Global talent management and multinational subsidiaries' resilience in the Covid-19 crisis: Moderating roles of regional headquarters' support and headquarters-subsidiary friction.

Resilience is central to developing organizational capability to respond to global crises such as the Covid-19 pandemic. To date, few studies have examined the relationship between global talent management (GTM) and the organizational resilience of multinational enterprise (MNE) subsidiaries during crises. This study contributes to the GTM and crisis management literature by examining the role of GTM in MNE subsidiaries' resilience during the Covid-19 crisis. Based on the sample of 166 Korean MNEs and their 1227 foreign subsidiaries, including 293 regional headquarters, operating in 49 host countries, this study shows a positive and significant impact of GTM on the resilience of Korean MNEs subsidiaries. Drawing on agency problems at the subsidiary level and bounded rationality at the headquarters level as factors that may bring about failure in TM in MNEs, as well as a core competence perspective, we shed light on the importance of the three-layered governance structure of MNEs in the analysis of GTM and MNE subsidiaries' resilience. The study has practical implications for Korean MNEs and the organizational resilience of MNEs' subsidiaries during crises more generally.

INTRODUCTION

Global crises such as the Covid‐19 pandemic have profound impacts on society and present significant challenges to organizations (Bailey & Breslin, 2021). This is especially the case for multinational enterprises (MNEs), whose subsidiaries are exposed to different levels of emergencies and resource constraints across the world and therefore require different responses from the strategic human resource management (HRM) perspective (Collings et al., 2021). The notion of resilience is central to developing organizational capability “to respond to external threats,” which “depends on the strength of employees” (Bailey & Breslin, 2021, p. 2; Linnenluecke, 2017). According to Avey, Luthans, and Jensen (2009, p. 682), resilience is “arguably the most important positive resource for navigating a turbulent and stressful workplace,” for example, in the context of cross‐border mergers and acquisitions (Khan et al., 2020) and global pandemic (Bailey & Breslin, 2021). Linnenluecke (2017) observed that developing organizational resilience requires considerable investment during normal times in order to prepare for and overcome challenges that emerge as the crisis unfolds in a variety of ways in different parts of the world.

It follows that organizations need to adopt a strategic approach to developing their human resources in general and talent more specifically to sail through the storm of crises (Lengnick‐Hall et al., 2011). This perspective suggests that organizational resilience can be fostered by developing the resilience of its workforce through specific HRM interventions, such as training and development, to elicit positive employee and organizational outcomes (Moenkemeyer et al., 2012; Wang et al., 2014). Bardoel et al. (2014) also suggest that a coherent and employee‐oriented set of HRM practices (e.g., development of social support at work, employee assistance programs, employee development programs, flexible work arrangements, occupational health and safety systems) helps enhance employee resilience.

Extant research has shown the crucial role of organizational engagement in global talent management (GTM) in MNEs (Collings et al., 2015, 2019). As Heinan and O'Neill (2004, p. 67) argue, “talent management practices can create the most enduring competitive advantages.” This can be achieved by developing a pool of high‐potential and high‐performing talents available to fill key strategic positions (Collings & Mellahi, 2009). These talents offer a high level of value‐added competencies and contribute to firms' business success (McDonnell et al., 2010; Tarique, 2021). In times of crisis, these talents are the backbones for MNEs to survive and thrive. However, GTM has become increasingly challenging due to the worsening talent shortage, especially in non‐Western markets (Vaiman et al., 2019). In this vein, it has become a critical and challenging task to coordinate GTM strategy, policy, and practice between the headquarters (HQ), regional headquarters (RHQ), and subsidiaries of MNEs to leverage their talents in the global network effectively (Collings et al., 2021; Sparrow et al., 2013).

For this paper, we define GTM as “the strategic integration of resourcing and development at the international level … [which] involves the proactive identification, development and strategic deployment of high performing and high potential strategic employees on a global scale” (Collings & Scullion, 2008, p. 102). The most common GTM practices identified are: (a) workforce planning, talent gap analysis, recruiting, and staffing; (b) training and development; and (c) talent reviews, succession planning, and retention (McDonnell et al., 2010; Stahl et al., 2007). In line with Collings et al. (2021), we also treat GTM as part of the MNE's strategic HRM.

Extant literature on TM primarily adopts the knowledge‐based and resource‐based views. Moreover, the emerging body of GTM literature is rather Western‐centric, with some exceptions (e.g., Collings & Isichei, 2018; Vaiman et al., 2019). It provides broad‐level analyses to shed light on implications for research and practice (Collings & Isichei, 2018; Meyer & Xin, 2018). However, few studies have investigated GTM across MNE levels (i.e., HQ, RHQ and subsidiaries), especially in non‐Western MNE settings. As Preece, Iles, and Jones (2013, p. 3457) argue, the “role of regional structures (i.e., RHQ) has been neglected in research on international management, and until recently, especially in research on HRM and TM.” MNE subsidiaries in different locations may play different roles strategically for the MNE, they may face distinct institutional constraints and TM challenges, especially during a global crisis, and have their own agendas and priorities. Accordingly, the HQ's local knowledge and ability to control the subsidiaries may be circumvented (Yahiaoui, 2015; Yahiaoui et al., 2021), thus creating difficulties for the MNE in GTM, which may, in turn, undermine business performance.

There is a heightened need for effective GTM to build up organizational resilience at times of high levels of uncertainty and crisis such as the Covid‐19 pandemic. Due to the geographic spread, MNE subsidiaries may be affected differently in the timing and severity of the pandemic. The specific requirements for responses to a crisis may also differ for diverse reasons. This has strong implications for GTM. To date, few studies have examined the relationship between GTM and the organizational resilience of MNE subsidiaries during crises. Collings et al. (2021, p. 1) argue, “the COVID‐19 pandemic is a human crisis” and HR leaders “have been central to the response in organisations globally,” which has “substantive implications for HR globally.” For MNEs, this involves the HR function at the HQ, RHQ, and subsidiary levels and makes their coordination critical as part of the MNEs' strategic management to combat the impact of crises. Covid‐19, therefore, provides a unique opportunity for researchers to investigate MNEs' HR responses to this global challenge to extend HR theories (Harney & Collings, 2021).

Our study fills the above research gap by exploring GTM in large Korean MNEs through a two‐time survey study at the HQ, RHQ, and subsidiary levels before and after the Covid‐19 crisis. It makes four main contributions to the GTM literature. First, the crisis literature has focused primarily “on the financial and physical resources fostering resilience” (Torres et al., 2019, p. 168). Therefore, research on GTM in a turbulent business context in general (Farndale et al., 2010; Vaiman & Collings, 2013) and the role of GTM in MNE subsidiaries' resilience, in particular, are scarce. This study fills these gaps by accentuating the critical role of GTM as part of the strategic management for MNEs' to develop organizational resilience to weather global crises. Second, we address the gap wherein GTM studies have not conceptually and empirically distinguished the multilayered governance nature of MNEs, i.e., HQ, RHQ, and subsidiaries. This three‐layered governance structure is crucial for organizational resilience, as well as the performance and survival of the foreign subsidiary during a global crisis. In particular, we accentuate the linchpin role of RHQ in global talent development (GTD). Third, most of the GTM studies have focused on the positive side of GTM. By contrast, the dark side of GTM, such as frictions between the HQ and its foreign subsidiaries in talent development, caused by the bounded rationality of HQ and the agency problem in subsidiaries, remains largely unexplored. There is little knowledge of how friction may influence the positive relationship between HQ‐led GTM and the organizational resilience of MNE subsidiaries. This study sheds light on these issues by highlighting the mutual influence of the HQ strategic orientation, the HQ–subsidiary relationship, and GTM. These factors present challenges to the MNE internationally and their dynamic relationships co‐evolve, which underpins various aspects of GTM. Fourth, our sample is unique, since it is based on Korean MNEs belonging to business groups with footprints across the globe. It provides insights into how MNEs affiliated with business groups behave in terms of GTM and maintain resilience before and after the Covid‐19 crisis.

THEORETICAL BACKGROUND AND HYPOTHESES DEVELOPMENT

Organizational resilience in pre‐and post‐Covid‐19 crisis

There are two approaches to organizational resilience (Teo et al., 2017). The first approach assumes resilience as the ability to bounce back from a disturbing temporary context to a state of equilibrium (Holling, 1973). The second treats organizational resilience as a learning process, moving from adjustment to a current turbulent context to the continuous development of the ability to overcome forthcoming hazards and crises (Teo et al., 2017; Wildavsky, 1988). In this paper, we adopt the second approach and consider resilience as “the capacity of individuals, households, businesses and communities to adapt to external shocks and ultimately thrive” (Torres et al., 2019, p. 178), and organizational resilience as an “ongoing process of protection, assessment, and improvement” (Mithani, 2020, p. 522).

