Global HRM integration: a knowledge transfer perspective

The Authors

Adam Smale, University of Vaasa, Vaasa, Finland

Abstract

Purpose – The purpose of this paper is to review the literature on HRM in MNCs from a knowledge transfer perspective, to identify some of the key weaknesses in extant research and to outline a future research agenda.

Design/methodology/approach – In this paper key contributions to the literature on HRM in MNCs are reviewed and discussed in connection with the knowledge transfer literature. The review and discussion culminate in the presentation of an agenda for future research.

Findings – The paper finds that, by viewing global HRM integration in MNCs as a process of knowledge transfer, three weaknesses in the extant literature are identified: inadequate consideration of HRM's knowledge characteristics at different levels within the organisation – the “what”; insufficient attention paid to the organisational mechanisms through which HRM is transferred – the “how”; and a lack of rigour in assessing when HRM transfer can be viewed as successful – the “when”.

Originality/value – In viewing global HRM integration as a process of knowledge transfer, the paper extends the HRM-knowledge link and provides an alternative point of departure from which to study HRM in MNCs. In highlighting some of the weaknesses in extant research and in proposing a research agenda, it is also hoped that this paper can assist other scholars in making incremental improvements to a field reportedly in need of further theoretical development.

Article Type:

Research paper

Keyword(s):

Knowledge transfer; Human resource management.

Journal:

Personnel Review

Volume:

Number:

Year:

2008

pp:

145-164

Copyright ©

Emerald Group Publishing Limited

ISSN:

0048-3486

Introduction

Contributions to the literature on HRM in MNCs have proliferated over the past decade and have come to comprise a large number studies on the attempts by MNCs to determine, by influence or design, the HRM practices of foreign subsidiaries (Kamoche, 1996; Lu and Björkman, 1997; Bae et al., 1998), or more specifically the transfer of HRM practices across borders (Beechler and Yang, 1994; Hannon et al., 1995; Tayeb, 1998; Hetrick, 2002; Myloni et al., 2004). However, the growing field of HRM in MNCs has not escaped criticism. Cited weaknesses include an over-emphasis on empirical investigation at the cost of theoretical development (Jain et al., 1998), too strong a focus on explaining grand tendencies rather than complex processes that lead to HRM convergence or divergence (Quintanilla and Ferner, 2003), as well as inadequate conceptualisations of national effects, the lack of attention attributed to organisational politics and insufficient focus on the internal division of labour (Edwards and Kuruvilla, 2005).

More recently in the field of HRM in MNCs, the organisational challenges that have transpired out of the processes of globalisation have engendered a call for effective global HRM and an efficient global HR function (Evans et al., 2002; Sparrow et al., 2004). Insofar as MNCs must make important strategic choices about whether to globally integrate or locally adapt HRM (Schuler et al., 1993; Rosenzweig and Nohria, 1994), forces of globalisation would appear to be turning the spotlight, for academics and MNCs alike, onto the global integration answer to the global – local question. Recurring themes in the espousal of global HRM and thus global HRM integration is the continued strategic realignment of HRM as an enabler of the overall business strategy, and HRM's key role in identifying and facilitating the flow of knowledge and talent throughout the firm (Roberts et al., 1998; Brewster et al., 2005).

Limited in quantity and scope, most of the extant literature linking HRM and knowledge has concentrated on the HRM-related challenges emerging from the increase in knowledge-intensive organisations (e.g. Sparrow, 2006). This paper extends the present discussion on HRM and knowledge by viewing global HRM integration itself as a process of knowledge transfer. Indeed, to the extent that global HRM integration within the MNC involves the internal transfer of knowledge that is embedded in HRM practices, then the knowledge-based literature provides an instructive point of departure. More specifically, the emphasis on an organisation's capacity to integrate and replicate specialised knowledge (Zander and Kogut, 1995; Grant, 1996a) as well as the factors that render the resulting transfer process “sticky” (Szulanski, 1996), draw particular attention to the characteristics of knowledge, the means through which it is transferred and an appreciation of when knowledge transfer has or has not been successful. All these themes are argued here to be pertinent yet lacking in studies of global HRM integration in MNCs.

Accordingly, the aim of the present paper is to review the literature on HRM in MNCs from a knowledge transfer perspective. Based on this review, some of the key weaknesses in the extant research are identified, which go on to form the foundations of a proposed agenda for future research. The paper seeks to contribute to the literature on HRM in MNCs by offering a review of extant research and by extending the HRM – knowledge link to include how global HRM integration can be viewed as a process of knowledge transfer. Furthermore, it is hoped that the paper can assist other scholars to make incremental improvements to a field that exhibits an imbalance in the volume of empirical research compared to parallel, theoretical development.

The next section discusses the drivers behind global HRM integration and presents the case for applying a knowledge transfer perspective. Subsequent sections go on to review the literature on HRM in MNCs along themes informed by the knowledge transfer discipline. After highlighting the key weaknesses in the extant literature the concluding section outlines an agenda for future research.

