an adaptation of overall HRM practices to local circumstances, since both imply an emphasis on local responsiveness/embededness. Rosenzweig and Nohria’s (1994) study provides a strong evidence for this link, since in their study the presence of expatriates is negatively related to the extent of which subsidiary HRM practices resemble local practices.

Our study has several limitations. First, although we feel that our research method provides a valid alternative to interviews and questionnaires, we had to exclude US companies from the sample to limit the chance of misclassifications. We might therefore not be able to generalize the results to US MNCs. In addition, the fact that some independent variables were available only in the smaller-scale mail survey sample reduces our ability to draw any firm conclusions based on these variables. Finally, our study was cross-sectional while some of the variables would clearly benefit from a longitudinal analysis, e.g. are expatriates replaced by local managers if a subsidiary becomes more established and profitable or if the level of education and political stability in a country increases.

However, our study did provide the first large-scale, international study that systematically investigated a large number of factors that might influence executive staffing practices in foreign subsidiaries. Future studies might attempt to explore a more limited number of these factors in further detail using more strictly controlled samples. Once we succeed in describing accurately which factors do affect staffing decisions, we can start to investigate whether certain choices are associated with a higher effectiveness, thus providing companies with a prescriptive model for executive staffing practices.

Literature

Banai, M (1992): The Ethnocentric Staffing Policy in Multinational Corporations: A Self-Fulfilling Prophecy, The International Journal of Human Resource Management, vol. 3 no. 3 (December), pp. 451-472.

Barkema, H.; Vermeulen, F. (1997): What Differences in the Cultural Background of Partners are Detrimental for International Joint Ventures?, Journal of International Business Studies, vol. 28, no. 4, pp. 845-864.

Bartlett, C.A., Ghoshal, S. (1989): Managing Across Borders. The Transnational Solution, Boston, Mass.: Harvard Business School Press.

Borg, M. (1988): International Transfers of Managers in Multinational Corporations, Uppsala: Acta Universitatis Upsaliensis.

Borg, M., Harzing, A.W.K. (1995): Composing an international staff, in Harzing, A.W.K.; Van Ruysseveldt, J.: International Human Resource Management, pp. 179-204, London: Sage.

Boyacigiller, N. (1990): The role of expatriates in the management of interdependence, complexity and risk in multinational corporations, Journal of International Business Studies, vol. 21 no. 3, pp. 357-381.

Brewster, C. (1991): The Management of Expatriates, London: Kogan Page.

Briscoe, D.R. (1995): International Human Resource Management, Englewood Cliffs, NJ: Prentice Hall.

Derr, C.D., Oddou, G. (1993): Internationalizing Managers - Speeding Up the Process, European Management Journal, vol. 11 no. 4, pp. 435-442.

Directory of Corporate Affiliations: International public and private companies (1996): vol. 5, National Register Publishing.

Dowling, P.J., Schuler, R.S. (1990): International Dimensions of Human Resource Management, Boston: PWS-Kent.

Edstrom, A., Galbraith, J.R. (1977): Transfer of Managers as a Coordination and Control Strategy in Multinational Organizations, Administrative Science Quarterly, vol. 22 (June), pp. 248-263.

Erramilli, M.K. (1996): Nationality and Subsidiary Ownership Patterns in Multinational Corporations, Journal of International Studies, vol. 27, no. 2, pp. 225-248.

Fatehi, K. (1996): International Management, New Jersey: Prentice Hall.

Franko, L.G. (1973): Who Manages Multinational Enterprises?, Columbia Journal of World Business, vol. 8 (Summer), pp. 30-42.

Ghoshal, S., Nohria, N. (1993): Horses for Courses: Organizational Forms for Multinational Corporations, Sloan Management Review (Winter), pp. 23-35.

Groenewald, H., Sapozhnikov, A. (1990): Auslandentsendungen von Führungskräften. Vorgehenweisen internationaler Fluggesellschaften, Die Unternehmung, vol. 44 no. 1, pp. 28-42.

Hair, J., Anderson, R.E.; Tatham, R.L.; Black, W.C. (1995): Multivariate Data Analysis with Readings, Englewood Cliffs, N.J.: Prentice Hall.

Hamill, J. (1989): Expatriate Policies in British Multinationals, Journal of General Management, vol. 14 no. 4, pp. 18-33.

Harvey, M.G. (1983): The Multinational Corporation’s Expatriate Problem: An Application of Murphy’s Law, Business Horizons, January-February, pp. 71-78.