To analyze this process, we need to consider organizational status both pre‐and post‐crisis (Torres et al., 2019). Some researchers relate organizational resilience to the conditions that persist in operation at the post‐crisis stage (Marshall & Schrank, 2014), whereas others highlight the need to assess business survival over time (Chang, 2010). In a crisis context, a business must firstly survive to recover and then recover to be resilient (Torres et al., 2019).

Furthermore, the level of organizational resilience will depend on the conditions and the crisis life cycle, and whether it is a sudden or recurrent crisis (Branzei & Abdelnour, 2010; Tang, 2006). In the MNE context, a threat that has an impact on talents in a particular locale may require a different response than one that impacts all the organization. Even the same threat requires the assessment of its potential impacts across the organization and the formulation of tailored regional and national plans based on collaborative forums and joint solutions (Mithani, 2020). Thus, talents across the world, while sharing an enemy such as Covid‐19, could react and manage their critical contexts differently (Caligiuri et al., 2020).

To measure businesses crisis recovery between pre‐ and post‐crisis stages, there are different types of indicators: the objective (e.g., gross revenue, ROA, growth in sales, number of employees, debt level, industry, size, business age, industry experience, emergency planning, pre‐crisis success, and crisis experience) and the subjective ones (e.g., customer satisfaction, leaders' perceptions of the business success, and managerial skills) (Marshall et al., 2015; Torres et al., 2019). To examine MNEs, we add the international experience of the company since its development with reference to its first foreign direct investment.

Mithani (2020) highlights a range of organizational reactions to life‐threatening events. Based on a list of previous studies on organizational resilience, he pointed out five resilience modes (avoidance, absorption, elasticity, learning, and rejuvenation) and two mechanisms (static and dynamic) operating at the individual (well‐being, dealing with frequent challenges, new experiences, and new beliefs) or organizational level (prior planning and design, active engagement, and being reluctant to accept simple solutions). In parallel to these two levels, we mention static resilience (with effective deployment of resources) or dynamic resilience (with the reassessment of resources). Managers are then asked to choose the most appropriate resilience mode with the right mechanism to use. To make the right choice, MNEs need culturally agile leaders who are able to adapt to the context norms and demands to nurture confidence with local staff; persuade and influence decisions; integrate different perspectives and demands according to the team cultures; set clear directions even if they are unpopular ones; and communicate about them effectively (Caligiuri et al., 2020). This critical role of leadership in activating organizational resilience was studied by Teo et al. (2017) during the severe acute respiratory syndrome (SARS) crisis in Singapore. They showed “how leaders can leverage upon their social influence to harness cognitive, social, and emotional reservoirs inherent within social networks, to activate organizational recovery” (Teo et al., 2017, p. 146). Hence, the development of leadership competencies among high‐potential talents is central to the overall organizational resilience and competitiveness (Khoreva et al., 2017). The Covid‐19 crisis showed the need to emphasize the types of talent portfolios and their contribution to corporate resilience in the context of global uncertainty (Caligiuri et al., 2020). This current crisis changes the requirements for a talent pool and the need to redefine talent within MNEs by accentuating its evolutionary nature (McDonnell et al., 2010), especially from the HQ and its RHQ side.

GTM and organizational resilience of MNE subsidiaries

The critical role of the leadership by talented managers in overcoming the challenges posed by external crises discussed above suggests that the establishment of an effective GTM system by MNEs would enhance the resilience of their global operations to major external crises having global impacts. Different organizational levels within the MNEs, however, are likely to play different roles in GTM. Specifically, the HQ, RHQ and local subsidiaries of MNEs would play different roles in the identification, development, and strategic deployment of high‐performing and high‐potential employees, which constitute the major dimensions of GTM in Collings and Scullion's (2008) definition thereof. Considering that other definitions of GTM often include the attraction and selection of high‐potential employees as major elements (e.g., Collings et al., 2019; Tarique & Schuler, 2010), we view that identification encompasses the recruitment of managers with high levels of relevant human capital, as well as the identification of strategic positions—“those organizational roles that can have an above‐average impact” (Caligiuri et al., 2020, p. 704)—to be filled by these managers. Strategic deployment, in turn, denotes staffing these strategic positions on the global stage with high‐performing and high‐potential employees. We summarize the different roles of HQ, RHQ, and local subsidiaries in GTM and the different types of human capital that should be fostered at each level in Table 1.

TABLE 1

A summary of the different roles of HQ, RHQ, and subsidiaries in GTM

Levels	Roles in GTM	Types of human capital fostered at each level a
HQ	Identification and development of top talents and their deployment at strategic positions across the MNE's global network	Corporate
RHQ	Identification and development of regional entrepreneurs with skills in intra‐regional cooperation and nurturing them for higher positions within the global corporation	International/regional
Subsidiary	Attracting, selecting, and retaining host country talents with high levels of local human capital and developing subsidiary human capital	Local and subsidiary

Based on the classification by Morris et al. (2016).

Specifically, at the HQ level, corporate human resource (CHR) management at HQ plays “the key role … in GTM for the top talent across the company” (Farndale et al., 2010, p. 165, original emphasis). Specifically, the role of CHR in GTM includes senior management development, succession planning, and cultivating a cadre of global managers. In terms of the dimensions of GTM, HQ thus plays a pivotal role in the identification and strategic deployment of top talents on the global stage.

At the regional level, RHQ plays an important role in connecting the MNE's HQ and overseas subsidiaries and facilitating cooperation and coordination within the region (Preece et al., 2013). Specifically, RHQ's functions regarding GTM include “enhancing interaction through such HRM activities as workshops, meetings, and intra‐regional teams,” and “engaging in intra‐regional training and development” (Preece et al., 2013, p. 3462). Through such activities, RHQ endeavors to nurture “regional entrepreneurs” with skills in intra‐regional coordination. RHQ thus plays a significant role in the identification and development of senior managers who would facilitate regional collaboration and constitute potential candidates for higher positions in the corporation.

Subsidiaries play a key role in recruiting and developing talented managers in host countries. According to Morris et al. (2016), four types of human capital are relevant in GTM: local, subsidiary, corporate, and international. TM at subsidiaries thus focuses on attracting, selecting, and retaining host country managers with high levels of local human capital and developing their subsidiary human capital.

HQ‐led GTM and organizational resilience

As noted above, HQ plays a key role in the identification and strategic deployment of top talents across the MNE's global network. Effective GTM led by HQ would thus make it more likely that pivotal positions at MNE's affiliates across the globe are staffed by top talents and thereby enhance the performance of these affiliates (Collings et al., 2019). While having the right people in the right places would matter in normal times, it could be more crucial in times of crisis (Caligiuri et al., 2020). Studies generally indicate that top managers play an especially important role during an organizational crisis (e.g., Dowell et al., 2011; König et al., 2020; Lafley, 2009; Wooten & James, 2008). Having the right people at strategic positions would increase the potential for the subsidiaries to weather the crisis by enhancing the organizational preparedness for external shocks and enabling effective leadership in crisis management (Bundy et al., 2017).

Under stable environments, the firm's operation would proceed in a more or less routinized manner, following policies and practices that have been proven effective through experience and thereby have come to be accepted as norms. In a turbulent environment, however, conventional policies and practices are likely to become less effective or rendered useless, and firms thereby face the challenge of developing new ways of managing their operations. In a similar vein, Garcia (2006, p. 4) suggests that whether “an organization survives a crisis … is determined less by the severity of the crisis than by the timeliness and effectiveness of the response.” Organizational crises require innovative, “out‐of‐the‐box” responses (König et al., 2020, p. 132), which in turn require effective leadership. Whether key positions at subsidiaries are staffed by top talents, therefore, would be more crucial in times of crisis, as their competencies, motivations, and initiatives are very much needed for the organization to respond flexibly to the volatile situations and explore novel ways of navigating through uncharted territories.