Global HRM integration in MNCs

For the purpose of clarity, this paper distinguishes between global “standardisation” and “integration” which risk being used interchangeably. One likely cause of confusion is the inconsistent references to the Bartlett and Ghoshal (1989) typology regarding organisational forms. Whilst the “global” firm is represented in their framework by firms that centralise facilities and standardise products and procedures, the term “global” today has assumed a more generic meaning to reflect the scale and scope of a firm's operations. Global integration is thus referred to here as the coordination and control of business operations across borders (Cray, 1984). Accordingly, whereas coordination refers to the forging of linkages between geographically dispersed units to reach an agreed goal (Martinez and Jarillo, 1989), control refers to the regulation of subsidiary activities by headquarters in an attempt to align them with pre-set expectations (Prahalad and Doz, 1981). Global HRM integration does not, therefore, refer to the complete standardisation of HRM, but rather to attempts by MNCs to achieve consistency in the delivery of HRM in foreign affiliates.

Evans et al. (2002) describe the link between global strategic pressures and implications for global HRM in terms of their shared goals of sustainable competitive advantage. In particular, they identify shifts towards greater operational effectiveness, an increased focus on the intangible assets of human capital, the development of distinctive organisational capabilities and the transferring and recombining of knowledge. Similarly, Sparrow et al. (2004) identify a set of “HR Enablers” based on the organisational challenges responsible for the globalising of HRM. These consist of HR affordability, a central HR philosophy comprising global themes, and HR excellence and knowledge transfer.

It would appear, therefore, that a recurring theme in the globalising of HRM is the continued strategic realignment of HRM as a means of delivering overall global business strategy, and HRM's key role in identifying and facilitating the flow of knowledge and talent throughout the firm (Roberts et al., 1998; Brewster et al., 2005). Building on this existing link between HRM and knowledge transfer, a stronger justification for adopting a knowledge transfer perspective on global HRM integration is turned to next.

A knowledge transfer perspective

When the resource-based theory of the firm focuses specifically on the management of human intangible assets, the subject of individual and organisational knowledge is argued to represent an increasingly critical resource of the firm (Grant, 1996b). Despite the reportedly fraught process of transferring knowledge about best practices throughout the firm (Szulanski, 1996), the strategy literature nevertheless presents that the ability to transfer and integrate knowledge will significantly affect organisational performance (Grant, 1996a; Martin and Salomon, 2003). However, the knowledge transfer literature has prescribed a largely supervisory role to the HR function. Indeed, certain HRM practices have been identified as key enablers in the knowledge transfer process in general (Conn and Yip, 1997; Laursen and Mahnke, 2001; Minbaeva et al., 2003), and more specifically in their capacity to remove knowledge sharing barriers amongst employees (Swart and Kinnie, 2003) and to retain expertise (Bender and Fish, 2000).

This paper extends the current discourse linking HRM and knowledge by viewing global HRM integration as a process of knowledge transfer. It is argued here that a knowledge transfer perspective is pertinent to the field of HRM in MNCs in general, and to global HRM integration in particular, through a shared focus on the ability of firms to integrate dispersed and specialised knowledge throughout their business operations (Zander and Kogut, 1995; Grant, 1996a) and the range factors which render the transfer problematic or “sticky” (Szulanski, 1996). Furthermore, since HRM is regarded as a context- and culture-specific resource (Tayeb, 1998), that increases in value as it is applied and learned in particular organisational settings (Tyson, 1999), and resides in a firm's routines, practices, norms and values (Flood et al., 2003), then HRM can be assumed to possess knowledge characteristics.

In essence, the knowledge transfer literature draws specific attention to the characteristics of knowledge, the means through which it is transferred and when such transfers can be considered successful. Accordingly, the review of extant literature on HRM in MNCs is structured along three themes. First, HRM knowledge transfer content (the “what”) looks at the characteristics of HRM knowledge, where HRM knowledge resides within organisations and how this applies to the research on global HRM integration. Second, the HRM knowledge transfer process theme (the “how”) explores the extent to which the literature has incorporated studies on the mechanisms of global HRM integration. Lastly, HRM knowledge outcomes (the “when”) reviews how global HRM integration performance has been conceptualised and assessed. The structure of the literature review is represented diagrammatically in Figure 1.

HRM knowledge transfer content

The first theme pertaining to the knowledge transfer perspective refers to the knowledge characteristics of HRM (the “what”), where these characteristics manifest themselves within organisations and how HRM knowledge characteristics have relevance for studies on HRM in MNCs.

In the global context of cross-border transfers of knowledge, whilst cultural variations are inevitably going to affect such transfers, it is nevertheless important to consider the type of knowledge being transferred in its own right when assessing transfer effectiveness (Bhagat et al., 2002). The often-cited definition of knowledge, which is directly relevant to the issue of transferability, was proposed by Polyani (1962) and later developed by Nonaka and Takeuchi (1995). They distinguish between “explicit” and “tacit” forms of knowledge, where “explicit” refers to knowledge that is documented, structured and easily transferable, and “tacit” refers to knowledge that resides in the human mind, manifests itself in behaviour and perception and is hard to transfer. Owing to the largely contradictory pressures for knowledge to be inimitable and transferable, tacit knowledge transfers to foreign locations have been described as “promising and problematic” (Martin and Salomon, 2003: 360). Empirically, albeit relating to marketing know-how, Simonin (1999) reaffirms that the degree of tacitness is the most significant determinant of knowledge transferability.

