Headquarters parenting advantage in Chinese MNEs: The moderating role of top managers ’ political and International experience

This study explores the relevance of top management teams ’ experience to support the headquarters parenting advantage in the context of Chinese multinationals. Specifically, it studies how the political and international experience of headquarters ’ top management teams moderates the relationship between headquarters involvement in knowledge transfer processes – a key aspect of value creation in the parenting advantage logic – and the extent of reverse knowledge transfer from subsidiaries. Based on the data from two complementary surveys of senior managers in 99 Chinese multinationals and managers in their 177 subsidiaries, our analysis indicates a contrasting effect of top managers ’ experience as their political experience weakens, but their international experience strengthens the positive effect of headquarters involvement in reverse knowledge transfer. This study contributes to the parenting advantage logic, by introducing the relevance of different top managers ’ experiences, and to our understanding of top management teams in the context of both reverse knowledge transfer and Chinese multinationals, particularly by showing the important implications of top management teams ’ experience for Chinese enterprises ’ international strategies.

Inspired by Chandler's (1991) view, the notion of parenting advantage suggests that HQs play an essential role in MNEs' competitive advantage and success (e.g., Foss, 1997;Goold et al., 1998).This is because HQs have the responsibility for formulating and judging MNEs' overall strategic orientation and for selecting core investments and organizational mechanisms to obtain a higher value than the simple sum of each MNE unit, and they also have the ability to fulfill these responsibilities (Collis & Montgomery, 1998).The advocates of parenting advantage acknowledge that HQs create or add value through such activities as synergy management and knowledge transfer.However, this rather optimistic view is challenged by recent studies which focus on specific value-creating activities (e.g., HQ involvement in subsidiary innovation development/transfer) (Ciabuschi et al., 2011;Ciabuschi, Forsgren et al., 2017;Forsgren & Holm, 2010;Lind & Kang, 2017).These studies take a relatively more pessimistic view of HQs' role and criticize their ability to add value within MNEs because HQs's knowledge of subsidiary activities is often constrained, thereby hindering fulfillment of their parental role (Ciabuschi et al., 2011;Forsgren & Holm, 2010).Thus, the ongoing discussion in the literature does not present a consensus on how HQs are able to fulfill their parenting advantage.Moreover, the research on HQ parenting advantage has focused mostly on Western MNEs and little is known about this in the context of emerging markets' MNEs.The emerging-markets MNEs context is unique as HQs typically face a double hurdle of liabilities, i.e., a liability of emergingness and a liability of newness (Ramamurti, 2012;Wilkinson, Wood, & Demirbag, 2014).Hence, they may encounter more difficulties, compared with Western MNEs, in fulfilling the parenting role.
In this paper we choose to focus particularly on Chinese MNEs, which are clearly attracting growing research interest in accordance with the sharp rise in the number of Chinese MNEs operating worldwide and the importance of China as (the largest) emerging market.Specifically, we lack an understanding of the extent to which, and under what conditions, Chinese MNE HQs are actually able to create value when managing their international operations.Against this background, this paper aims to extend the literature on parenting advantage in the context of Chinese MNEs by assessing the effect of HQ involvement on reverse knowledge transfer from subsidiaries and by looking at how this process is moderated by two important contingency factors: the political experience and the international experience of the HQ's top management team (TMT).In other words, in this paper, we study HQs' ability to fulfill a parenting advantage from a TMT perspective in the context of Chinese MNEs.
So, on one side, we need to consider the fact that MNEs are worldwide learning organizations with a network of geographically dispersed subsidiaries that possess unique, locally created knowledge (Håkanson & Nobel, 2001;Nohria & Ghoshal, 1997).This applies particularly to Chinese MNEs aspartly driven by an innovation-led growth model of the Chinese governmenttheir strategic asset-seeking internationalization (e.g., the acquisition of foreign firms and the establishment of foreign R&D centers) is a deliberate initiative, substantially adopted by Chinese firms for innovation catch-up through accessing local advanced knowledge (e.g., Anderson, Sutherland, & Severe, 2015;Deng, 2009;Elia & Santangelo, 2017;Luo & Tung, 2007, 2018;Mathews, 2006).On the other side, we argue an HQ's value-creating activities and parenting advantage depend on its knowledge situation which is determined by its TMT's knowledge (Ciabuschi et al., 2011;Forsgren & Holm, 2010), which is influenced, in turn, by the TMT's experiences and backgrounds.
A TMT is defined as a group of top-ranking executives in an enterprise who provide an interface between the enterprise and its environment, and who are relatively powerful in critical organizational decisions; therefore their choices and actions are likely to have effects on the enterprise (Hambrick, Finkelstein, & Mooney, 2005).Scholars whose ideas are rooted in upper echelons theory (Hambrick & Mason, 1984) have striven to confirm that TMT characteristics can be used to predict the success of the strategy choices adopted (Díaz-Fernández, González-Rodríguez, & Simonetti, 2019) and to shape corporate value-adding (Grillo, Ferreira, Marques, & Ferreira, 2018).As Hutzschenreuter and Horstkotte (2013) highlight, TMT characteristicsparticularly prior experiencecreate value when they can be applied to a specific task.Thus, HQs' TMT characteristics seem to be an influential factor in shaping their parenting advantage.Yet the potential value of heterogeneity in TMT composition has so far been neglected in the literature on parenting advantage, constituting a critical omission in our understanding of HQ parenting advantage.Owing to the high-power distance and uncertainty avoidance dominating Chinese firms' decision-making process (Carpenter & Fredrickson, 2001;Hofstede, 1980), the influence and power of TMTs in corporate strategies and behaviors are likely to be stronger in China than in other contexts (Li, 2017).Given the difference in composition and value between the TMTs of Chinese MNEs and those of advanced-market MNEs (Li, 2017), such a context may be particularly well-suited for observing the effects of TMTs' specific experiences (Cui, Li, Meyer, & Li, 2015;Hutzschenreuter & Horstkotte, 2013).
In this study we focus on HQs' TMTs' political and international experience, two critical managerial resources in Chinese MNEs.Specifically, TMT political experience in Chinese companies has been acknowledged in the literature as a key influencing factor in Chinese firms' performance and international growth (e.g., Wu & Cheng, 2011;Wei, Clegg, & Ma, 2015), while TMT international experience has long since been an established key factor in international business literature (e.g., Sambharya, 1996), central to the management of multinationals, but not yet properly investigated in the context of Chinese MNEs.Moreover, it has been argued that TMTs' experience and knowledge base is a crucial component of HQ parenting advantage logic (e.g., Ciabuschi et al., 2011, Ciabuschi, Forsgren et al., 2017), driving value creation, but still with very limited empirical evidence, particularly in relation to reverse knowledge transfer.
On the one hand, top managers with political experience, referring to top managers currently or previously serving as government bureaucrats in the political system, are prevalent in China (Fan, Wong, & Zhang, 2007;Francis, Hasan, & Sun, 2009;Lu, Liu, Filatotchev, & Wright, 2014), as, in China, state capitalism has a strong presence and there are insufficient formal regulatory frameworks (Peng & Luo, 2000;Xin & Pearce, 1996).This political experience cultivates top managers' political knowledge and political connections, which have a profound effect on Chinese firms' value-adding activities.Firm-level political connections have been well investigated in the research on Chinese MNEs' internationalization, where they have been found to be conducive to firms' international expansion (e.g., Duanmu, 2014;Luo, Xue, & Han, 2010;Wei et al., 2015), but, conversely, TMTs' political experience is under-explored in research on Chinese MNEs' management.
On the other hand, China has recently experienced a substantial inflow of returnee executives with foreign education and/or work experience (Cui et al., 2015;Filatotchev, Liu, Buck, & Wright, 2009;Lu et al., 2014), leading to a growing number of top managers with international experience in Chinese MNEs.The international experience of a firm's TMT is regarded as one of the prerequisites for that firm becoming truly multinational (Sambharya, 1996).This is because internationally experienced top managers can be better at dealing with diverse cultural and institutional environments and at making strategic decisions that result in superior performance (Carpenter, Sanders, & Gregersen, 2001;Heijltjes, Olie, & Glunk, 2003;Nielsen & Nielsen, 2011;Tan & Meyer, 2010).The political and international experience of TMTs is extensively viewed as a strategic resource that enables value creation (Carpenter et al., 2001;Cui et al., 2015;Luo et al., 2010;Schmid & Dauth, 2014;Xin & Pearce, 1996).Owing to a lack of specific research in this area, it is so far unclear if and how these two specific types of TMTs' experiences can support Chinese MNEs' HQ parenting advantage in cross-border knowledge acquisition activities.
The empirical analysis of this study is based on data from two complementary surveys of senior managers in 99 Chinese multinationals and managers in 177 of their subsidiaries (located in 34 advanced and developing markets).To capture HQ involvement in reverse knowledge transfer practices, our sampling MNEs are required to be Chinese MNEs with at least one subsidiary operating in an advanced market, ensuring that every sampling MNE is able to provide at least one advancedmarket subsidiary for investigation.This accords with the notion that strategic asset-seeking investments undertaken by Chinese MNEs mostly target advanced markets where there is a comparative advantage in high-tech R&D (Elia & Santangelo, 2017;Luo & Tung, 2007).However, the research on R&D internationalization also reveals an increasing trend in R&D investments in developing markets (Awate, Larsen, & Mudambi, 2015;Gammeltoft, 2006) and, consistently, the literature on reverse knowledge transfer based on Western MNEs always covers analysis of subsidiaries in developing markets (e.g., Håkanson & Nobel, 2001;Rabbiosi, 2011;Yang, Mudambi, & Meyer, 2008).This means that subsidiaries located in developing markets may also have product or production competence and may engage in reverse knowledge transfer (e.g., cases of cost innovation).Following this logic, our sample includes subsidiaries in both advanced and developing markets.
This study makes several theoretical contributions.First, it contributes to the literature on HQs' role, particularly in the context of Chinese MNEs, a context that has so far been largely neglected in the literature.Specifically, our findings suggest that HQs' direct involvement enhances the extent of reverse knowledge transfer from foreign subsidiaries, verifying the potential for HQ parenting advantage in Chinese MNEs.Second, we contribute specifically to parenting advantage logic, by showing the relevance of different top managers' experiences, and to our understanding of TMTs in the context of both reverse knowledge transfer and Chinese multinationals, particularly by showing the important implications of TMTs' experience for Chinese companies' international C. Su et al. strategies.In doing so, the study builds theoretical links between TMT experience and HQ parenting advantage and confirms that TMT experience is pivotal in explaining and determining HQ parenting advantage.Third, this study takes an important step forward by identifying new factors to explain reverse knowledge transfer in Chinese MNEs when linking HQs' TMT experience and reverse knowledge transfer.Specifically, our findings imply that, in the logic of HQ parenting advantage, TMTs' international experience can mitigate the negative effect of political experience.These findings are therefore important in pushing forward the current research on reverse knowledge transfer and the international strategies of Chinese MNEs.