For example, during the current crisis, Covid‐19 reshapes the workplace, creating a “new normal” of work by breaking the constraints of time and place, limiting international travel and mobility, creating a new remote workplace, and leading to cross‐border and intra‐firm distancing. To maximize effectiveness in this context, MNEs require talents to be culturally agile, tolerant of ambiguity, resilient, and curious (Caligiuri et al., 2020). These competencies, in addition to the technical skills, are highly relevant to manage global activity and virtual teams across cultures.

In sum, GTM at HQ renders it more likely that strategic positions at subsidiaries are filled by top talents. Having high‐potential managers in the right place is especially important in times of crises, as they could facilitate flexible and innovative responses to a turbulent environment. HQ‐led GTM would thereby enhance the resilience of overseas subsidiaries to major external crises. Accordingly, we hypothesize as follows:

HQ‐led GTM is positively related to organizational resilience of the foreign subsidiary.

The moderating roles of RHQ‐assisted GTD and HQ–subsidiary GTM friction

Besides having top talents placed at pivotal positions, the availability of a pool of senior‐ and middle‐level managers with high levels of human capital would also be important for the subsidiaries to respond effectively to external shocks. Availability of talent per se, however, is of little strategic value if it is not nurtured and used effectively (Mellahi & Collings, 2010, p. 144). Thus, the HQ‐led GTM will be more effective in enhancing the resilience of subsidiaries when it is augmented by proactive programs to nurture and develop the senior and middle‐level managers at subsidiaries.

Talent development constitutes “an important component of the overall talent management process” (Garavan et al., 2012, p. 5); it “focuses on the planning, selection and implementation of development strategies for the entire talent pool to ensure that the organization has both the current and future supply of talent to meet strategic objectives and that development activities are aligned with organisational talent management processes” (Garavan et al., 2012, p. 6). Talent development is context‐specific, that is, “on‐going talent development processes need to be flexible, adaptable, and capable of scalability and in tune with the evolving context” (Garavan et al., 2012, p. 8). This leads to the reflection that RHQ could play a more important role in the talent development process, as they are more familiar with and can adapt more flexibly to the evolving context within the region. Tarique and Schuler (2010, p. 126) similarly suggest that “many of GTM challenges are region […] specific” and thus point to the important role of RHQ in identifying and developing talents within the region. In particular, through organizing intra‐regional teams, workshops, and meetings, and engaging in intra‐regional training and development (Preece et al., 2013), RHQ plays an important role in developing the global competencies of subsidiary managers.

Global competencies refer to “the abilities to operate in culturally, linguistically, spatially, and temporally distant environments” (Collings et al., 2019, p. 556). Having a pool of talents with global competencies would enhance the subsidiary's capacity to respond more effectively to external crises having global impacts, as such crises are likely to necessitate closer coordination and cooperation between the MNE's affiliates in different countries. Studies also indicate that the ability of managers to work effectively with others in distant places with different linguistic and cultural backgrounds becomes more important under turbulent environments (e.g., Caligiuri et al., 2020; De Cieri et al., 2019; Sit et al., 2017). As RHQ plays a pivotal role in the development of such competencies, the GTD programs implemented by RHQ would increase the efficacy of the GTM led by HQ in enhancing the subsidiaries' capacities to respond more effectively to external crises. Accordingly, we propose the following hypothesis.

RHQ‐assisted GTD for its HQ strengthens the positive relationship between HQ‐led GTM and organizational resilience of the foreign subsidiary.

Mellahi and Collings (2010) have identified agency problems at the subsidiary level and bounded rationality at the HQ level as factors that may bring about failure in TM in MNEs. The agency problem arises due to the potential incongruence between the goals and objectives of the MNE HQ—the principal—and those of the subsidiary—the agent—and the potential for the latter to pursue its own interests at the sacrifice of those of the former (Gong, 2003; O'Donnell, 2000). Bounded rationality is attributable to “the cognitive limits experienced by individuals in their ability to process and interpret a large volume of pertinent and complex information in their decision‐making process” (Mellahi & Collings, 2010, p. 146). Accordingly, the decision‐making tends to be rationally bounded rather than being perfectly rational (Simon, 1979).

Due to the agency problem and bounded rationality, frictions between the HQ and subsidiaries may arise in identifying, developing, and deploying high‐potential employees in subsidiaries. The agency problem may give rise to conflicts in the identification, deployment, and development of talented subsidiary managers between the HQ and subsidiaries in the following respects. First, it is likely to be in the best interest of the subsidiaries to keep their best talent within the subsidiaries rather than promoting and transferring them to other units within the MNE. The self‐interest seeking by subsidiaries would thus hinder the potential for HQ to identify high‐potential employees in subsidiaries effectively and deploy them in other important strategic positions within the MNE as a whole. Second, subsidiary managers may have little incentive to develop the human capital of the subsidiary's high‐potential employees beyond that applicable within the host country context. Such a lack of incentives would impair the effectiveness of the MNE's efforts to nurture and develop the global competencies of its high‐potential employees.

At the same time, bounded rationality constrains the HQ's ability to make an accurate judgment about talented managers working in distant overseas subsidiaries. Thus, “management at the centre are less likely to effectively identify talented people, and develop appropriate strategies to nurture and retain them in physically distant locations” (Mellahi & Collings, 2010, p. 147). Moreover, high‐potential employees working in distant locations have limited interactions with decision‐makers in the center, which may lead to the “marginalization of subsidiary talents” (Mellahi & Collings, 2010, p. 147) and, consequently, under‐utilization thereof on a global scale.

Frictions arising from the agency problem and bounded rationality would thereby undermine the potential for HQ to effectively identify, develop, and deploy high‐potential and high‐performing employees in overseas subsidiaries. While the friction would be detrimental under normal conditions, it would be especially so in crises, as it likely undermines organizational preparedness and hinders effective leadership for crisis management. Thus, the effectiveness of HQ‐led GTM in enhancing the resilience of subsidiaries to external shocks would be impaired to the extent that there exist frictions between HQ and subsidiaries regarding talent management. Accordingly, we hypothesize as follows.

Friction in GTM between HQ and its subsidiary weakens the positive relationship between HQ‐led GTM and the organizational resilience of the foreign subsidiary.

Based on the above discussion for our hypotheses, we summarize our conceptual model in Figure 1.

FIGURE 1

A conceptual model of GTM in MNEs

METHOD

Empirical context

The empirical context of this study is relatively novel because we deal with group‐affiliated MNEs that belong to business groups from South Korea (hereafter Korea) and their foreign subsidiaries operating in 49 host countries. Korea is frequently categorized as one of the “advanced emerging economies” or “advanced Asian economies” (Kim et al., 2010; Witt & Redding, 2013), since it has the characteristics of both developed economies and capitalist emerging Asian countries (Carney et al., 2009; Witt, 2014). HRM practices in Korean MNEs were traditionally paternalistic, heavily affected by Confucianism and Japanese HRM practices, as well as committed to internal labor markets and a seniority‐based promotion system (Bae & Rowley, 2003). However, the 1997–1998 Asian financial crisis impacted the traditional HRM practices in Korean MNEs (Debroux et al., 2018), and since then, there have been dramatic changes, such as the introduction of “individual performance‐based approaches and more flexible labor market policies” (Chung et al., 2020, p. 540). These practices are generally based on the US HRM models, which were considered to be the global standards (Horak & Yang, 2019).

Yet, since the late 1960s, the Korean economy has been largely dominated by the Chaebol, defined as “a Korean business group of peer group‐affiliated companies (GACs) clustered under a single administrative and financial controlling entity” (Lee et al., 2010, p. 587). Thus, the main characteristics of the Chaebols have been deeply embedded in the HRM practices and styles of Korean MNEs and their foreign subsidiaries (Zou & Lansbury, 2009). Moreover, since Korean MNEs have expanded into international markets, they have developed their IHRM practices. However, many Korean MNEs are still experiencing the challenge to “[manage] the potential tensions between the espoused ‘best practice’‐centered HRM policies and the traditional hierarchical management style” (Chung et al., 2020, p. 541). In particular, the crisis of Covid‐19 pandemic presents an unprecedented environmental disruption, and even displacement, of established managerial norms and HR practices. This has led many Korean MNEs to rethink their traditional way of organizing IHRM (Caligiuri et al., 2020), as well as to reconsider how their GTM can help develop the resilience of their foreign subsidiaries in the light of Covid‐19 crisis.

