Approaches to international industrial relations in Chinese Multinational Corporations Management Revue Provided in Cooperation with :

Using a longitudinal case study approach this study examines the approaches to international industrial relations (IIR) in eleven Chinese multinational corporations MNCs. It reveals that the Chinese MNCs adopted an integrative approach to IIR, combining both the home and host country industrial relations (IR) systems. The extent of home-based or host-based IR was dependent on the MNC’s bargaining power, which was determined by the size of the subsidiaries, their abilities to transfer knowledge and technology, and their reliance on the host market. International experience and industry also affected IIR approaches. The practical implications of the findings are discussed.


Introduction
International industrial relations (IIR) are related to international business success (Gunnigle/Collings/Morley 2003). The management of industrial relations (IR) by multinational corporations (MNCs) is challenged by cultural diversity and geographical dispersion (Shen/Edwards 2006). MNC managers cannot assume that they can simply transfer their home IR system to their overseas subsidiaries: political, legal, economic and socio-cultural differences between the home and host countries complicate the transfer (Shen/Edwards/Lee 2005). For example, although the post-war German IR system contributed to the success of the German economy, it put German MNCs at a competitive disadvantage in a global market with diverse national IR regulations and traditions (Streeck 1997). National differences in political, legal, socio-cultural and economic systems produce markedly different IR systems (Dowling/Welch 2004). Thus, the managements of MNCs are required to choose their approaches to their IIR and their choices are influenced by different factors. However, the literature lacks consensus on what these choices are and how they are made.
Although there has been an increased interest in IIR since the early 1990s, it lacks conceptual coherence (Shen and Edwards 2006), and such studies as have been undertaken have focused on Western MNCs rather than those from developing and transitional economies, such as China. Whether the findings on IIR in Western MNCs are generally applicable to non-Western market economies remains a matter of conjecture.
Labour disputes in MNCs are common and the relationship between MNCs and host countries' trade unions are often tense. The dominant role of MNCs in internationalization has led to IIR having strategic importance for MNCs' performance and even their survival in host markets. Consequently, MNCs' investment decisions, such as American MNCs' investment in Europe, are largely based on the character of IR in the host countries and the MNCs' IIR systems (Gunnigle et al. 2003).
A strong desire for natural resources and growing competition at home has resulted in an increasing number of Chinese companies expanding overseas to more than 2000 in 2004 (Shen/Edwards 2006). By the end of 2003, China had invested US $33.2 billion in establishing 7,470 non-financial enterprises in more than 160 countries (China Development Bank 2004). Since economic reform in the late 1970s, the scope of Chinese overseas investment has widened from trading, shipping and catering to processing, manufacturing, mining, engineering, farming, and research and development. Although most Chinese MNCs are small and not influential, the accession of China to the World Trade Organization (WTO) in December 2001 has led to further significant Chinese investment abroad. As a result, more and more Chinese companies are going oversea and growing fast. IBM's Institute for Business Value recently identified 60 Chinese companies as potential global players in the next decade (IBM 2006). Success in international business will require Chinese MNCs to develop effective IIR. To this end, an understanding of how Chinese MNCs manage IIR and what factors determine Chinese MNCs' decisions on IIR policies and practices is necessary. Such an understanding will serve to build a theory of IIR in small and medium-sized, newly-internationalized MNCs from developing economies. This study, therefore, poses two research questions: Research question 1: What are the approaches of Chinese MNCs to the management of IIR? Research question 2: What are the factors associated with the approaches to IIR of Chinese MNCs? The locus of decision-making on IIR, trade unionism, collective bargaining, labor relations, and labor dispute resolution are the predominant issues in the IIR literature and have been the focus of the past IIR studies, such as Dowling andWelch (2004), Hamil (1984), Gunnigle et al. (2003), Mueller and Purcell (1992), Shen and Edwards (2006) and Veersma (1995). In a similar vein this study examines these important IIR policies and practices in order to identify Chinese MNCs' approaches to IIR. This introduction has defined the research objectives. The remainder of the paper is structured as follows: a review of the literature on the IR system in China; a discussion of the literature on IIR approaches and their determinants; a description of the research methodology; a report of the research findings; and a conclusion that revisits the research questions and discusses the implications for practitioners of the findings.