Theoretical background
This section departs from the current discussion of HQ parenting advantage based on the HQ knowledge situation and introduces the link between TMT experience and parenting advantage.This theoretical framing will then be discussed in the context of Chinese MNEs to highlight the relevance of TMT experience for Chinese MNEs and the unique nature of parenting advantage in this context.

HQ parenting advantage
Corporate HQ is defined by Collis, Young, and Goold (2007: 385) as "staff functions and executive management with responsibility for, or providing services to, the whole (or most of) the company, excluding staff employed in divisional headquarters".An HQ is perceived as an entity with ultimate decision rights within a corporate body (Chandler, 1991) and as those "functions governing the development, allocation, and deployment of valuable corporate resources within the hierarchy" (Collis et al., 2007: 388) that are supposed to lead to the creation of a corporate advantage, i.e., "parenting advantage" (Collis & Montgomery, 1998;Foss, 1997;Goold et al., 1998).In other words, as Markides (2002) emphasizes, HQs should add value to business units; otherwise, there is no economic rationale for HQs' existence.
Parenting advantage suggests that HQs should not only formulate a successful overall strategy, but should also be the best possible owner of each business unit in the corporate portfolio (Goold et al., 1998).HQs have four direct ways to create or add value for business units: stand-alone influence, linkage influence, central functions and services and corporate development activities (Kruehler, Pidun, & Rubner, 2012).Clearly, HQs have a potential role of creating or adding value when involving themselves in subsidiary activities such as innovation development and transfer (Goold & Campbell, 2002;Poppo, 2003).One of HQs' parenting functions is to determine intra-organizational knowledge sharing to establish and support to create synergistic effects (Ciabuschi et al., 2011;Ciabuschi, Forsgren et al., 2017;Ferreira, Ratten, & Dana, 2017).It follows that HQs are able to identify and transfer subsidiary knowledge that is strategically important within the whole MNE; benefits can thus be spread beyond the single subsidiary to the whole MNE.That is, good parenting should enhance the outcome of these intra-firm knowledge transfers and maintain MNEs' innovation capability.Apparently, the parenting advantage logic fits well with those studies focusing on innovation strategy and on cross-border knowledge transfer which maintains the competitive advantage of MNEs (Ciabuschi, Forsgren et al., 2017;Ferreira, Fernandes, Alves, & Raposo, 2015, 2017).
The literature reveals the contrasting perspectives of bounded rationality and sheer ignorance regarding parenting advantage.The traditional perspective, taking an optimistic view, assumes that HQs are boundedly rational and have a reasonable possibility of strategically directing the value-creation activities (e.g., contributing expertise to subsidiaries) of MNEs and of making rational decisions (e.g., staffing, resource allocation) when managing and controlling subsidiaries (e.g., Buckley & Hashai, 2009;Egelhoff, 2010;Nohria & Ghoshal, 1997).The authority of an HQ, in combination with its potential to structure an MNE and the information systems in an efficient way, can not only make it knowledgeable about every subsidiary of the MNE, it can also enable it, when appropriate, to intervene in subsidiary operations.It will, at least, be able to assess, and have a sufficient understanding of, what knowledge or information it does not have and it will be able to design structures, rules and control systems and to carry out its own role optimally in accordance with these.This bounded rationality perspective is in line with the knowledge-based view which assumes that an HQ, to some extent, has the capability to understand and support value-creating processes within the firm (Hymer, 1970).Several studies have confirmed that MNE HQs act as a resource allocator involved in subsidiary innovation activities (Dellestrand & Kappen, 2012), as the network orchestrator and as the integrator of intra-organizational knowledge flows within MNEs (Awate et al., 2015;Mudambi & Navarra, 2004).Therefore, HQs are assumed to have the ability to effectively design, and involve itself in, subsidiary innovation transfer, which augments MNEs' innovativeness, even though the environment may be uncertain and complex.
Although the bounded rationality perspective has dominated in the literature on MNEs, a contrasting perspective, called "sheer ignorance", has recently emerged and holds a less optimistic view of parenting advantage.The core message of the "sheer ignorance" perspective is that HQs suffer from genuine uncertainty and limited knowledge, leading to severe limitations to value creation within MNEs (Ciabuschi et al., 2011;Ciabuschi, Forsgren et al., 2017;Vahlne, Schweizer, & Johanson, 2012) as knowledge is more socially embedded and therefore not controlled by anyone in its entirety.This perspective is rooted in the Austrian school's view which regards MNEs as distributed knowledge systems containing several interrelated sub-themes in the strongest sense, with decentralized systems that lack an overseeing mind (Kristensen & Zeitlin, 2001;Mintzberg, 1990;Tsoukas, 1996).This idea is more in line with the strategic concept of differentiated-network MNEs (Nohria & Ghoshal, 1997).This concept asserts that a subsidiary is a semi-autonomous actor with its own local network embeddedness, capable of making its own strategic choices, which makes the MNE extremely "distributed" when it comes to knowledge and control.In this sense, HQ is always an outsider in terms of the specific network context in which subsidiaries are embedded, with a limited possibility of understanding subsidiary operational activities, which makes its parental role much more difficult to carry out (Ciabuschi et al., 2011;Forsgren & Holm, 2010;Vahlne et al., 2012).Under this more pessimistic perspective, HQ basically does not know what it does not know, and is consequently unable to make a rational decision regarding the extent of its possible interventions in subsidiary activities.For instance, HQ may be unable to assess who should be involved in a subsidiary knowledge transfer project, what kind of resources are relevant and, furthermore, how they can best support the transfer process.This is highlighted by the empirical study of Ciabuschi, Forsgren et al. (2017), in which HQ involvement hampers subsidiary innovation development and transfer, supporting a situation of sheer ignorance.In short, according to this perspective, HQ is likely to act as an absentee landlord "who is not only ignorant, but who destroys rather than creates value" (Barner-Rasmussen, Piekkari, Scott-Kennel, & Welch, 2010: 99), and who "grope(s) in the darkness" when involving itself in subsidiary activities and designing its own role in these processes (Ciabuschi et al., 2011: 962).
It is clear from the aforementioned discussion that the discrepancy regarding HQ parenting advantage largely mirrors the different views of what can be called the HQ knowledge situation (Ciabuschi et al., 2011;Forsgren & Holm, 2010;Lind & Kang, 2017).The core of parenting advantage logic is based on HQs having sufficient knowledge to make rational decisions regarding value-adding activities.Hence, the ability of HQs to play a positive role in cross-border value-adding activities is highly dependent on the quantity and quality of relevant knowledge.
C. Su et al.