Sample

Due to the complexity of the phenomenon in this study, we selected a two‐wave survey method, although we combined the primary data with secondary data. The sample of this study consists of 54 Korean business groups: all are involved in the Chaebols that undertook international diversification and are on the monitoring list of the Korea Fair Trade Commission (KFTC). According to the Korean Fair Trade Act, any business group in Korea that owns over 5 trillion won (KRW) worth of group‐level total assets should disclose their main trades between GACs in the Electronic Disclosure System of the Korean Financial Supervisory Service (KFSS). And any business group that owns over 10 trillion won of group‐level total assets should not cross‐subsidize, engage in circular shareholding, or guarantee a debt between sister GACs within the same group.

Based on previous studies (e.g., McDonnell et al., 2010; Torres et al., 2019), we developed the first survey questionnaire before, and after the Covid‐19 crisis, we developed the second survey questionnaire. The first questionnaire was pretested for face and construct validity with 28 senior HR managers of 25 largest Chaebol MNEs. We interviewed these senior HR managers and asked them about the appropriateness of our questionnaire for each MNE's GTM, GTD, and HQ–subsidiary friction. There was a high correspondence between the perceptions and the concepts from the prior study (McDonnell et al., 2010). The second questionnaire was also pretested for face and construct validity with 25 CEOs of foreign subsidiaries belonging to the Korean MNEs, which were included in the sample of the first survey. We interviewed these CEOs and asked them about the appropriateness of our questionnaire for the organizational resilience of foreign subsidiaries, and there was a high correspondence between the perceptions and the concept from the prior study (Torres et al., 2019) as well.

As suggested by the Chaebol MNE literature (Lee et al., 2010), during our data collection of the first survey in the 4 months from August to November 2019, we employed three approaches to increase the response ratio and reliability: (1) an off‐line visit to the firm; (2) brief face‐to‐face interviews with our respondents; and (3) on‐site collection of respondents' questionnaires; for some cases, we received the completed questionnaires by email. As a result, the response rate for this first survey was substantially high, like other survey‐based studies for Korean MNEs (e.g., Lee et al., 2010: 94.5%; Lee et al., 2014: 93.6%) that used the same data collection approach. Altogether, 171 responses from 193 requested senior HR managers at the HQ (as well as 336 complete responses from 375 requested ones at the RHQ and 1785 complete responses from 1956 requested ones at foreign subsidiaries) were received, so the response rate was 86.0% at the HQ level. Thus, our sample in the first survey included 171 Chaebol MNEs belonging to 54 Korean business groups and their 1785 foreign subsidiaries. After the Covid‐19 pandemic outbreak, we conducted the second survey for 5 months from June to October 2020. This time we targeted the 1785 foreign subsidiaries and only performed an email survey of CEOs of these subsidiaries, due to the unique conditions of the Covid‐19 crisis. We received 1227 complete responses from these subsidiaries' top management teams; therefore, the final sample in this study consists of 166 MNEs affiliated to 54 Korean business groups and their 1227 foreign subsidiaries, including 293 RHQ, operating in 49 host countries (see Table 2). We can justify that this half a year gap was sufficient to measure our context of the foreign subsidiary's organizational resilience due to the four macro‐economic‐business evidence. First, according to KFTC (2021), the average sales growth rate for Chaebol firms during the third and fourth quarters of 2019 was −1.56%, but after the Covid‐19 shock during the first quarter of 2020, the average sales growth rate for Chaebol firms during the third quarter of 2020 has recovered rapidly. And even during the fourth quarter of 2020, the average sales growth rate for Chaebol firms reached 6.63%. Also, average return on assets (ROA) for these large firms had shown a similar path of average sales growth rate, so during the first quarter of 2020, average ROA was decreased to less than 0.5% from 1.05% of average ROA during the third and fourth quarter of 2019, but the third and fourth quarters of 2020, it was recovered to around 1.0%. Second, according to KFTC (2021), during our first survey period of 2019, the total number of Chaebol firms was 2103 and the total designated number of Chaebol groups was 59, but after the Covid‐19 shock, the macro‐business condition of these Chaebols has rapidly recovered. So, during our second survey period of 2020, the total number of Chaebol firms was increased to 2284 and the total designated number of Chaebol groups was increased to 64. Third, according to the Korea Stock Exchange (2021), on November 29, 2019, the Korea Composite Stock Price Index (KOSPI) showed 2087.96 points before the Covid‐19 crisis, but the Covid‐19 shock hit hard on March 20, 2020, with 1566.15 points, but after the shock, the KOSPI was recovered to (and even exceeded the 2019 record) 21,881.87 points on June 5, 2020, and 2391.96 points on October 8, 2020. Fourth, the Korean Ministry of Economy and Finance (2021), during our first survey period of 2019, the gross domestic product (GDP) and GDP growth rate of Korea were 497,064.8 billion won and 2.3% respectively, and during the first, second, and third quarters of 2020, the GDP was changed to 458,202.4 billion won, 472,328.1 billion won, and 491,181.5 won, respectively; while during these quarters of 2020, GDP growth rate was −1.3%, −3.2%, and 2.2%, so that revealed a clear recovery.

TABLE 2

The description of questionnaire distributions and complete responses in each survey

Level	First survey		Second survey
	Questionnaire distribution	Complete responses	Questionnaire distribution	Complete responses
HQ	193	171	None	None
RHQ	375	336	None	None
Foreign subsidiary	1956	1785	1785	1227

In addition, we examined the responses for non‐response bias by employing five parameters (subsidiary size and age, parent firm size and age, industry classification, early respondents, and late respondents). We found no significant differences between the responding and non‐responding subsidiaries and parent companies with regard to two key parameters, or between the early respondents and the late respondents (Deming, 1990; Fowler, 2009). Based on these results, we conclude that non‐response bias is minimal.

Variables and measurements

Dependent variable

In line with the previous literature (Torres et al., 2019), our dependent variable, the foreign subsidiary's organizational resilience, was measured by an ordinal variable. Hence, the foreign subsidiary's recovery was measured by comparing gross revenues before and after the Covid‐19 crisis. From the Korean Chaebol MNEs' side, the Covid‐19 pandemic is still going on even currently, but the crisis has been solved for the Korean Chaebol MNEs mostly as can be seen in the above sample section. Accordingly, we can justify the 6‐month time interval based on our four macro‐economic‐business evidence when we measured this dependent variable. Specifically, “revenues went down (y = 0), stayed about the same (y = 1), or went up (y = 2)” (Torres et al., 2019, p. 173). Our sample includes 846 cases for “0,” 127 cases for “1,” and 254 cases for “2.”

Independent variable

Extending the previous literature (McDonnell et al., 2010), our independent variable, (HQ‐led) global talent management (GTM), was also measured by an ordinal variable. Thus, GTM was operationalized: “Is there both a global succession planning system and formal management development program, which is led by your headquarters, for high potentials covering each host country's (e.g., Chinese, US, UK, etc.) operation? Yes (= 2); either global succession planning or management development (= 1); neither system (= 0).” We included the 220 cases with “0,” as those cases indicate MNE HQs that do not implement any formal GTM programs. On the other hand, cases with “2” indicate those implementing the most comprehensive GTM programs. Thus, our independent variable reflects the extent to which the MNE HQ implements formal GTM programs, ranging from none (0) to most comprehensive (2). Our sample includes 220 cases for “0,” 768 cases for “1,” and 239 cases for “2.”

Moderating variables

Extending the prior literature (McDonnell et al., 2010; Mellahi & Collings, 2010), our moderating variables, RHQ‐assisted GTD and HQ–subsidiary GTM friction, were also measured by two ordinal variables. Specifically, RHQ‐assisted GTD was operationalized by counting the number of talent development programs being implemented by RHQ for HQ within a Korean MNE, such as “The mechanisms, being supported by your regional headquarters for the headquarters, are short and long‐term international assignments, formal global management training, performance assessment against global management competencies and external qualifications. Six categories are incorporated; those that include all five mechanisms (=5), those including four mechanisms (=4), those including three mechanisms (=3), those including two mechanisms (=2), those including one (=1) and those including none (=0).” Our sample includes 55 cases for “0,” 63 cases for “1,” 45 cases for “2,” 265 cases for “3,” 278 cases for “4,” and 521 cases for “5” for these six categories.

