Compensation practice in China: A configurational analysis of the impact of regional institutions and MNC decentralisation on adaptation

Working Paper 2017-86 Compensation practice in China: A configurational analysis of the impact of regional institutions and MNC decentralization on adaptation Wei Lu Department of Management Studies Aalto University, School of Business Ayse Saka-Helmhout Nijmegen School of Management Radboud University Rebecca Piekkari International Business Department of Management Studies Aalto University, School of Business Abstract: This article poses the question under what conditions do foreign subsidiaries in China locally adapt their compensation practice transferred by MNC headquarters? Earlier work suggests that practice adaptation is primarily due to a decentralized management approach or high institutional pressure. Our configurational Qualitative Comparative Analysis reveals that decentralization needs to be considered in combination with regional regulation for it to explain practice adaptation. Decentralization facilitates adaptation of compensation when it co-exists with lax regional regulatory institutions and low portability of skills in the region. In other words, despite low institutional pressures, subsidiaries still adapt their compensation practice when facilitated by decentralization. On the other hand, subsidiaries also adapt their compensation practice within regional governance systems when faced with tight regional labour regulations and high portability of skills in the region regardless of the management approach. Our findings of Finnish MNCs in China advance the impact of regional institutions on adaptation of compensation practice. Keywords: adaptation of compensation practice, regional institutions, China, MNC decentralization, international HRM, portability of skills in the region,, regional labour regulation, fsQCA COMPASSS Working Paper 2017-86 Available at http://www.compasss.org/wpseries/LuSakaHelmhoutPiekkari-2017.pdf COMPASSS Working Paper Series Managing Editor, Claude Rubinson Wei Lu International Business Department of Management Studies Aalto University, School of Business P.O. Box 21230 FI-00076 AALTO, Finland Email: wei.lu@aalto.fi Ayse Saka-Helmhout Nijmegen School of Management Radboud University P.O. Box 9108, 6500 HK Nijmegen the Netherlands Email: a.saka-helmhout@fm.ru.nl Rebecca Piekkari International Business Department of Management Studies Aalto University, School of Business P.O. Box 21230 FI-00076 AALTO, Finland Email: rebecca.piekkari@aalto.fi 1 Compensation practice in China: A configurational analysis of the impact of regional institutions and MNC decentralization on adaptation Abstract: This article poses the question under what conditions do foreign subsidiaries in China locally adapt their compensation practice transferred by MNC headquarters? Earlier work suggests that practice adaptation is primarily due to a decentralized management approach or high institutional pressure. Our configurational Qualitative Comparative Analysis reveals that decentralization needs to be considered in combination with regional regulation for it to explain practice adaptation. Decentralization facilitates adaptation of compensation when it co-exists with lax regional regulatory institutions and low portability of skills in the region. In other words, despite low institutional pressures, subsidiaries still adapt their compensation practice when facilitated by decentralization. On the other hand, subsidiaries also adapt their compensation practice within regional governance systems when faced with tight regional labour regulations and high portability of skills in the region regardless of the management approach. Our findings of Finnish MNCs in China advance the impact of regional institutions on adaptation of compensation practice. Key words: Main key words: adaptation of compensation practice, regional institutions, fuzzy-set Qualitative Comparative Analysis (fsQCA), China. Additional key words: MNC decentralization, international HRM, portability of skills in the region, regional labour regulation. 2 Introduction From the labour market point of view, the situation is about heavy competition for talent… Many companies, not only foreign but also domestic… have tried to poach people from us. They offer very high salary. We have had such cases where our managers and experienced engineers have left for our competitors. - China HR Director, Finnish multinational corporation As the above quotation suggests, multinational companies (MNC) face fierce competition for talent in China and compensation is an important HR practice in attracting and retaining employees. The lack of especially managerial and professional talent is a key challenge for MNCs operating in China. According to the IMD World Talent Report 2014, China ranks only 43rd among 60 countries in terms of its ability to develop, attract and retain talent for corporate needs. Chinese managerial/professional employees have high expectations concerning pay/benefits which have partially contributed to high labour mobility. This, coupled with the mismatch between supply and demand, has made China one of the most competitive labour markets in the world. Our focus on highly educated employees is particularly topical, as China has become one of the leading economies in the world and provides substantial business opportunities for foreign MNCs. Compensation tends to be a highly centralized and standardized HR practice in MNCs because it supports and enhances the strategic objectives of the firm (e.g. Nohria and Ghoshal, 1997). We refer to compensation practice as the sum of elements concerning pay mix, pay level and benefits for managerial employees at foreign subsidiaries of MNCs, which serve as our focal group. A globally aligned compensation system can create a sense of equality among employees all over the world and facilitate the building of a talent pool at the company level (Rosenzweig, 2006). However, MNCs are also likely to experience counter-pressures on their compensation practices. For instance, the study by Lu (2014) shows that among the six HR 3 practices (recruitment, compensation, benefits, performance management, training and employee representation), compensation faced the strongest competing pressures of global integration and local responsiveness between the MNC HQs and the Chinese local context. Existing research on compensation practice captures these competing pressures. Studies highlight, on the one hand, the role of institutional/cultural influences and the strategic position of subsidiaries in the adaptation of compensation practice (e.g. Festing Eidems and Royer, 2007; Marín, 2008; Myloni, Harzing and Mirza, 2004; Tosi and Greckhamer, 2004). On the other hand, studies demonstrate the pressure for global integration when subsidiaries, operating in institutionally less regulated and highly competitive markets, mimic the practices of parent companies operating in institutionally developed settings, to gain or sustain a competitive advantage. HR practices of Western MNCs, in particular outcome-driven and merit-based performance and pay, tend to be seen as universal best practices in China. This encourages subsidiaries to adopt compensation practices that are similar to the practices of parent companies (e.g. Björkman and Lu, 2001; Farley, Hoenig and Yang, 2004). Although scholars acknowledge that various influences can be seen to act together (Aguzzoli and Geary, 2014), the emphasis is on the independent effects of these influences on compensation practice. In addition, despite calls for an investigation of the impact of sub-national institutions on firm behaviour (e.g., Ferner and Tempel, 2006; Almond, 2011), existing analysis focuses primarily on the effects of national institutions. We seek to address this significant gap by taking a configurational approach (see Fiss, 2009) and showing how firm-level conditions interact with regional institutional conditions in China. We examine the adaptation of compensation practice as a configuration (see Grandori and Furnari, 2013), meaning that compensation practice is not considered in isolation but rather as a bundle of constituent aspects which are aligned with one another. We ask under what conditions subsidiaries of Finnish MNCs locally adapt compensation practices in China. 4 Particularly, our interest lies in explaining how regional institutions and portability of skills in the region combined with MNC decentralization influence the adaptation of compensation in China. Our contributions are twofold. First, contrary to much evidence concerning location decisions of MNCs in environments with lax regulations (e.g. Fuller and Phelps, 2004; Dunning, 2009), we demonstrate that MNCs seek to adapt their compensation practice in regions where labour regulations are of lower standard. Subsidiaries mimic the higher standards of firms in neighbouring regions in their compensation. Rather than aiming to keep costs low by implementing standard packages, the Chinese subsidiaries in regions of lax regulation adapt their compensation to attract higher skills. Second, we advance our understanding of the impact of regional institutions on adaptation of compensation practice. Previous studies adopting an institutional perspective have primarily focused on the national rather than the regional level (see Almond, 2011; Ferner and Tempel, 2006). We show that a subsidiary’s conformity to regional labour regulations is not simply explained by the constraining effects of national institutions on MNC activity. Our findings point to the significance of the autonomy of regional governments in setting their own labour standards that offer the opportunity to MNCs to display strategic choice in adaptation (e.g. Almond, 2011). The pressure stemming from high portability of skills in the region and tight regional labour regulations were strong enough for local adaptation. Hence, whether the management approach to foreign subsidiaries was centralized or decentralized did not make a difference in adaptation. International Business research has traditionally associated adaptation with a decentralized management approach (Bartlett and Ghoshal, 1989). The next section introduces the theoretical background to the institutional and organizational conditions that influence, in combination, the adaptation of compensation 5 practice and how this configurational thinking extends previous research. This is discussed in the context of China. We then present our methodology, qualitative comparative analysis (QCA) followed by our findings. We position them in the extant literature on HR practices and draw implications for future research on compensation practice. Theoretical background We draw on comparative institutionalism and the global integration-local responsiveness (IR) framework (Prahalad and Doz, 1987) to specify the configuration of fundamental conditions, i.e. portability of skills in the region, regional labour regulations and MNC decentralization. We examine the impact of