The relevance of TMT experience with parenting advantage
An HQ's knowledge situation is decided by its TMT's knowledge, which shapes parenting advantage (Ciabuschi et al., 2011;Forsgren & Holm, 2010).As Conner and Prahalad (1996: 480) state, a firm is created by the fact that "the managers' understanding (present and future) is believed to be of superior value compared with corresponding elements of the employees".The behavioral theory of firms (Cyert & March, 1963) recognizes that top managers, to some extent, have bounded rationality in making decisions.However, the upper echelons theory (Hambrick & Mason, 1984) argues that, in situations of complex and uncertain information, managerial decisions are not always rational, but are rather influenced by top managers' limited knowledge and by their biases.Indeed, top managers may suffer not only from incomplete knowledge but also from the situation that they do not know what they do not know (Tsoukas, 1996), resulting in the sheer ignorance problem, because knowledge is context-specific and action-oriented (Giddens, 1984;Weick & Roberts, 1993).As experiential learning drives knowledge accumulation (Johanson & Vahlne, 1977), top managers' experiences throughout their lives shape their perception, cognitive value, biases and preferences, which inform a TMT's knowledge and skills (Carpenter et al., 2001;Sambharya, 1996).The variations in different firms' knowledge depends on different TMTs' experiences.TMTs' experience seems to be a potential factor explaining parenting advantage and one which is rarely studied.Most studies (e.g., Bouquet & Birkinshaw, 2008;Ciabuschi, Forsgren et al., 2017;Decreton et al., 2017;Nell & Ambos, 2013) employ bottom-up processes to explore HQ's role by incorporating a subsidiary perspective (e.g., subsidiary local embeddedness, subsidiary competence and contribution) rather than a top-down administrative perspective.This leads to us having a limited understanding of HQ parenting advantage.

Parenting advantage in Chinese multinationals
MNEs differ in terms of their organizational structure, strategy and resources, which affects the type, frequency and locus of activities conducted by HQs to create better value (Ciabuschi et al., 2011).Previous literature on the HQ's role, primarily focusing on advanced-economy MNEs, assumes that HQ is a purely economic actor pursuing profit maximization.Unlike advanced economies with more market orientation and more well-developed institutions, China has a mixture of state capitalism, persistent government power and underdeveloped institutions (Peng & Luo, 2000;Xin & Pearce, 1996).This results in a prevalence of politician managers in Chinese MNE HQs (Fan et al., 2007;Francis et al., 2009;Lu et al., 2014), which leads to HQs pursuing both economic and political interests instead of profit maximization (Fan et al., 2007;Meyer, Ding, Li, & Zhang, 2014).The underlying reasoning is that executives strongly connected with government tend to conform to government expectations in an acquiescent manner in order to attain institutional legitimacy and government support (Peng & Luo, 2000;Xin & Pearce, 1996) and to secure their power and careers in the political arena (Arocena & Oliveros, 2012).Moreover, compared with MNEs in advanced economies, Chinese MNEs, as "latecomers" in global markets, lack accumulated internationalization experience and have little international knowledge (Lu et al., 2014;Ramamurti, 2012;Sahasranamam, Rentala, & Rose, 2019).The prevalence of HQs' top managers with political experience and HQs' latecomer identity may complicate the parenting role of Chinese HQs, which may bring many challenges in terms of the process of cross-border value-adding activities.
For instance, as "latecomers" to internationalization, Chinese MNEs typically lack firm-specific advantages (e.g., advanced technology) that could be deployed in international operations (Luo & Tung, 2007;Sahasranamam et al., 2019) and hence commonly undertake strategic asset-seeking foreign investments as a channel to augment their innovative assets and to catch up with their Western counterparts (Deng, 2009;Elia & Santangelo, 2017;Luo & Tung, 2007).Achieving such catch-up strategies needs Chinese HQs to identify, acquire and integrate advanced knowledge from overseas subsidiaries, which in turn requires foreign-market knowledge, international management know-how and entrepreneurial leadership that Chinese HQs generally lack (Cui et al., 2015;Luo & Tung, 2018).Achieving these catch-up strategies therefore carries a high potential risk.Whether Chinese MNE HQs have a parenting advantage in international activities (e.g., reverse knowledge transfer) or not becomes a concern and this has not attracted scholars' attention.Because of the unique conditions that Chinese MNE HQs confront, employing the traditional view to evaluate and analyze HQ parenting advantage in Chinese MNEs may not be appropriate.
TMTs in advanced-economy firms benefit from greater diversity in terms of gender, nationality and ethnicity compared with Chinese firms (Heijltjes et al., 2003).The decision-making process in Chinese firms is more likely to exhibit characteristics of high-power distance and uncertainty avoidance (Carpenter & Fredrickson, 2001;Hofstede, 1980).Hence, the influence and power of TMTs are more likely to be stronger and more apparent in Chinese firms (Li, 2017).Further, Chinese MNEs, as latecomers, typically have limited managerial and human resources and therefore their TMTs' knowledge and experience become essential for international value-adding activities such as reverse knowledge transfer (Tan & Meyer, 2010).Despite TMTs' importance in Chinese MNEs, the role of TMTs remains underdeveloped in the Chinese MNE literature (Cui et al., 2015;Li, 2017;Lu et al., 2014).