Next, the HQ–subsidiary GTM friction was operationalized to count the number of mechanisms facing the friction between the HQ and its foreign subsidiary within a Korean MNE, such as “The mechanisms facing the friction between the headquarters and your subsidiary are: short and long‐term international assignments, formal global management training, performance assessment against global management competencies, and external qualifications.” Six categories are incorporated: those that include all five mechanisms (=5), those including four mechanisms (= 4), those including three mechanisms (=3), those including two mechanisms (=2), those including one (=1), and those including none (= 0). Our sample includes 239 cases for “0,” 111 cases for “1,” 251 cases for “2,” 188 cases for “3,” 139 cases for “4,” and 299 cases for “5” for these six categories.

Control variables

We controlled for a group of factors that are also predicted to affect the foreign subsidiary's organizational resilience. First, because younger subsidiaries suffer from the liability of newness (Benito, 1997) and are less likely to be resilient to environmental disruptions like the Covid‐19 crisis, we included foreign subsidiary age measured as the logarithm of the time interval between the first year of establishment and the year of our observation. Foreign subsidiary ownership reflects the MNE's involvement in the investment and effective control over the subsidiary's operation. As the percentage of a foreign subsidiary's ownership share by its parent firm increases, that subsidiary is likely to be regarded to have a lower failure ratio and to be more resilient (Hennart et al., 1998). In addition, foreign subsidiary size is predicted to affect the organizational resilience of the subsidiaries, since these foreign subsidiaries with larger resources or capabilities are likely to possess the additional “human capital” that helps these subsidiaries cope with environmental disruptions (Caligiuri et al., 2020; Chung et al., 2015). Thus, larger subsidiaries (Caligiuri et al., 2020) are assumed to be more resilient against environmental disruptions. Subsidiary size was measured as the logarithm of the number of total employees for each foreign subsidiary.

Second, the same situation for the parent firm‐level variables was also predicted to affect the organizational resilience of foreign subsidiaries, because foreign subsidiaries with larger parental resources or capabilities have additional “human capital” that help the subsidiaries address environmental hardships (Caligiuri et al., 2020; Chung et al., 2015). Specifically, larger firms (Caligiuri et al., 2020; Dhanaraj & Beamish, 2004) are forecast to be less likely to fail and are thus likely to be more resilient. Parent firm size is measured as the logarithm of the number of total employees for each foreign subsidiary's parent firm. Next, a parent firm's previous overall experience in business and its prior experience operating internationally help its foreign subsidiaries respond better to the environmental disruptions in home and host countries (Chung et al., 2010). We measure parent firm age as the logarithm of years since the establishment of each parent firm and measure each parent firm's international experience by calculating a count‐year of experience based on the sum of the operating years for all foreign subsidiaries until our observation (Delios & Beamish, 2001). In addition, foreign subsidiaries with larger parental R&D capabilities have extra innovation capabilities that help the subsidiaries respond to environmental uncertainties; hence these subsidiaries are more likely to be resilient against the Covid‐19 crisis. Parent firm R&D intensity is measured as the percentage of each parent firm's R&D expenses divided by its total sales (Chung et al., 2015).

Third, we considered host country effects by including each host country's GDP growth rate (measured as the percentage of GDP growth) and the logarithm of each host country's GDP based on the International Monetary Fund (IMF) database for the host country GDPs. The cultural distance between Korea and each host country is also controlled and measured by using Kogut and Singh's (1988) method, but with the data of the GLOBE (House et al., 2004) cultural index, instead of Hofstede (1980), to overcome the weaknesses of Hofstede's cultural index (Shenkar, 2001). Cultural distance is predicted to increase foreign subsidiaries' management costs and negatively influences their resilience; while, alternatively, we assimilate cultural distance to “cultural diversity.” The latter is assumed to create synergy in cross‐border operations, cross‐border knowledge‐sharing, and learning, and unleash the creative potential of diverse teams (Shenkar, 2001; Stahl et al., 2010; Tung & Verbeke, 2010; Zaheer et al., 2012); thus it positively influences the subsidiaries' resilience. In addition, each host country's Covid‐19 case fatality rate, Covid‐19 cases, and Covid‐19 test capacity need to be controlled, since our dependent variable is the foreign subsidiary's resilience by comparing its gross revenues between before and after the Covid‐19 crisis. Host country Covid‐19 case fatality rate is measured as the ratio of the cumulative confirmed Covid‐19 death tolls divided by the cumulative confirmed Covid‐19 cases for each host country, host country confirmed Covid‐19 cases are measured as the cumulative confirmed Covid‐19 cases per million people of each host country, and host country Covid‐19 test capacity is measured as cumulative Covid‐19 tests per 1000 people of each host country. As the ratios of these two Covid‐19 crisis‐related variables are increased, the foreign subsidiaries operating in these host countries where the Covid‐19 driven environmental disruptions are greater are likely to be less resilient.

Last, we included parent firm dummies and business group dummies to control for each organization's potential impacts of unobserved factors in association with differential firm‐ and business group‐level characteristics on their foreign subsidiaries' resilience. Also, we included industry dummies using the two‐digit Korean standard industry code to consider the potential effects of unobserved differentials in capital intensity or competition related to differential industrial characteristics on the foreign subsidiaries' resilience (Caligiuri et al., 2020; Song & Lee, 2017).

Statistical estimation approach

In line with the previous literature, wherein dependent variables are ordinal variables, such as small business resilience (Torres et al., 2019) or MNE's team performance (Huang & Cummings, 2011), we employed an ordered logit regression model to examine the relationship between GTM and the foreign subsidiary's resilience and the moderating effects of GTD and HQ–subsidiary GTM friction on this aforementioned relationship. The ordered logit regression model is a statistical estimation model preferable to ordinary least squares (OLS) regression, due to our observed dependent variables that have an ordinal scale (Greene, 2003), as such gross revenues of foreign subsidiaries after the Covid‐19 crisis could be lower, the same, or higher than those before the Covid‐19 crisis. Because the foreign subsidiary's gross revenue is originally a continuous variable, the rating method can follow a naturally ordered scale. We assumed the recovery process of a foreign subsidiary has a natural order (from low to high); yet, the distances between adjacent degrees of operating subsidiaries are unknown (Greene, 2003). Thus, we used the “ologit” command to estimate an ordered logistic regression model in our analysis by Stata 15 (Long & Freese, 2006).

RESULTS

Common method bias

Our key variables were based on self‐reported data. As such, there might be the possibility of bias stemming from common method variance (CMV) (Podsakoff et al., 2003). However, our survey could not be undermined by CMV because, by conducting the two‐wave surveys, we collected separate responses with time gaps between answering independent/moderating variables and answering a dependent variable. This was a procedural remedy “to create a temporal separation by introducing a time lag between the measurement of the predictor and criterion variables” (Podsakoff et al., 2003, p. 887). Moreover, we employed another procedural remedy, in that the sources of responses for the independent/moderating variables and a dependent variable were different (Podsakoff et al., 2003); thus, we eliminated the CMV through the design of our study.

Correlation analysis and multicollinearity diagnosis

The descriptive statistics and correlation coefficients for all variables are summarized in Table 3. As shown in Table 3, GTM is positively and significantly correlated to the foreign subsidiary's resilience (r = 0.12, p < 0.01), which is consistent with H1. RHQ‐assisted GTD is positively and significantly correlated with the foreign subsidiary's resilience (r = 0.06, p < 0.05), whereas HQ–subsidiary GTM friction is negatively and significantly correlated with the foreign subsidiary's resilience (r = −0.07, p < 0.05). Hence, the correlations between these independent and moderating variables and a dependent variable are relatively weak. To make a diagnosis for any potential multicollinearity among variables, we checked the variance inflation factor (VIF) for each variable; a VIF over 10 is indicative of a multicollinearity problem (Kennedy, 1992; Studenmund, 1992). Our results show that the VIFs for our explanatory variables did not exceed 2.0, therefore we concluded that our sample was devoid of multicollinearity.

