The liability of mimicry: Implementing “global human resource management standards” in United States and Indian subsidiaries of a South Korean multinational enterprise

There is increasing evidence that multinational enterprises (MNEs) from less dominant economies tend to mimic and disseminate human resource management (HRM) practices sourced from a dominant economy, usually the United States, to overcome their “ liabilities of origin. ” However, our understanding of the specific challenges involved in the implementation of such practices by firms across different national and subsidiary contexts remains limited. Drawing on evidence from a case study of a South Korean MNE, we examine the extent to which, and ways in which, global HRM policies mimicking U.S. practices are implemented across its sales, manufacturing, and research and development subsidiaries in the United States and India. We find discernible differences in the implementation of the global policies both between the two host country sites and across the three function-specific subsidiaries in each country, identifying a range of national and subsidiary-specific factors that inform these variable implementation outcomes. In addition to legitimacy challenges related to the source, appropriateness, and process of transfer, we note a unique form of legitimacy challenge —“ the liability of mimicry ”— whereby local actors can challenge head office policies on the basis of a claim to superior expertise in the dominant practices, as a particular concern of MNEs from emerging economies.

. The intention is that such mimicry brings the MNEs' HRM policies in line with modern practice and makes transfer, even to subsidiaries in more developed countries, simpler. In some cases, subsidiaries in advanced economies are set up in a polycentric and adaptive approach precisely for the capture and diffusion of such global best practices (Patel, Sinha, Bhanugopan, Boyle, & Bray, 2018), but the mimicry also widely involves dominance effects that are diffused globally through multiple channels .
While EMNEs' attempts to distribute "best practice" HRM policies across their subsidiaries may contribute to the global convergence of policies, it is still unclear whether and how HRM policies sourced from a non-home-country origin are translated into practices in subsidiaries, given the mixed evidence on transfer outcomes (Aguzzoli & Geary, 2014;Demirba g et al., 2016). As the externally sourced policies are not embedded in the home country, headquarter's (HQ) actors potentially have limited experience of implementing and utilizing them. This raises distinct challenges that may be different from the case of parent-practice transfer in developed country's MNEs, where HQ actors have the most expertise in the transferred practices. As evidence on the transfer of externally sourced dominant practices grows with the increasing number and influence of EMNEs (Thite, Wilkinson, & Shah, 2012), the dynamics of such transfer demand more attention.
To extend our understanding of this emerging phenomenon, we examine the extent to which, and the manner in which, HRM policies sourced from a dominant economy are implemented by an MNE across its subsidiaries in economies that are both more and less developed than its home country. Our findings offer important insights into international HRM in two respects. First, they build on efforts to gain a better understanding of how the transfer of HRM practices in EMNEs, which are becoming increasingly "prominent" (Wilkinson, Wood, & Demirba g, 2014, p. 835), remain "distinctive and different from" (ibid.: 841) what has hitherto been observed in developed country's MNEs. More specifically, they demonstrate how the formulation and implementation of "global best practices" of EMNEs have complex, multivariate, and even idiosyncratic dynamics which cannot be uniformly linked to the tendency toward global convergence or national divergence. Second, our findings also draw attention to some distinct factors that affect the transfer of practices, but which have previously received little attention.
Our study uses evidence from a relatively novel research setting-an MNE in the global automotive industry, from South Korea, and its sales, manufacturing, and research and development (R&D) subsidiaries in the United States and India. While South Korean MNEs are from an "advanced" emerging economy that has reached developed economy standards by certain criteria (Kim, Kim, & Hoskisson, 2010), they are still impacted by rapid institutional changes in employment practices in their home country with limited status in the global economy, like many other EMNEs (Debroux et al., 2018).
We contend that cases of South Korean MNEs deserve particular attention in theorizing the practices of EMNEs. Their experiences can not only tell us how EMNE practices differ from those of advanced economy MNEs, but also help us anticipate the changing nature of MNE activity in general, as further EMNEs come of age (Khanna & Palepu, 2006). In addition, our research setting allows for a comparison of global policy implementation in subsidiaries in two host countries with different positions in the hierarchy of national economies (Smith & Meiksins, 1995) relative to the MNE's home country. While comparative national contexts may involve differences in institutional maturity (Khanna, Palepu, & Sinha, 2005;Thite et al., 2012) that may affect the implementation of "best practice" HRM policies across the subsidiaries, this has not been exhaustively studied empirically. Furthermore, studying a sales, manufacturing, and R&D function subsidiary in each host country enables us to highlight how local adaptation of global HRM policies may vary across subsidiary functions even within the same host country.
Our study makes several contributions to the theory of HRM practice transfer in MNEs. First, the case of transferring best practicebased global policies by the South Korean MNE contributes to our understanding of the importance of legitimacy in the transfer of HRM practices by illuminating the multiple legitimacy challenges that arise.
Despite the mimicking and adoption of "global best practices," the firm still faced considerable challenges in implementing these across subsidiaries, partly due to a lack of professional legitimacy, as well as the hierarchical management style deeply ingrained in its home country. In particular, we find a novel form of local resistance based on the claimed asymmetry of expertise between HQ and subsidiary actors.
We propose that a framework of multiple legitimacy challenges based around legitimacy of source, appropriateness, transfer process, and expertise-could offer a theoretical lens to better understand the transfer of HRM practices in wider populations of MNEs.
Second, we suggest that the home country effect in the context of certain MNEs, particularly those from less advanced economies, needs to be understood more broadly than the formal practice dimension. The liabilities of origin-the legitimacy challenges created by the country of origin-manifest across multiple dimensions including formal practices as well as informal processes of transfer.
Finally, the study sheds light on the multi-layered nature of local contexts by highlighting the different extents of implementation of the global policies, depending on the level of development of a host country relative to the home country and the pervasive influence of subsidiary-specific factors such as function.

