A Review of Theories on Transnational Transfer of HR Practice within Multinationals

This article discusses the process of transferring human resource (HR) policy and practice internationally within multinational companies (MNCs), and the factors that influence the transfer process. The first section thoroughly surveys the literature on why MNCs transfer HR practices across borders and generalizes three lines of arguments. The second section looks at “what to transfer” with regard to particular HR issues, and points to a gap in the literature. The next section briefly reviews three main methods of transfer adopted by MNCs. Finally, the results of transfer are discussed both prescriptively and descriptively in light of Kostova’s (1999) multilevel model. The arguments presented in this article have two main implications which are summarized in the conclusion.

This section examines the reason for MNCs to transfer HR practice.The author believes it worth a detailed analysis, for it largely defines the scope for analyzing the following issues of the transfer process.Three generalizations of the reasons for transfer emerge from the literature.They are: international competition pressure, international integration and strategies, and organizational politics.

International competition pressure
As to the reason for transfer, one of the most developed arguments is that competition in the global economy on the basis of competitive advantages is the incentive for MNCs to transfer and recombine new knowledge and practices across borders (e.g., Bartlett and Ghoshal, 1995;Kogut and Zander, 1993;Taylor et.al., 1996;Kostova, 1999).As Kostova (1999, p.308) put it, "for purposes of synergy and efficiency, organizations often engage in cross-unit transfers of business practices that reflect their core competencies and superior knowledge and that they believe to be a source of competitive advantage".HR policies and practices are often considered by top management to be one of those sources.Moreover, they may stem from both national and organizational contexts.
Firstly, HR practices being transferred originate from national business system.Hall and Soskice (2001) believe that the existing social system of a nation generates national comparative advantage.Thus one can expect that MNCs tend to possess those 'superior' HR practices that are developed in support of their national comparative advantage and to transfer them to their subsidiaries worldwide.This can be regarded as a demonstration of the "country-of-origin" effects (Ferner, 1997;Edwards, 2004).
Secondly, HR practices may stem from particular organizational contexts and convey organizational strategies.It is obvious that the competences of different MNCs of the same country-of-origin are not identical.They may develop core competences which are highly firm-specific.Indeed, a number of Japanese MNCs have been identified with different core competencies in the literature (Bartlett and Ghoshal, 2000;Evans et.al., 2002).Similarly, Kostova (1999, p.309) has stressed the influence of "an organization's history, people, interests, and actions" on the strategic organizational practices being transferred.
It is noteworthy that transfers of HR practices can occur in various directions within the MNC, including transfers from parent companies to foreign subsidiaries, from subsidiaries to parent companies, or from one subsidiary to another.For example, German companies have been observed to adopt some 'Anglo-Saxon' tradition from their "vanguard" subsidiaries in UK (Ferner and Varul, 2000).This is what Edwards (1998) terms 'reverse diffusion', in which firms may look to 'reverse-transfer' policies associated with companies and countries that are dominant in the global economic system, since such policies are likely to be seen to be linked with international competitive advantage.This phenomenon could also be explained by the theory of dominance effects (Smith and Meiskins, 1995), which is consistent with the argument of competition pressure.The logic of this theory is that countries in dominant positions in the international economic system may also dominate HR practices at the international level as a result of international competition.That may explain why the post-war period saw US and Japan, in turn, act as the most influential power in the international HR areas.

International Integration and international strategy
A second general conclusion from the literature concerning the reason for transfer is that the degree of international integration and the corresponding strategies of MNCs largely determine the extent to which MNCs are likely to transfer their HR practices (e.g., Porter, 1986;Taylor et. al., 1996;Edwards, 2004).In other words, the transfer of HR practices is often linked to certain industry sectors.For example, those sectors such as automobiles, IT services and pharmaceuticals, with a high degree of international integration, are easy to transfer HR practices across borders.Furthermore, Edwards (2004, p.401-402) distinguishes between two forms of international integration: standardization and segmentation.He argues that "In those sectors in which MNCs have developed standardized operations, the transfer of employment practices is likely to be more attractive to management"; whereas "in MNCs which have segmented their international operations", even where the degree of integration is high, "there will be little incentive to transfer practices across borders".This is evident in many relevant studies.Examples include Japanese MNCs in Asian subsidiaries making low value-added products, which exhibit no transfer of Japanese employment relations (Dedoussis, 1995).
One can expect a close linkage between MNCs' degree of international integration and their international strategies.Porter (1990) identifies two generic MNC strategies: multidomestic and global.MNCs in sectors like foods and clothing tend to choose multidomestic strategies; whereas MNCs in sectors such as automobiles and pharmaceuticals are likely to choose global strategies.Linked with Porter's theory of MNC strategies, Taylor and her colleagues (1996, p.966-968) define three strategic international HRM (SIHRM) orientations: adaptive, exportive and integrative.They propose that MNCs following a multidomestic strategy will adopt an adaptive SIHRM orientation, hence making no transfer of HR practices; in contrast, MNCs following a global strategy will adopt an integrative SIHRM or an exportive SIHRM orientation, hence are more likely to transfer HR practices.Taylor et. al. (1996, p.969-970) also point to the importance of top management's perception of the context generalizability of parent company's HR competence, i.e. whether top management believe the competence can be used in other contexts outside the home country of the firm.If not, they argue, there will be no incentive to transfer its HR system across borders, regardless of the international strategy of the firm.Figure 1 illustrates Taylor's (1996) arguments.