TABLE 3

Means, standard deviations, and Pearson correlations

		Level	Mean	SD	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16
1.	Subsidiary's organizational resilience	Subsidiary	0.52	0.82
2.	HQ global talent management	HQ	1.02	0.61	0.12**
3.	RHQ global talent development	RHQ	3.80	1.39	0.06*	0.19**
4.	HQ–subsidiary GTM friction	Subsidiary	2.63	1.81	−0.07*	0.06*	0.05
5.	Subsidiary size (log)	Subsidiary	8.75	2.95	0.04	0.09**	0.13**	−0.01
6.	Subsidiary age (log)	Subsidiary	5.84	0.62	0.07*	0.08**	0.16**	−0.05	0.35**
7.	Ownership share (%)	Subsidiary	71.05	30.56	0.04	−0.02	−0.01	0.00	0.06*	−0.04
8.	Cultural distance (GLOBE Kogut & Singh)	Country	2.42	1.07	0.07*	0.01	0.05	0.00	0.35**	0.05	0.09**
9.	Host country GDP (log)	Country	30.88	2.07	−0.03	0.03	0.12**	−0.02	0.12**	0.01	0.05	0.07*
10.	Host country GDP growth	Country	3.45	5.82	−0.00	−0.01	−0.01	0.03	0.03	0.07*	−0.03	0.09**	0.14**
11.	Host country Covid‐19 case fatality rate	Country	0.16	1.09	−0.04	−0.01	0.03	−0.05	0.04	−0.01	0.03	−0.01	0.00	−0.14**
12.	Host country confirmed Covid‐19 cases	Country	1062.01	1847.88	−0.04	0.01	0.00	−0.04	−0.04	0.01	−0.18**	−0.20**	0.23**	−0.40**	0.09**
13.	Host country Covid‐19 test capacity	Country	30.472	19.801	−0.06	−0.04	−0.03	−0.06*	0.00	−0.08**	−0.06*	−0.09**	0.32**	−0.21**	−0.00	0.49**
14.	Parent company size (log)	Parent	31.96	2.04	0.07*	0.29**	0.21**	0.18**	0.36**	0.02	0.05	0.13**	0.18**	0.04	0.01	0.01	−0.07*
15.	Parent company age (log)	Parent	6.65	0.85	0.12**	0.05	0.01	0.18**	0.02	0.03	0.02	0.04	0.06*	0.01	−0.11**	0.00	0.00	−0.04
16.	Parent company R&D intensity (%)	Parent	2.64	2.87	0.25**	0.09**	0.08**	0.07*	0.14**	0.09**	0.05	−0.02	0.04	0.05	−0.02	0.10**	0.07*	0.19**	0.16**
17.	International experience (log)	Parent	5.82	1.30	0.02	0.12**	0.23**	−0.04	0.27**	0.22**	0.01	0.05	−0.10	0.02	0.03	−0.02	−0.04	0.23**	0.08**	0.11**

Note: N = 1227. *p < 0.05; **p < 0.01 (two‐tailed significance levels). GTM refers to global talent management.

Hypothesis testing

The ordered logit analysis provides the analytical results of our hypotheses (H1–H3) shown in Table 4. We included control variables only in Model 1 and added independent and moderating variables (main effects) in Model 2. We added the interaction term between GTM and RHQ‐assisted GTD in Model 3, and we added the interaction term between GTM and HQ–subsidiary GTM friction in Model 4. Finally, Model 5 is a full model in which all the variables and interaction terms are included.

TABLE 4

Results of ordered logit regression models predicting foreign subsidiary's organizational resilience

Variables		DV: Foreign subsidiary's organizational resilience (subsidiary level)
	Hypotheses	Model 1		Model 2		Model 3		Model 4		Model 5
		B	SE	B	SE	B	SE	B	SE	B	SE
Subsidiary age (log)		0.340*	0.156	0.340*	0.156	0.348*	0.156	0.340*	0.156	0.350*	0.156
Subsidiary size (log)		0.021	0.030	0.019	0.030	0.015	0.031	0.019	0.031	0.016	0.031
Ownership share (%)		0.001	0.002	0.001	0.002	0.001	0.002	0.001	0.002	0.001	0.002
Parent firm age (log)		0.059	0.083	0.004	0.095	0.019	0.095	0.003	0.097	0.039	0.099
Parent firm size (log)		0.086*	0.042	0.084†	0.046	0.082†	0.046	0.084†	0.046	0.080†	0.046
Parent firm R&D intensity (%)		0.220***	0.028	0.217***	0.028	0.201***	0.029	0.216***	0.028	0.200***	0.029
International experience (log)		0.176	0.125	0.193	0.128	0.182	0.128	0.192	0.128	0.184	0.128
Cultural distance (GLOBE Kogut & Singh)		0.204*	0.084	0.210*	0.084	0.212*	0.084	0.211*	0.084	0.211*	0.084
Host country GDP (log)		−0.021	0.064	−0.021	0.064	−0.021	0.064	−0.021	0.064	−0.024	0.064
Host country GDP growth		−0.025	0.020	−0.026	0.020	−0.026	0.020	−0.026	0.020	−0.026	0.020
Host country Covid‐19 case fatality rate		−0.169	0.109	−0.170	0.109	−0.168	0.108	−0.170	0.109	−0.167	0.107
Host country confirmed Covid‐19 cases		−0.000	0.000	−0.000	0.000	−0.000	0.000	−0.000	0.000	−0.000	0.000
Host country Covid‐19 test capacity		−0.008†	0.005	−0.008†	0.005	−0.008†	0.005	−0.008†	0.005	−0.008†	0.005
HQ global talent management	H1(+)			0.503***	0.151	1.640***	0.359	0.604**	0.216	1.579***	0.363
RHQ global talent development				0.123*	0.056	0.145†	0.077	0.058	0.057	0.163*	0.080
HQ–subsidiary GTM friction				−0.087*	0.039	−0.021	0.040	−0.144*	0.072	−0.126†	0.075
HQ global talent management × RHQ global talent development	H2(+)					0.328***	0.089			0.371***	0.098
HQ global talent management × HQ–subsidiary GTM friction	H3(−)							−0.149*	0.062	−0.132†	0.069
N		1227		1227		1227		1227		1227
Pseudo R 2		0.110		0.161		0.210		0.203		0.254
−2 log‐likelihood		1316.461		1846.353		1843.300		1846.351		1842.782

Notes: Unstandardized coefficients are reported, and standard errors are in parentheses. All tests are two‐tailed. Parent firm‐, business group‐, and industry‐fixed effects are included but not reported. GTM refers to global talent management.

p < 0.10.

p < 0.05.

p < 0.01.

p < 0.001.

As can be seen in Model 1 of Table 4, foreign subsidiary age (b = 0.340, p < 0.05), parent firm size (b = 0.086, p < 0.05), and R&D intensity (b = 0.220, p < 0.001), as well as cultural distance (b = 0.204, p < 0.05), are positively and significantly associated with the foreign subsidiary's organizational resilience, while host country Covid‐19 test capacity is negatively and significantly associated with the foreign subsidiary's organizational resilience, and these results are consistent through all models. In particular, the positive and significant result for cultural distance suggests that cultural distance can be interpreted as the positive side of “cultural diversity” in terms of at least our most recent data and the Covid‐19 context in relation to the MNE subsidiary's organizational resilience.

H1 predicts that GTM is positively associated with the foreign subsidiary's organizational resilience. As can be seen in Model 2 of Table 4, GTM is positively and significantly related to the foreign subsidiary's organizational resilience (b = 0.503, p < 0.001), this result is consistent with Models 3–5 as well (e.g., Model 5: b = 1.579, p < 0.001), thus, supporting H1. In addition, in Model 2, RHQ‐assisted GTD is positively and significantly related to the foreign subsidiary's organizational resilience (b = 0.123, p < 0.05); whereas HQ–subsidiary GTM friction is negatively and significantly related to the foreign subsidiary's organizational resilience (b = −0.087, p < 0.05). These results are also consistent in a full model (Model 5); but in Model 3, HQ–subsidiary GTM friction is negatively but insignificantly related to the foreign subsidiary's resilience, while in Model 4, RHQ‐assisted GTD is positively but insignificantly to the foreign subsidiary's resilience.

H2 assumes that RHQ‐assisted GTD positively moderates the positive relationship between GTM and the foreign subsidiary's resilience. Model 3 shows that the interaction term between GTM and RHQ‐assisted GTD is positive and significant (b = 0.328, p < 0.001), and this result is consistent with that in Model 5 (b = 0.371, p < 0.001). To facilitate the interpretation of this interaction effect in Model 5, we graphed the interaction effect in Figure 2 (Aiken & West, 1991), which consistently supports H2.

FIGURE 2

The interaction effect between HQ GTM and RHQ GTD

Finally, H3 forecasts that HQ–subsidiary GTM friction negatively moderates the positive relationship between GTM and the foreign subsidiary's resilience. Model 4 reports that the interaction term between GTM and HQ–subsidiary GTM friction is negative and significant (b = −0.149, p < 0.05), and this result is consistent with that in Model 5 (b = −0.132, p < 0.10). We plotted this interaction effect in Figure 3, which consistently supports H3.