| TRANSFERRING HRM PRACTICES IN MNEs: THE STRUGGLE FOR LEGITIMACY
Institutional theories have provided an important foundation for studies of the transfer of managerial practices, including HRM practices, within MNEs (Björkman, 2006). A "universal" approach based on the assumption that the same effective practices can be applied across different national settings has been extensively challenged by institutionalist arguments (Demirba g et al., 2016). Earlier institutionalist studies have convincingly proposed that the MNE is defined by a "institutional duality" (Kostova, 1999), where its efforts to transfer standardized practices are subject to local adaptation due to isomorphic pressures emanating from a host country (Ferner et al., 2005). It has since become commonly agreed that coercive, mimetic, and normative pressures (DiMaggio & Powell, 1983) that inform the legitimacy of firms in given organizational fields regularly force MNE subsidiaries to follow practices embedded in the host country rather than following parent practices (Rosenzweig & Nohria, 1994;Westney, 1993).
Comparative institutional theories take a somewhat different approaches to characterizing the impact of institutions on MNE practice and practice transfer (Aguzzoli & Geary, 2014;Boussebaa, Morgan, & Sturdy, 2012;Brewster, Wood, & Brookes, 2008). Though variants exist, they broadly aim to explain how distinct national institutional configurations shape a particular pattern of practice in a country. In the context of MNE organizations, Morgan et al. suggest a multi-layered institutional framework to understand practice transfer (Boussebaa et al., 2012;Morgan, 2012;Morgan & Kristensen, 2006). Accordingly, not only national institutional configurations of the home and host countries, but also an MNE's "transnational logic of appropriateness" and its "transnational embeddedness" in a meta-institutional field beyond national spheres help shape practice transfer in its global network (Boussebaa et al., 2012, p. 469). The agency of a range of actors inside the "contested terrain" of the multinational matters critically, too. The transfer and implementation of HRM practices within MNEs are legitimacy-seeking endeavors involving various actors across multiple institutional contexts (Dörrenbächer & Geppert, 2011;Edwards & Belanger, 2009;Morgan & Kristensen, 2006). In particular, local managers in a subsidiary can play a critical role as "interpreters" of the limitations and possibilities of corporate action and may mount resistance based on power resources such as knowledge of distinct local practices and relationships with local stakeholders (Demirba g et al., 2016;Ferner et al., 2005;Klimkeit & Reihlen, 2016).
HQ actors' views on the appropriate HRM practices to be transferred are themselves influenced by the institutional context of the home country Ferner, 1997). When established developed-country's MNEs attempt to transfer HRM practices to subsidiaries, they may by default draw on their parent practices as the key source . This country-of-origin-effect has been found in various empirical studies of MNEs from developed countries: For example, in the UK subsidiaries of U.S. MNEs (Ferner et al., 2004); UK and Spanish subsidiaries of German MNEs (Ferner, Quintanilla, & Varul, 2001), Chinese subsidiaries of Japanese MNEs (Furusawa & Brewster, 2018;Gamble, 2010), and Chinese subsidiaries of UK MNEs (Gamble, 2006).
While the legitimacy of parent policies could be challenged by subsidiary actors in terms of their appropriateness in a particular local context, some MNEs question them as a legitimate source for practice transfer and instead choose to adopt and transfer third-party practices-"global best practice"-sourced from a dominant economy as a template for their global policies (Aguzzoli & Geary, 2014;Boussebaa et al., 2012;Demirba g et al., 2016). National economies with greater power and influence in the global political economic hierarchy exert a "dominance effect" (Smith & Meiksins, 1995), which means "global best practices" are shaped by the institutional legacies of powerful countries-at present, mainly the United States (Pudelko & Harzing, 2007).
This tendency to transfer third-party practices, rather than exporting parent practices, has often been reported in the cases of MNEs from non-traditional origins including newly developed economies, such as South Korea or Taiwan, or emerging economies, such as China or Brazil (Chang, Mellahi, & Wilkinson, 2009;Cooke, 2014;Edwards, Schnyder, & Fortwengel, 2019;Geary, Aguzzoli, & Lengler, 2017;Glover & Wilkinson, 2007). These EMNEs face distinct challenges in internationalizing their businesses (Luo & Tung, 2007), encapsulated as the "liability of origin" (Ramachandran & Pant, 2010) or the "liability of emergingness" (Madhok & Keyhani, 2012). These include their attempts to internationalize without strong managerial resources (Peng, 2012) to manage, with limited organizational capabilities, the demands of highly developed host countries (Barnard, 2010), and a lack of experience and established practices from their home base-plus their selfdoubt regarding the legitimacy of their own practices (Ramachandran & Pant, 2010). The home country institutional links can also often lead to liabilities for EMNEs, including those due to their close links with home country governments (Khan, Wood, Tarba, Rao-Nicholson, & He, 2018), their perceived low standards of corporate governance and transparency (Ouyang, Liu, Chen, Li, & Qin, 2019), and institutional voids that constrain the development of critical competencies (Meyer & Xin, 2018). To overcome the liability of origin, they may pursue rapid internationalization to catch-up quickly with MNEs from developed economies (Luo & Tung, 2007;Madhok & Keyhani, 2012) by exploiting available resources from the globalized economy (Ramamurti, 2012), such as HRM practices.
Indeed, most research into EMNE subsidiaries in mature, advanced markets have focused on the strategic purpose of these units for organizational learning (Jackson, 2014), and especially their "best practice" capture for reverse diffusion (Zhang & Edwards, 2007).
Are these examples of mimicry a way for MNEs to overcome the legitimacy challenge? Arguably, the transfer of such policies within MNEs may contribute to global convergence around dominant practices, if they are accepted and implemented in the subsidiaries (Geary et al., 2017). A recent study illustrates how a Brazilian MNE in mining and metals refining could transfer its global HRM policies-influenced by U.S. firms-to institutionally well-established economies in North America and Europe (Geary et al., 2017).
However, a similar level of HQ dominance across subsidiaries is less likely in EMNEs that have recently established themselves in an industry, such as automobiles, that still has very powerful players from advanced economies, including the United States (Lee, Paik, Cave, & Jung, 2018). An MNE's global position in an industry and the extent of host countries' reliance on foreign direct investment might affect its ability to disseminate policies across subsidiaries (Geary et al., 2017).
There is considerable variation across industries, particularly when the third-party practices are neither strongly embedded in an MNE's home country nor extensively applied at HQ before the transfer (Cooke, 2014).
Notwithstanding some recent studies, there remains a dearth of research examining subsidiary responses to, and the implementation of, "global best practice" in MNEs. We contend that the following aspects of local contexts shape the dissemination of global HRM policies in MNEs. First, national institutional conditions for a subsidiary and power relations between HQ and the subsidiary vary depending on the host country's relative status in the hierarchy of national economies (Khanna et al., 2005;Smith & Meiksins, 1995). The level of economic development is correlated with significant institutional differences between countries (Thite et al., 2012), with unique, relatively mature, and stable national institutional arrangements in developed economies. These have considerable influence on subsidiary HRM practices of MNEs . Less developed economies undergoing rapid transitions can be expected to have less stable institutional infrastructures in, for example, educational and legal systems. In such circumstances, MNEs tend to have greater power, and hence more autonomy, but may also face challenges such as skill shortages and strong competition for available talent from other MNEs and local firms (Beamond, Farndale, & Härtel, 2016).
Second, recent institutionalist analysis has recognized that the subsidiary's function also matters in HRM practice transfer (Almond, 2011;Monaghan, Gunnigle, & Lavelle, 2014), with different levels of HQ monitoring between higher-and lower skill- Considering these two particular local contexts, we aim to develop our understanding of third-party HRM practice transfer by addressing the following research questions: To what extent, and how, are "best-practice" global HRM policies transferred and implemented by an EMNE in subsidiaries (a) in developing and developed economies; (b) across multiple subsidiaries in each country such as R&D, manufacturing, and sales. Although one may expect the particular local contexts will affect the transfer of HRM practices, we aim to examine whether an MNE from a newly industrialized economy in a competitive industry can overcome the legitimacy challenges in terms of "source" and "appropriateness" of the practices in such local contexts by mimicking and disseminating "best-practice" HRM policies. We suggest that the findings in relation to these two research questions also offer important insights into whether the adoption of "best-practice" policies by MNEs might significantly contribute to the global convergence of HRM practices.

