تجزیه و تحلیل هزینه های معامله از تصمیم گیری نیروی انسانی در عملیات بین المللی

This paper analyzes staffing decisions in foreign subsidiaries from the perspective of transaction cost theory. We focus on the ex post transaction costs of the employment relation. Specifically, we look at the monitoring, bonding, maladaptation, and bargaining costs of conducting activities in specific subsidiaries in a foreign country. We hypothesize that the transaction costs of using expatriates are lower than those generated by local employees, especially in the higher managerial echelons of foreign subsidiaries, but also that costs can be reduced as individuals become more experienced. We also conjecture that ex post transaction costs are influenced by cultural differences between the host and the home countries, and by characteristics of the companies and their subsidiaries. The framework is empirically corroborated by survey data on a sample of 145 Norwegian MNCs.

One of the central decisions that multinational companies (MNCs) must make is whether to select expatriate or local personnel to manage and work in their foreign subsidiaries. In spite of the alleged globalization of businesses, many MNCs continue to be ethnocentric in their staffing policies (Mayrhofer & Brewster, 1996). This is effectively the case in Japanese companies: according to a survey by Kopp (1994), 74 percent of their subsidiaries’ directors are expatriates, whereas European and US MNCs are less ethnocentric, with expatriates occupying 54 percent and 48 percent of senior management posts, respectively. Similar results were obtained by Tung (1982), and more recently by Harzing (2001) and Peterson, Napier, and Shim (2000) who also find that while there is substantial cross-national variation regarding staffing practices in foreign subsidiaries, US and European MNCs in particular seem to have reduced their use of expatriates somewhat over time. There are problems associated with an ethnocentric posture (Perlmutter, 1969). It has been noted that it creates problems of adaptability to foreign environments and cultures, leads to high failure rates, has a discouraging effect on local management morale and motivation, increases the “foreignness” of the subsidiary, and may involve high salary costs (see e.g. Black & Gregersen, 1999; Welch, 1994). Overall, international assignments seem to be an expensive option for most companies, and it is estimated that the costs of expatriates can amount to more than three times that of local employees (Gates, 1996). If we focus on costs, we must ask ourselves the following question: If local personnel are a cheaper recruitment option, then why is it that, in a business context defined by the need to cut costs, multinational companies continue to make extensive use of international assignments? Research on foreign assignments dates back to the seminal study of international transfers by Edström and Galbraith (1977), but previous studies have tended to be descriptive and/or to have an operational and practical orientation. They lack analytical rigor and as pointed out by Harzing (2001, p. 140) “…few authors offered a theoretical framework for their investigation”. To advance our understanding of the complexities of staffing patterns in MNCs, we believe that it is appropriate to base such examinations on solid theoretical ground. The paper analyzes that question from the perspective of transaction cost economics (Williamson, 1985). Although the organization of labor has been seen as a key topic in transaction cost theory (Masters & Miles, 2002; Williamson, Wachter, & Harris, 1975; Williamson, 1985), it is only an emergent theoretical lens for analyzing the use of expatriates within MNCs (Tan & Mahoney, 2003). As far as we know, the present study is the first to use transaction cost theory as the basis for an empirical analysis of staffing decisions in MNCs. After explaining why this theoretical perspective is particularly suited to examining this question, we analyze the overall costs that are implicit in managing personnel in foreign subsidiaries, above and beyond the salary-related costs to which the traditional expatriate literature limits its attention. In this paper we focus on the ex post transaction costs of the employment relation. Specifically, we look at the monitoring, bonding, maladaptation, and bargaining costs of conducting activities in specific subsidiaries in a foreign country. We propose that the costs of using expatriates are lower than those generated by local employees, especially in the higher managerial echelons of the foreign subsidiary, and we develop a testable model of factors that influence the level of intra-company transaction costs. The framework is empirically corroborated by survey data on a sample of 145 Norwegian MNCs. Altogether, the study makes a contribution towards a transaction cost explanation for why companies sometimes apply ethnocentric staffing policies. The remainder of this paper is organized as follows. The next section develops a theoretical argument for transaction costs differentials between local employees and expatriates, and discusses those factors that may have an effect on levels of intra-company transaction costs. The data collection and measurement procedures are then described, followed by a report of the empirical results. The final section discusses the findings of the study, points out its limitations, and suggests directions for further research.