FIGURE 3

The interaction effect between HQ GTM and HQ–subsidiary GTM friction

In addition, Pseudo R values for Models 1–6 range from 0.110 to 0.254, which are relatively small, but in an ordered logit regression model Pseudo R values are simply reference information (Allison, 2012).

Robustness tests

We also conducted robustness tests, because we wanted to confirm that even with employing our alternative variable and statistical estimation approach, our results are still supported. First, in line with Torres et al. (2019, p. 173), instead of using the foreign subsidiary's “economic business resilience” as an alternative dependent variable, we used the foreign subsidiary's “attitudinal business resilience,” based on the comparison of the respondent's perceptions of business success between before and after the Covid‐19 crisis as “less successful (y = 0), similarly successful (y = 1), or more successful (y = 2).”. Second, as an alternative statistical estimation approach, we used the ordered probit model analysis instead of the ordered logit model analysis. Nevertheless, despite employing the alternative dependent variable and statistical estimation, the results were qualitatively unchanged. Therefore, our results are strongly supported.

DISCUSSION AND CONCLUSION

In this study, we show clearly that GTM creates a real competitive advantage for MNEs (Heinen & O'Neill, 2004) and could lead them to global success (McDonnell et al., 2010). Based on the results of this study, we highlight that one of the GTM competitive advantages is its significant effectiveness in supporting MNEs subsidiaries' resilience during a disrupted environmental context such as the Covid‐19 crisis. This confirms our first hypothesis (H1). We corroborate that HQ with GTM policies and practices are more capable to cope with a turbulent environment. Our results show that MNEs with high‐potential talents at their foreign subsidiaries (Heinen & O'Neill, 2004; Piansoongnern et al., 2008), and with the requisite global competencies needed to cope with the challenges stemming from a crisis (Caligiuri et al., 2020) and to operate virtually (Collings et al., 2019), are successful in surviving and maintaining the same performance level or even a better performance level at their subsidiaries (Torres et al., 2019). We noticed that although these Korean MNEs belonged to the Chaebol system, they succeeded in exceeding their traditional HRM practices by adopting new practices, such as GTM, which is more adaptable to the current disrupted context (Caligiuri et al., 2020).

Moreover, this study finds support for the second hypothesis (H2), suggesting that it is essential for RHQ to assist their HQ with talent development programs for the group's survival. This is important as it helps to strengthen the positive relationship between HQ‐led GTM and their subsidiaries' resilience. In fact, the moderating role of RHQ support strengthens this positive relationship, whereas the moderating role of HQ–subsidiary friction is the opposite. This confirms to a certain extent that GTM opportunities and challenges are often regional (Tarique & Schuler, 2010) and context‐related (Napathorn, 2020). It also adds insight into the local context embeddedness of managers' reactions to the crisis (Caligiuri et al., 2020).

RHQ, as the intermediator between its HQ and subsidiaries, is more aware of the local job market, the availability of local competencies, and the subsidiaries' need for talents. Thus, RHQ could help in the global competencies development of subsidiaries' managers (Collings et al., 2019) through intra‐regional training (Preece et al., 2013), and locate high potentials more efficiently and retain them (Scullion & Collings, 2006).

Morris et al. (2016) highlight that GTM is context‐related and depends on the corporate strategy. Our research goes beyond this confirmation and shows that the multi‐level interactions between GTM and RHQ‐assisted GTD and between GTM and HQ–subsidiary friction will largely depend on the international strategy alignment of the HQ but also the support of the RHQ for its international assignments, global management training, and performance assessment (Cerdin & Brewster, 2014; McDonnell et al., 2010; Mellahi & Collings, 2010). This supports Rothwell's (1994) research about the requirement for implementing such practices in each subsidiary for a successful GTM. It also shows the importance of GTD in the international experience of high‐potential talents within foreign subsidiaries (Cerdin & Brewster, 2014), in training on cross‐cultural relationship formation across borders, and in the improvement of their cross‐cultural skills to be able to manage virtual teams across cultures (Caligiuri et al., 2020; Collings et al., 2019; De Cieri et al., 2019), especially during the Covid‐19 context of remote and virtual working.

We add to this the level of HQ centralization of the GTM, which could limit the subsidiaries in assigning the right local talents, offering adequate management training, and international development. This goal incongruent situation could lead to competing interests between the HQ and its foreign subsidiary (Björkman et al., 2004; Jiang & Yahiaoui, 2019; Yahiaoui, 2015; Yahiaoui et al., 2021) and a strained relationship, which is not conducive in a disrupted context. This confirms the third hypothesis (H3), that GTM friction between the HQ and its subsidiary weakens the positive relationship between HQ‐led GTM and the subsidiaries' resilience. Indeed, within such a turbulent environment, there is a heightened need for organizational cohesion and shared professional identity across the different units (Caligiuri et al., 2020; O'Leary et al., 2014; Zimmermann & Ravishankar, 2011). This organizational cohesion cannot be possible if there is a GTM friction between the HQ and its subsidiary, as shown in our results. Consequently, subsidiary resilience will be very difficult to achieve.

Lastly, the correlation coefficients between the independent and moderating variables and the dependent variable (shown in Table 3) and the Pseudo R values of the models (shown in Table 3) seem relatively low. Firstly, the correlation coefficients between the explanatory variables and the dependent variable are still statistically significant with p < 0.01 or p < 0.05, and this is mainly due to our very large sample size. We also presume that the relatively low scores may have been caused by our data collection procedure, collected at two different time points, before and after the Covid‐19 outbreaks. Secondly, when we used an ordered logit regression model, we employed the logit link function to model the probability as a function of covariates (e.g., logistic regression). The purpose of the logit link is to take a linear combination of the covariate values (which may take any value between ±∞) and convert those values to the scale of a probability (Hardin & Hilbe, 2002). Pseudo R is similar to R in the ordinary least squares regression model, indicating an explanatory power of explanatory variables for a dependent variable. However, in the logit regression model, Pseudo R generally has small values. Accordingly, in the logit regression model, Pseudo R does not constitute an important criterion for evaluating the model fit, but it is used as reference information only (Allison, 2012).

Theoretical and empirical contributions

This paper investigates GTM from a multilevel perspective and whether alignment between the multiple levels contributes to the resilience of the MNE on the subsidiaries' level. It makes four contributions to the literature of international HRM and GTM in the context of global crises.

First, This research contributes to crisis management literature review (Marshall et al., 2015; Mithani, 2020; Teo et al., 2017; Torres et al., 2019) by showing the implications of environmental disruption to organizational HR systems in general and GTM in particular. The environmental disruption due to the Covid‐19 crisis played an important role in the HQ‐led GTM process by highlighting its importance for resilience and also the need to reconsider the GTM activities and policies according to the RHQ and subsidiaries' context. This research extends the literature by revealing the significance of GTM as a strategic part of developing MNE subsidiaries' resilience to withstand global crises and external shocks like the Covid‐19 pandemic (Caligiuri et al., 2020; Torres et al., 2019). Minbaeva and De Cieri (2015) argue that organizational resilience is a crucial outcome variable for international HRM in the context of large‐scale crises and scholars need to rethink their key dependent variable—organizational performance. Hence, our study of HQ‐led GTM in MNE subsidiaries' resilience contributes to the international HRM and GTM literature in the context of crises.

Second, this study highlights the importance of considering the three‐layered governance structure of MNEs (the HQ, the RHQ, and the subsidiary) while analyzing GTM and its friction. We argue that HQ, RHQ, and subsidiaries play different roles in GTM and that alignment of understanding and strong coordination among them is critical. In particular, adopting an evolutionary approach to GTM, this paper points to the important role of RHQ in assisting its HQ in talent development as the intermediator between the HQ and subsidiaries. Our study shows that RHQ‐assisted GTD is a support to HQ‐led GTM, thus leading to MNE subsidiaries' resilience. We, therefore, make a contribution by responding to the call for more research on the role of RHQ in HRM in general and TM in particular (Preece et al., 2013).

This contribution also questions the strategic orientation of the HQ (Jiang & Yahiaoui, 2019; Yahiaoui et al., 2021) and highlights the importance of integrating the RHQ in the decision process of GTM implementation. The RHQ is well‐positioned in the local context to have deep knowledge about its features and thereby knows better whether and how the interplay between talents and resilience could occur. The RHQ is also closer to the regional social networks and has access to influential senior leaders who could provide key business information. These social capitals can thus enhance subsidiary resilience by providing know‐how on how to solve particular business problems encountered during the crisis (Adler & Kwon, 2002).