Specifically in relation to HRM, Tayeb (1998) asserts that since HRM falls into the category of a context- and culture-specific resource (as opposed to it possessing universal qualities), its successful transfer will be problematic but not impossible. However, it is pertinent to make the distinction here between the characteristics of knowledge residing in individual HRM practices and those knowledge characteristics residing in configurations or systems of HRM. One common distinction derived from the resource-based view is the strategic issue of resource inimitability and its implications for competitive advantage. In this sense, whilst certain individual HRM practices can be replicated to some degree in other organisational settings, it is the tacit and inimitable knowledge inherent in systems of HRM that can provide firms with an important source of competitive advantage (Lado and Wilson, 1994; Kamoche, 1996). Flood et al. (2003) further theorise about the various characteristics of knowledge that are manifest in HRM systems or bundles of HRM practices, concluding that “the knowledge involved in the diffusion of a human resource management system is likely to be complex, tacit and systemic and hence, it will be more difficult to transfer and absorb” (p. 62).

Fundamental to this knowledge-based perspective is the acknowledgement that HRM practices are based on HRM knowledge, which can be both explicit and tacit in nature. Indeed, Cerdin (2003) interprets one source of this HRM knowledge tacitness as residing in knowing how to apply and implement the more explicit forms of HRM knowledge manifest in routines and practices. Similarly, Tyson (1999) argues that HRM knowledge cannot be separated from its creation since it represents the learning of how to apply theories and concepts (i.e. “process skills”) in an organisation-specific context. Collectively, this serves to highlight important manifestations of HRM knowledge tacitness which can refer to the degree of difficulty in not only transferring the know-what (e.g. recruitment procedures), but also in transferring the know-how (use of multiple recruitment channels) and the know-why (e.g. attracting skilled individuals from a specific talent pool).

Based on the previous discussion of HRM knowledge characteristics, it can be concluded that studies on global HRM integration should account for differences in HRM knowledge characteristics:

across organisational levels; and
across and within individual HRM practices.

Accordingly, two important assumptions are made in that first, HRM represents a collection of individual HRM practices which are transferable in their own right, and second, that these practices manifest themselves at different organisational levels ranging from a high-level HRM philosophy to a specific HRM process.

HRM knowledge at the organisational level

An important distinction regarding HRM knowledge content is the organisational level at which global HRM integration might take place. Schuler (1992), for instance, suggests that an HR function operates at multiple organisational levels, namely “philosophy[rdquo ],; “policy”, “program” “practice” and “process”. Similarly, Tayeb (1998) argues that in addition to the internal differentiation of HRM functions, there exists a qualitative difference between a company's overall HRM “philosophy” and “policies” compared to its HRM “practices”. Supported empirically, Tayeb (1998), demonstrates that whilst company-wide HRM policies may typically be easier to implement, MNCs are often not able to avoid cultural intricacies and local adaptation at the lower organisational level of HRM practices and behaviour.

Ferner (1997), in developing an analytical framework for examining country-of-origin issues of HRM practices in MNCs, also argues for the need to focus on different levels within the corporation, in particular between the global corporate level where strategic decision-making takes place and the operational level where significant influences exist from the local business culture. This is necessary, it is argued, since corporate systems tend to overstate the strength of the country-of-origin factor, whereas operational systems tend to understate it. Bae and Rowley (2001) comment further on the tendencies for convergence studies to concentrate on macro-level variables of HRM systems (e.g. structure and technology), whereas divergence studies have tended to highlight micro-level variables (e.g. individual behaviour). Consequently, in order to illustrate the effect that organisational levels of HRM might have on transferability or the interpretation of convergence and divergence, they distinguish between three levels of HRM systems; “System Architecture”, “Policy Alternatives” and “Practice Process”.

From a knowledge transfer perspective, the organisational level of HRM is pertinent to efforts at its global integration since HRM at these different organisational levels is likely to possess different knowledge characteristics. More specifically, if HRM at the lower organisational and operational levels of “practice” and “process” is more susceptible to the influences of local business cultures, institutions and individual behaviour, then its knowledge characteristics can be said to be more tacit and causally ambiguous. The global integration of HRM at these lower organisational levels will therefore be rendered more problematic or “sticky”. Equally, HRM situated at the higher, corporate levels such as “policy” and “philosophy” is less susceptible to the influences of culture and behaviour, is likely to possess more codifiable and explicit knowledge characteristics and will therefore be less problematic to integrate throughout the MNC.