This paper has looked at staffing decisions in foreign operations from a transaction cost perspective. While recognizing that staffing decisions are complex and that they involve social as well as economic considerations, we focus in this paper on the ex post transaction costs of the employment relation. We expected the costs of using expatriates to be lower than those generated by local employees, especially in the higher managerial echelons of foreign subsidiaries, but also that the managerial experience of individuals would help to reduce transaction costs. Furthermore, we expected transaction costs to be influenced by cultural differences between the host and the home countries, and by characteristics of the companies and their subsidiaries. To the best of our knowledge this is the first systematic empirical analysis of the transaction cost ramifications of international staffing decisions. We demonstrate that transaction cost economics provides a solid theoretical foundation for examining human resource decisions, arguably among the most important issues that internationalizing firms must deal with. Our empirical findings indicate that transaction cost economics is indeed a useful approach to such issues. Nevertheless, it should be noted that our analysis is limited. Our emphasis is on ex post transactions costs. We concur that ex ante costs (e.g. selection and training costs) as well as other components (e.g. wages, turnover, promotion) of staffing decisions also deserve attention, but a more complete analysis must be left for future studies. Another main contribution of this study is that it actually measures intra-company transaction costs. Even though transaction cost economics has spurred a tremendous amount of research activity over the last three decades, the empirical and measurement aspects of that research are still being developed. Transaction cost constructs have been operationalized and measured before (e.g. Dahlstrom & Nygaard, 1999 ), but this is the first time data on such costs have been collected in an international HQs-subsidiary context. Structural equations analysis substantiates by and large that the measurements are satisfactory as regards validity and reliability. Nonetheless, there is scope for improvement on the operationalization of bonding costs, and the measurement of maladaptation costs—albeit based on existing measures ( Dahlstrom & Nygaard, 1999 )—may not fully tap the theoretical construct. In terms of empirical results, the following findings emerge as noteworthy. First, our data indicate that cultural differences drive up ex post transaction costs. It is almost a credo in international business studies that operating in foreign contexts exposes companies to an array of challenges, and that these tend to increase, the more unfamiliar the business setting in question. Our findings support those views. Second, our study suggests that the managerial experience of the CEO works in the opposite direction: CEO experience drives down ex post transaction costs. This finding makes intuitive sense, but it remains to be investigated in more detail exactly why transaction costs are inversely related to managerial experience. One possibility is that managers, as they get more experience, become better at reducing intra- company conflicts and make more effective use of various formal and informal organizational mechanisms. Developing managerial competence also means that individuals gradually develop a valuable reputational asset, which they do their best to protect: the more valuable it is, the harder they work to safeguard it. Another likely explanation is that in most countries, especially outside the Nordic cultural sphere, age and experience are qualities to be expected of a manager. Novice managers may be seen as lacking the necessary legitimacy and authority to perform well in their positions, to a great extent regardless of whether they are home, host or third country nationals. The issue then is not so much that older managers, as individuals, are generally more competent, or have a better understanding of managerial issues than younger ones, but rather that those being managed as individuals and/or as a collective have their preferences tinted by norms based on seniority. A third possible explanation is that experience-related performance differences are simply due to a selection process. Many individuals may be given a chance to take a managerial position at some point in their careers. Some will fail outright, others will perform moderately well, and a few will show highly satisfactory results. Since it is these last that are the most likely to be repeatedly depended upon to take on managerial positions, experience becomes a proxy for differences between individuals regarding their managerial potential from the start. Third, our study did not provide clear-cut empirical results regarding the implications of staffing decisions for ex post transaction costs. From our framework we deduced that transaction cost differentials should increase at higher hierarchical levels. Hence, it was not unexpected that the coefficients for the dummy capturing the expatriate component of the overall work workforce were not significant. However, most of the results for higher-level positions were also non significant. One explanation is that expatriate CEOs are sometimes relied upon precisely because a subsidiary is experiencing or anticipating troubled times. Our data provide some support for this line of reasoning although alternative explanations cannot be ruled out. Fourth, we find that while larger subsidiaries generate fewer intra-company transaction costs than smaller ones, the opposite applies to the larger corporations of which they are a part. We propose that whereas larger companies generally tend to incur additional costs due to their complexity, large subsidiaries enjoy a privileged position within their companies, thus receiving more resources and attention than smaller, less important subsidiaries. Overall, the present study demonstrates that transaction cost economics can elucidate a phenomenon, namely international staffing policy, which has been regarded as lying outside the traditional sphere of economics. Nevertheless, our results must be interpreted with caution. As the first to examine the transaction cost effects of staffing decisions in MNCs, this study is exploratory in its nature. One important limitation of the study is its empirical context. The use of data on Norwegian MNCs made good sense in terms of collecting data of high quality, but the low degree of staffing ethnocentricity among Norwegian MNCs constrained our empirical analysis. There were relatively few cases of expatriates, especially at the lower hierarchical levels. Lack of variation may possibly explain why some of theresults were inconclusive. Further, the use of Norwegian data obviously limits the extent to which our findings can be generalized to other contexts. Future research could proceed in various directions. One would be to extend the analysis, theoretically and empirically, to examine other national contexts and to include not only the ex post but also the ex ante transaction costs of staffing, as well as other economically relevant decision parameters such as wages, output, and turnover rates. Another is on the measurement side. More valied and reliable measures could be tried out, especially for the bonding and maladaptation costs constructs. Moreover, in the present study (ex post) transaction costs were taken for the most part as a summary construct, despite the recognition (as evidenced in their measurement) that there are several different types of transaction costs. An implicit assumption in our analysis is that the different types of transaction costs can be meaningfully added into a concept of total costs. This would mean among other things that the various types of ex post transaction costs should carry identical weights when they are incorporated in a total cost concept. This cannot of course be taken for granted.