| THE INSTITUTIONAL CONTEXT OF HRM IN A SOUTH KOREA MNE
We outline the context of our research sites: The South Korean background of the MNE and the U.S. and Indian subsidiaries. South Korea is often categorized as an advanced Asian economy (Witt & Redding, 2013) or an "advanced emerging" economy (Kim et al., 2010(Kim et al., , p. 1148, featuring characteristics of a developed economy as well as of an emerging Asian capitalism (Carney, Gedajlovic & Yang, 2009;Witt, 2014). On the one hand, since the Asian financial crisis in 1997, the Korean business system has shown tendencies of deregulation and liberalization, moving toward a more market-oriented economic model, with some evidence of the growth in performance-driven managerial practices. On the other hand, the economy still retains important elements of the traditional government-led developmental model, with continued dominance of family-controlled conglomerates, the significance of informal and social control, and top-down decision making practices with limited delegation (Witt, 2014;Witt & Redding, 2013).
The characteristics of the Korean business system strongly affect HRM practices in Korean firms. HRM practices in Korean firms were traditionally paternalistic, heavily influenced by Japanese practices and the Confucian tradition, and built on internal labor markets and a seniority-based approach (Bae & Rowley, 2003). The Asian financial crisis in 1997 led to dramatic changes in HRM as these traditional approaches were blamed for causing the loss of competitiveness of Korean firms (Debroux et al., 2018). Many firms introduced individual performance-based approaches and more flexible labor market policies, seen as "global standards," largely modeled on U.S.-based ideals (Horak & Yang, 2019;Kim & Bae, 2017). However, the implementation of these practices varied significantly across firms, and changes are still in progress (Horak & Yang, 2019;Kim & Bae, 2017;Rowley, Bae, Horak, & Bacouel-Jentjens, 2017).
The rapid growth of the Korean economy since the late 60s has been largely attributed to the significant role of dominant business organizations known as Chaebols-large, diversified, and familycontrolled conglomerates-with strong support from the Korean government (Whitley, 1991). One of the key characteristics of Chaebols is an authoritarian management style based on a highly centralized and hierarchical mode of control with personal authority and domination by the founders' families (Whitley, 1991;Whitley & Zhang, 2016;Witt, 2014). Most large Korean MNEs are part of Chaebols and thus the key features of Chaebols are deeply ingrained in the management practices and styles of Korean MNEs, including their foreign subsidiaries (Glover & Wilkinson, 2007;Zou & Lansbury, 2009).
The Korean economy has internationalized rapidly in recent decades (UNCTAD, 2016) and Korean MNEs have begun to develop their approaches to international HRM. There is widespread adoption of U.S.-influenced "global standards" in firms' IHRM policies (Yang, 2015), in part due to the lack of strongly established home country practices. Despite this, Korean MNEs have largely maintained the traditional hierarchical and authoritarian management style, including in their foreign operations (Glover & Wilkinson, 2007;Horak & Yang, 2019). Managing the potential tensions between the espoused "best practice"-centered HRM policies and the traditional hierarchical management style is a key challenge for Korean MNEs (Horak & Yang, 2019;Kang & Shen, 2015, 2016.