Whilst distinguishing between organisational levels in studies on global HRM integration would seem to be logical both on theoretical and practical grounds, such distinctions have remained the exception rather than the rule in the extant literature on HRM in MNCs. Instead, the majority of large-scale research reduces HRM to an aggregate whole or to sets of individual HRM practices with little or no consideration for the organisational hierarchy behind HRM and how it could provide a more nuanced account regarding patterns of global HRM integration and local adaptation. As has been argued, the absence of these distinctions and thus not comparing like with like, may then lead to conflicting conclusions regarding the strength of forces such as the country-of-origin effect (Ferner, 1997) or patterns of HRM convergence and divergence (Bae and Rowley, 2001). It would seem, therefore, that there is scope for more in-depth studies on the characteristics of HRM knowledge across organisational levels in order to reveal more accurately where global HRM integration is most likely to place as well as encounter most difficulties.

HRM knowledge at the practice level

Whilst recent research has focused on the business function level of global integration (Kim et al., 2003), the discipline of HRM in MNCs has also placed emphasis on studying the individual practices within the HR function. Indeed, it has been argued that the HR function should be regarded as comprising of distinct HRM practices that will vary in terms of their susceptibility to be globally standardised or locally adapted (Rosenzweig and Nohria, 1994; Lu and Björkman, 1997). In line with this approach, Bae et al. (1998) reaffirm that the transfer of HRM is therefore less about opting for total standardisation or localisation, but is rather a choice about which practices and how much.

Braun and Sparrow (2003) distinguish between “input-related” HRM practices (e.g. selection, training and development) and those that are “performance management-related” (e.g. appraisal and bonus systems). “Narrow” functional HRM practice integration subsequently refers to MNCs which only transfer a single practice, input-related or performance management-related, whereas “broad” integration focuses on transferring all practices. Whilst they argue that “broad” HRM integration possesses greater potential for achieving competitive advantage, empirically only few MNCs in China were shown to achieve this, with “narrow” integration occurring more commonly. A different approach again is adopted by Gooderham et al. (1999) in their classification of “calculative HRM” that refers to efficiency-seeking devices aimed at assessing and rewarding employee contribution, and “collaborative HRM” that refers to efforts at creating and communicating a culture of partnership between employer and employee.

In short, distinctions between individual HRM practices have too seldom been included in the research undertaken on HRM in MNCs. Those contributions that have made such distinctions, however, have done so by using diverse sets of criteria that often do not speak to each other. Partly as a result of this inconsistency, there still exist sizeable contradictions regarding precisely which HRM practices are easier or more difficult to transfer across borders (Myloni et al., 2004). To date, explanations for why certain HRM practices are more easily transferred have been attributed to, amongst others, socio-cultural factors (Lu and Björkman, 1997; Tayeb, 1998), institutional pressures (Rosenzweig and Nohria, 1994), resource-based considerations (Taylor et al., 1996) and to issues of resource dependency and relative bargaining power (Hannon et al., 1995; Björkman and Lu, 2001). Whilst these approaches imply that transferability rests on the extent to which HRM practices fit with the external environment or the context of parent-subsidiary relations, they speak little to the relationship between transferability and the characteristics of the HRM practices themselves. It is here that a knowledge-based perspective could make a valid contribution.

Liu (2004) proposes that different HRM practices may be perceived to vary in terms of their explicit-tacit, simple-complex and independent-systemic knowledge attributes, but does not go on to specify which HRM practices might possess which attributes. The way in which a knowledge-based perspective could be applied to this discussion is demonstrated here through a couple of examples. First, research on HRM in MNCs has generally indicated that parent performance appraisal systems have managed to be transferred fairly extensively to foreign subsidiaries (see, e.g. Björkman and Lu, 2001; Myloni et al., 2004). The knowledge-based explanation for these findings might then refer to the typically high codifiability of appraisal systems, which usually outline the performance review procedures in explicit detail. This is frequently reflected in the considerable investments made by MNCs to develop standardised forms and systems to ensure that the performance appraisal process is simple and transparent enough to be understood and accepted by all parties. Appraisals are arguably more systemic than independent, however, especially when they are linked to other areas such as bonuses and personal development plans. In this sense, performance appraisal's relatively high levels of codifiability and simplicity could help to explain its relatively high levels of transfer within MNCs. A second example might be that of recruitment, which unlike appraisals has more often been shown to resemble local practices and thus is not so extensively transferred (see, e.g. Björkman and Lu, 1999; Namazie, 2003). Here, the knowledge-based explanation might focus on the more tacit or experience-based characteristics of recruitment that are required to find and attract the best talent. Furthermore, knowledge regarding what makes an effective job advertisement or which is the best way to promote the company as an employer is likely to be complex and systemic in nature. Describing HRM practices in knowledge terms like this also draws much needed attention to the important within-practice differences that exist in HRM. When viewed in this way, it could follow that a more accurate account of what is happening is the greater global integration of appraisal procedures, for instance, but more local adaptation in appraisal criteria.