these three conditions on the adaptation of compensation practice. Adaptation of compensation practice The competing pressures of global integration and local responsiveness explain, to a large extent, when compensation practice varies across borders (see Fay, 2008; Sparrow, 2004). Comparative institutionalists have focused on home and host institutional differences to justify the adaptation of compensation practice in the host context (e.g. Almond et al., 2006). They argue that subsidiaries adapt their practices to address the regulative structures, and normative and cognitive understandings of the host country in order to gain legitimacy. For instance, their case study of a European MNC, Festing et al (2007) revealed that while the HQ standardized rules in basic pay, long-term and short-term incentive, some aspects of compensation, such as the range of bonuses, fringe benefits and currency bases were adapted to the local conditions. They demonstrated that adaptations are the result of host institutional and cultural demands, as well as the strategic position of the subsidiaries in question. Similarly, Festing and Sahakiants (2013) highlighted the local adaptation of practices ranging 6 from pay-for-performance and bonuses to social benefits as MNCs aimed to comply with host regulations or cultural norms. In contrast to the above contributions, several studies highlight the need for firms to increase competitive advantage by adopting universal ‘best practices’ where these firms face severe institutional and competitive pressures. Subsidiaries benchmark or tap into the MNC’s ‘learning network’ to emulate the compensation practice of HQ (e.g. Edwards et al., 2010; Pudelko and Harzing, 2007). The ‘dominance effect’ is noted, in particular, in China where the weak host institutional pressures and strong MNC influence via expatriate management result in the global integration of subsidiaries (e.g. Björkman and Lu, 2001). The two broad approaches point to mixed research findings. We believe this is related to the study of independent effects of relevant influences on adaptation of compensation practice. Therefore, we propose a configurational research as an alternative approach to accounting for the interdependencies between sub-national institutional conditions and organizational condition. Portability of skills in the region We expect portability of skills in the region to influence the adaptation of compensation practice. Comparative institutionalists identify two forms of skills formation across societies: one relying on external labour markets, outside the firm, and another within the firm emphasizing long-term commitment to developing firm-specific employee capabilities (Estevez-Abe, Iversen and Soskice, 2001). Skills formation outside the firm, or general skills, reduces the dependence of the firm on employees, and the capacity of employees to influence management decisions (Aguilera and Jackson, 2003). Where skills are portable across firms, or skills are general with theoretical content as in academic professional skills, employees tend to choose exit over voice in their response to grievances. In contrast, skills that are firmspecific (acquired through on-the-job training) or industry-specific (acquired through 7 apprenticeships and vocational schools) are less portable. They are valuable to a given employer or an industry but not to other employers or industries. Industry-specific skills, especially when they are not authoritatively certified, are not as well recognized by other employers as are general skills. ‘General skills, recognized by all employers, carry a value that is independent of the type of firm or industry’ (Estevez-Abe et al., 2001: 148). Studies on skills development in China are primarily focused at national level. For instance, Asuyama (2009) shows that 42 per cent of students in upper secondary education enter vocational schools. Similarly, OECD (2010) reports that about half of the student cohort in upper secondary education, i.e. about 20 million students, is enrolled in vocational schools which affects the formation of industry-specific skills. Although China provides vocational education and training to 8.6 per cent of the entire workforce, about 65.2 million people (Asuyama, 2009), it lacks cooperation between employers and vocational schools. Furthermore, there are few regional or sectoral bodies to engage employers and link them with the vocational education system (OECD, 2010). In contrast, the higher education in China has experienced a dramatic expansion since 2000 which produces general skills and increases the portability of skills in a region. In 2010 the total number of students in higher education institutions amounted to 31 million (People’s Daily Online, 2011), and the percentage of higher education attainment accounted for 8.93 per cent of the total population in China (China Population Census, 2010). Although research on skills formation at the sub-national level is rare, there is some evidence to show that there is variation across regions which can influence alterations in employment strategies and practices (Whitley, 1999). It is observed that provinces which have larger percentage of higher education attainment tend to have a lower percentage of vocational education attainment, pointing to differences across provinces in the local pool of skills, vocational or higher education, available to employers (e.g., China Statistical Yearbook, 2010-2013). 8 Companies tend to be reluctant to invest in employee training in China because they run the risk of losing employees after training. This is fuelled by the fundamental changes that China’s overall economic system has experienced from a command economy to the direction of a free market economy (see HFW Holman Fenwick Willan, 2013). This has brought with it changes in the form of employment from life-time, known as the “iron bowl” within one organization to fixed-term employment based on labour contracts. The use of short term contracts has prevailed, in particular, in the private sector (among Chinese domestic and foreign firms) (Gallagher, 2004) where the turnover rate of 20 per cent to 40 per cent has been fairly common (Kettunen, Lintunen, Lu and Kosonen., 2008; Kilian et al., 2012). The ongoing social reform in relaxing the residence registry system (‘Hukou’), especially in recent years, has greatly reduced the barriers for labour mobility across regions and facilitated a rapid labour movement. In addition, China has been facing serious shortages of managers as a result of continuous high economic growth over the past three decades. The competition for talent has extended to that for workers (Das and N'Diaye, 2013; Farrell and Grant, 2005). This has accelerated competition in the labour market, which further encourages employees to move between firms and actualize their market value, creating low abundance of and high demand for skills as our opening quotation suggests. At the same time, firms facing scarcity of skills are aware of the need to invest in in-house training to develop their own capabilities. These pressures support the significance of analysing the portability of skills in a region in the adaptation of compensation. Regional labour regulation Institutional analysis of compensation practice has widely examined the impact of national regulations and trade unions (see Almond et al, 2006; Marín, 2008). This national approach is 9 based on the premise that most important institutional structures, namely labour market regulations, education and training, and corporate governance, depend on the regulatory regimes of nation states. More recently, however, comparative institutionalists studying international HRM have acknowledged the potential effects of regional governance on HR practices (e.g. Almond, 2011; Monaghan et al., 2014; Almond et al., 2015). They recognize that there is considerable diversity within most national economies in terms of culture and regulations, in particular in large countries (Edwards and Kuruvilla, 2005). This leads to differences across regions in the regional government’s ability to shape skills development (Almond, 2011). This is most apparent at the level of higher education in terms of the competencies of graduates and in the capacity of regional governments to create exploitable knowledge. Consequently, sub-national variations in institutional arrangements provide considerable space for firms to choose from among alternative paths or actions in a national institutional setting. Sub-national actors such as regional development agencies have the power to adapt, amend and, to an extent, contradict elements of the national institutional system in ways that create local forms of comparative (dis)advantage (Peck and Theodore, 2007). They can strategically couple between the resources of different places within the region to meet the changing strategic needs of global production networks (Almond et al., 2015). This has implications for our understanding of what happens to compensation practices as they are implemented in an institutional setting with regional variation. In the context of China, national laws and regulations are defined broadly to enable regional governments flexibility in their implementation. Since the introduction of the first national labour law—the Labour Law in 1995— there has been a proliferation in the number of labour laws and regulations and a diminishing role of the national state in labour administration relative to the local state (Knight and Song, 2005). For instance, pay related 10 issues (salaries and benefits) are currently stipulated by national labour laws as well as local labour regulations. In addition to the flexible nature of national laws, local governments can devise their own regulations to suit local conditions (Cooke, 2011). For instance, given the vast size and considerable disparity in social development, China exhibits considerable variation in the levels of minimum wage and standards of social benefits across regions and cities. Although regional governments are also responsible for surveilling and monitoring laws, they tend to prioritise economic growth and lack the incentive to strictly enforce labour laws (Gallagher, 2004). This suggests variation in the regulatory institutional environment of firms located in different regions. The setting of managerial pay illustrates well the heterogeneity of the institutional environment across Chinese regions. He and Fang (2015) argue that since the setting of executive pay is an institutionalized process in which businesses respond to local social constraints to gain legitimacy, regional institutional differences can influence executive compensation. More developed provinces tend to have better functioning and more mobile executive labour markets (Cordeiro et al., 2013). Such a labour market condition may prompt firms to offer larger compensation so as to better motivate and retain their executives. We focus on minimum wage regulation given that the setting and adjustment of local minimum wage