Hypotheses development
A necessary strategy in the value-adding role of corporate parenting is an actual hands-on strategy (Goold & Campbell, 2002;Goold et al., 1998) through the more active involvement of HQ in subsidiary-level activities (Ciabuschi, Forsgren et al., 2017;Egelhoff, 2010;Lind & Kang, 2017), particularly when competition has intensified and uncertainty has become more prevalent (Campbell et al., 1994;Chandler, 1991).Acknowledging the bounded rationality perspective that argues an HQ has a superior ability in terms of selective intervention based on a realistic understanding of its own expertise (or lack thereof) (Conner & Prahalad, 1996;Egelhoff, 2010;Foss, 1997), studies dealing with parenting theory (e.g., Goold & Campbell, 2002;Goold et al., 1998) indicate that there is a high likelihood of HQ involvement in subsidiary-level activities and that such involvement brings a positive influence.However, the empirical studies by Ciabuschi, Forsgren et al. (2017) and Lind and Kang (2017) suggest the possibility of HQ involvement having detrimental effects on knowledge transfer in Western MNEs.The reasoning behind this is that these HQs suffer from genuine uncertainty and limited knowledge when involving themselves in subsidiary activities owing to subsidiary local embeddedness (Ciabuschi et al., 2011;Forsgren & Holm, 2010;Vahlne et al., 2012).Thus, we do not yet have full understanding of what may (or may not) support the parenting advantage potential of headquarters.This is true for Western MNEs and even more so for Chinese MNEs where research is still scant.
However, HQ involvement can vary in degree and may be more or less intertwined with different ongoing subsidiary activities and projects.HQ can be involved in subsidiary knowledge transfer in different ways, for instance: by taking the initiative to begin a knowledge transfer project from a subsidiary; by supporting the subsidiary knowledge transfer process through providing specific resources (e.g., funding, talents) and expertise; by intense communication with the actors involved in the process; by closely guiding and supervising different parts of the process.For Chinese MNEs, one primary motive of outward foreign investment is to seek and obtain advanced knowledge overseas to achieve innovation catch-up (Anderson et al., 2015;Deng, 2009;Elia & Santangelo, 2017;Luo & Tung, 2007, 2018;Sahasranamam et al., 2019).Thus, Chinese MNE HQs tend to get substantially involved in subsidiary knowledge transfer practices as this fulfills their international C. Su et al. strategy.This is an important difference compared with Western MNEs which typically already possess core assets in the home organization.
According to parenting advantage logic, HQ has an accurate and comprehensive overview of subsidiaries' operations and therefore a reasonable capability to get involved in subsidiary knowledge transfer.First, through HQ involvement, HQ is in a position, by being an insider, to allocate the right resources and to provide suitable guidance and expertise at the right time in the knowledge transfer process, ensuring the implementation of reverse knowledge transfer, despite its lack of knowledge and understanding of the transfer context (e.g., Dellestrand & Kappen, 2012;Lind & Kang, 2017).For example, HQ can evaluate and identify the most suitable units within HQ to receive and absorb the knowledge from the sending subsidiary.HQ's resource allocation, guidance and expertise can also support both sender and receiver in overcoming the potential difficulties related to knowledge transfer.Second, HQ has the formal authority to persuade and instruct both senders and receivers to participate in, and make efforts to achieve the knowledge transfer process in an effective manner, thereby increasing the success of reverse knowledge transfer.For example, HQ can design incentive and evaluation systems to obtain desired behaviors (e.g., subsidiary willingness and commitment to transferring knowledge) at the subsidiary level (e.g., Gupta & Govindarajan, 2000;O'Donnell, 2000).This can somehow alleviate or overcome the problem of subsidiaries' unwillingness to share knowledge, particularly from more advanced host countries (Ciabuschi, Kong, & Su, 2017).HQ involvement also has a signaling function.Specifically, it can confer legitimacy and importance on knowledge transfer within the MNE, attracting more attention from other units and employees and encouraging them to become engaged in, and to support, the implementation of knowledge transfer (Birkinshaw, Hood, & Jonsson, 1998).In particular, this is applied in Chinese firms which are commonly embedded in high-power distance and in the hierarchy of authority (Hofstede, 1980).On the basis of all the above reasoning, we argue that the Chinese HQ will add value to the transfer process through its involvement and will consequently improve the extent of reverse knowledge transfer.Thus, the following hypothesis is formulated: Hypothesis 1.In Chinese MNEs, HQ involvement in knowledge transfer positively affects the extent of reverse knowledge transfer.
Many firms in China are led by politically experienced TMTs (Fan et al., 2007).Governments routinely appoint their own officials as top managers of Chinese state-owned firms in order to exercise control and implement government policies (Li & Xia, 2008;Nee & Opper, 2010).Meanwhile, other Chinese firms are interested in appointing politicians to join TMTs (Agrawal & Knoeber, 2001;Fan et al., 2007;Boubakri, Cosset, & Saffar, 2012;Wu & Cheng, 2011) because their political knowledge and ties can help to obtain government resources and preferential treatment in China, which are valuable to firms (Li & Atuahene-Gima, 2001;Peng & Luo, 2000).Apparently, top managers' political experience is regarded as a unique type of managerial resource for adding value in Chinese firms (Li & Atuahene-Gima, 2001;Xin & Pearce, 1996).However, such political experience is not always beneficial for firms, as it can be a double-edged sword (Wu & Cheng, 2011).Many empirical studies, particularly in finance research, have confirmed that firms with politically experienced top managers have worse business efficiency and performance (e.g., Fan et al., 2007;Fisman, 2001;Li, Liu, & Ren, 2007).
Experience produces knowledge; hence, knowledge is contextspecific and action-oriented (Giddens, 1984;Weick & Roberts, 1993).This implies that political experience enables Chinese top managers to have sufficient knowledge of the political scene (e.g., a government subsidy system) and to have insight in terms of predicting government actions (Agrawal & Knoeber, 2001).But it easily leads these Chinese politician managers to have low levels of professionalism in corporate management and a lack of business intelligence (Boubakri et al., 2012;Fan et al., 2007).Once knowledge is constructed inside a person/team, it becomes entrenched over time and is often retained for interpreting future issues (Weick, 1979).In this sense, the Chinese politician managers have certain business experience from their current jobs and such managers, particularly those in state-owned firms, focus on political connections as a source of their advantage and are more concerned with their appointment, dismissal and future promotion (Arocena & Oliveros, 2012;Li et al., 2007), motivating them to pursue political learning and to undermine the learning and expertise development associated with their managerial jobs.Hence, Chinese top managers with political experience are more likely to have incomplete knowledge.More specifically, they are more likely to lack managerial knowledge and innovation-related knowledge, which are important to value-adding activities, although they may not know what they lack (Tsoukas, 1996).Also, the knowledge associated with top managers' political experience is locally bound and embedded in the institutional context of the home country.Therefore, it may not be transferable to an international context (Lu et al., 2014).At HQ level, Chinese HQs with many politically experienced TMT members are keen to rely on the advantages of their political connections in order to compete and are less motivated to take advantage of innovation resources as efficiently as they should do under market competition (Li & Xia, 2008;Li et al., 2007;Lu et al., 2014).This constrains HQs' learning and capability development associated with domestic business and leads to less market-based and innovation-based knowledge (Lu et al., 2014).
This deficit in the commercial, innovation and international knowledge of politically experienced Chinese TMTs and their HQs becomes a principal obstacle to the formulation of rational strategies and decisions that will manage subsidiaries effectively and lead to subsidiary knowledge transfer.Politically experienced Chinese TMTs may even be unaware of what kind of knowledge they lack, resulting in the possibility of a sheer-ignorance situation as they are always over-confident because of their powerful authority and high-power distance dominating in Chinese firms' decision-making process (Carpenter & Fredrickson, 2001;Hofstede, 1980).Facing more complexity and uncertainty resulting from a lack of knowledge, Chinese HQs with many politician managers will not know how to deal with subsidiaries, or how to get involved in their activities, consequently leading to light-touch integration with subsidiaries (Liu & Woywode, 2013;Su & Kong, 2020), or even to subsidiaries being out of control (Ciabuschi, Kong et al., 2017).This happens particularly in those subsidiaries with a strategic asset-seeking purpose as they are more powerful because of the fact that knowledge is power (Mudambi & Navarra, 2004).In essence, these Chinese HQs are basically outsiders vis-a`-vis the operational context of each subsidiary, leading to a lack of knowledge about subsidiary context and about the knowledge controlled by the subsidiary.However, HQs' ability to add value to local subsidiaries' knowledge transfer processes requires both knowledge and understanding of subsidiaries' local contexts (Ciabuschi et al., 2011;Forsgren & Holm, 2010).Hence, Chinese HQs with a politically experienced TMT are more likely to have a limited ability to make rational decisions (e.g., determining what knowledge to be transferred, or assigning each unit to their role in the knowledge transfer process) and to be less able to provide useful and reliable knowledge and support during their involvement in subsidiary knowledge transfer.This probably complicates the whole reverse knowledge transfer process.Because of a demand for the exercise of authority (Ciabuschi et al., 2011) and hierarchical and bureaucracy culture that dominates in China, even when these Chinese politically connected HQs suffer from sheer ignorance, they may tend to interfere substantially in subsidiary knowledge transfer processes.This may even lead to misguided benevolence and to over-commitment, which are harmful and which result in less effective implementation efforts in terms of reverse knowledge transfer (Ciabuschi, Forsgren et al., 2017).Apparently, too many TMT members with political experience are more likely to reduce HQ parenting advantage and to limit the benefits of HQ involvement.Thus, we expect that: Hypothesis 2. In Chinese MNEs, the positive effect of HQ involvement in knowledge transfer on the extent of reverse knowledge transfer is negatively moderated by the HQ TMT's political experience.
International experience has widely been acknowledged as being instrumental in developing top managers' global mindsets (Piaskowska & Trojanowski, 2014) and the knowledge and ability to navigate cross-cultural differences (Athanassiou & Nigh, 2002;Tihanyi et al., 2000), thus playing an essential role in those firms eyeing international involvement.Specifically, international experience through overseas education and/or work is a surrogate for accumulating international knowledge (Sambharya, 1996).TMT members with international experience are more likely to possess deeper knowledge and understanding of foreign environments and are more aware of cultural differences and ambiguities than those without international experience, thereby augmenting their MNEs' existing knowledge base (Athanassiou & Nigh, 2002;Tihanyi et al., 2000).This makes them better equipped to address cross-border management issues.Also, international experience provides TMTs with a valuable network of international contacts that facilitates acquisition and access to useful and diverse information (Athanassiou & Nigh, 2002;Cui et al., 2015;Herrmann & Datta, 2005), thus mitigating the liabilities of foreignness (Johanson & Vahlne, 1977) and enabling inter-firm coordination and cooperation (Daily, Certo, & Dalton, 2000).Further, TMT international experience shapes cognitive orientation and helps to develop a global mindset "that allow executives to think locally and act globally" (Piaskowska & Trojanowski, 2014: 43).This broadens their cognitive horizons and increases their awareness of overseas business environments, thereby enabling them to recognize and capture international business opportunities (Herrmann & Datta, 2005;Tan & Meyer, 2010).All these can contribute to organizational learning and to building a reservoir of expertise and competence, the better to cope with the uncertainties and complexities associated with operating internationally (Dang, Jasovska, & Rammal, 2020) and to overcome the potential detrimental effects of liabilities of foreignness (Hutzschenreuter & Horstkotte, 2013;Sambharya, 1996).
Apparently, highly internationally experienced TMTs are more bounded rationally, having the knowledge and ability to formulate and implement more rational international strategies (Azam et al., 2018) and to deploy organizational resources efficiently to succeed in foreign operations (Carpenter & Fredrickson, 2001;Tihanyi et al., 2000).This is stressed by Azam et al. (2018), who found that TMTs' international experience enhances the rationality and effectiveness of international strategic decision-making in Pakistani IT firms.The benefits of TMTs' international experience are more distinctive in emerging markets whose executives lack knowledge about how to act in a highly sophisticated market and a mature institutional system and internationally experienced managers remain in short supply (Cui et al., 2015;Farndale, Scullion, & Sparrow, 2010).The global mindset emanating from TMTs' international experience helps firms to learn to adapt to an international context and thus develops market-orientation expertise in emerging-market firms (Cui et al., 2015).Hence, emerging markets, particularly China, seek managerial know-how from returnees to overcome domestic shortages of skilled talent, increase local firms' managerial capabilities and help their international expansion (Filatotchev et al., 2009).
HQ TMTs' international experience supplies knowledge and capabilities that allow TMTs to be well-placed in terms of managing, and coordinating with, subsidiaries in Chinese MNEs (Hutzschenreuter & Horstkotte, 2013;Sambharya, 1996), helping them to obtain wide-ranging and detailed knowledge about subsidiaries and their activities (Daily et al., 2000) and to have an accurate overview of subsidiaries' operations.In such cases, an HQ TMT has more scope to shape subsidiary knowledge transfer and is in a better position to determine the appropriate level of involvement and the best approach to take.This is because the knowledge situation of HQ regarding subsidiaries dictates the HQ's actions and involvement (Ciabuschi et al., 2011;Lind & Kang, 2017).An HQ TMT with substantial international experience is able both to identify the nature of its disadvantage (e.g., the lack of some knowledge) and to achieve more effective involvement (e.g., allocating the right resources at the right time) to improve the implementation of reverse knowledge transfer, leading to better parenting advantage.Hence, for HQ TMTs with high proportions of executives with international experience, HQ involvement in knowledge transfer will create a better result.In line with the positive view of TMTs' international experience in internationalization dominant in the literature based on advanced-market MNEs, we expect that: Hypothesis 3. In Chinese MNEs, the positive effect of HQ involvement in knowledge transfer on the extent of reverse knowledge transfer is positively moderated by the HQ TMT's international experience.