HRM knowledge transfer process

The knowledge transfer perspective now turns our attention towards how integration is achieved. At this juncture it is pertinent to address the issue of semantics regarding the transfer of knowledge. In the literature, terms such as “transfer” (e.g. Beechler and Yang, 1994; Zander and Kogut, 1995), “diffusion” (e.g. Ghoshal and Bartlett, 1988; Martin and Beaumont, 1998) and “flow” (e.g. Gupta and Govindarajan, 1991) have been used when discussing a firm's adoption of new practices or technology. Galbraith (1990) distinguishes between the terms transfer and diffusion by referring to transfers as a “distinct experience” as opposed to the “gradual process” (p. 70) of diffusion. Whilst it is acknowledge that HRM is also likely to be globally integrated via gradual processes of diffusion, the term “transfer” has been adopted here since it not only best describes the often deliberate attempts by parent firms to globally integrate their HRM into foreign affiliates, but also focuses our attention on the organisational mechanisms through which such transfers are facilitated.

Knowledge-based theory regards the mechanisms of integration to be of key significance in creating organisational capability (Grant, 1996a). Ghoshal and Bartlett (1990) argue that in response to the increasingly complex organisational practices, including HRM, that need to be transferred to ensure effective control and coordination amongst foreign affiliates, MNCs must correspondingly develop effective forms of transfer mechanisms. At the functional level, Kim et al. (2003) empirically show that different business functions, due to their specific characteristics, orientations and requirements, possess a unique ideal profile of effective modes for global integration. However, since research of this kind has not included the HR function it is unclear how these findings extend to the global integration of HRM.

With regard to integration mechanisms themselves, Martinez and Jarillo's (1989) study on the characteristics of mechanisms in MNCs indicated a significant increase in their degree of informality and subtlety over time. A further empirical study by Martinez and Jarillo (1991) identified that these informal and subtle mechanisms play a more significant role than structural, formal mechanisms in integrating foreign subsidiaries with their corporate parents. Evans et al. (2002) describe this development as a movement from “vertical” to “horizontal” mediums of knowledge transfer. “Vertical” or hierarchical forms of transfer are thus described as comprising top-down rules and operating procedures, where decisions are made centrally and expatriates represent agents of headquarters control. However, Evans et al. (2002), comment on how vertical mediums of transfer have been coming under pressure due to the time it takes for policies or innovations to filter their way through to practice and how important context-specific aspects are lost in this process. Consequently, they draw attention to the increasingly sophisticated “horizontal” mediums of knowledge transfer such as best practice workshops and the transfers of personnel. A knowledge transfer perspective on the mechanisms of global HRM integration is now discussed using this classification.

Vertical HRM transfer mechanisms

Referring to the “vertical” mechanisms of global HRM integration, centralised decision-making by headquarters on HRM issues has been shown to be prevalent in MNCs, especially when the issue in question carries significant financial implications such headcount and senior executive pay (Edwards et al., 1996; Martin and Beaumont, 1999). This form of global HRM integration is also typically accompanied by the design and implementation of global HRM policies, structures and guidelines (Tempel, 2001). In the international HRM literature, however, the most common “vertical” transfer mechanism studied is the use of expatriates. Such studies have used quantitative designs to test for the presence of expatriates in explaining the similarity between home and host HRM practices (Rosenzweig and Nohria, 1994; Björkman and Lu, 2001). Qualitative studies, however, describe the central and multifaceted role that expatriates play in the transfer of HRM practices, including “role models”, “boundary spanners” and “coaches” (Hetrick, 2002), including their capacity to communicate explicit and tacit forms of HRM-related knowledge (Gamble, 2003).

Horizontal HRM transfer mechanisms

With regard to the “horizontal” mechanisms of HRM transfer, the literature on HRM in MNCs has touched on a number of additional means through which HRM practices are integrated, including global expertise networks and HR centres of excellence (Sparrow et al., 2004), host management training and development (Harzing, 1999), the careful selection of managerial employees (Bonache, 2000), formal and informal communication (Rosenzweig and Nohria, 1994), global cultural management (Nohria and Ghoshal, 1994), benchmarking (Martin and Beaumont, 1998) and the trend towards more technology-driven HRM integration via HR information systems (HRIS) (Shrivastava and Shaw, 2003; Ruta, 2005). The development of these more horizontal-type HRM transfer mechanisms has also challenged the traditional paradigm that knowledge flows are one-directional, from parent to subsidiary. Indeed, in line with the view that an MNC is a differentiated network, the HRM knowledge accumulated by subsidiaries can be shared and contribute to the improvement of parent HRM via the process of “reverse diffusion” (Edwards, 1998).

Reflecting on the vertical and horizontal mechanisms of global HRM integration, and when applying a knowledge transfer perspective, the literature on HRM in MNCs appears to place much more emphasis on uncovering the structural and contextual determinants of HRM standardisation and local adaptation in foreign subsidiaries, than on trying to understand how the deployment of organisational transfer mechanisms may also affect these outcomes. Whilst some empirical work has already been conducted in this area, it has either focused on a narrow selection of mechanisms (e.g. expatriation) or on the use of a broader range of mechanisms but within single cases – predominantly western MNCs. In sum, the extant research on the mechanisms of global HRM integration remains fragmented. Furthermore, since the designs of these studies often lead to a lack of generalisability, we are left without a holistic and therefore more realistic appreciation of how HRM is globally integrated in MNCs.