considers, among other variables, levels of economic development across regions (Fang and Lin, 2015), hence it can influence compensation practice. We, therefore, argue that regional labour regulations offer another explanation for why firms adapt their compensation practices. Decentralization in MNCs 11 Given our interest in the adaptation of compensation practice, decentralization in MNCs is of particular interest to us. Decentralization refers to “the allocation of decision rights to lowerlevel managers” in foreign subsidiaries in areas such as human resource management or technology investments (Williams and Triest, 2009: 156). In decentralizated MNCs, foreign subsidiaries have more scope to flexibly adapt their compensation practice and respond to local needs. The typology of international, multidomestic, transnational, and global MNCs suggests that these types vary in terms of the degree of centralization vs. decentralization and the way they deal with pressures stemming from local responsiveness and global integration (Bartlett and Ghoshal, 1989). For us, the multidomestic and the transnational type are especially relevant. In the multidomestic MNC (low level of global integration, high level of local responsiveness), assets and capabilities are decentralized, and foreign subsidiaries have the autonomy and the decision-making power to exploit local opportunities. Similarly, in the transnational MNC (high level of integration, high level of responsiveness), the resources and capabilities are dispersed across the subsidiary network, which provides subsidiaries with more freedom to develop themselves while simultaneously being integrated into worldwide operations. MNCs tend to align their HR systems, including compensation, with corporate strategy and structure (Taylor et al., 1996). Bloom et al. (2003) offer a typology of three approaches to compensation design which echoes the MNC types by Bartlett and Ghoshal (1989). Bloom et al (2003) show that MNCs with locally adaptive strategies commonly use a decentralized, multiple market approach in their compensation design. Like in the multidomestic MNC, this allows subsidiaries to develop unique compensation practices that meet the specific conditions of each local market. In contrast, MNCs that follow a global strategy typically design a unified compensation system across locations and subunits. Under such a centralized strategy, subsidiaries are compelled to implement HQ-transferred compensation practices with limited 12 adaptation to the local context. The third type is in-between these two extremes and resembles the transnational MNC. While the MNC employs highly integrated compensation systems to facilitate the implementation of one global strategy, its subsidiaries are at the same time highly locally responsive allowing them to also tailor compensation practices at the local level (Bloom et al., 2003). The typology offered by Bloom et al. (2003) suggests that decentralization in MNCs is integral to explaining the adaptation of compensation practice. In summary, we expect to observe multiple paths to the adaptation of compensation practice as we consider the combinatorial influence of institutional (portability of skills in the region and regional labour regulations) and organizational (decentralization) conditions. Methodology In order to address our research question, we apply Qualitative Comparative Analysis (QCA) (Ragin, 2008; Rihoux and Ragin, 2009) to our study of the adaptation of compensation practice. QCA represents an ideal method for systematically comparing cases, because it treats each case as configurations of qualitative attributes rather than disaggregates them into analytically separate aspects (Fiss, 2009). Each case is represented by a combination of the presence and absence of these conditions, which in turn leads to a given outcome. We have identified, in the theoretical background, portability of skills in the region, regional labour regulation and MNC decentralization as our conditions. We examine the combinations of these conditions under which adaptation of compensation practice occurs. This conjunctural causation sought by QCA differs from conventional statistical methods where causal effect tends to be investigated with independent variables. QCA permits multiple causations, i.e. different causally relevant conditions may combine in a variety of ways to produce a given 13 outcome (Ragin, 2008). This is likely owing to the complementary ways in which balancing or contrasting institutions affect compensation adaptation. The other advantage of QCA is its ability to combine features of both quantitative and qualitative methods. It can embrace the complexity of case analysis and offer significant theoretical leverage so that generalization is possible. The QCA method is particularly suited for a small-N case study such as ours. We adopt fuzzy-set analysis, a variant of QCA techniques (see Ragin, 2008; Schneider and Wagemann, 2012). The core analytical principle of QCA techniques lies in the set-theoretic methods. Crisp-set dichotomizes a case as either belonging to or not belonging to a set. Fuzzy-set, however, allows for treating membership of cases with gradual degrees. A case cannot only be fully in or fully out of a set, but also partially belong to a set. We are interested in the variations of adaptation of HQ compensation practice across subsidiaries. Adaptation is seldom an either-or issue. By applying fuzzy-set QCA, we can capture subsidiaries with partial adaptation, and avoid categorizing them as extreme cases. Selection of cases The study included ten subsidiaries (case companies) of Finnish MNCs operating in China. As a comparative method, an essential feature of QCA is to have as comparable cases as possible at the outset. By “controlling” most of the “variables”, different outcomes may be attributed to the remaining factors (Rihoux and Ragin, 2009). The ten case companies were selected following the “homogeneous sampling” strategy in case research (see Fletcher and Plakoyiannaki, 2011:179). The parent companies of the ten cases had all transferred similar compensation practice which consisted of three main elements: salary standard, position grading, and bonus scheme. The bonus schemes could be used to compensate for any shortfalls in pay levels. The cases were all wholly owned subsidiaries of Finnish MNCs, which eliminates the ownership effect. 14 The case subsidiaries located in five cities of China: Tianjin, Beijing, Shanghai and two cities in Jiangsu Province. Greater Shanghai (Shanghai and Jiangsu) and Beijing-Tianjin area had the most China operations of Finnish MNCs (Finland Chamber of Commerce, 2011). The variation of locations of the cases allowed us to examine the regional impact of labour regulations and skills portability on adaptation. The subsidiaries came from different Finnish MNCs. This offered variation in corporate structure in terms of decentralisation versus centralisation. The case subsidiaries also differed in industries where they operated- engineering, equipment, machinery, metal, chemistry, paper and electronics. More detailed information on the cases is offered in Appendix 1. We used pseudonyms to present the case companies. For instance, the three subsidiaries operated in engineering sector were named EngCoSZ, EngCoSH and EngCoBJ. The two subsidiaries in equipment industry were named EquCo1 and EquCo2. The subsidiaries in machinery and electronics were named MacCo and ElecCo, respectively. The subsidiaries in metal, chemicals and paper were named MetalCo, CheCo and PaperCo, respectively. Collection of data The data were collected from interviews, documents such as annual reports and internal company policies, archives such as the Labour Law providing information on minimum wage standards and benefits standards, and census statistics (Table 1 summarizes the data). A total of 35 semi-structured interviews were conducted between 2008 and 2010 at case subsidiaries in China. The interviewees were senior business managers, functional managers and HR managers who had the necessary knowledge about compensation practice that applied to managers at their companies. The interviews lasted between one and two hours, and were all 15 transcribed verbatim. The interview transcriptions were coded along the conditions of MNC decentralization and adaptation. -----------------------------Insert Table 1 about here -----------------------------The interviews on decentralization and adaptation were complemented by documentary data of relevant Chinese laws and regulations at both the national level and the regional level, as well as company documents such as HQ policy documents, annual reports, company presentations and organizational charts. The examination of relevant national laws was to confirm the scope of regional regulations concerning standards of minimum wage. The municipal regulations were found to have explicit standards on minimum wage, hence were used to code the condition regional labour regulations. HQ policy documents supplemented interviews in checking the consistency of compensation practice that was transferred across the HQs of the case MNCs to their subsidiaries. Company presentations and organizational charts added necessary information to the interviews in coding the MNC decentralization condition. Calibration of conditions and outcome Our fuzzy-set scale is defined with knowledge of theory and relevance to the cases (Ragin, 2009). Research on human resource management practices with QCA application is rather limited. A rare piece of work by Greckhamer (2011) addresses pay inequality at national level, while our study examines compensation practice adopted at firm level. We use the four-level scale following the studies on multinational enterprise response to internal and external environment (e.g. Crilly, 2011). The numerical values 0, 0.33, 0.67, and 1 indicate fully out, more out than in, more in than out, and fully in, respectively, in the set of causal conditions 16 (Ragin, 2009; see also Crilly, 2011). The conditions were attributed values by one rater and discussed with another project group member for potential discrepancies in calibration. There was an agreement on the calibration of all the conditions. Based on the literature review we derived three conditions: two conditions at regional level- portability of skills in the region and regional labour regulations, and one firm level condition- MNC decentralisation. Portability of skills in the region. We operationalize portability of skills in the region in terms of high-level formal educational qualifications, including the degrees that require a minimum of three years in higher education. This is a measure of general skills with theoretical content that is relevant for managers (see Estevez-Abe et al., 2001). In other words, a high percentage of high-level educational degrees indicates high portability of skills in the region, which requires the adaptation of compensation practices to a lesser extent than those attaining low-education, including