| The United States and Indian subsidiary contexts
The United States is often seen as a paradigmatic institutional context, particularly in relation to HRM practices. The American business system imposes few restrictions on employment practices (Ferner et al., 2005), allowing firms the freedom to develop HRM practices intended to enhance workforce flexibility and performance in order, ultimately, to serve the interests of shareholders (Beer, Boselie, & Brewster, 2015). The rhetoric and, to some extent, the practices of HRM in the United States were introduced widely across the world, including in South Korea and India.
U.S. subsidiaries of Korean MNEs therefore occupy a contradictory context: They are in an inferior position to HQ in the organizational hierarchy, but they are located in a national economy that occupies a dominant position in the hierarchy of national economies (Smith & Meiksins, 1995), which remains the source of the "global HRM practices" that Korean parent companies are eager to adopt.
The other host country in our study, India, provides a contrasting context of a major emerging economy. As in the South Korean case, HRM practices in India have historically been subject to external influences. While some commentators have emphasized the cultural specificity of management styles and practices (Budhwar, 2000;Cappelli, Singh, Singh, & Useem, 2010), formal personnel functions have broadly reflected those of the ex-colonial power and recent economic liberalization and FDI have further precipitated rapid, if patchy, change (Budhwar & Varma, 2012). Foreign companies have traditionally opted to adapt to local practices in India (Björkman & Budhwar, 2007), but have recently contributed to change in HRM practices where certain sectors such as information technology and business process outsourcing now emerge as relying widely on "best practices" aligned with those in developed economies (Patel et al., 2018).