Harzing, A.W.K. (1995) The persistent myth of high expatriate failure rates, The International Journal of Human Resource Management, vol. 6, May, pp. 457-475.

Harzing, A.W.K. (1997) Response rates in international mail surveys: Results of a 22 country study, International Business Review, Vol. 6, no. 6, pp. 641-665.

Harzing, A.W.K (1999) MNC Staffing policies for the CEO-position in foreign subsidiaries. The results of an innovative research method, Brewster, C.; Harris, H. (eds.) International HRM: Contemporary Issues in Europe, Routledge, pp. 67-88.

Harzing, A.W.K.; Hofstede, G. (1996) Planned change in organizations: the influence of national culture, special issue of the Research in the Sociology of Organizations series, JAI Press, pp. 297-340.

Hendry, C. (1994): Human Resource Strategies for International Growth, London: Routledge.

Hillman, F., Rudolph, H. (1996): Jenseits des brain drain, WZB Discussion Paper no. FS I 96-103.

Hodgetts, R.M., Luthans, F. (1994): International Management, New York: McGraw-Hill.

Hofstede, G. (1980): Culture's Consequences. International Differences in Work-Related Values, London: SAGE Publications.

Hofstede, G. (1991): Cultures and Organizations: Software of the Mind, London: McGraw-Hill.

Howell, L.D.; Chaddick, B. (1994): Models of Political Risk for Foreign Investment and Trade, Columbia Journal of World Business, Fall,70-91.

Kalt, D., Gutmann, M., Groff, A.(eds) (1997): Prices and Earnings Around the Globe. An international comparison of purchasing power, Zurich: Union Bank of Switzerland. Economic Research Department.

Kenter, M.E. (1985): Die Steuerung ausländischer Tochtergesellschaften. Instrumente und Effizienz, Frankfurt am Main/Berlin/New York: P. Lang.

Kobrin, S.J. (1988): Expatriate Reduction and Strategic Control in American Multinational Corporations, Human Resource Management, vol. 27 (Spring), pp. 63-75.

Kogut, B., Singh, H. (1988): The Effect of National Culture on the Choice of Entry Mode, Journal of International Business Studies (Fall), pp. 411-432.

Kopp, R. (1994): International Human Resource Policies and Practices in Japanese, European and United States Multinationals, Human Resource Management, vol. 33 no. 4 (Winter), pp. 581-599.

Kumar, B.N., Scheimann, H. (1986): Japanische Fuhrungskrafte zwischen ensandten und lokalen Fuhrungskraften in Deutschland, Zeitschrift fur Betriebswirtschaftliche Forschung, vol. 38 no. 6, pp. 493-516.

Kumar, B.N., Karlshaus, M. (1992): Auslandseinsatz und Personalentwicklung. Ergebnisse einer empirischen Studie über den Beitrag der Auslandsentsendung., Zeitschrift für Personalforschung, vol. 6 no. 1, pp. 59-74.

Lane, C. (1995): Industry and society in Europe: Stability and Change in Britain, Germany and France, Aldershot: Edward Elgar.

Macharzina, K. (1992): Internationaler Transfer von Führungskraften, Zeitschrift fur Personalforschung, vol. 6 no. 3, pp. 366-384.

Martinez, J.I., Jarillo, J.C. (1991): Coordination demands of international strategies, Journal of International Business Studies (Third Quarter), pp. 429-444.

Mead, R. (1994): International Management. Cross Cultural Dimensions, Oxford: Blackwell.

Mayrhofer, W., Brewster, C. (1996): In Praise of Ethnocentricity: Expatriate Policies in European Multinationals, The International Executive, vol. 36, no. 6, pp. 749-778

Negandhi, A.R., Welge, M. (1984): Beyond Theory Z: Global Rationalization Strategies of American, German and Japanese Multinational Companies, Advances in International Comparative Management, Greenwich, Connecticut: JAI Press.

Negandhi, A.R. (1987): International Management, Newton, Massachusetts: Allyn and Bacon.

Pausenberger, E., Noelle, G.F. (1977): Entsendung von Fuhrungskraften in auslandische Niederlassungen, Zeitschrift fur betriebswirtschaftliche Forschung, vol. 29 no. 6, pp. 346-366.

Phatak, A.V. (1989): International Dimensions of management, Boston.

Political Risk Services (1996): International Country Risk Guide, July 1996, NY: Political Risk Services.

Prowse, S. (1994): Corporate Governance in an international perspective: a survey of corporate control mechanisms in the United States, the United Kingdom, Japan and Germany no. 41 (July), Basle: Bank for International Settlements, BIS Economic Papers.