Methodology
We test our research hypotheses using data from a survey consisting of 99 Chinese MNEs and their 177 subsidiaries in 34 foreign markets.

Data sampling
The survey has two questionnaires, administered within the same MNE to the HQ and to overseas subsidiaries.The sampling MNEs are required to be Chinese MNEs which have at least one subsidiary that has been operating in advanced markets for a minimum of three years.The sampled subsidiaries need to meet three criteria: (1) the subsidiary needs to be any entity in which the sampled MNE holds at least 50 % of the equity, in accordance with definition of subsidiary (Voxman, 1992); (2) it must have been owned by the MNE for a minimum of three years; and (3) it must be important to the sampled MNE's business in order to avoid sales offices or minor subsidiaries.These criteria support the selection of suitable subsidiaries in which there is a high possibility of conducting reverse knowledge transfer within the sampled MNEs.
We derive our sample MNEs from the 2,679 Chinese enterprises listed on the Shenzhen Stock Exchange, the Shanghai Stock Exchange and the Hong Kong Stock Exchange.This is because: (1) the listed enterprises are representatives of large Chinese enterprises that have a high possibility of being MNEs; (2) it is difficult to obtain public detailed information for Chinese non-listed enterprises to assess whether they meet the sampling criteria.Considering the sampling criteria and the known difficulties in the collection of survey data among large firms in emerging markets (Hoskisson, Eden, Lau, & Wright, 2000), a purposive sampling procedure (Adams & Brace, 2006) is necessarily employed.This procedure enables us to select the MNEs that meet the criteria and to obtain a "difficult-to-identify" sample through our judgment based on secondary data and public information (Saunders, Lewis, & Thornhill, 2007).This procedure also ensures that we obtain the necessary support from the sampled MNEs to complete the questionnaires and avoid C. Su et al. time-consuming access issues (Saunders et al., 2007).
From the 2,679 Chinese listed enterprises, we identify a list of 370 Chinese enterprises that fulfill our sampling MNE criteria based on public information.We attempt to contact them for the survey, and eventually, 106 out of these 370 enterprises agree to participate in the survey.We then obtain 106 filled-out HQ questionnaires and 185 filledout subsidiary questionnaires from those 106 Chinese MNEs.The response rate is 28.6 %.After dropping the MNEs and subsidiaries that do not fulfill the criteria mentioned above, we end up with a valid sample of 99 Chinese MNEs and 177 subsidiaries.The 99 MNEs are located in 20 out of a total of 31 provinces and municipalities directly controlled by the central government in mainland China.For the 177 subsidiaries, 141 are located in advanced markets (mainly in the USA, Germany, Australia and the UK) and 36 in developing markets (mainly in Brazil, India, Thailand and Vietnam).A total of 49 of the subsidiaries are acquired, while 128 are greenfield.Table 1 presents basic information about the sampled MNEs and subsidiaries, exhibiting good variance across key demographic variables.