| METHODS
In order to explore the complex interactions between multiple contextual influences and various actors' involvement in the implementation of global HRM policies across the subsidiaries, we adopted a multi-site case study of a large Korean MNE in the highly competitive auto industry. This is a global industry which is more internationally coordinated and interdependent than most (Schlie & Yip, 2000), and hence a globally oriented approach to HRM might be expected. The case study approach is appropriate for our purposes as it permits the detailed examination of complex relationships among lesser known factors in a novel context. Our study responds to calls for a return to the use of case studies to provide the rich data necessary to understand the complexity of international business (Doz, 2011).
We refer to our case company using the pseudonym "K-Auto." K-Auto is affiliated with one of the largest Chaebols in South Korea, focusing on auto-related industries from steel manufacturing to auto financial services. It consists of two auto makers which have their own brands and designs but share a common platform for each type of car and utilize shared R&D functions. K-Auto shows a high degree of internationalization as a major global player in the industry, with a sales volume of US$55.1 billion and with an overseas sales ratio of 75% in 2010. It has achieved significant growth in emerging as well as developed economies. In 2011, the company was the second largest automotive firm in India, while in the United States, it was one of the top five auto companies by sales volume. A corporate HRM function is responsible for HRM in both car makers, developing IHRM strategies and policies for all their subsidiaries. "Globalizing" HRM activities have been a key corporate-level strategic initiative in K-Auto and the firm has made intensive efforts to implement global HRM policies across all their subsidiaries in recent years.
At the corporate level, data were collected at the HQ of K-Auto in South Korea to identify their strategic approach to subsidiary HRM policies, particularly in terms of standardization versus localization and the rationales behind their approach. To examine the implementation of HRM policies at the subsidiary level, we chose India and the United States as the host country sites for two reasons. First, they represent an emerging and a developed economy, respectively. Second, the company has multiple subsidiaries of three functions, sales, manufacturing, and R&D, each in a separate location, in both countries (Table 1). All the subsidiaries were established by K-Auto. Comparisons between the six subsidiaries allowed us to examine the Profile of subsidiaries of the case company   only the quotes used in this paper were translated by the first author, who is competent in both languages and did the data analysis.
For the structured components of the interviews, data coding was straightforward. After coding the data into the predefined categories, we identified patterns of similarities and differences in the degree of implementation of the global HRM policies in the six subsidiaries across the 44 components of HRM practices. For the semistructured components, we followed an analysis procedure, informed by principles and practices of inductive data analysis (Gioia, Corley, & Hamilton, 2013). First, accounts of the development and implementation of global HRM policies were categorized into broad categories such as HQ intentions and acceptance, adaptation, and rejection in subsidiary responses. Second, each account was given an emergent code describing and summarizing the key themes of each account.
Third, the emergent codes were grouped into codes of higher-level concepts through inductive reasoning. Fourth, the codes from the analysis of individual interview data were aggregated at HQ or subsidiary level and reviewed. By comparing the codes from multiple respondents in each organizational unit, the initial codes were validated and refined to reflect common views at HQ or at a subsidiary.
Finally, we compared HQ and the subsidiaries, as well as the different countries and functions, in order to examine common or distinct patterns in relation to the implementation of global HRM policies. Specific conditions of local adaptation were identified by examining which local institutional factors were involved in the local adaptation process and how those factors could be leveraged by local actors to influence HRM policies.

| FINDINGS
We first analyze K-Auto HQ's approach to the development and transfer of global HRM policies and the home country influences on the content and manner of transferring the global policies. Then, we present the subsidiaries' responses to the global HRM policies, identifying patterns of variation in the way these are locally adapted in each host country and across the three function-specific subsidiaries.

| Initial subsidiary responses to "third-party" global HRM policies
The GHRS was distributed as a detailed guidance document with regard to global HRM policies and desirable practices by the corporate HRM team. Overall, the initial reactions from subsidiaries to the GHRS were not positive, and there was a lack of substantive implementa- This indicates that regardless of the compatibility or similarity of practices in the institutional context of host countries, a legitimacy challenge may stem from a lack of "proper" transfer process, at least in certain locations.