Roessel, R. von (1988): Führungskräfte-Transfer in internationalen Unternehmungen, pp. div., Köln: Wirtschaftsverlag Bachem in Köln.

Ronen, S. (1986): Comparative and multinational management, John Wiley and Sons

Root, F.R. (1986): Staffing the overseas unit, in Famularo, J.J.: Handbook of Human Resources Administration, New York: McGraw-Hill.

Rosenzweig, P.M., Nohria, N. (1994): Influences on Human Resource Management Practices in Multinational Corporations, Journal of International Business Studies, vol. 25 no. 2 (Second Quarter), pp. 229-251.

Roth, K., Schweiger, D.M.; Morrisson, A.J. (1991): Global Strategy Implementation at the Business Unit Level - Operational Capabilities and Administrative Mechanisms, Journal of International Business Studies, vol. 22 no. 3, pp. 369-402.

Schaffer, R.A.; Rhee, J.H. (1996): Should an expatriate or a host-national be utilized to establish a foreign, wholly-owned subsidiary: an integration of economic theory and organizational theory concepts, Paper presented at the annual conference of the Academy of Management, Cincinnati, Ohio, 9-14 August.

Schuler, R.S., Dowling, P.J.; DeCieri, H. (1993): An Integrative Framework of Strategic international Human Resource Management, The International Journal of Human Resource Management, vol. 4, pp. 717-764.

Shane, S. (1993): Cultural Influences on National Rates of Innovation, Journal of Business Venturing, vol. 8, no. 1, pp. 59-73.

Shane, S. (1994): The Effect of National Culture on the Choice between Licensing and Direct Foreign Investment, Strategic Management Journal, vol. 15, pp. 627-642.

Scullion, H. (1991): Why Companies Prefer to Use Expatriates, Personnel Management, vol. 23 no. 11, pp. 32-35.

Søndergaard, M. (1994): Research Note: Hofstede's Consequences: A Study of Reviews, Citations and Replications, Organization Studies, vol. 15, 3, pp. 447-456.

Taylor, S., Beechler, S.; Napier, N. (1996): Toward an integrative model of Strategic International Human Resource Management, Academy of Management Review, vol. 21 no. 4 (Oct.), pp. 959-985.

Tung, R.L. (1981): Selection and Training of Personnel for Overseas Assignments, Columbia Journal of World Business, vol. 15 (Spring), pp. 68-78.

Tung, R.L. (1982): Selection and Training Procedures of U.S., European, and Japanese Multinationals, California Management Review, vol. 25 no. 1 (Fall), pp. 57-71.

Tung, R.L. (1987): Expatriate Assignments: Enhancing Success and Minimizing Failure, Academy of Management Executive, vol. 1 no. 2, pp. 117-125.

UNESCO (1997): Statistical Yearbook 1996, Lanham: UNESCO Publishing & Bernan Press.

Union Bank of Switzerland (1997) Prices and Earning Around the Globe, An international comparison of purchasing powerI, Zurich: UBS, Economic Research Department.

Van den Bulcke, D. (1984): Autonomy of decision making by subsidiaries of multinational enterprises, in Vandamme, J.: Employee Consultation and Information in Multination Corporations, pp. 219-251, London: Croom Helm.

Youssef, S.M. (1973): The Integration of Local Nationals into the Managerial Hierarchy of American Overseas Subsidiaries: An Exploratory Study, Academy of Management Journal, vol. 16 no. 1, pp. 24-34.

Welge, M.K. (1980): Management in deutschen multinationalen Unternehmungen. Ergebnisse einer empirischen Untersuchung, Stuttgart: Poeschel Verlag.

Wolf, J. (1994): Internationales Personalmanagement. Kontext - Koordination - Erfolg, Wiesbaden: Gabler, MIR Edition.

Zeira, Y.; Harari, E. (1977): Genuine Multinational Staffing Policy: Expectations and Realities, Academy of Management Journal, June, pp. 327-333.