The IR system in China
Economic reform has brought about a dramatic transformation of China's IR system. Under the planned economy , the State rather than the enterprise was the employer, and therefore IR was between employees and the State as manifested by the Government under the control of the Chinese Communist Party (CCP). IR was unproblematic so long as there were an 'iron-rice-bowl' (tie fan wan), an 'iron wage' (tie gong zi) and workplace 'cradle-to-grave' welfarism . Labour disputes were rare and resolved by politics and administration.
The control of trade unions was through the All-China Federation of Trade Unions (ACFTU) that linked the CCP to the masses (Zhu 1995). Union membership was almost 100 per cent (Ding, Goodall and Warner 2002). The functions of the trade unions were to mobilize workers, organize production campaigns, promote enterprise economic goals, allocate welfare benefits, and exercise some responsibility for training (Ng and Warner 1998).
Enterprise reforms (such as the "Fourteen Rights of Enterprise Managers" issued by the State Council in 1992) shifted a range of 'control rights' from the State to enterprises. The former 'workers-as-masters' status has been replaced by a managementemployee relationship, and management power has expanded considerably with little concern for workers' rights (Zhu/Warner 2005). As a result, harmonious labour relations have been replaced by employer-employee tensions, labour disputes and unequal employment relations (Chan 1998;. The increasing incidence of industrial conflict is largely due to the absence of free collective bargaining and independent trade unions (Shen 2007). Constitutionally, trade unions (i.e. the ACFTU and its affiliates) represent workers and protect their interests, and the ACFTU has had some influence on national policy-making, for example, helping to shape the 1994 Labour Law (Baek 2000;Chen/Chan 2004;Ding/ Warner 1999;Taylor 2000;White 1996), but the ACFTU has continued to act as a transmission belt from the CCP to the workforce. Many Chinese trade unions at the enterprise level have been established by employers and thereby cannot pursue labour management revue, volume 18, issue 4, 2007 413 grievances independently. Employers and the government appoint the most trade union cadres, some from the ranks of former senior CCP or government officials. Not surprisingly the unionization rate has fallen, as low as 4 per cent in privately-owned enterprises (POEs) (Chang 2000;Shen, in press).
Constitutionally, Chinese workers have rights of freedom of assembly, of organization and to demonstrate, but there is no a right to strike although strikes are not prohibited. Permission to strike is never granted and a request normally results in numerous negotiations and the intervention of the government. A complex labour dispute settlement mechanism, through which the number of disputes settled has been growing rapidly, involves local government, union councils, tripartite consultation, arbitration and litigation.