vocational, degrees. We use the higher education attainment as the measure for portability of skills in the region. The case companies were all located in the east coastal area of China. We obtained percentages of population with higher education in nine coastal cities from the statistical yearbooks of respective municipalities (See Appendix 1). These statistics, which were based on the National Population Census of China in 2009, matched the time period of our data collection (2008-2010). The mean value of the nine cities was 18.35 per cent. This served as the crossover point of maximum ambiguity. -----------------------------Insert Table 2 about here ------------------------------ As Table 2 above shows, the case companies located in Beijing (31.50% attainment of higher education) were calibrated with a full membership value of 1. The cases located in Shanghai (21.95% attainment of higher education) were assigned a partial membership value of 0.67. The education standard of Suzhou is nearer to Tianjin than to Wuxi. Hence, we 17 assigned Suzhou a partial membership value of 0.33 to both Suzhou and Tianjin. Wuxi was assigned with the full non-membership value of 0. Regional labour regulation. We operationalize regional regulation in terms of minimum wage standards at the municipal level (Cordeiro et al., 2013; Fang and Lin, 2015). Similar to the calibration logic underlying portability of skills in the region (higher education attainment), we rely on the minimum wages of China’s coastal area as external standards. We draw on data from 2008-2009 to keep consistent with the qualitative interviews which were conducted mostly in this period. The municipal regulations on minimum wage standards were retrieved from Baidu (http://wenku.baidu.com/view/6a50f2aedd3383c4bb4cd236.html), the largest and most authentic internet search engine in China. The data on minimum wages in ten coastal cities of China are illustrated in Appendix 3. The highest minimum wage was 1000 yuan while the lowest was 800 yuan. The mean value of 885 yuan was set as the cross-over point of 0.5. Following Ragin’s (2009) guidelines, we assigned the score of 1 to the cases located in Shanghai with a minimum wage of 960 yuan. Suzhou and Wuxi, with a minimum wage of 850 yuan, below the average of the ten coast cities were assigned a partial membership value of 0.33. Tianjin and Beijing with minimum wage of 820 yuan and 800 yuan respectively were assigned with full non-membership. Table 3 below summarises the calibration. ------------------Insert Table 3 here ---------------------MNC decentralization. We follow Bartlett and Ghoshal’s (1989) taxonomies— multidomestic, global, international, and transnational— to conceptualize our organizational condition. We draw on Williams and Triest (2009) to capture the condition of decentralization in terms of subsidiary participation in decisions over investment and market and product 18 responsibility. We sought further evidence of this in interviews, annual reports and internal company documents. We assigned the subsidiary a full membership value of 1 where the decision making over investment and product development was decentralized in the MNC. A partial membership value of 0.67 was assigned where the case MNC had strategic assets across the subsidiaries. This type of coordination found in some of our cases that resembles the transnational model (Bartlett and Ghoshal, 1989). Some degree of centralization was evident in those cases. A partial membership score of 0.33 indicated that the case MNC was largely centralized in the distribution of its strategic assets and resources. Table 4 shows the illustrative quotations and document extracts on the evidence used for measuring this condition. -----------------------------Insert Table 4 about here ------------------------------ Adaptation. The outcome was calibrated with interview data collected from the case subsidiaries (see, Crilly, 2011). We draw on the understanding of “adaptive” integration by Taylor et al. (1996) to conceptualize the adaptation of compensation practices by subsidiaries. We regard adaptation as the integration by subsidiaries of HQ practice to address host environmental demands. The condition receives a full membership (value of 1) when the subsidiary has primary emphasis on responding to the local conditions hence substantially modifies HQ compensation practices. An example quotation: “The priority [in our compensation system] is to ensure the skills and competence we need for China operation. Now we have this kind of employee saving plan just for China which is different from the Corporation. This is China-specific” (R & D Director of EngCoSZ). A higher partial membership score of 0.67 is assigned when the subsidiary emphasizes local adaptation or engages in adaptive actions while aligning with the MNC organization to some extent. An 19 illustrative quotation: “When the unified practice comes from the top we look at how it can be implemented here. We modify what is necessary meanwhile try to maintain some standardised elements” (Business Director of EquCo1). A subsidiary that emphasizes alignment with the MNC organization while adjusting the HQ practice to some extent is coded a lower partial membership of 0.33. An illustrative quotation: “We have global compensation which is implemented in all locations. We basically follow the same. Only in exceptional situation we make some changes” (HR Manager of ElecCo) A full non-membership 0 is assigned to the subsidiary when alignment with the MNC organization was the priority of subsidiary; hence, no adjustment was made to HQ practice. An example quotation: “The corporate [compensation] policy is world-wide. We are part of the corporation and want to be consistent with the policy” (General Manager of EngCoBJ). The assignment of fuzzy-set membership scores to causal conditions and the outcome facilitated the construction of a matrix table of fuzzy-set membership (see Table 5). -----------------------------Insert Table 5 about here -----------------------------Using fsQCA 2.5 (Ragin and Davey, 2014), we first transformed the fuzzy-set membership scores in Table 5 into a truth table (Appendix 4). There are five logically possible combinations, three for the positive outcome and two for the negative outcome. The truth table also contains three logical remainders (the last three rows) that do not have real cases. Following the recommendation by Ragin (2009: 118) on fuzzy-set QCA, we chose 1 as the consistency threshold. As the number of cases is small, we chose 1 as the frequency threshold (Ragin, 2006). We sought the necessary conditions before conducting the analysis of sufficient conditions in the adaptation of compensation practice. A threshold consistency score of 0.90, which displays the proportion of cases consistent with the outcome, was used to judge whether a 20 condition was necessary (Ragin, 2006). Our analysis did not result in any necessary conditions (see Appendix 5). We followed Standard Analysis for identifying combinations of sufficiency (Ragin and Davey, 2014). This procedure can produce three solutions: a complex solution (no logical remainder is used), a parsimonious solution (all logical remainders are considered) and an intermediate solution (some logical remainders are included). The last solution requires the researcher’s substantive and theoretical knowledge as for what logical remainders may be included in the analysis (Ragin, 2009). While government regulations are likely to create coercive pressures for foreign subsidiaries to adapt their practices to the host country environment (DiMaggio and Powell, 1983), recent research has shown that foreign companies may circumvent or even shape host country institutions (e.g. Regnér and Edman, 2013). Comparative institutionalism suggests that a high portability of skills reduces the need for adaptation. However, the cases show that adaptation occurs in the presence as well as the absence of high portability of skills in the region. Therefore, one directional assumption could not be made for the conditions of regional labour regulations and portability of skills. According to the predominant view in the literature on MNC structure (Bartlett and Ghoshal, 1989), decentralised MNCs can grant subsidiaries autonomy for local adaptation of HRM practices (see, Bloom et al., 2003). Based on the above-mentioned grounds, we obtained the intermediate solution (presented in Table 6). The solution exhibits an overall consistency score of 100 per cent, considered ideal for fuzzy set analyses (Fiss, 2011), and explains more than two thirds of the adapted compensation practices (73.97 per cent coverage). The complex solution is the same as the intermediate solution. The parsimonious solution is shown in Appendix 6. The intermediate solution is displayed in the Findings section and further elaborated thereafter (see, Ragin, 2009). The parsimonious solution is also outlined in the Discussion section. 21 -----------------------------Insert Table 6 about here ------------------------------ We also performed the analysis of negative outcome (see Schneider and Wagemann, 2012). We obtained one intermediate/ complex solution ~mnc decentralisation*~regional labour regulations*portability of skills (see Appendix 7) which suggests that centralisation of MNC combined with lax regional labour regulation and high regional portability of skills is associated with non-adaptation of compensation practice. This additional analysis suggested that causation is not symmetric (Ragin, 2008), i.e., the negative outcome is not explained by the reverse role of the same conditions as for adaptation. Findings The first configuration shows that the adaptation of compensation practice is likely when subsidiaries operate within a decentralized MNC organization, and are located in a region of lower standards of wage and low portability of skills. This configuration is important because it highlights the interplay between institutional forces and the structural attributes of the corporate parent. It also indicates that operating in an environment with lax regulations and less transferable skills, i.e. lower education including vocational skills, does not necessarily mean the absence of or weak institutional pressure on compensation practice. On the contrary, our interviews demonstrate that subsidiaries are pressed to follow the higher standards of the neighbouring region in the compensation offered to managers. In other words, regional differences suggest that there are normative pressures to comply with the standards set by the region with the tightest regulations. However, the effect of this pressure actualizes when the subsidiaries have sufficient decision-making power granted by decentralization in the MNC. 22 The finding that the adaptation of compensation practice tends to occur in regions of lax regulations is contrary to FDI debates where lax labour regulations are seen as encouraging investment of a standardized nature in which practices do not contribute to local labour market