| Dealing with "rejection" and recalibrating GHRS implementation
The initial reactions from subsidiaries made the corporate HRM team reconsider their approach. They recognized that they had focused exclusively on the content of the new global best practice-based policies and paid insufficient attention to the process side of the transfer.
The credibility of the policies did not outweigh a perceived lack of procedural legitimacy.
Their reflection triggered a second round of HQ's attempts and, crucially, the mobilization of various resources to gain procedural legitimacy and increase their influence on the implementation process.
First, the corporate HRM team made visits to subsidiaries to explain the rationale behind the policies and support the implementation process. Second, key members who had been involved in the design of the global policies were sent to subsidiaries, particularly in developed countries, to help local HRM managers to implement the policies.

| The key areas of local adaptation in India and the United States
Despite some reported progress in the implementation of GHRS, it took much longer-over 3 years-to implement global HRM policies across the subsidiaries. It was a costly process, too, as K-Auto had to mobilize considerable resources to get subsidiaries to implement the policies. Interviewees at HQ reported that their extensive efforts paid off, with the mandatory standardization components being widely adopted (Table 3). However, interviews at the subsidiaries revealed that certain policies had nevertheless been subject to significant negotiation with local actors. Two areas where such negotiation took particularly salient forms were the forced distribution of performance appraisals and the grade system. We examine in greater detail these two policies and the responses to them by the U.S. and Indian subsidiaries.

| Forced distribution in performance appraisal
The global standards included the forced distribution of performance appraisals, whereby managers were required to distribute ratings for certain portions of their employees into predefined categories. Based on the consultancy's recommendations, the corporate HRM team believed that this was "normal" practice in the United States. However, in K-Autos' U.S. subsidiaries, local actors rejected the application of the forced distribution practice (

| Grade system
For the grade system, the adoption pattern in the two host countries was the other way around (Table 3). The global policies mandated a five-level grade structure for all subsidiaries to facilitate the movement of employees across organizations, a key objective of the global HRM policies. However, the standard grade structure was adapted significantly in the India subsidiaries, where it was seen as insufficient to accommodate the levels of hierarchy necessary to run the organization. It was argued that a highly differentiated hierarchical grade structure is the norm in Indian labor markets.
There are more grades, but the grade is for the purpose of giving ego satisfaction, sheer employee appeal, like "I got a promotion in two years." (Local HRM Manager, Indian manufacturing subsidiary).
The standard grade system was seen as a core element of the global HRM policies. However, HRM managers in India persuaded corporate HQ that they should be allowed to implement their own local version of the grade system and introduced ten levels of grading, relying on a logic of "talent retention": It might be inevitable to follow the local norm; otherwise, we may lose people. (Expatriate in HRM, India sales subsidiary, India) Retention of key employees is a critical issue in India, and local managers argued that adopting the local, widely used practice of multigrading would help avoid losing local talent. Here too, the resistance to the top-down global standards from HQ was tied to local (informal) institutional norms, but only through the arguments posed by the local actors rather than being directly impacted by formal institutional rules.
After a series of debates with subsidiary HRM managers, including the expatriates, the corporate HRM team agreed to the locals' proposed grade system on the condition that each alternate grade level was matched with one of the five global grades: In the end, we thought that this is a reasonable resolution as the Indian grade system can be aligned to our global grade system, while it can accommodate the local needs, which we cannot ignore. We understood retention is a critical business issue in India, one of our strategic markets. (Corporate HRM executive) By contrast, the five-level standard grade structure was adopted as it stood in the U.S. subsidiaries. No serious issue was reported, though there was an amendment in the U.S. R&D center, splitting one level into two to reflect a unique situation in the local labor market. It was claimed that the easy adoption of the standard was largely attributable to the similarity of the structure to general practices in the

| Subsidiary functions and local adaptation
Responses to the global standards differed not just across countries, but also across functions. In both host countries, the subsidiaries undertaking research and development showed relatively low levels of implementation of the GHRS (Table 4).
One common factor affecting the less favorable responses to the global policies from the R&D subsidiaries, in India as well as the United States, was the heavy reliance on highly skilled members of local workforces, and the need to attract and retain talented employees in the local labor markets. The R&D subsidiaries are located in areas where many MNEs and local firms have their R&D functions. It was often mentioned, in both subsidiaries, that the retention of key employees was a top priority. Subsidiary HRM managers felt considerable pressure to adopt practices familiar to that labor market, since they believed this affected their ability to attract and retain talented engineers: We have a serious retention issue here…many firms in this area use particular practices such as flexi-time. … You don't find them in the Korean culture…but we adopted the practices, otherwise we might lose our talented people. (Local HRM manager, India R&D subsidiary); Well, we're a [Korean] company, and we're trying to compete in North America, in the United States, and we're in the automotive industry, so we're competing with the American Big Three, the Japanese Big Three, we fight to be competitive in this area. We need to consider the practices which work in this competitive environment. (Local HRM manager 2, U.S. R&D subsidiary) As employees tend to move across firms within local labor markets, they spread a good or bad reputation of the HRM practices of particular firms. approach, made the situation worse: They established the first Global HRM team a couple of years ago…. What happens then, and it happens repeatedly, is they will create projects, they will create programs, they will create policies, and they will try to just implement them overseas….