Organizational politics
A third approach explaining the incentives for MNCs to transfer HR practices looks at political relationships within organizations (e.g., Edwards, 2004;Kostova, 1999).Edwards (2004, p.393) terms this explanation "the political approach", indicating that "actors in organizations can be willing to engage in the process of transfer as a way of obtaining legitimacy and to advance their own interests".Furthermore, Kostova (1999, p.319) bases this approach on resource dependence theory and institutional theory.She argues that MNC subsidiaries "may develop perceptions of dependence on the parent" due to various resources such as technology, capital, and promotion of the subsidiary staff, etc.She suggests that "under such conditions of dependency and intraorganizational competition", a subsidiary will try to implement parent company's practices as a way of gaining internal legitimacy.
A key contribution of this approach is to recognize the influence of organizational actors' personal motivations on MNC's decision to launch a transfer process.It is noteworthy that such concerns should not be confined to senior managers in organizations.Indeed, 'shopfloor' workers may also play a significant role in the process.Under certain circumstances, they can even suffocate a potential transfer at an early stage.In a further exploration, Edwards (2004, p.404) points out that "actors at plant level may be reluctant to share their expertise with their counterparts for fear of undermining their performance within the group".As a result, they may prefer to keep those practices instead of letting it transfer.
In summing up the research on the reasons for transfer, one can conclude that the transfer of HR practices within MNCs is the result of either external competition pressures or internal politics, or both, and is often linked with certain industry sectors and organizational strategies.

What to Transfer
Based on the above studies on why transfer, it is possible to generalize that MNCs will tend to transfer those HR competence that they look on as sources of their competitive advantage (e.g., Bartlett and Ghoshal, 2000) and at the same time, those they believe to be context generalizable (Taylor et. al., 1996).
Another line of argument draws attention to particular HR issues concerned.For instance, Rosenzweig and Nohria (1994, p.232-233) hypothesize that HR practices will tend to be transferred in the following order, starting with the one that will most likely to be transferred: (1) Participation; (2) Executive Bonus; (3) Training; (4) Gender Composition; (5) Benefits; and (6) Time off.This hypothesis has been generally supported by their following empirical test.They argue that this is because "practices for which there are diffuse and poorly defined local norms, or which are seen as being critical to maintaining internal consistency or arriving at critical decisions", are more likely to be transferred.Conversely, "practices for which there are well-defined local norms and which affect the rank-and-file of the affiliate organization are likely to conform most closely to the practices of local competitors" (Rosenzweig and Nohria, 1994, p.233).This argument highlights the internal differentiation of HR management practices within a MNC, instead of viewing it in terms of an overall orientation.It recognizes the significance of host country effects as well as organizational strategies.
Moreover, this study hints at some fact behind the scene.Since Rosenzweig and Nohria's (1994) study was conducted in U.S., whose employment system overall offers less paid time off and provide a lower percentage of benefits, MNCs from different nations in this study exhibit varied degrees of local isomorphism in their U.S. affiliates with regard to particular HR issues.For instance, close adherence to U.S. practices in benefits and time off makes German and Swedish MNCs far different from the very generous policies of their home countries.Whereas Japanese MNCs adhere to US practices much less, as they follow their home policies which offer even less paid time off than U.S.
Another particular HR issue indicating a similar problem is union recognition.U.S. MNCs are known to have sought to avoid collective employee representation even where there is strong institutional support for it (see Royle, 2002, for example).This is seen as the result of their long history of hostility towards unions.However, German MNCs also turn to non-unionism in their U.K. subsidiaries (e.g., Guest and Hoque, 1996), rather than transferring their own traditions of close cooperation with unions, which are thought to be a resource of German's national comparative advantage (Hall and Soskice, 2001).Though this behavior is seen as a result of their devolution of HR policies to local management decision, German MNCs' intentions in this regard deserve further consideration.
One possible proposition which can be induced from the above phenomena is, whenever and wherever possible, MNCs' subsidiaries, especially in relatively low-skilled sectors, would tend to pursue "low-road" HR practices, such as less paid time off, lower benefits, non-unionism, etc. in order to maximize their profits.Therefore, MNCs are likely to take this "principal" into consideration when they decide whether to transfer their home country practices, or to adapt to local polices with regard to certain HR issues.This proposition needs to be tested in deregulated host countries, especially in developing countries such as China.But existing research has failed to systematically explore this area.Thus the author believes it an interesting topic for further study.