Questionnaires and data collection
We design two questionnaires: within the same MNE, one questionnaire is answered by HQ representatives and the other is filled out by subsidiary representatives.This allows us to gain insights from both HQs and subsidiaries in the reverse knowledge transfer dyad.The questionnaires are developed on the basis of the literature and a case study of one Chinese state-owned MNE.Prior to developing the questionnaires, through in-depth interviews with managers and engineers from both HQ and subsidiary, we have undertaken a case study in a Chinese stateowned MNE and its four subsidiaries in Germany and the Netherlands to investigate reverse knowledge transfer and its facilitators and barriers.The questionnaires are originally designed in English and then translated into Chinese.In the translation procedure, the conventional and well-accepted back-translation process (Brislin, 1986) is followed to avoid cultural bias, detecting in particular any significant misunderstandings due to translation, thereby ensuring validity.The questionnaires are pretested in four Chinese MNEs and four subsidiaries in order to ensure that the questions are clear and that they capture the desired information as well as checking for information sensitivity.
Common method variance bias (Chang, Van Witteloostuijn, & Eden, 2010) is of limited concern in this study.The data on the independent (HQ involvement in knowledge transfer) and dependent (the extent of reverse knowledge transfer) variables are collected from HQ and subsidiary respondents respectively to avoid using single-source data.In addition, the sets of questions and items measuring the investigated variables and the controls are placed in different sections, with different types of response formats and scales (illustrated in the Sub-section 4.3).Moreover, we conducted a questionnaire pretest to avoid the use of confusing, vague or unfamiliar terms in the wording of questions and the content of the items.Further, as personal face-to-face interviews can enhance the reliability and quality of the data (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003), the majority of the HQ questionnaires are collected via face-to-face interviews.We also assure anonymity to HQ and subsidiary respondents.
The data-collection process starts with contacting and arranging an interview with HQ managers.The HQ questionnaires are administered on-site in order to secure valid responses and to obtain access to subsidiaries for subsidiary questionnaire collection.HQ questionnaires are answered by MNE HQs' top managers (e.g., vice-CEO, board secretary and board-office manager) and HQ divisions' heads (e.g., director of the international business department), who have sufficient knowledge about the management of overseas subsidiaries.These HQ respondents are asked to nominate up to five relevant overseas subsidiaries in HQ questionnaires.They further helped us to contact the top managers of these matching subsidiaries to complete the subsidiary questionnaire.We received most of the subsidiary questionnaires via online social tools (e-mail, QQ, Wechat, etc).The majority of subsidiary respondents are Chinese expatriate managers, and a few are local top managers, the assistants of CEOs and functional/divisional managers such as the head of R&D.These respondents are expected to possess a comprehensive knowledge of the subsidiary's management, the subsidiary-HQ relationship and knowledge flows.

Measures
Measurement items of the dependent, independent, moderating and control variables are illustrated in Table 2. 4.3.1. Dependent and independent variables 4.3.1.1. HQ involvement in knowledge transfer.This reflects a more "hands-on" strategy which is intended to capture the level of attention to, the direction of, and participation in the subsidiary knowledge transfer process by MNE HQs (Lind & Kang, 2017: 584).Adapted from the studies of Ciabuschi, Forsgren et al. (2017) and Lind and Kang (2017), we employed three items, as shown in Table 2, to measure HQ involvement in knowledge transfer, based on a seven-point scale ranging from 1 (not at all) to 7 (very much).

The extent of reverse knowledge transfer.
Reverse knowledge transfer is commonly referred to as the transfer of knowledge (e.g., know-how about products, processes, technologies and markets) from a subsidiary to its HQ (Gupta & Govindarajan, 2000;Rabbiosi, 2011;Yang et al., 2008).Developing from Gupta and Govindarajan (2000) and Yang et al. (2008), it is measured by the extent to which a subsidiary transferred three different types of technology-related knowledge to the MNE HQ over the previous 12 months, based on a seven-point scale ranging from 1 (not at all) to 7 (very much).We focus on technology-related knowledge transfer because technological innovation is strategically important for the competitive advantage of MNEs in global competition (Ferreira et al., 2015;Li & Atuahene-Gima, 2001).This is particularly pronounced for Chinese MNEs that are latecomers and under pressure to catch up with their Western counterparts (e.g., Anderson et al., 2015;Deng, 2009;Luo & Tung, 2007).

Moderating variables
Following the upper echelons research, a TMT is identified as being the top two tiers of an organization's management (e.g., CEO, chairman, CTO, CFO, COO, CMO, president and the next highest management tier) (e.g., Carpenter & Fredrickson, 2001;Carpenter et al., 2001).This is expected to capture the dominant coalition for all MNE HQs in the sample (Cyert & March, 1963).

HQ TMT's political experience.
This captures a proportion of members with political experience (current or previously serving as government bureaucrats) in the top management team.Adapting from the studies of Francis et al. (2009) and Lu et al. (2014), a TMT's political experience is measured as the percentage of top managers who are, or who have been, government officers, members of the Chinese People's Congress or members of the Chinese People's Political Consultative Conference at county level or above in the total number of top managers and the question is answered by HQ respondents.

HQ TMT's international experience.
This generally reflects the experiences that top managers have had in an international context, including international education and international work (Azam et al., 2018;Hutzschenreuter & Horstkotte, 2013;Lu et al., 2014;Tihanyi et al., 2000).Many studies use the ratio of managers with international experience to operationalize TMT international experience (e.g., Cui et al., 2015;Lu et al., 2014;Nielsen & Nielsen, 2011;Tihanyi et al., 2000).Following these studies, we measure TMT international experience as the percentage of top managers with international experience (studying or working abroad for more than one year) in the total number of top managers and the question is answered by HQ respondents.

Control variables
Other likely predictors of reverse knowledge transfer are controlled for to avoid spurious effects.Based on the review of the literature on the determinants of reverse knowledge transfer, we identify seven control variables commonly operationalized in the literature and incorporate them in our empirical analysis.4.3.3.1.Subsidiary size.This is often considered to be a proxy of a subsidiary's knowledge stock, which affects reverse knowledge transfer through scale effects (Gupta & Govindarajan, 2000;Rabbiosi, 2011).In general, larger subsidiaries have a greater pool of resources and knowledge assets dedicated to reverse knowledge transfer compared with smaller ones (Gupta & Govindarajan, 2000).In line with previous studies (Björkman, Barner-Rasmussen, & Li, 2004;Simonin, 1999), we measure subsidiary size as the total number of employees.

Establishment mode.
This captures whether or not a subsidiary is acquired.Following the study by Björkman et al. (2004), a subsidiary that is an acquisition takes on the value of 1 and the value of 0 otherwise.Compared with greenfield subsidiaries, acquired subsidiaries are likely to possess knowledge that is unique or different within the MNE context and thus to transfer more knowledge to their HQs (Björkman et al., 2004;Borini, de Miranda Oliveira, Silveira, & de Oliveira Concer, 2012;Gupta & Govindarajan, 2000;Rabbiosi, 2011).Therefore, we include the establishment mode as a control variable.

The weight of Chinese in relation to the total number of subsidiary
employees.This variable indicates the potential for socialization between HQ and subsidiary (Björkman et al., 2004;Guo, Jasovska, Rammal, & Rose, 2020), which is a stimulus for reverse knowledge transfer (Gupta & Govindarajan, 2000).4.3.3.4.Subsidiary R&D intensity.This is a proxy of subsidiary innovativeness and of the subsidiary's technological knowledge stock (Yanadori & Cui, 2013), which is central to reverse knowledge transfer (Gupta & Govindarajan, 2000).Subsidiaries with greater R&D intensity are expected to have higher technological competence and thus can transfer and contribute more innovative knowledge to their HQs.

MNE R&D intensity.
Similarly, this reflects MNE innovativeness and knowledge stock (Yanadori & Cui, 2013).MNEs with greater R&D intensity are more likely to have stronger entrepreneurial orientation (Borini et al., 2012) and greater knowledge stock, which leads to a stronger intention to learn subsidiary knowledge and a better capability to learn and integrate subsidiary knowledge.
4. 3.3.7.Organizational distance between HQ and subsidiary.This captures the degree of dissimilarity, in terms of business practices and organizational culture, between an HQ and a subsidiary (Simonin, 1999), commonly regarded as a barrier to reverse knowledge transfer (Simonin, 1999;Su & Kong, 2020).It is measured by a two-item scale developed by Simonin (1999).