 

  1. see e.g. Banai, 1992; Borg & Harzing, 1995; Briscoe, 1995; Dowling & Schuler, 1990; Edström & Galbraith, 1977; Fatehi, 1996; Hamill, 1989; Hendry, 1994; Hodgetts & Luthans, 1994; Mead, 1994; Negandhi, 1987; Phatak, 1989; Root, 1986; Tung, 1981, 1982, 1987.
  2. A handful of other studies has empirically investigated the choice between expatriate and local managers (Franko, 1973; Youssef, 1973; Zeira & Harari, 1977; Hamill, 1989; Scullion, 1991). Unfortunately, these studies focus on limited samples (e.g. two subsidiaries of one American MNC), provide very fragmented empirical information and do not usually discuss factors influencing staffing practices in any detail. Franko’s study did investigate a larger sample of companies, but discussed only one factor – the stage of internationalisation – as an explanation of staffing practices and was conducted more than 25 years ago.
  3. Although one could question the use of such a global descriptor at company level, previous studies (e.g. Barkema & Vermeulen, 1997; Erramilli, 1996; Kogut & Singh, 1988; Shane 1993 & 1994) have shown that national cultural variables measured as the countries’ scores on Hofstede’s dimensions have a considerable explanatory power for decisions taken at company level. Although Hofstede’s work has elicited some criticism, it is largely accepted as a helpful, though crude way to quantify cultural differences (see Harzing & Hofstede (1996) for a discussion of the various critiques and the extensive use of Hofstede’s dimensions in other studies; see Sondergaard (1994) for a summary of reviews, replications and citations)
  4. In this case it might be very difficult to asses the causal relationship: are under-performing subsidiaries attracting PCNs or do these subsidiaries perform under par because they are headed by a PCN? In addition, it might be difficult to measure this relationship in a cross-sectional analysis, since it would assume that PCNs would leave as soon as the subsidiary becomes profitable again. However, although this relationship would benefit from a longitudinal analysis, in a cross-sectional analysis, we would still expect a minor negative effect to be present.
  5. A comparable model derived from the literature was provided by Hamill (1989). He did not, however, provide any empirical tests for the proposed relationships and many of our home country/company and host country variables were not included in this model and nor was industry.
  6. In our smaller-scale mail survey sample, however, the nationality of the managing director was significantly related to the number of PCNs among top-5 managers in the subsidiary. In subsidiaries with a PCN as managing director, the average number of PCNs among top 5 managers was 2.54, while for subsidiaries with a HCN as managing director it was 0.51. This difference is significant at the p = 0.000 level, t-value –15.921. Therefore the presence of a PCN as managing director can be regarded as a proxy for the presence of PCNs among top managers in general.
  7. The three coders consisted of two MA-students and the author of the manuscript. Between them they speak 8 different languages and have lived in five different countries. Furthermore, all three coders studied International Management, a study with linguistic elements, have extensive experience in interacting with foreign students and share a keen interest in languages.
  8. The reason for comparing profitability with other subsidiaries rather than with other firms in the same industry is related to the purpose of the study for which the data in the small-scale mail survey were originally collected. However, this way of measurement might be very suitable for the staffing decision as well, reasoning that headquarters would be distributing a limited resource (expatriate managers) between their subsidiaries, based on their relative performance.
  9. Mayrhofer & Brewster cite two studies (Ondrack, 1985 and Derr & Oddou., 1993) that indicate that expatriation is still an important phenomenon. This statement, however, does not unambiguously show that a majority or even a major part of management positions in subsidiaries of European MNCs is taken by PCNs. The fact that all major European airlines use PCNs in top management positions in their sales branches and the fact that financial directors in Philips subsidiaries are nearly always PCNs, might very well be an industry or company specific anomaly. The last two cited references also provide rather contentious support. The fact that one particular large European company has only one nationality represented at the Board level (Bournois & Chauchat, 1990) hardly proves that European MNCs generally have ethnocentric staffing policies where their subsidiaries are concerned. Finally, Mayrhofer & Brewster report that Banai (1992) is convinced that "many MNCs, whether American, European or Japanese, still assign a large number of PCNs to key managerial positions in their subsidiaries and affiliates" (1992:451). Apart from the fact that assigning a large number of PCNs can still mean that the majority of subsidiary managers are local nationals, the publications that Banai cites (Tung, 1981, 1982 and Root, 1986) are rather outdated and even do not unambiguously support the statement. Tung shows that in American MNCs a substantial majority of senior management in subsidiaries is of local origin, while the same is true for Western subsidiaries of European MNCs. Root does not even offer figures about the extent of localisation of management, he only cites a 1978 Forbes publication which indicated that there were 100,000 American managers in subsidiaries of American MNCs. This of course hardly gives a clue about ethnocentricity in Europe if we don’t have the figures about non-American managers.
  10. In order to be able to compare our data with the only other large scale survey (Tung, 1981, 1982, 1987) that included data at host country level, we used broadly the same country clusters as Tung did.