How to Transfer
This section briefly reviews the literature on the mechanisms of transfer, focusing on three main methods adopted by MNCs.
One method of transfer is through formal policies and management control systems, allowing parent policies to be propagated and implementation monitored.This is common in U.S. companies.But these formal systems may allow practices transferred from subsidiaries to parent companies as well.For example, Marginson and his colleagues (1995) report that North American-based companies usually establish worldwide personnel policy committees.And these committees hold meetings of personnel managers from different countries, on an ad hoc basis or a regular basis.Labour performance comparisons at different sites would appear to be very much on the agenda of such meetings.Thus one can expect a two-way transfer of HR policies and practices between HQ and subsidiaries under such mechanisms.
Another mechanism is called "best practice" or "coercive comparisons" (e.g., Ferner and Edwards, 1995).As Marginson et. al. (1995) note, labour performance comparisons are getting important in management decisions on investment or divestment.Furthermore, Martin et.al. (1998) argue that internal and external benchmarking has been the most important and most widespread mechanism for the transfer of best practices, especially in MNCs with the same or similar operations at sites in different countries.However, it has been pointed out that coercive comparisons may intensify the pressures for non-managerial employees and cause tensions between management and labours (Ferner and Edwards, 1995;Marginson et. al., 1995;Martin et.al., 1998).
A third mechanism is through socialization of parent company's culture (Edstroem and Galbraith, 1977).Edstroem and Galbraith (1977) suggest that socialization represents the most complex in the control strategies used in MNCs.Evans (1995) has outlined a number of means by which multinational firms have attempted to integrate their corporate cultures, including mission statements, communications of the CEO, the process of building corporate charters, and management education, etc.While a most widely used method among these may be the transfer of personnel, or expatriates, between parent and subsidiaries.Japanese companies are said to rely heavily on the use of expatriate managers to monitor implementation of systems and cultures, hence obviating the need for very formal international control systems that are common in U.S. companies.Therefore, one may argue that the distinctions of the main method adopted by MNCs of different nations also exhibit some "country-of-origin" effects (Ferner, 1997).

The Results of Transfer
This section discusses first the factors that may influence the results of the transfer as a prescriptive analysis.Kostova's (1999) multilevel model is critically examined.Then the actual results of transfer are described in light of the multilevel model and finally the meaning of "hybridization" is discussed.

Factors that may influence the results of transfer
It has been noted that diverse factors, like national culture, employment law, IR institutions, firm strategy, etc., in both the home country and host countries of MNCs, have an impact on the results of HR transfer.One could assume that these factors roughly correspond to country-of-origin and host-country effects (e.g., Ferner, 1997;Edwards, 2004).Moreover, as Schmitt and Sadowski (2003, p.410) put it, "these factors are interrelated and they determine what package of personnel and labour practices is appropriate in a given situation".In order to study such complex organizational phenomena as the results of cross-national transfer of HR practices, Kostova (1999) develops a multilevel model, in which she proposes that three sets of factors at three levels ---country, organization, and individual ---affect transfer success reflecting social, organizational, and relational embeddedness.
Kostova's multilevel model is systematic and integrated.First, it incorporates the interaction of country-of-origin and host-country effects at country level, namely "institutional distance between home and recipient" (Kostova, 1999, p.313).For instance, it is proposed that the success of transfer of HR practice is negatively associated with the institutional distance between the countries of the parent company and the recipient unit.Second, it stresses the abilities and motives of subsidiaries to adopt the transferred practices at organizational and individual levels.For example, it is hypothesized that the success of transfer is positively associated with the recipient unit's commitment to, identity with, and trust in the parent company.
One more contribution of the model is, it points out that, transfers based on power/dependence relationships lead to implementation rather than internalization of the practice transferred.As discussed previously, one potential motive for transferring HR practice is the organizational politics (i.e., the actors in organizations can be willing to engage in the process of transfer as a way of obtaining legitimacy and to advance their own interests, see Edwards, 2004).Kostova (1999) argues that such motive usually leads to formal and ceremonial adoption, and is highly unlikely to achieve positive attitudes of the employees at the recipient unit toward the practice.This argument explains, in part at least, why the transfer practices may actually be implemented in form rather than in substance in the host environment.
The limitation of Kostoya's (1999) model lies in two facts.First, it defines the success of transfer as "the degree of institutionalization of the practice at the recipient unit" (Kostova, 1999, p.311).However, this definition might be narrow in some ways.For instance, as mentioned earlier, it is inferred in this model that the success of transfer of HR practice is negatively associated with the institutional distance between the countries of the parent company and the recipient unit.But a considerable body of evidence exists to suggest that practices transferred between countries with a high institutional distance may act as a trigger to change, hence be internalized with a different form or structure within the recipient unit (e.g., Scarbrough and Terry, 1998).Thus the transfer may not be looked on as a 'failure' in that it brings about some improvements within the company.Second, this model is highly prescriptive, and may not apply to all circumstances, especially in heterogeneous social contexts.This is evident in transition economies such as China, where the old institutional systems are collapsing while the new ones have not been fully established yet.