IIR approaches and their determinants
IR deals with the relationship between management and workers (Ackers/Wilkinson 2003;Salamon 2000), and is often defined as unionized employment relations in which collective bargaining is the main means of determining terms and conditions of employment (Kaufman 2001). IIR refers IR in the overseas operations of MNCs. The major IR issues in an MNC's home operations are also found in the MNC's overseas subsidiaries (Dowling/Welch 2004;Gunnigle/Collings/Morley 2003;Shen/Edwards 2006). However, IIR is complicated by differences between the political, legal, economic and socio-cultural systems in the home and host countries and between host countries. IIR may also involve some issues that may not occur in home operations, e.g., the locus of decision-making, which indicates the extent to which IIR is centralized or decentralized, or is influenced by the home IR system (Hamill 1984;Shen/ Edward 2006;Veersma (1995). The transfer of the home IR system to overseas subsidiaries is a major concern of MNCs management (Ferner 1997;Royle 2002;Shen/Edward 2006;Tuselmann/McDonald/Heise 2002). The literature distinguishes the home-based and the host-based approaches to the management of IIR. International human resource management (IHRM) is concerned with issues, functions, policies and practices that result from the strategic activities of multinational enterprises and impact on the international concerns and goals of enterprises (Schuler/ Dowling/De Cieri 1993). Although often studied separately, IHRM generically covers IIR (see, for example, Brewster/Scullion;Dowling/Welch;2004;Schuler/ Budhwar/Florkowski;2002;Shen/Edwards 2006) A substantial amount of the literature deals with the country of origin effect on MNCs' IIR policies and practices (Ferner 1997;Royle 2002;Shen/Edward 2006;Tuselmann et al. 2002). Royle (2002) claimed that the majority of MNCs tended to adopt home-based IR. For example, the MNCs from Western Europe and North America operating in Germany were unlikely to support workers' councils, showing little regard for German institutional arrangements. An advantage of the home-based approach is that the MNC's familiarity with the home IR system enables it to better handle its IIR. Another is that the MNC can exercise control over IIR, and coordinate its overseas operations.
On the other hand, because it is designed for particular domestic institutional arrangements, the home-based IR system may not work with the host country's quite different institutional arrangements. Although scholars have observed that MNCs' IIR is primarily shaped by the host country experience and to varying degrees adapted to the host environment (Gunnigle et al. 2003;Tuselmann et al. 2002), increasingly MNCs are aware of the need for flexibility in adapting to local experience. A complete host-based IR approach may ensure acceptance by employees, unions and other stakeholders, but the MNC risks losing control over its IIR and may find it difficult to retain its overall organisational culture. Indeed, institutional and cultural varieties make it highly unlikely for an MNC to maintain either an entirely home-based or a wholly host-based IIR.
The IIR literature only conceptualises the home-based/host-based dichotomy, although some researchers recognise MNCs' employment of an integrative IHRM approach that is influenced by, but different from, home-based and host-based systems (Schuler/Dowling/De Cieri 1993;Shen/Edward/Lee 2005;Shen/Edwards 2006;Taylor/Beechler/Napier 1996). An integrative IHRM approach best fits MNCs' internal and external environments and reflects the effects of both the country of origin and the host country. As IIR is included in generic IHRM, it is therefore likely that MNCs adopt an integrative IIR approach.
Under the integrative IIR approach the transfer of IR overseas and its adaptation to local practices are contingent on a range of firm-specific and host-contextual factors. One is senior management's preference (Ferner 2000;Hamill, 1984;Hayden/ Edwards 2001;Taylor et al. 1996), which may change from a home-based to more host-based as it gains international experience. Another is the MNC's bargaining power within the host IR systems (Gunnigle et al. 2003). The MNC's bargaining power derives from its importance to the host country, including its contribution to the host economy and to its capacity to transfer knowledge and technology. If the MNC is important to the host country, it will be able to challenge host country IR arrangements for union recognition and collective bargaining. Wal-Mart, for example, threatened to relocate from China if it was required to set up enterprise unions in its operations there (www.aboluowang.com 2006). In the event, the lure of China's huge market and potential competition from other MNCs weakened Wall-Mart's power to insist on union-free workplaces (Shen/Edwards 2006). Elsewhere, such as in Ireland, anti-union American MNCs have successfully resisted the unionisation of their employment relations (Gunnigle et al. 2003).
An MNC's bargaining power is also a function of the size of its international operations and its degree of reliance on the host market. Large MNCs usually contribute significantly to the local economy and have stronger bargaining power but an MNC that is heavily reliant on the host market usually has weaker bargaining power. The contribution of the MNC is specifically related to its ability to transfer know-how and technology to local enterprises. An MNC from a mature market entering a growing market is expected to transfer technology and knowledge to local enterprises. This may give it an advantage in bargaining on IR issues. Also, when the overseas subsidiary is important to the whole company the headquarters may get directly involved in its IIR (Hamill 1984).
The political pressure from host governments and IR legislation and regulations determine much of the character of IR in modern states, and the basic framework management revue, volume 18, issue 4, 2007 415 within which MNCs practice IIR (Bean 1985;Rosenzweig/Noria 1994), but the deregulation of labour markets in many industrialised countries over the past three decades has provided MNCs with greater freedom to choose to practice their own IIR (Tuselmann et al. 2002). However, they often comply with the host country's IR tradition by mimicking host companies' practices in order to be responsive. For example, German MNCs operating in the United Kingdom tend to adopt the local approach to IR (Ferner/Varul 2000).
The effect of the host economic factor is mainly embedded in labour standards, including wages and working conditions that MNCs set for host country nationals. Many MNCs may provide low wages and maintain poor working conditions in developing nations, although they might be expected to be exemplary in setting high labour standards and demonstrating social responsibility in host environments. Failure to meet expectations in the host-country is often a major cause of labour disputes. The following framework depicts the relationship between IIR and the major determinants.