In this regard, applying a knowledge transfer perspective not only places greater emphasis on the mechanisms of HRM integration and their respective transfer qualities, but also highlights the need to take these qualities into account to ensure that there is an appropriate fit between transfer mechanism and the type of knowledge being transferred. Indeed, matching the transfer mechanism with knowledge characteristics (Leonard-Barton, 1995), and ensuring that these mechanisms reflect the different locations (“reservoirs”) of knowledge (Argote and Ingram, 2000) are argued to be important considerations.

Sparrow et al. (2004) argue that transfer mechanisms emphasising person-to-person contact will be more effective in facilitating HRM knowledge transfers than attempts at codification using information systems. This is due to the highly tacit nature of HRM knowledge and the complex social processes involved in internalising new practices. Bonache and Brewster (2001) concur in their proposition that if the knowledge to be transferred is tacit then international assignments, due to their capacity to transfer more effectively the knowledge embedded in employees' skills and experiences, will be adopted as the preferred transfer mechanism. However, if explicit forms of knowledge are being transferred then other means of transfer (e.g. manuals, IT) will be used due to expatriation's high costs, difficult management and the problems associated with measuring value.

Taking the knowledge transfer perspective still further, there are also characteristics within an individual transfer mechanism that render it more or less effective that may not arise in assessments of the mechanism as a whole. For example, whilst expatriation might be considered an effective means of transferring tacit knowledge, it is more specifically the HR competence of the expatriate (who is rarely an HR specialist) and the skills they possess in teaching and implementing HRM knowledge to an unfamiliar audience that should determine the mechanism's true potential. Similarly, HRIS might be casually regarded as ineffective at transferring tacit forms of HRM knowledge. However, as a result of large company investments and rapid technological developments, some existing corporate HRIS have made significant developments in, for example, establishing global networks of HR professionals which have helped facilitate effective knowledge sharing (Sparrow, 2006).

HRM knowledge transfer outcomes

The third theme which arises from the application of a knowledge transfer perspective on global HRM integration addresses the issue of performance, or more specifically, when the transfer of HRM knowledge can be said to have been successful. A knowledge transfer perspective draws our attention here to performance indicators that can either be measures of effectiveness in transferring HRM knowledge content (content-based) or effectiveness in the HRM knowledge transfer process (process-based).

Content-based

In assessments of how effectively HRM content has been globally integrated within the MNC, one possible measure of performance could be the extent to which the HRM knowledge of the source (the parent) is homogenous in the recipient unit (the foreign subsidiary). Indeed, from a knowledge-based perspective, Zander and Kogut (1995) operationalised the transfer success of organisational capabilities in terms of accurate imitation of source knowledge. Empirical research along these lines in the field of HRM in MNCs is commonplace and often couched in terms of the global-local question (for a detailed review see Edwards and Kuruvilla, 2005). However, this approach could be criticised for its over-simplification in measuring levels of HRM integration from only one dimension, namely the degree of resemblance. Whilst this provides a relatively hard measure of global HRM integration performance from a content-based perspective, it neither addresses the extent to which the HRM content is embedded and applied in the recipient's operations nor fits with our definition of global integration that does not include the wholesale standardisation of HRM.

The stages approach to the adoption of a transferred organisational practice, however, potentially represents a more insightful measure of HRM integration performance, not least because it addresses more explicitly the issue of defining when a transfer can be deemed successful. In this regard, whilst Davenport and Prusak (1998) suggest that a successful knowledge transfer involves the transmission, absorption and use of that knowledge, Kostova (1999) argues that transfer success should be defined in terms of institutionalisation, which requires recipient employees to have both implemented and internalised the practice. At these two levels of institutionalisation, “implementing” refers to the adoption of formal rules, whereas “internalising” means attaching symbolic meaning and value to the practice. Kostova and Roth (2002) further refine the measure of successful practice transfers by identifying various patterns of adoption. Their empirically-based classification of “active”, “minimal”, “assent” and “ceremonial” patterns of adoption by MNC subsidiaries highlight the caution required when attempting to measure accurately the success of a practice transfer.

Although much of the research on HRM in MNCs approaches the issue of HRM content transfers in fairly simplistic terms of parent or host resemblance, it has made some progress in addressing this weakness through the adoption of host-country perspectives (e.g. Monks, 1996; Tayeb, 1998; Myloni et al., 2004) and in-depth, qualitative research designs (e.g. Gamble, 2003; Ferner et al., 2004). Nevertheless, studies conducted on HRM in MNCs seldom make the distinction between levels of HRM adoption in foreign subsidiaries. Indeed, Khilji and Wang (2006) argue that the failure to acknowledge differences between “intended” and “implemented” HRM represents a key methodological weakness in this respect and has arisen in part from the tendency to investigate these issues exclusively amongst HR policy makers such as top management and HR departments. This lack of rigour in assessing the extent to which HRM content has been successfully integrated therefore presents an opportunity for the knowledge transfer perspective to make a contribution. When applied, the successful global integration of HRM might then refer not only to the degree of resemblance with parent HRM, but also to the implementation and internalisation of parent HRM in the MNC's foreign subsidiaries.