Analysis and results
The hypotheses are assessed by the partial least squares (PLS) modelling (Wold, 1982), a powerful variance-based structural equation modelling (SEM) technique, via SmartPLS (Ringle, Wende, & Will, 2005) software program.Variance-based SEM makes it possible to examine measurement and structural models and to consider measurement error.This is considered as the appropriate data analysis technique for our study in view of the research objectives, the relatively small sample size and the non-normal distribution of most indicators (Hair, Hult, Ringle, & Sarstedt, 2016).
The measurement model is firstly assessed in terms of item and construct reliability and convergent and discriminant validity.Table 3 presents the parameter estimates for the measurement model.All the loadings of each indicator used to measure the multiple-item constructs (see column 2 in Table 3) are over the 0.7 threshold (Carmines & Zeller, 1979).Construct reliability, measured as composite reliability (Werts, Linn, & Jöreskog, 1974), is over the suggested 0.7 boundary for all the constructs, ranging between 0.88 for "HQ absorptive capacity" and 0.92 for "organizational distance between HQ and subsidiary".The AVE estimates (Fornell & Larcker, 1981) are over the minimum acceptance threshold of 0.5 for all the constructs, implying convergent validity.The square root of each construct's AVE is greater than its correlation with the rest of the constructs (see Table 4), and the Heterotrait-Monotrait Ratio (HTMT) matrix (Henseler, Ringle, & Sarstedt, 2015) is also satisfactory, suggesting that each construct meets the requirement of discriminant validity and is sufficiently different from the others.Thus, we conclude that the measurement model has suitable metric properties.
Turning to the structural model, based on bootstrap tests with 5000 resamples, as illustrated in Table 5 and Fig. 1, the structural model indicates that there is a positive relationship between HQ involvement in knowledge transfer and the extent of reverse knowledge transfer (β = 0.52, p < 0.001), providing empirical support Hypothesis 1.To test the moderating effects of TMT's political and international experience, this study applies a two-stage approach suggested by Hair et al. (2016).As shown in Table 5 and Fig. 1, the estimations confirm the negative moderating effect (β=-0.14, p < 0.01) of TMT's political experience and the positive moderating effect (β = 0.20, p < 0.001) of TMT's international experience on the relationship between HQ involvement in knowledge transfer and the extent of reverse knowledge transfer.Hypotheses 2 and 3 are therefore empirically confirmed.Figs. 2 and 3 depict these two significant moderating effects.Kenny's (2015) effect-size criteria for interaction terms are used in order to detect and estimate the strength of the moderating effect.According to Kenny (2015), f 2 ≥ 0.025 represents large effect size.The effect sizes of TMT's political experience and TMT's international experience moderating on Hypothesis 1 is 0.055 and 0.096 respectively, thus indicating large moderating effects.
Continuing with the assessment of the structural model, the variance explained by the model (R 2 ) is 0.68 for the dependent variable -the extent of reverse knowledge, which shows a satisfactory explanatory power of the model.The model has predictive relevance, as the Stone-Geisser cross-validated redundancy (Q 2 ) statistic for the dependent variable computed using a "blindfolding" technique (with the omission distance set at 8) is 0.45.This technique assesses the validity of the paths by repetitively estimating the model parameters with random data points omitted (hold-out samples).

Discussion and concluding remarks
Given the debate surrounding parenting advantage and the related limited research in the context of Chinese MNEs, this study brings some novel insights.First, demonstrating a positive relationship between HQ involvement in knowledge transfer and the extent of reverse knowledge transfer from subsidiaries, we confirm the logic of parenting advantage in the context of Chinese MNEs (Foss, 1997;Goold et al., 1998).To some   a Diagonal values in bold are the square roots of the variance shared between the constructs and their measures.For discriminant validity to be established, the diagonal elements must be greater than the off-diagonal elements in the corresponding rows and columns.unwillingness to share knowledge, particularly from more advanced host countries, with China (Awate et al., 2015;Ciabuschi, Kong et al., 2017).Hence, we can argue that, in the context of Chinese MNEs, broadly EMNEs, it could be worth pursuing a proactive HQ strategy to stimulate subsidiary knowledge transfer, instead of operating the active non-action (wu wei) strategy commonly adopted by Chinese MNEs (Li, Zhou, & Zhou, 2016) to achieve innovation catch-up.However, our empirical results further reveal that HQ TMT experience, specifically political and international experience, are essential factors that moderate, negatively and positively respectively, the beneficial effect of HQ involvement on reverse knowledge transfer in Chinese MNEs.The impact of top managers' political experience is quite controversial as, although it is commonly regarded as a managerial resource in China (Li & Atuahene-Gima, 2001;Peng & Luo, 2000;Xin & Pearce, 1996), contributing to the growth of Chinese MNEs' foreign investment (e.g., Duanmu, 2014;Luo et al., 2010;Wei et al., 2015), our results show a negative view of such in terms of extracting benefits from cross-border knowledge transfer.This is in line with the findings of studies (e.g., Fan et al., 2007;Li et al., 2007) which identify that a TMT's political background is detrimental to a firm's value and efficiency.The fundamental reasoning is that the knowledge accumulated from a TMT's political experience in Chinese MNEs is local-specific and context-specific, which is neither useful nor transferable to international operations (Lu et al., 2014).
Consistent with an extensive body of research which demonstrates that a TMT's international experience is conducive to internationalization (e.g., Athanassiou & Nigh, 2002;Herrmann & Datta, 2005;Hsu et al., 2013;Li, 2017;Lu et al., 2014;Sambharya, 1996), this study also highlights the importance of a TMT's international experience in cultivating HQ parenting advantage in terms of generating reverse knowledge transfer in Chinese MNEs.We argue that the knowledge and advantage built on the international experience of the TMT is fungible  across national borders, giving the TMT a more enabling role in subsidiary operations and activities.This applies particularly to Chinese MNE HQs as HQ TMTs' extensive international experience augments organizational knowledge and capabilities and helps to compensate for a lack of the international knowledge and expertise needed for international value-adding activities in Chinese MNEs.It also seems to mitigate the detrimental effects of a TMT's political experience when a TMT is involving itself in subsidiary activities.International experience seems to be inimitable and non-substitutable, providing a person with skills that are not easily developed through other means (Sambharya, 1996).In other words, firms should hire talents with international experience to join TMTs and this strategy is apparently effective in Chinese MNEs, broadly in EMNEs.For EMNEs, "grafting" -through the recruitment of returnee managers and the utilization of repatriatesis a more effective way than knowledge accumulation through conventional experiential learning and is more complete than knowledge accumulation through imitation (Cui et al., 2015) as EMNEs are typically latecomers in a "catch-up game" in the competitive pressures of the global market (Deng, 2009;Luo & Tung, 2007, 2018;Mathews, 2006).
Moreover, as we can assume that many top managers in Chinese MNEs at the time of internationalization already possess political experience, because of the nature of their home markets, we identify a crucial issue not only in terms of the contrast between the political and the international experience of top managers, but also in the probable time period in which these experiences are accumulated.Political experience is known to trigger and support Chinese firms' internationalization, but, according to our results, later on, in the management of more established MNEs, it becomes a problem to create synergies and to practise knowledge transfer, a key element of many Chinese MNEs' competence-upgrade process.Thus, it is the top managers' international experience that, in time, should compensate for the effect of their political experience.This corroborates the view that a TMT's international experience typically complements the role played by its other experiences (Reuber & Fischer, 1997;Sambharya, 1996).
These opposing moderating impacts are in line with the findings of Lu et al. (2014), which identify that the positive effect of domestic diversification on international diversification is reduced by TMT political connections and reinforced by TMT international experience in Chinese MNEs.However, alongside the discussion of parenting advantage from a knowledge perspective, we argue that the political and the international experience of TMTs in Chinese MNEs shape the HQ-knowledge situation vis-a`-vis subsidiaries differently: the former hinders HQ knowledge, increasing the risk of an ignorant HQ, while the latter enhances HQ knowledge, enabling the HQ to be more rational and in a position to create value.Hence, as verified by our results, a TMT with extensive international experience lays out the ground for developing parenting advantage, while a TMT with extensive political experience generates a weaker HQ knowledge position.These new insights also contribute to the "parenting paradox" (Ciabuschi, Forsgren et al., 2017: 56) and to "complex parenting situations" (Decreton et al., 2017: 859) as, in addition to the matter of how much knowledge HQs possess, we highlight the effects of different types of knowledge, i.e., political vs. international experience.