development (e.g. Fuller and Phelps, 2004; Dunning, 2009; Beugelsdijk et al., 2010). The configuration suggests that lax regional labour regulations can facilitate the adaptation of compensation if foreign subsidiaries have the latitude and are under normative and competitive pressure from neighbouring regions. What is of significance here is that subsidiaries can adapt their compensation practice to attract high levels of skills rather than look to minimize costs by globally integrating this practice. Such subsidiary behaviour can lead to higher labour standards in a region (Ashton et al., 2010). The second configuration that we identified in the adaptation of compensation practice is high portability of skills in the region coupled with tight regional labour regulation standards. Subsidiaries adapted their compensation practice in compliance with regional labour regulations regardless of whether the management approach was centralized or decentralized. However, this cannot be interpreted simply as the constraining effect of institutions on MNC activity (Jackson and Deeg, 2008). Rather the finding points to the significance of regional autonomy in invigorating economic activity (e.g. Phelps, 2000). In the context of this study, high percentage of skilled labour, or high portability of skills in the region, encouraged regional governments to use higher regulatory standards in wages and social benefits, rather than heavy investment in education and training, to leverage MNC activity. It offered opportunities for multinationals to pursue adaptation of compensation within local systems of governance (e.g. Phelps and Tewdwr-Jones, 2001; Almond, 2011). We will now discuss these two configurations in more detail. 23 Configuration 1: MNC decentralization coupled with lax regional labour regulation and low portability of skills in the region is associated with the adaptation of compensation practice In our data set, three cases—EngCoSZ, MacCo and MetalCo—demonstrate the first configuration. For example, EngCoSZ was located in a smaller city, Suzhou adjunct to Shanghai. The differences in living standards and quality of life were remarkable. Shanghai served as a point of reference for job candidates in terms of work conditions. It was evident that Suzhou was in a disadvantageous position. As the R&D Director of EngCoSZ commented, “If you need a manager, an expert, or a specialist, the location is a challenge, difficult.” The HQ standardized salary policy was to apply the median salary level of the cities where subsidiaries were located. This practice apparently underestimated the big difference in median salary between Suzhou and Shanghai. The variations in social standards posed a threat to EngCoSZ in terms of attracting and retaining talents. As a strategic unit of the Finnish MNC, EngCoSZ was responsible for manufacturing, R&D, sales and the supply chain of the MNC. The compensation practice of EngCoSZ was oriented to strategic implementation. The decentralized structure enabled EngCoSZ to adjust HQ compensation in response to local conditions. Apart from benchmarking its salary standard against that of Shanghai, EngCoSZ initiated some local elements. As the business director of EngCoSZ explained: The priority [in our compensation system] is to ensure the skills and competence we need for China operation that is part of our global strategy. We need the best people working for us… Our problem has been how to keep good people. Now we have this kind of employee saving plan just for China [in order to retain these employees ]. … The employee saving plan was a retention scheme for selected key employees based on the assumption that the employee stays with the company for an agreed number of years. EngCoSZ also emphasized coaching, training and career advancement for employees. The performance appraisal was an effective tool for identifying employee needs and for 24 monitoring their personal development. This training practice enhanced the role of rewards in employee retention. MacCo’s challenge came from the location of Tianjin that had lower social standards compared with Shanghai. Both Tianjin and Shanghai are mega cities located in the east coast of China, the most developed region of the country. The relative difference between these two big cities resulted in a higher number of job seekers in Shanghai than in Tianjin. The difference in minimum wage between these two cities was 17 per cent (see Table 2). In addition, MacCo was situated in a district in the outskirts of Tianjin, about 40 kilometres from the central city area. The district was a recent upgrade from an agricultural county to a municipality. The living standards and social life stood at stark contrast to the districts in the city. This intra-city difference also posed a challenge to recruiting managers and engineers. MacCo modified the HQ practice of a pay rise. The yearly salary increase defined by the HQ was perceived by the subsidiaries as too low to meet the local labour market conditions. As the HR manager of MacCo noted, “The head office [in Finland] defined the policy that the salary rise could not exceed the inflation rate of the host country. So, the rise of salary has been some 3-4 per cent in the past.” This increase, followed the official inflation rate (the consumer price index (CPI)), was much lower than the actual salary rise in China. The general manager of MacCo revealed, “The salary has been rising around 10 per cent yearly. In the past, the HQ set the rate of salary rise. For a couple of years, we have been giving a proposal and asking them to approve. They generally agree with our suggested rate.” Additionally, MacCo provided some managers and key professionals with subsidies for using personal cars and trains in commuting to work. The transportation allowance became a local element in the compensation package. The parent company of MacCo had three business areas with two product lines and a service function. MacCo was the strategic unit of Asia among other strategic units in Europe, 25 America and the Pacific. The heads of such units were the members of the top management team at the HQs. While Europe and America were the mature markets, China and wider Asia were the key development targets of the MNC. MacCo was the only international manufacturer of its sector that had the whole value chain— sales, production and maintenance services—locally present in China. The potential of this cutting-edge advantage of MacCo over its competitors could materialize only by building such a value chain. The decentralization in the MNC enabled the subsidiaries to adjust HQ compensation practices accordingly. Configuration 2: High portability of skills in the region coupled with tight regional labour regulation standard is associated with the adaptation of compensation practice Four cases—EquCo1, EquCo2, EngCoSH and PaperCo—located in Shanghai had the highest minimum wage among the five cities in which our case companies were based. The high minimum wage pushed the general salary level upwards. Shanghai was categorized as the first-tier city in China while all its surrounding cities were second-tier or even third-tier cities. The difference in median salary varied between 20 to 50 per cent among these cities. In addition, three of the four subsidiaries operated in sectors where labour turnover was higher than the average of all sectors. In an environment of highly portable skills and high salary standard, keeping pace with the market was crucial in maintaining competitiveness in employee recruitment. EquCo1 and EquCo2’s strategy was to adjust the general pay rise to a level higher than the HQ standard. Salary increase for the managers was higher than that specified by the global policy. As the HR director of EquCo2 commented, “In the highly dynamic and fast changing labour market, we must closely follow the market trend in order to stay competitive in retaining employees.” 26 Apart from adjustments in salary, subsidiaries tried to integrate social benefits into compensation package. The total contribution by employers of social benefits (pension insurance, medical insurance and unemployment insurance, and housing fund in some cities) in Shanghai ranked the highest at the rate of 43.5 per cent of the salary among the five cities in the sample. This level was regarded even higher than the Finnish level which was about 30 per cent. Adapting to the high standards of social benefits did not seem to be a problem for these case subsidiaries. It has to be noted that because the general salary level was much lower in Shanghai than in Finland, the absolute value of social benefits contribution was much less in Shanghai than in Finland albeit the percentage was higher. Social benefits standards not only varied between regions but also stipulated some flexibility within the same region/city. The housing fund was a typical example. For instance, the minimum contribution by enterprises was seven per cent in Shanghai while the maximum was 22 per cent. Companies had considerable room to choose between these two limits taking into account other components in the compensation package. For example, EngCoSH applied the maximum rate of housing fund for its employees. EquCo1 adopted a moderate housing fund rate above the minimum standard. Both companies had commercial insurance and a provision for health checks in the total compensation package. Despite the adaptation of financial rewards, EquCo1 and EquCo2 acknowledged the importance of non-financial rewards as an extension to the compensation system. For example, the regular recreational activities organized at EquCo2 created and maintained close ties between the company and its employees as well as among the employees themselves. The general manager described, “We organize weekend outings outside the city from time to time. First, there are some light programmes [entertainments]. Then, we sit down. People really open their mouths. They have a place to tell about their work, achievements and challenges. They feel that their work is recognized by peers, and they get a feeling of being rewarded.” 