Third, the findings of this study contribute to the GTM literature by showing the presence of a negative side of GTM and highlighting the importance of considering the HQ's bounded rationality and the agency cost for the subsidiaries (Collings et al., 2019; Mellahi & Collings, 2010) when contemplating GTM, which could lead to friction between the HQ and its foreign subsidiary. We also found that this friction can influence the positive relationship between HQ‐led GTM and the resilience of the MNE subsidiary. Thus, the strategic alignment of the HQ and the fit of the GTM program to the subsidiary context (labor market, culture, etc.) are conditions to consider, especially to get prepared to encounter the challenges of a disrupted environment. Furthermore, this dark side of HQ–subsidiary friction related to GTM points to the mutual influence of the HQ strategic orientation, the HQ–subsidiary relationship, and GTM. The connection between these factors constitutes multiple challenges internationally that evolve closely together. This dark side of GTM influences the expatriation process, the key decision process in the subsidiaries, the leadership of talents and their development internationally, and, ultimately, the subsidiaries' trust toward their HQ (Cappelli, 2008; Mellahi & Collings, 2010).

Fourth, this study provides insights to researchers who are conceptually interested in business group‐affiliated MNEs in terms of their unique ways of GTM, and of fostering their subsidiaries' resilience during crises. Business groups now exist widely in both developed and emerging economies, some of which have internationalized successfully; for example, Huawei, Tata, Samsung, and LG. Effective GTM of these business group‐affiliated MNEs can be one of the important factors in understanding how they might have built their global empires rapidly in recent years (Chung et al., 2020; Son et al., 2020; Vance et al., 2013). It could also inspire other Asian MNEs successful in the Asian context to develop further their GTM program to succeed internationally. Moreover, due to the size and complexity of business groups, the structure and governance of MNEs and subsidiaries offer unique opportunities for conceptualization that smaller MNEs may not present. Hence, this study contributes to the GTM literature in the context of business groups that operate internationally.

Managerial implications

This research provides managerial implications for HR professionals and managers working within MNEs. First, HR professionals and managers should reconsider the role of an effective GTM system in helping their subsidiaries to survive and be more resilient to face crises such as the Covid‐19 pandemic. Second, they need to adopt the appropriate TM policies and practices for global succession planning and GTD in all their units, since a multilayered governance structure is crucial for the overall GTM led by the HQ. They have to consider the importance of assigning high potentials with the requisite skills to the right place in all their units, with the optimal balance among HQ, RHQ, and subsidiaries. Thus, companies' focus on GTM strategy alignment to the host countries' units is very important. To set up an effective GTM program, managers should integrate into their policies the subsidiaries' labor market features, the local economic situation, the local culture, the local infrastructure, and so on. Moreover, to develop international high potentials, HR professionals need to set up a training program of global competencies for their talents.

Additionally, in the case of MNEs from emerging economies or business group‐affiliated MNEs, HR professionals and managers should maintain the HQ–RHQ relationship efficiently and encourage support from HQ and RHQ for subsidiary GTM, to avoid unnecessary friction between HQ and subsidiaries. Specifically, these HR professionals and managers need to nurture the relationship between the three‐layered structure of MNEs and ensure that global talents fit within all the units, to alleviate any friction between the HQ and its subsidiaries. Finally, these HR professionals and managers need to develop a good understanding of both the positive and negative sides of GTM to overcome the “double‐edged effect” of GTM in practice in these MNEs (c.f., Son et al., 2020).

Limitations and future research directions

Despite several theoretical contributions and practical implications, this study has several limitations. First, this study's sample is based on MNEs from a single‐parent country, Korea, which limits the generalizability of our findings. In particular, given that Korea is an advanced emerging economy with a relatively strong hierarchical culture, it remains to be tested whether our model and results are still valid for MNEs headquartered in advanced economies or less developed countries. Moreover, a comparison with other group‐affiliated MNEs from other countries would be beneficial in highlighting the similarities and differences across national settings.

Second, our sample includes only business group‐affiliated MNEs, the factors related to the corporate governance and HRM of MNEs from other developed economies can be different from those of our sample MNEs, and thus the strategic orientation of their GTM may be very different. In this regard, future research in the area of GTM and organizational resilience should explore these likely variations to deepen the understanding of MNE subsidiaries' resilience during global crises and external shocks.

Third, the dependent variable of this study—the resilience of foreign subsidiaries—was measured through self‐report of respondents rather than objective data, and thereby may lack reliability. While it would have been desirable to test the robustness of our results by using the dependent variable based on objective data, the data on the financial performance of subsidiaries of Korean MNEs are extremely hard to access due to Korean laws and regulations stipulating the confidentiality of such data. Hence, we had no choice but to use the self‐report measure of organizational resilience based on the existing literature (Torres et al., 2019). To the best of our knowledge, this confidentiality of the financial performance of MNE foreign subsidiaries applies to not only Korean ones but also those in developed countries that have the same or similar practice with Korea. In addition, we had no choice but to use the self‐report by respondents as the measure of other explanatory variables because of the data access limitation. Hence, future research is needed to test whether using the alternative measures of resilience and GTM based on objective data produces consistent results.

Fourth, we need to discern whether our somewhat different operationalization of GTM is appropriate when we operationalize independent and moderating variables. In particular, there is some difference between HQ‐led GTM versus RHQ GTD and HQ–subsidiary GTM friction. Yet, although all HQ, RHQ, and subsidiaries can be involved in GTM, a global succession planning system is usually orchestrated and masterminded by HQ rather than RHQ or subsidiaries in particular in the Korean MNE case. Therefore, to reflect the status quo, we opted to operationalize a formal management development program for RHQ and subsidiaries. As such, both a global succession planning system and a formal management development program are included in the HQ‐related relationship so that we can use the term HQ‐led GTM and HQ–subsidiary GTM friction. We suggest that future research adopt a qualitative method to scrutinize whether and how both a global succession planning system and formal management development program can be linked to the sides of RHQ and subsidiaries to build up the consistency of operationalization.

In addition, our measurement of RHQ‐assisted GTD is more fine‐grained than that of McDonnell et al.'s (2010) original measurement, because McDonnell et al. (2010, p. 155) used only three categories for this variable, i.e., “those that use all five mechanisms,” “those using three to four mechanisms,” and “those using two or less” rather than our six categories for this variable in the present study. We had to adjust McDonnell et al.'s original measurement to ours because the one by McDonnell et al. (2010) cannot capture the detailed differences of the five mechanisms of GTD. Also, conceptually, based on McDonnell et al. (2010) and Mellahi and Collings (2010), we had to adjust McDonnell et al.'s (2010) original measurement of GTM to our measurement because there has been no previous study that empirically measured this variable. Yet, we believe that our measurement fits the context of HQ–subsidiary GTM friction because the five mechanisms in McDonnell et al.'s (2010) measurement, which were used by us, are the most important ones in an MNE's GTM. Therefore, our fine‐grained and novel measurement for RHQ‐assisted GTD and HQ–subsidiary GTM friction opens up future research avenues in GTM.

Finally, since local situations differ significantly in managing the Covid‐19 crisis, survey studies cannot capture in‐depth what is happening, why, and what the impact may be on the ground across nations. Future studies should use qualitative methods to develop a deeper and local understanding of how MNE subsidiaries are impacted by the crisis and what specific challenges they face in GTM to develop organizational resilience. In this regard, the following questions could be raised: how are MNEs responding to the changes in the international labor market and labor restrictions related to the Covid‐19? What GTM practices and international strategies have been adopted by the MNEs to respond to the local situations? How can future research integrate GTM research as part of strategic HRM more broadly?

More research questions could also arise following the Covid‐19 crisis. For example, are there any evolvement of HQ–subsidiary relationship and HQ–RHQ relationship before and after the Covid‐19 crisis? More specifically, will there be less friction across these units in terms of GTM? How would these relationships evolve and influence the GTM practices of the MNEs internationally? Examining the HQ consciousness level in terms of RHQ‐GTM integration and knowledge of the GTM transfer process would add value to the IHRM literature. These insights can be best developed through qualitative studies within the RHQ and subsidiaries, by analyzing the interactions between the stakeholders in different MNE units, their perception of their role evolvement following the Covid‐19 crisis, and their needs and expectations in terms of GTM.