Process-based

Performance indicators relating to the HRM transfer process are distinctive from content-based indicators insofar as they focus on issues related to the effectiveness of transfer mechanisms and of the transfer process as a whole. Accordingly, global HRM integration success from a process-based perspective can be evaluated in terms of the cost, speed and ease of the transfer.

First, the costs of facilitating HRM knowledge transfers might understandably dominate management considerations about selecting appropriate mechanisms of transfer. What “costs” might refer to here will of course vary, but broadly could adopt the definition by Teece (1977) in referring to “the costs of performing the various activities which have to be conducted to ensure the transfer […]” (p. 245). The more obvious costs incurred by the HRM transfer process could include, for example, capital investments in HRIS, the rotation of personnel including travel expenses or the use of expatriates including relocation packages. However, by adopting a broad definition of transfer costs, attention is also drawn to the fact that some costs cannot be ascribed a precise monetary value, such as those involved in host units internalising new HRM knowledge. Furthermore, one might consider the issue of opportunity costs in not transferring HRM. Referring specifically to the transfer of HRM practices, Schmitt and Sadowski (2003) attest to the usefulness of this kind of rationalistic cost minimisation approach to how MNCs balance centralisation costs and decentralisation costs in explaining patterns of HRM/IR practices in foreign subsidiaries.

A second indicator of process-based performance can be the speed in which HRM knowledge transfer takes place. The knowledge transfer literature has taken various approaches to this measure ranging from the time to imitate (Zander and Kogut, 1995) to the “velocity” that knowledge flows through an organisation (Davenport and Prusak, 1998). However, in line with the earlier discussion on levels of HRM adoption, a distinction needs to be made that speed refers not only to the transfer time between source and recipient, but also to the subsequent internalisation of the knowledge. For example, whilst HRIS possess the capacity to transfer large volumes of explicit HRM knowledge very quickly, if no further mechanisms are used then the relatively slow expected speed of internalisation is likely to compromise the overall speed of the transfer.

Lastly, the ease of HRM transfer could represent a process-related performance indicator, evaluated through an assessment of how many problems arise throughout the global integration process. An often-cited example of this approach to measuring knowledge transfer outcomes is Szulanski's (1996) notion of knowledge “stickiness” as reflected in the eventfulness or perceived difficulty of a transfer, including problems in communication and the extent to which expectations have not been met. When applied specifically to expatriation (Riusala and Suutari, 2004), for example, it is then possible to identify the presence of stickiness and thus the ease of knowledge transfer through this mechanism.

Towards a research agenda

The aim of this paper was to review the literature on HRM in MNCs from a knowledge transfer perspective and to highlight some of the key weaknesses in extant research on that basis. The paper concludes by outlining some of the relevant issues that should be considered in future research on global HRM integration in MNCs. The research agenda is structured around the weaknesses identified in the literature and suggests ways in which the knowledge transfer perspective could be applied to help advance a field reportedly in need of further theoretical development (Jain et al., 1998).

First, whilst the current discourse on HRM and knowledge is mostly concerned with identifying the HRM practices that support knowledge-related activities, significantly less attention has been paid to the knowledge-based characteristics of HRM practices themselves and how these characteristics are likely to shape patterns of global HRM integration within MNCs. Due to the extensive research within the knowledge transfer discipline, there are several ways the characteristics of knowledge could be applied to studies on HRM. For instance, Zander and Kogut's (1995) operationalisation of knowledge characteristics into dimensions such as codifiability, teachability, complexity and system dependence would present opportunities to identify precisely which aspects of HRM are considered to be the most tacit, and should go on to be applied across both organizational levels (e.g. philosophy or process) and HRM practice levels (e.g. recruitment or training) to explain the patterns of global HRM integration. However, for such research to facilitate meaningful comparisons across national and MNC settings, the way in which HRM practices are defined and categorized needs to become more consistent. At this juncture it is pertinent to note that this kind of knowledge-based interpretation of global HRM integration is not intended to be mutually exclusive of those interpretations, which draw on, for example, cultural issues and the nature of parent-subsidiary relationships. Indeed, any explanations contributed by a knowledge transfer perspective should complement rather than replace existing approaches. In line with the conclusion of Myloni et al. (2004), this is because one framework is unlikely to identify the multitude of influences on the transfer of HRM practices. Future research should therefore strive to employ multi-disciplinary designs which see the knowledge transfer perspective, better suited to analyses of HRM practice characteristics, integrated with those such as cultural (e.g. Bhagat et al., 2002) or institutional theory (e.g. Riusala and Suutari, 2004) that are better suited to investigating the characteristics of the international context in which global HRM integration is taking place.