The actual results of transfer
Consistent with the implications of Kostova's (1999) multilevel model, existing field studies suggest that the results of transfer exhibit combined effects from all three levels of contexts.First, they provide substantive support for the notion that results differ at country level.For instance, Doeringer et.al. (2003, p.266-267) contend: "patterns of transfer and accommodation differ systematically from country to country in ways that suggest new management practices are blended with traditional practices to create distinctive national 'hybrid' management regimes".Second, there is evidence to support organizational and relational embeddedness.As Scarbrough and Terry's (1998) study on the British response to Japanese-style management practices indicates, in the process of "hybridization", managements have been influenced by a number of factors; not only by the context of national institutions, but also by company and plant-level path dependencies, and by the persistence of trade unionism as an active force amongst their workforces.
With regard to the different results of transfer at country level, the transfer of Japanese-style high-performance management practices -usually termed "lean production"-may have drawn the most intensive interest in the existing field studies (e.g., Scarbrough and Terry, 1998;Doeringer et.al., 2003).Surveys (Doeringer et.al., 2003) shows that the overall adoption rates of Japanese-style practices are higher in the US than in the European countries.It is explained partly by the relatively laissez-faire US industrial relations system which poses fewer barriers to the transfer of efficient management practices compared with European countries.Whereas in the UK, despite a significant decline in their influence, unions remain an obstacle to adopting "lean production" (Doeringer et.al., 2003).This finding is consistent with Scarbrough and Terry's (1998) study of the continuing influence of the institutions of collective bargaining over events at plant-level in UK.
The term "hybridization" depicts the way transferred practices are actually implemented in the host environment.From this perspective, differences in national IR systems and in organizational relations may block the transfer of particular management practices, or they may force modifications in practices to make them compatible with other parts of the national IR system (e.g., Doeringer et.al., 2003).One can expect that the "modification" is not merely to tinker with the transferred practice, but a process that demands creation and innovation.As mentioned earlier, Scarbrough and Terry (1998) point out that lean production may act as a trigger to workplace change.This argument acknowledges the creative side of the transfer process in the host environment, that is, the tendency for transferred innovations to be selectively re-invented within different social and organizational contexts.Therefore, they argue that "the pattern of change in established British producers should be viewed not as a diluted form of Japanisation nor as a minor set of system modifications but as a creative process of adaptation" (Scarbrough and Terry, 1998, p.235).In other words, recipient units would evolve their historical legacy of employee relations and managerial practice into new regimes.This argument might add to the meaning of "hybridization".

Conclusion
This article has attempted to develop an analytical framework for examining the process of transferring HR practice internationally within MNCs by looking at the incentives, the contents, the mechanisms, and the results of the transfer in turn.The arguments presented in this essay have two principal implications.First, when looking at the pictures as a whole, a multilevel approach may be appropriate for studying the process of cross-national transfer of HR practices.Indeed, one can conclude from the previous sections that the reason for transfer may stem from national, organizational and relational contexts.And the contents, methods and results of transfer can also be analyzed from the three levels.Second, HR practices in MNCs' subsidiaries in developing countries or transition economies deserve more systematical exploration.For instance, one may hypothesize that MNCs tend to pursue "low-road" HR practices in their subsidiaries in deregulated countries, and this needs further empirical tests in developing countries.Moreover, a different framework may be needed to study the cases in heterogeneous social contexts like transition economies.