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.