| DISCUSSION
While the tendency of mimicking "global best practices" from a dominant economy and disseminating them to subsidiaries has been well evidenced in the cases of EMNEs (Andreeva et al., 2014;Geary et al., 2017;Zhang & Edwards, 2007), the specific challenges of implementing this sort of approach have received little attention.
Recent studies of EMNEs have examined the implementation of such practices across subsidiaries in industries where EMNEs enjoy dominant positions, or in particular kinds of host countries (Andreeva et al., 2014;Geary et al., 2017). MNEs from newly industrialized or emerging economies may "cast aside their prior routines to compete in the open global environment" and adopt more ambitious approaches (Madhok & Keyhani, 2012, p. 35). Mimicking "best practices" in developing and transferring global HRM policies could be understood as an endeavor to overcome the liability of origin (Chung, Sparrow, & Bozkurt, 2014). However, purportedly "context-free" third-party HRM policies do not eliminate the scope for local resistance and adaptation. Rather, they generate further tensions between HQ and subsidiaries, particularly due in this case to the Korean hierarchical and authoritative managerial style and a lack of professional legitimacy. The Korean MNE needed to mobilize substantial resources in order to manage the tensions arising from the implementation process, even in the subsidiaries in the United States, which the HQ actors regarded as the source of the "best practices." While not without evidence of progress, the implementation of the global policies still showed considerable cross-national as well as cross-functional variation across the subsidiaries.
Our research setting of a single country of origin and two host countries in developed and developing economies clearly has limitations for generalizability. We were not able to examine the "internalization" dimension of practice transfer, but only cover the implementation dimension (Kostova, 1999 context), of process (the use of extensive consultation process with local stakeholders), and of expertise (the perceived expertise of HQ actors in relation to transferred practices) (Figure 1).
Previous studies have mainly revealed the legitimacy challenge of appropriateness Ferner et al., 2005;Gamble, 2010) or the legitimacy challenge of source (Andreeva et al., 2014;Geary et al., 2017). We observed a novel form of local resistance by U.S. R&D subsidiary actors, based on claims of superior expertise in the relevant HRM practices, bolstered by drawing on a strong professional identity prevalent in the U.S. context, where the HRM profession is more highly organized and formalized than in many other countries (Ferner et al., 2005) including Korea. Rather than the "dominant" U.S.-influenced global policies being an immediate fit with its origin country, we observe that claims of HQ's lack of professional expertise can emerge as an additional form of legitimacy challenge.
Where the parent firm has not experienced the full implementation and internalization of the dominant system-influenced policies at home, the adoption of "global best practice" could be considered at best "ceremonial adoption" (Kostova, 1999). Due to their limited experience with the practices in question, HQ actors may lack sufficient competence to guide subsidiaries. The claimed asymmetry of competence between HQ and subsidiary actors enabled and emboldened the latter to challenge HQ mandated policies as "poor quality" or "irrelevant." Based on these findings, we suggest that the framework of multiple legitimacy challenges could be used as a theoretical lens to better understand the transfer of HRM practices in MNEs in future research.
F I G U R E 1 A conceptual framework of human resource management (HRM) practice transfer in multinational enterprises (MNEs) Previous studies based on neo-institutional theory (DiMaggio & Powell, 1983) have focused on the isomorphic patterns in types of practices transferred (e.g., home-country isomorphism, local isomorphism, and transnational isomorphism) (Björkman, 2006;Pudelko & Harzing, 2007;Rosenzweig & Nohria, 1994;Westney, 1993). We argue that practice transfers should be viewed as legitimacy struggles in MNEs and thus researchers need to examine various forms of legitimacy challenges and the ways of legitimacy construction more broadly (Meyer & Rowan, 1977) rather than just examining the patterns of isomorphism. In this regard, the multi-layered comparative institutional approach (Boussebaa et al., 2012;Morgan, 2012;Morgan & Kristensen, 2006) is particularly relevant to examining the varied sources of legitimacy challenges across transnational and national institutional contexts.
Second, our findings suggest that the home country effect in the context of MNEs from newly industrialized or emerging economies needs to be understood more broadly than the formal practice dimension. Some of the neo-institutionalist literature has considered home country practice isomorphism as a key indicator of the country of origin effects, but the adoption of third-party practices by some MNEs brings this into question. We find two distinct forms of home country effects in terms of formal practices and informal managerial style. Our case suggests an interesting dynamic for MNEs, particularly those many cases that emanate from corporate traditions with highly con-  (Barnard, 2010;Fiaschi, Giuliani, & Nieri, 2017), including issues around "managerial ethos" (Zhu, Zhu, & De Cieri, 2014) which our findings also corroborate.
Asian Business Systems theories provide further insights into the nature of home country effects in the case of Asian EMNEs by highlighting the importance of considering informal, cultural, and historical developments of home country institutional configurations (Carney et al., 2009;Witt, 2014;Witt & Redding, 2013).
Third, the study sheds light on the multi-layered nature of local contexts in relation to HRM practice transfer in MNEs. We find different responses across subsidiaries in India and the United States in terms of the specific areas for local adaptation of the global policies.
As one would expect, the variation in subsidiary responses was partly explained by the institutional contexts of the two host countries, such as the dynamics of local labor markets and the employment-related legal systems. However, our study also sheds light on the combinative effects of national and subsidiary function-specific contexts on the implementation of global HRM policies. We find that the national institutional influences were highly selective in a narrow scope of HRM practice areas (e.g., the adoption of a hierarchical grade system in the Indian subsidiaries; the local rejection of forced distribution rating practice in the U.S. subsidiaries). By contrast, our study reveals a more pervasive influence of function-specific contexts giving considerable variations across subsidiaries within a host country. As evidenced by the R&D subsidiaries in both countries, the firm's relative position in the local labor markets and its reliance on the highskilled labor in them offered subsidiary managers an opportunity for leverage and led to similarities in the resistance to the adoption of the global HRM policies from HQ. This is a novel contribution to the studies of local effects in MNEs, as extant IB literature tends to focus on institutional differences or distances between MNE home and host countries (Kostova, 1999;Xu & Shenkar, 2002), paying less attentions to the multi-layered local contexts (Lu, Saka-Helmhout, & Piekkari, 2019).
Our findings also offer managerial insights on how and when subsidiary actors can exercise their potential agency to reduce the knowledge asymmetry between corporate HQ and subsidiaries with regard to local applicability of global policies. While we note that the emergence of local agency is dependent upon the presence of institutional distinctiveness in that context (Kostova, Roth, & Dacin, 2008;Oliver, 1991), we found that local adaptation was possible through local actors' abilities to create legitimacy by arguing a compelling business case (Ferner et al., 2005;Geppert & Matten, 2006;Klimkeit & Reihlen, 2016). Subsidiary managers did this by, for instance, presenting evidence of competitors' practices or pointing out the risk of losing talent. They were able to align their localization logic to a more universal business logic that could more easily be associated with the common interest shared by the HQ and the subsidiary. Without local actors' active involvement and voice with regard to potential benefits and risks entailed by the implementation of the global policies at particular locations, the required local adaptation may not occur.