27 For EquCo2 the adaptation was to use differentiated rather than standardized position grading methods for the units in China. The China HR manager reported: The business units in China are complicated. A general manager of a small BU [business unit] may be equivalent to a middle manager or even lower manager of a large BU, but the small BU requires a general manager. So, we simply cannot use the standardized position grading system for all the units. At EngCoSH, the emphasis in compensation practice was on achieving a balance between financial rewards and employee wellbeing. The HR manager commented, We cannot always use salary as a retention strategy. We don’t even want to. We keep our salary slightly above the market average. Then, we try to improve employee wellbeing by e.g., providing higher standard housing fund, extended medical insurance and some help to employees who have familial difficulties [which are not mandated by regulations]. Thus, the configurational approach allowed us to analyze institutional and organizational conditions in combination to better explain when foreign subsidiaries of Finnish MNCs in China adapted their compensation practices. We believe that our findings have important theoretical and practical implications to which we will now turn. Discussion and conclusion Previous research on the adaptation of compensation practice within MNCs has typically focused on competing pressures of global integration and local responsiveness. They have discussed institutional/cultural influences (e.g. Festing et al., 2007; Myloni et al., 2004) and mimetic behaviour triggered by dominant institutional logics and competitive pressures (e.g. Björkman and Lu, 2001; Farley et al., 2004). More often than not, these explanatory conditions are studied individually rather than as a configuration. Therefore, we adopted fs/QCA to consider simultaneously a range of institutional and organizational conditions at 28 two levels of analysis to explain when adaptation of compensation practice occurs. Earlier work commonly associates practice adaptation with decentralization. It is assumed that such a management approach to subsidiaries provides them with the necessary autonomy and flexibility to undertake adaptation (e.g. Bartlett and Ghoshal, 1989). However, the configurational analysis of compensation practice indicates that decentralization needs to be considered in combination with the regional regulatory condition to explain adaptation. It facilitates the adaptation of compensation when it co-exists with lax regional regulatory institutions and low portability of skills in the region (configuration 1). These low regional institutional (regulatory and labour market 1) pressures would, in theory, suggest a low likelihood of adaptation. However, their association with the adaptation of compensation practice in our analysis indicates that MNCs, despite low institutional pressures, still adapt their compensation practice to the local context. This adaptation is facilitated by decentralization in the MNC, which is further confirmed by the parsimonious solution term 1 (Appendix 7). Upon inspecting our interviews, we noticed that the variation in salary levels and benefits between cities and provinces in China directly affects the availability of workforce and the size of the local labour pool. Subsidiaries benchmark their salary and benefits against the city with higher standards in order to enhance their image as an employer and to attract talent. A lax regulatory environment does not necessarily suggest an evasion of local institutions or mimicry of HQ practices. Variations in regulations between regions may create normative pressure for the subsidiaries to conform to higher levels of regulations than what is required by the city in which they are located (DiMaggio and Powell, 1983). Regions may operate as ‘residential communities’ where firms are located in a defined geographical space and through their dependence develop a social identity (Freeman and Audia, 2005). The development of 1 This pressure is manifested in the degree to which labour may easily exit the firm without sanctions in the labour market (Aguilera and Jackson, 2003). 29 social identities in regions can exert pressure on other regions to adopt stringent labour regulations. Where these regional identities are strong, pressure on regionally-concentrated firms will be high (Greenwood et al., 2010). Firms under such pressure are inclined to learn from each other on how to become better at what they do or to minimize the competitive risk of losing a market or a source of supply (e.g. Güler et al., 2002). Therefore, our study offers a more holistic explanation of when subsidiaries adapt HQ compensation practice to the local context than what is commonly focused upon in extant research on adaptation of practices due to an emphasis on decentralization. This challenges previous research suggesting that adaptation depends on the management approach of the MNC (Bloom et al., 2003; Taylor et al., 1996). Our second configuration highlights the adaptation of compensation practice within regional governance systems when subsidiaries face tight regional labour regulations and high portability of skills in the region regardless of their management approach. It points to the significance of regional institutions as enablers of adaptation by firms. Although there have been calls to incorporate regional institutions into the study of adaptation of practices within MNCs (Almond, 2011; Almond et al., 2015; Monaghan, 2012), empirical research has remained at the national level (see, e.g. Almond et al., 2006; Festing et al., 2007, 2013; Myloni et al., 2004). We advance the impact of regional influence by uncovering how municipal regulations influence compensation practice. The parsimonious solution term 2 (Appendix 6) indicates regional labour regulations to be a core condition. This further supports the significance of regional institutions within a country. Our finding challenges, to some extent, the basic assumption of much of the research in international business that institutional pressures are coercive pressures that increase transaction costs, and hence constrain firm behaviour (e.g. Buckley and Casson, 1976; Delios and Beamish, 1999). Rather, it lends support to the more recent developments that highlight 30 the dynamic interaction between MNCs and their multiple and often contradictory institutional environments that enables strategic action (Kostova et al., 2008; Saka-Helmhout and Geppert, 2011). Our findings support the significant role of regional institutions in actively shaping inward investment insidership or local embeddedness of MNCs to a foreign market (e.g. Monaghan et al., 2014). Where regional institutional pressures are high, a less empowered subsidiary can also adapt its compensation practice to the local context. The equifinality (same outcome associated with different configuration of conditions) attained in the study suggests that examining different combinations of compensation adaptation mechanisms may be a more fruitful way to advance our understanding than examining net effects (see also McGaughey and de Cieri, 1999). Even works that reconcile the two competing pressures—global integration vs. local responsiveness—on adaptation of compensation overlook the intersection between configuration of firm-level conditions and institutional factors (Pudelko, 2005). They argue for the transformation of practices as they move across ‘space’ (Spicer, 2006), which has implications for the process of legitimating compensation practices in new settings. Institutionalists claim that the process of attaining legitimacy in a host context occurs at nested levels (Deephouse and Suchman, 2008). Hence, organizational conditions that are recognized and understood within widely-held cognitive structures of an institutional environment (Sanders and Tuschke, 2007) need to be considered in conjunction with host country regulatory and normative institutional conditions for an assessment of the legitimation of new practices. Although our study was conducted in China, we believe that the results have implications for other transition economies that share similar characteristics with China in terms of labour market institutions and regional differences such as Vietnam and Eastern Europe. Our chosen method—QCA—offers control over unwanted causal inferences in case comparison. It deals 31 simultaneously with capturing complexity of interactions and attaining causality in patterns of activities for generalizability (Saka-Helmhout, 2011). One limitation of our study is that the interviews solely covered subsidiary managers’ viewpoints. Subsidiary managers might have over-emphasized the role of their unit or might have had a hidden agenda to promote their own interests. We attempted to mitigate this limitation by triangulating the interview data with information from annual reports, company presentations, company websites, internal policy documents and organization charts received from the HQs of the case subsidiaries. Another limitation lies in the small number of conditions included in the model. In order to ensure the validity of our model, we had to consider the ratio of conditions to cases. According to Marx (2006), models drawing on 10 cases must include at most three conditions to reduce to likelihood of fi nding a model at random. Hence, we could examine only three conditions – portability of skills in the region, regional labour regulations and MNC decentralization. Future research could be conducted with a larger number of cases, e.g., around 40 (see Crilly, 2011). This would allow for the inclusion of other relevant factors on the MNC side such as organizational culture and the composition of top management team. The regional institutional level could also include more dimensions such as socio-economic development and level of education. In addition, it would be worthwhile to research further from where foreign subsidiaries draw their resources to respond to strong institutional pressures. Nonetheless, the configurations presented here offer an initial model for future studies. They challenge the common argument that HR practices from less developed institutional contexts would draw upon ‘international best practice’, hence serving as a mere medium of diffusion for hegemonic Western practices (Aguzzoli and Geary, 2014). Our study suggests that, in China, compensation practice is closely associated with employee retention. MNCs should carefully and continuously align their pay standard with the local labour market so as to 32 remain competitive and attractive as an employer. In particular, MNCs that operate in smaller cities tend to face bigger challenges in attracting managers and experts. Foreign subsidiaries in China should be allowed to adjust the HQ pay standards with a high margin. Moreover, salary alone is not a decisive factor for employees to stay or leave. A good compensation package that consists of a balanced combination of various elements such as salary, bonus and benefits is important in China. Once compensation is aligned with opportunities for career development and training as well as work environment, employee commitment tends to be enhanced. Acknowledgement: We would like to thank Claude Rubinson, the editor of COMPASSS WP Series, and the two anonymous reviewers for their constructive comments. REFERENCES Aguilera R, Jackson G (2003) The cross-national diversity of corporate governance: Dimensions and determinants. Academy of Management Review 28(3): 447–465. Almond P (2011) The sub-national embeddedness of international HRM. Human Relations 64(4): 531–551. Almond P, Muller-Camen M, Collings D and Quintanilla J (2006) Pay and performance. In: Almond P and Ferner A (eds) American Multinationals in Europe: Managing Employment Relations Across National Borders. Oxford: Oxford University Press, 119–145. Almond P, Ferner A and Tregaskis O (2015). The changing context of regional governance of FDI in England. European Urban and Regional Studies 22(1): 61-76. Aguzzoli R and Geary J (2014) An ‘emerging challenge’: The employment practices of a Brazilian multinational company in Canada. Human Relations 67(5): 587–609. 