Second, future research needs to address better the question of how HRM is globally integrated which, in turn, could contribute to the discussion on why it does or not take place. More specifically, research is merited on the mechanisms MNCs use to transfer HRM to foreign subsidiaries, seeking answers to such questions as which mechanisms are most commonly used to integrate HRM, are certain HRM practices integrated best through certain mechanisms, and what factors influence the selection of these mechanisms? Furthermore, does the use of certain mechanisms, or combination of mechanisms, lead to superior global HRM integration outcomes? Whilst research addressing these kinds of questions has been conducted on other key business functions (e.g. Kim et al., 2003), there is scant empirical evidence to suggest whether the findings extend to the HR function. As demonstrated in recent case study research on HRM in MNCs, the literature on control and coordination as well as the application of power and political perspectives (e.g. Martin and Beaumont, 1999; Ferner et al., 2004) would appear to be instructive in this regard. However, whilst theories of control and power relations may increase our understanding of how and why global integration is enforced and negotiated, the knowledge transfer perspective offers scope for analysing the qualities of individual integration mechanisms and the extent to which they fit the characteristics of HRM being integrated. Although this form of knowledge-based theorizing exists on the use of expatriation (e.g. Bonache and Brewster, 2001) it addresses neither the transfer of HRM specifically, nor the plethora of other organisational mechanisms that lie at the disposal of MNCs. Particularly interesting in this respect would be, for example, how the increasing use of globally integrated HRIS (e.g. Tansley et al., 2001) will impact the global integration of HRM in MNCs given the tendency for HRM practices to be complex, systemic and causally ambiguous.

Lastly, the performance question in the context of global HRM integration needs to be subject to more rigour and critical reflection. One way in which more rigour might be applied is to include more critical evaluations of HRM integration content outcomes. Whilst an approach similar to Kostova and Roth's (2002) patterns of practice adoption could be useful here, a key methodological implication of this approach is the need to draw on responses from diverse organisational groups, ranging from those who are in the best position to know what the corporate parent's HRM practices are (e.g. subsidiary general and HRM managers) to those previously under-represented groups who are best placed to attest, first hand, to whether HRM has been successfully integrated (e.g. line managers and shop floor employees). Cross-national research amongst the latter group, however, will likely present cultural and linguistic challenges that can only best be overcome through cross-national researcher collaboration. In terms of the performance question and the global HRM integration process, Szulanksi's (1996) concept of “stickiness” could serve as a useful tool in highlighting how effective MNCs are at facilitating global HRM integration. For instance, where internal stickiness most frequently occurs in transfers of HRM knowledge and at which stage (e.g. implementation or integration) may provide important insights into the under-researched “issues of process” concerning HRM in MNCs (Gunnigle et al., 2003). Furthermore, by extending the analysis of stickiness to include the characteristics of HRM knowledge source and recipient, it could be explored, for example, whether the absorptive capacity and motivational disposition of expatriate subsidiary managers increases HRM knowledge transfer stickiness given their typically non-expert HRM status and the potentially low priority they assign to HRM issues.

Limitations

The weaknesses identified in the literature on HRM in MNCs from a knowledge transfer perspective must be viewed, however, in light of the present paper's own limitations. Important in this respect are the underlying assumptions behind the use of the terms “knowledge” and “transfer” and whether these assumptions accurately reflect the way in which HRM is globally integrated in foreign subsidiaries. With regards to the notion of knowledge, an implicit assumption made in this paper is that HRM knowledge denotes some form of ownership (typically by the parent) and is an entity that can be captured and moved. However, knowledge has been argued to be more of a social construction, which should take into account the mutual and complex processes of learning and interaction, which include the contributions of the learner as well as the teacher (e.g. Hollinshead and Michailova, 2001). Such interpretations not only highlight that global HRM integration will seldom be as simplistic as is perhaps described here, but also require research to place equal emphasis on the inputs of parent and host. Similarly, “transfer” in this paper has implied a unilateral process in which parent HRM is integrated into foreign subsidiaries. Whilst this may be the intention from the MNC parent perspective, it is suggested that the reality involves the far more complex use of power relations and organisational politics in resisting such top-down transfers (e.g. Martin and Beaumont, 1999; Ferner et al., 2004). In addition, descriptions of the cross-national transfer process such as a “translation” (e.g. Bedward et al., 2003) and the “travel of ideas” (Czarniawska and Joerges, 1996) are suggested to be more accurate in their acknowledgement that culturally embedded knowledge first needs to be dis-embedded and then re-embedded in host contexts. However, despite these limitations, it is hoped that the weaknesses identified in the literature together with the suggested avenues of future research will encourage scholarly endeavours to improve a field reportedly in need of further theoretical development.

Figure 1Global HRM integration: a knowledge transfer perspective

References

Argote, L., Ingram, P. (2000), "Knowledge transfer: a basis for competitive advantage in firms", Organizational Behaviour and Human Decision Processes, Vol. 82 No.1, pp.150-69.