Research contributions
The research contribution of this paper is threefold.First, we extend the parenting advantage logic to the context of Chinese MNEs.In doing so, we add to the literature on HQs' role in general and on EMNEs in particular.In support of parenting advantage logic (Foss, 1997;Goold et al., 1998), our findings demonstrate that HQ involvement in knowledge transfer is highly associated with reverse knowledge transfer in Chinese MNEs.This, to some extent, contrasts with the recent notion by several scholars who have a pessimistic view of parenting advantage (e. g., Ciabuschi et al., 2011, Ciabuschi, Forsgren et al., 2017;Lind & Kang, 2017).Our findings provide empirical evidence to verify the potential of HQ parenting advantage in Chinese MNEs, contributing to the flourishing research in this area.As Chinese MNEs have been characterized as latecomers that lack knowledge of how to do business internationally and manage subsidiaries, they are assumed to have parenting disadvantage in dealing with international affairs.Our results suggest that HQs' proactive involvement is of particular importance in facilitating reverse knowledge transfer, contributing to the ambitious growth and catch-up strategy of Chinese MNEs.
Second, this study integrates TMT literature with the parenting advantage perspective in order to elaborate on how, and under what, conditions can facilitate or inhibit HQ parenting advantage.This has not yet received due attention in HQ-role research.By the same token, TMT research focuses primarily on the role of TMT characteristics in determining firm internationalization, largely ignoring the impact on HQ parenting advantage.This study provides evidence that supports the relationship between TMT experience and HQ parenting advantage in Chinese MNEs.Specifically, we reveal that the strength of HQ parenting advantage actually depends on the political experience and the international experience of TMTs.Thus, our findings complement recent work challenging HQ parenting advantage (e.g., Ciabuschi et al., 2011, Ciabuschi, Forsgren et al., 2017;Lind & Kang, 2017) by showing how not only the extent of knowledge, but alsoin particularthe different types of knowledge (political vs. international experience) matter in further framing HQs' knowledge situation in relation to parenting advantage.In short, we suggest that the TMT is a pivotal perspective for further explaining HQ parenting advantage, which encourages further inquiry into the impact of other TMT characteristics on HQ activities.
Third, this study takes an initial step in understanding reverse knowledge transfer from the perspective of a TMT's experience, enriching the research on reverse knowledge transfer, particularly in regard to EMNEs (e.g., Awate et al., 2015;Borini et al., 2012;Ciabuschi, Kong et al., 2017;Nair et al., 2016).Our results, showing the opposing moderating effects, suggest that TMTs' international experience is necessary for Chinese MNEs in ensuring a superior parenting advantage and in enhancing reverse knowledge transfer.Meanwhile, TMTs' substantial political experience hinders HQ parenting advantage and eventually poses a barrier to reverse knowledge transfer.These new factors that we have identified provide novel insights into understanding reverse knowledge transfer, opening up a new potential research area.

Managerial implications
This study suggests that TMTs' political experience constrains Chinese MNE HQs in effectively performing their parenting role in terms of involvement in foreign knowledge acquisition and learning.Meanwhile, HQ TMTs' international experience helps to compensate for the lack of organizational experiential knowledge needed for international valueadding activities in Chinese MNEs and is conducive to HQ parenting advantage.In other words, TMTs' international experience is one of the critical managerial resources that complements TMTs' political experience and contributes to the successful transition of Chinese firms into an international environment.This highlights the importance for Chinese MNE HQs, particularly state-connected ones, of incorporating members with international experience into TMTs and of cultivating their top managers' knowledge of international markets in times of increased international competitive pressure.More specifically, for Chinese HQs, a more effective way could be to recruit returnee managers with international leadership experience in order to resort to grafting so as to accelerate learning and thus enhance parenting advantage in international operations.This is particularly useful when Chinese MNEs, as latecomers to global business, are under time pressure due to intense international competition.Moreover, Chinese firms which have a substantial international presence can promote repatriate executives to senior management positions at HQs to augment international knowledge and insights.In relation to this, one can also speculate that Chinese MNEs need to do a better job of selling the notion of international tasks to their executives as a stepping stone to entry into the upper echelon rather than an entry to obscurity.These approaches can reduce the dominance of politician managers in Chinese HQs' TMTs and enhance the heterogeneity of TMTs, thereby promoting HQ parenting advantage in cross-border management, as the benefits of TMTs with a diverse background for multiple perspectives, particularly knowledge accumulation, are more apparent in a Chinese context which typically emphasizes group orientation and interpersonal harmony (Tjosvold, Poon, & Yu, 2005).

Limitations and future research
This study has several limitations which give rise to future research areas.The first research direction is suggested by our single-country empirical setting, i.e., MNEs from China.While we believe that our empirical focus on Chinese MNEs provides us with a unique opportunity to contribute to this line of research, it would be worthwhile to examine the effects of TMTs' international and political experience on HQ parenting advantage in alternative empirical settings, optimally crosscountry ones.It is acknowledged that differences in societies, or in national systems, may influence the effect that top managers and their experience are able to exert (Hambrick, 2007).For example, as a result of one party dominating in Chinese political institutions, the effects of top managers' political experience in Chinese MNEs may be more apparent and stronger than in other emerging markets.An important area for future studies is research on multi-countries, including other large emerging markets such as India, Russia and Brazil, that are emerging from different contexts and which have different political systems.This may provide the confidence needed to generalize our findings into these settings as well.
In addition, although we focus on the moderating effects of TMTs' international and political experience, we do not argue that these are the only important moderators.Future studies could account for other characteristics of TMTs, such as functional experience, or educational background and age, in regard to the relationships investigated in this study.Furthermore, this study focuses only on investigating the impact of HQ involvement on knowledge transfer and reverse knowledge transfer in regard to testing HQ parenting advantage in Chinese MNEs.Future studies can expand the focus to examine the effects of different HQ involvement, or different HQ initiatives, on other subsidiary activities, such as subsidiary innovation development (Ciabuschi et al., 2011, Ciabuschi, Forsgren et al., 2017) in EMNEs, to improve understanding of HQ parenting advantage in EMNEs.
extent, by focusing on the actual extent of reverse knowledge transfer, and not on the efficiency of the process itself, this study adds to, and contrasts with, the more pessimistic view on parenting advantage proposed byCiabuschi et al. (2011), Ciabuschi, Forsgren et al. (2017) andLind and Kang (2017).HQ involvement, as a type of hands-on strategy, seems to be advantageous and important to Chinese MNEs.The effort by HQs to support reverse knowledge transfer directly can be interpreted as a possible measure to overcome problems related to the well-known issue of light-touch integration by Chinese MNEs(Liu & Woywode, 2013;Su & Kong, 2020) and to the problem of subsidiaries' .s. = not significant; * p < 0.05; ** p < 0.01; *** p < 0.001 (based on a two-tailed Student's t (4999) -distribution).

Table 2
Operationalization of the constructs and descriptive data.

of Chinese in relation to the total number of subsidiary
a To what extent, during the past 12 months, has the MNE HQ received different knowledge (listed below) from each subsidiary?b what extent do you agree with the following statements?creversecode.C.Su et al.

Table 3
Item and construct reliability and average variance extracted (AVE).

Table 4
Discriminant validity a : Correlations and square root of the average variance extracted (AVE).

Table 5
Dependent variable: effects, explained variances and the Stone-Geisser Q 2 test.