33 Ashton D, Brown P and Lauder H (2010). Skills webs and international human resource management: Lessons from a study of the global skill strategies of transnational companies. The International Journal of Human Resource Management 21(6): 836-850. Asuyama, Y (2009) Skill formation systems of China and India. In M, Ohara and K. Kimura, (eds). Interim Report on Comparative Study on Industrial Development Process in China and India. Chiba: Institute of Development Economics Bartlett CA and Ghoshal S (1989) Managing across Borders: The Transnational Solution. Boston: Harvard Business School Press. Beugelsdijk S, McCann P and Mudambi R (2010). Introduction: Place, space and organization —Economic geography and the multinational enterprise. Journal of Economic Geography 10(4): 485-493. Björkman I and Lu Y (2001) Institutionalization and bargaining power explanations of HRM practices in international joint ventures - the case of Chinese-Western joint ventures. Organization Studies 22(3): 491–512. Bloom M, Milkovich G and Mitra A ( 2003) International compensation. Learning from how managers respond to variations in local host contexts. International Journal of Human Resource Management14: 1350-1367. Buckley J and Casson MC (1976) The Future of the Multinational Enterprise. London: Macmillan. China National Bureau of Statistics, 2010. available at http://www.stats.gov.cn/tjsj/tjgb/rkpcgb/dfrkpcgb/index.html Cordeiro, J., He, L., Conyon, M., and Shaw, T. (2013). Informativeness of performance measures and Chinese executive compensation. Asia Pacific Journal of Management, 30(4): 1031–1058. 34 Crilly D (2011) Predicting stakeholder orientation in the multinational enterprise: A mid-range theory. Journal of International Business Studies 42(5): 694–717. Das M and N'Diaye P (2013) The end of cheap labor. Finance & Development 50(2): 34–37. Deephouse DL and Suchman MC (2008) Legitimacy in organizational institutionalism. In: Greenwood R, Oliver C, Sahlin K and Suddaby R (eds), The SAGE Handbook of Organizational Institutionalism. Thousand Oaks CA: SAGE, 49–77. Delios A and Beamish PW (1999) Ownership strategy of Japanese firms: Transaction, institutional and experience influences. Strategic Management Journal 20(10): 915–933. DiMaggio PJ and Powell WW (1983) The iron case revisited: Institutional isomorphism and collective rationality in organizational fields. American Sociological Review 48(2): 147– 160. Dunning JH (2009) Location and the multinational enterprise: A neglected factor? Journal of International Business Studies 40(1): 5-19. Edwards T and Kuruvilla S (2005) International HRM. National business systems, organizational politics and the international division of labour in MNCs. International Journal of Human Resource Management 16(1): 1–21. Edwards T, Edwards P, Ferner A, Marginson P and Tregaskis O (2010) Multinational companies and the diffusion of employment practices from outside the country of origin: Explaining variation across firms. Management International Review 50(5): 613–634. Estevez-Abe, M., Iversen, T. and Soskice, D. (2001). A social protection and the formation of skills: A reinterpretation of the welfare state. In P. A. Hall and D. Soskice (eds), Varieties of Capitalism: The institutional foundations of comparative advantage, 145-183. Oxford: Oxford University Press. Fang, T and Lin, C (2015) Minimum wages and unemployment in China, Institute for the Study of Labour (IZA) Discussion Paper No. 7813, http://ssrn.com/abstract=2370779. 35 Farley JU, Hoenig S and Yang JZ (2004) Key factors influencing HRM practices of overseas subsidiaries in China’s transition economy. International Journal of Human Resource Management 15(4/5): 688–704. Ferner A and Tempel A (2006) Multinationals and national business systems: A ‘power and institutions’ perspective. In: Almond P and Ferner A (eds) American Multinationals in Europe: Managing Employment Relations Across National Borders. Oxford: Oxford University Press, 10–34. Festing M, Eidems J and Royer S (2007) Strategic issues and local constraints in transnational compensation strategies: An analysis of cultural, institutional and political influences. European Management Journal 25(2): 118–131. Festing M and Sahakiants I (2013) Path-dependent evolution of compensation systems in Central and Eastern Europe: A case study of multinational corporation subsidiaries in the Czech Republic, Poland and Hungary. European Management Journal 31(4): 373–389. Finland Chamber of Commerce. 2011. http://kauppakamari.fi/en/2011/08/30/finnish-businesscouncils-in-china/ Fiss P C (2009) Case studies and the configurational analysis of organizational phenomena. In: Ragin C and Byrne D (eds) Handbook of Case Study Methods. London: SAGE, 424– 440. Fiss, P C (2011) Building better causal theories: A fuzzy set approach to typologies in organization research. Academy of Management Journal. 54: 393-420. Fletcher M and Plakoyiannaki E (2011) Case selection in international business: Key issues and common misconceptions. In: Piekkari R and Welch C (eds) Rethinking the Case Study in International Business and Management Research. Cheltenham, UK: Edward Elgar, 171–191. 36 Fuller C and Phelps NA (2004). Multinational enterprises, repeat investment and the role of aftercare services in Wales and Ireland. Regional Studies 38(7): 783-801. Gallagher ME (2004) Time is money, efficiency is life: The transformation of labor relations in China. Studies in Comparative International Development, 39(2): 11–44. Grandori, A and Furnari, S (2013) Configurational analysis and organization design: Towards a theory of structural heterogeneity. In Fiss, PC, Marx, A and Cambre, B (eds) Configurational Theory and Methods in Organizational Research: Research in the Sociology of Organizations, 38: 77-105. Greckhamer, T (2011) Cross-cultural differences in compensation level and inequality across occupations: A set-theoretic analysis. Organization Studies, 32(1), 85-115. Greenwood, R, Díaz, A M, Li, S X and Lorente, J C (2010) The multiplicity of institutional logics and the heterogeneity of organizational responses. Organization Science 21(2): 521-539. Güler I, Guillén MF and Macpherson JM (2002) Global competition, institutions, and the diffusion of organizational practices: The international spread of ISO9000 quality certificates. Administrative Science Quarterly 47(2): 207–232. He, L and Fang, J (2015) Subnational institutional contingencies and executive pay dispersion, Asia Pacific Journal of Management, forthcoming, DOI 10.1007/s10490015-9429-9. HFW Holman Fenwick Willan (2013) China’s future market economy status: Recent developments. Available at: http://www.hfw.com/downloads/HFW-China-future-marketbriefing-November-2013.pdf. Jackson G and Deeg R (2008) Comparing capitalisms: Understanding institutional diversity and its implications for international business. Journal of International Business Studies 39(4): 540–561. 37 Kettunen E, Lintunen J, Lu W and Kosonen R (2008) Suomalaisyritysten Strategiat Kiinan Muuttuvassa Toimintaympäristössä (Strategies of Finnish Companies in the Changing Business Environment of China), Helsinki School of Economics Publications B-98. Kilian HP, Schiller D and Kraas F (2012) Workplace quality and labour turnover in the electronics industry of the Pearl River Delta, China. Zeitschrift für Wirtschaftsgeographie 56(1–2): 58–79. Knight, J and Song, L (2005). Towards a labour market in China. Oxford University Press Kostova T, Roth K and Dacin M (2008) Institutional theory in the study of MNCs: A critique and new directions. Academy of Management Review 33(4): 994–1007. Lu W (2014) Localisation of human resource management practices in China: A qualitative comparative analysis approach. Ph.D. dissertation. Aalto University publication series: Doctoral Dissertations 6/2014. Marín GS (2008) National differences in compensation: The influence of institutional and cultural context. In: Gomez-Mejia LR and Werner S (eds) Global Compensation: Foundations and Perspectives. London and New York: Routledge. Marx A (2006) Towards more robust model specification in QCA results from a methodological experiment. Working paper. COMPASSS WP Series 2006-43. Available at: http:\\www.compasss.org\wpseries\Marx2006.pdf. McGaughey SL and de Cieri H (1999) Reassessment of convergence and divergence dynamics: Implications for international HRM. The International Journal of Human Resource Management 10(2): 235–250. Farrell D and Grant A (2005). Addressing China’s looming talent shortage, McKinsey Global Institute, retrieved on 28.8.2014 from http://www.mckinsey.com/global- themes/china/addressing-chinas-looming-talent-shortage. 38 Monaghan (2012). Attraction and retention of foreign direct investment (FDI): The role of subnational institutions in a small, highly globalised economy. Irish Journal of Management 31(2):45-61. Monaghan S, Gunnigle, P and Lavelle J (2014). ‘Courting the multinational’: Subnational institutional capacity and foreign market insidership. Journal of International Business Studies 45: 131-150. Myloni B, Harzing A-W and Mirza H (2004) Host country specific factors and the transfer of human resource management practices in multinational companies. International Journal of Manpower 25(6): 518–534. Nohria N and Ghoshal S (1997) The Differentiated Network. San Francisco, CA: Jossey-Bass. Peck, J. And Theodore, N. (2007). ’Flexible recession: The temporary staffing industry and mediated work in the United States’, Cambridge Journal of Economics, 31(2): 171192. OECD. 2010. Vocational Education and Training in China: Strengths, Challenges and Policy Options. http://www.oecd.org/education/skills-beyond-school/45494135.pdf People’s Daily Online (2011). China’s higher education students exceed 30 million, retrieved on 28.8.2014 from http://en.people.cn/90001/98649/7315789.html. Phelps NA (2000). The embedded multinational and institutional capture, Area 32: 169-178. Phelps NA and Twedwr-Jones M (2001). Globalisation, regions and the State: Exploring the limitations of economic modernization through inward investment, Urban Studies 38(8): 1253-1272. Prahalad CK and Doz Y (1987) The Multinational Mission: Balancing Local Demands and Global Vision. New York: The Free Press. 