| CONCLUSION
Our study highlights the significance of dominance effects, but also indicates that the global convergence toward "best practices" may still be limited by the multiple legitimacy challenges in implementing such practices. EMNEs have recently been identified as potentially important actors adopting and distributing practices from a dominant economy to their subsidiaries (Aguzzoli & Geary, 2014;Glover & Wilkinson, 2007). But our case company shows significant evidence of local adaptation in both developed and developing economy subsidiaries. This vulnerability we observe in the ability of an MNE to implement its global policies in subsidiaries is different from what may occur in other contexts where such an implementation may have been unproblematic due to a dominant position in the industry (Aguzzoli & Geary, 2014). This highlights the need for much more extensive inquiry into EMNEs in different circumstances, building especially on the rapidly growing research on Chinese MNEs in both developed economies (Ouyang et al., 2019) and less developed economies, notably those in Africa (Cooke, Wood, & Horwitz, 2015;Xing, Liu, Tarba, & Cooper, 2016). We note the variability in the way EMNEs may transfer home country practices and the form hybridization of HRM practices may take (Patel et al., 2018). We further argue that internationalization poses unique challenges for EMNEs, but it could also offer an opportunity to develop distinctive capabilities to manage the subtle differences across diverse locations (Amankwah-Amoah & Debrah, 2017) and strike a delicate balance in managing the multiple embeddedness of subsidiaries (Meyer, Mudambi, & Narula, 2011), since MNEs from a dominant economy may be less agile in responding to such a challenge.