39 Pudelko M (2005) Cross-national learning from best practice and the convergence-divergence debate in HRM. The International Journal of Human Resource Management 16(11): 2045–2074. Pudelko MA and Harzing A-W (2007) Country-of-origin, localization or dominance effect? An empirical investigation of HRM practices in foreign subsidiaries. Human Resource Management 46(4): 535–559. Ragin, C C (2006) Set relations in social research: Evaluating their consistency and coverage. Political Analysis 14: 291-310. Ragin CC (2008) Redesigning Social Inquiry: Fuzzy Sets and Beyond. Chicago, IL: University of Chicago Press. Ragin CC (2009) Qualitative comparative analysis using fuzzy sets (fsQCA). In: Rihoux B and Ragin CC (eds) Configurational Comparative Methods: Qualitative Comparative Analysis (QCA) and Related Techniques. Los Angeles, CA: SAGE, 87–121. Ragin CC and Davey S (2014) fs/QCA, Fuzzy-set/qualitative comparative analysis, version 2.5. Tucson, AZ: University of Arizona. Ragin C C and Fiss P C (2008) Net effects versus configurations: An empirical demonstration. In: C C Ragin (ed) Redesigning Social Inquiry, Chicago, IL: University of Chicago Press, 190-212. Rihoux B and Ragin CC (eds) (2009) Configurational Comparative Methods: Qualitative Comparative Analysis (QCA) and Related Techniques. Los Angeles, CA: SAGE. Rosenzweig PM (2006) The dual logics behind international human resource management: Pressures for global integration and local responsiveness. In: Stahl KG and Björkman I (eds) Handbook of Research in International Human Resource Management, Cheltenham, UK: Edward Elgar, 36–48. 40 Saka-Helmhout A (2011) Comparative historical analysis in international management research. In: Piekkari R and Welch C (eds) Rethinking the Case Study in International Business and Management Research. Cheltenham, UK: Edward Elgar, 383–407. Saka-Helmhout A and Geppert M (2011) Different forms of agency and institutional influences within multinational enterprises. Management International Review 51(5): 567–592. Sanders GW and Tuschke A (2007) The adoption of institutionally contested organizational practices: The emergence of stock options in Germany. Academy of Management Journal 50(1): 33–56. Schneider C and Wagemann C (2012) Set-Theoretic Methods for the Social Sciences: A Guide to Qualitative Comparative Analysis. New York: Cambridge University Press. Spicer A (2006) Beyond the convergence-divergence debate: The role of spatial scales in transforming organizational logic. Organization Studies 27(10): 1467–1483. Taylor S, Beechler S and Napier N (1996) Toward an integrated theory of international human resource management. Academy of Management Review 21(4): 959–985. Tosi, H L and Greckhamer, T (2004) Culture and CEO compensation. Organization Science 15(6): 657-670. Whitley R (1999) Divergent Capitalisms: The Social Structuring and Change of Business Systems. Oxford: Oxford University Press. Williams C & van Triest S (2009) The impact of corporate and national cultures on decentralization in multinational corporations. International Business Review 18: 156–1. 41 Appendix 1 Summary of the case companies in China Case Age No. of subsidiary (year) employees 5 4 9 10 7 6 9 10 4 7 60 100 150 600 400 450 80 1650 100 80 Ownership Sector/ industry Location (city) (pseudonyms 1 2 3 4 5 6 7 8 9 10 name) MetalCo EquCo1 CheCo EngCoSZ EquCo2 EngCoSH MacCo ElecCo EngCoBJ PaperCo wholly owned wholly owned wholly owned wholly owned wholly owned wholly owned wholly owned wholly owned wholly owned wholly owned 42 Metal Equipment Chemical Engineering Equipment Engineering Machinery Electronics Engineering Machinery Suzhou Shanghai Wuxi Suzhou Shanghai Shanghai Tianjin Beijing Beijing Shanghai Appendix 2. Higher education attainment in nine coastal cities of China in 2009 Municipality Beijing Shanghai Guangzhou Hangzhou Tianjin Shenzhen Suzhou Wuxi Changzhou Higher education attainment (%) 31.50 1 21.95 2 19.23 3 18.88 4 17.48 5 17.18 6 14.32 7 12.87 8 11.72 9 Mean 18.35 Source: Statistics Yearbook of respective municipalities 43 Appendix 3. Minimum wages of ten coastal cities in China 2008 – 2009 Municipality Minimum wage (Yuan) 1 Shenzhen Special Economic Zone 1000 2 Shanghai 960 3 Hangzhou 960 4 Shenzhe 900 5 Guangzhou 860 6 Suzhou 850 7 Wuxi 850 8 Changzhou 850 9 Tianjin 820 10 Beijing 800 Mean 885 Source: Respective municipal government regulations in 2008 and 2009. Appendix 4. Truth table Portabilit Regional MNC Numbe y of skills labour decentralisatio r of in the regulatio n cases region n 0 0 1 3 Adaptation 1 Cases MetalCo, EngCoS 44 Raw PRI SYM consist. consist. consist 1 1 1 Z, 1 0 0 2 0 MacCo EngCoBJ 1 1 0 2 1 , ElecCo PaperCo, 0.42 0.33 0.33 1 1 1 EngCoS 1 0 0 1 0 1 0 1 0 1 1 0 0 1 1 2 1 0 0 0 1 H EquCo1, 1 1 1 0 EquCo2 CheCo 0.74 0.66 0.66 45 Appendix 5. Necessity test for adaptation Outcome variable: adaptation Conditions tested: Consistency Coverage Portability of skills 0.630915 0.705467 ~ Portability of skills 0.577287 0.845266 Regional labour regulations 0.630915 0.801603 ~ Regional labour regulations 0.473186 0.598802 MNC Decentralisation 0.684543 0.868000 ~MNC Decentralisation 0.471609 0.598000 Appendix 6 Parsimonious solution of adaptation Configuration 1) MNC decentralisation 2) Regional labour regulations Solution coverage: 0.895899 Solution consistency: 0.811429 Raw Unique Consistency coverage 0.684543 0.630915 coverage 0.264984 0.211356 0.868 0.801603 Cases with greater than 0.5 membership in term RLR: PaperCo (1,0.67), EngCoSH (1,0.67), EquCo1 (1,0.67), EquCo2 (1,1) Cases with greater than 0.5 membership in term MNCDecen: MetalCo (1,0.67), EquCo1 (1,0.67), EngCoSZ (0.67,1), EquCo2 (0.67,1), MacCo (0.67,1) 46 Appendix 7 Intermediate solution of negative outcome: non-adaptation Assumptions: ~MNCDecen (absent) raw unique coverage coverage consistency ---------- ---------- ---------~Regional labour regulations*~MNC Decen 0.639344 0.639344 0.78 solution coverage: 0.639344 solution consistency: 0.78 frequency cutoff: 1.000000 consistency cutoff: 0.716738 Cases with greater than 0.5 membership in term ~Regional labour regulations*~MNC Decen: EngCoBJ (1,1), CheCo (0.67,0.67), ElecCo (0.67,0.67) 47 Table 1. Summary of the data Data collection Amount of data Type of data collected method Personal 35 interviews with business HQ compensation practice, the Interviews managers, functional extent to which compensation managers and HR managers, practice was adapted by the ranging from two to four subsidiary, subsidiary autonomy in interviews in each subsidiary decision making, and company background information such as Company 754 pages of annual reports, age, size and ownership. Organizational structure of sampled documents company presentations, MNCs company websites, internal policy documents and Public organizational charts 12 laws, regulations Standards of minimum wage and documents concerning wage and social benefits of the cities remuneration at national and Statistical regional level The 6th Population Census of Data of higher education in the databases People’s Republic of China total population of respective (China National Bureau of regions and cities Statistics, 2010) 48 Table 2. City Calibration of portability of skills in the region Higher education attainment (%) Calibration Case (population with higher education/ total population) ElecCo, Beijing 31.50 1 EngCoBJ PaperCo, EngCoSH , EquCo1, Shanghai Tianjin 21.95 17.48 0.67 0.33 EquCo2 MacCo EngCoSZ, Suzhou Wuxi 14.32 12.87 0.33 0 MetalCo CheCo Table 3. Calibration of regional labour regulation (revised) City Calibration Case PaperCo, EngCoSH, Equco1, Shanghai Suzhou Tianjin Beijing Wuxi 1 0.33 0 0 0.33 Equco2 EngCoSZ, MetalCo MacCo ElecCo, EngCoBJ CheCo Table 4. Calibration of decentralization and illustrative quotations Case Interview quotes and extraction of documentary data Calibration for Our China manufacture is one of our foreign operations in fuzzy- set QCA 1 MetalCo the group company. We run our local operation very independently. The technology development is also based 49 on local markets. For example, here we have a technology team that designs products suited to China. (General CheCo Manager) Our mother company decides these things [investment and 0 market development]. This kind of issues have to be considered at group level. Local subsidiaries modify the design to each country, like we do in China. (Vice PaperCo president, China) In our sector, an investment is a huge sum and long term 0.33 investment issues are centralized. Of course, local units give our voices in e.g. local market prospect and product development. Like us, we make proposal on these for EngCoSZ China. (China Business Director) We have been quite de-centralized, meaning subsidiaries of 0.67 local companies enjoy high degree of freedom in making decisions, business models and types of customers. However, we do have a number of standardized processes [e.g., IT process, employee information system]. (HR EngCoSH Director, Asia Pacific) Our corporation has been quite centralised. For example, 0.33 the issues we decide at China team have to be approved by the HQs (Vice President, HR , China) We have had a lot of expatriates sent by the HQs. My opinion is that two third of them should be replaced by EngCoBJ local Chinese. (Vice President, China) We rely on our group company for a lot of resources, for 0 example, capital for investment and technology. The decisions on these issues are normally taken by the headquarters, like the target country for investment and the EquCo1 core of technology. (China President) The MNC organization had three diverse business lines. 1 Each country offered unique service programme for customers. Business decisions were made at country level. EquCo2 (Company annual report and website) We need local expertise knowledge in our research and development. Although our parent in Finland plays an 50 0.67 important role we cannot have everything done there. We need a kind of flow of knowledge. As for key investments, head office in Finland is the key player, but subsidiaries are ElecCo involved actively in the process. (President, Asia) We do have some flexibility in the kinds of products for 0.33 local market. For example, we have recently adjusted our production lines to other electronics products due to the MacCo market situation. (HR Manager) Our group company has an overall strategy, broad directions and targets. The operational issues are based on local markets. In product and technology development we work together with Finland and other countries. (General Manager) 51 0.67 Table 5. Fuzzy-set membership of conditions and outcome Case Portability of Regional labour MNC decentralization skills in the regulations Outcome: Adaptation region MetalCo CheCo PaperCo EngCoSZ EngCoSH EngCoBJ EquCo1 EquCo2 ElecCo MacCo 0.33 0 0.67 0.33 0.67 1 0.67 0.67 1 0.33 0.33 0 1 0.33 1 0 1 1 0 0.33 1 0 0.33 0.67 0.33 0 1 0.67 0.33 0.67 0.67 0.33 0.67 1 0.67 0 0.67 1 0.33 1 Table 6. Intermediate solutions of the positive outcome: adaptation Assumptions: MNCDecen (present) Configuration Raw coverage 1) MNC decentralisation*~regional 0.369085 regulation*~portability of skills 2) Regional labour regulation*portability of skills 0.369085 Unique coverage 0.212934 Consistency 0.422713 1.000000 1.000000 Solution coverage: 0.791798 Solution consistency: 1.000000 Cases with greater than 0.5 membership in term regional labour regulation*portability of skills: PaperCo (0.67,0.67), EngCoSH (0.67,0.67), EquCo1 (0.67,0.67), EquCo2 (0.67,1) Cases with greater than 0.5 membership in term MNC decentralisation*~regional labour regulations: MetalCo (0.67,0.67), EngCoSZ (0.67,1), MacCo (0.67,1) 52