Figure 1: The Integrative Approach to International Industrial Relations
Methodology Ferner (1997) argued for a phenomenological approach to the exploration of the dynamics of organizational micro politics and the constraints within which the firm operates; to the evaluation of the influence of historical legacies; to the assessment of the subtle interactions between structure, strategy, and national and corporate culture. To investigate these phenomena and provide an interpretive understanding, a field study Home industrial relations  Welch (1994) believed that if the nature of the study requires a reliance on human memory or recall, it is important to use a corroboratory mode of data collection. Structured or semi-structured interviews with different people would yield different perceptions of the same phenomena, as well as possibly providing multiple measures of them. Longitudinal studies provide data about individuals and organisations at different points in time, allowing the researcher to track changes. This study uses a longitudinal qualitative approach in order to explore the changing pattern of IIR and the underlying causes. Semi-structured interviews involving eleven Chinese MNCs were conducted in 2001 and 2006. The reason for choosing this time period is that the first five years of the 21 st century have witnessed considerable changes in IR in China and rapid expansion of Chinese outward investments. An analysis of these changes may enable us to understand the causes of the changes in the IIR in growing Chinese MNCs.

Firm-specific factors
The case companies cover a range of industries, and differ in size, number of employees, economic ownership and history, thereby enabling an examination of the influence of different firm-specific factors on their IIR. They are largely representative in terms of the international strategy, industry pattern, international experience and the scale of international operations of Chinese outward direct investments. While the case Chinese MNCs directly invested in both developing and developed countries the investments of ElectronicsCo1 and ElectronicsCo2 were concentrated in developed economies, and that of SteelCo and MinmetalsCo in developing countries.
BankCo's overseas subsidiaries were all retail banks. AirlineCo directly employed and managed local employees rather than used sales agents. Seven case companies BankCo, AirlineCo, InstrumentTradeCo, PlantCo, MinmetalsCo, HealthCo and Steel-Co wholly owned their overseas subsidiaries. All overseas subsidiaries of the two manufacturing companies, ElectronicsCo1 and ElectronicsCo2, were joint-ventures (JVs). TechnicalTradeCo, and OceanShippingCo had established both wholly-owned enterprises and JVs. In all the overseas JVs the Chinese MNEs had majority ownership. The number of overseas subsidiaries and employees in the overseas subsidiaries changed over the study period. The case companies are profiled in table 1.
Thirty interviews in ten MNCs were conducted between March and November in 2001. The interviews involved ten HR/deputy HR managers and ten general/deputy general managers with responsibility for international HR/IR at headquarters and ten executive managers in the UK subsidiaries. The HR mangers were asked about overall organisational IIR policies and practices and executive mangers were asked about IR in specific overseas subsidiaries. To assess the extent of the change in IIR since 2001 another ten interviews were conducted in June 2006. HealthCo, one of the original companies, declined to participate in 2006 study and was replaced by SteelCo, the largest iron producer in China. SteelCo has acquired a copper mine and an iron mine, and established beneficiation plants in South America and Africa in the late 1990s.

The findings The locus of decision-making on and significance of IIR
By November 2001 the case Chinese MNCs had not been seriously concerned with nor paid much attention to IIR in their overseas subsidiaries. The managers interviewed commonly stated that they had not had major IIR-related troubles in their overseas subsidiaries, and that, 'IR issues have never become our major concern in our home operation nor have IIR issues in our overseas operations'; 'IR issues in host countries have not affected our decisions on overseas investment'. The 2001 interviews had revealed that there was no formal policy at either corporate or subsidiary levels concerning unionism, collective bargaining, labor standards and labor dispute resolution. The 2006 interviews, however, revealed that there have been significant changes in management's attitudes toward IIR. The managers from SteelCo, Minmet-alsCo, OceanShippingCo, ElectronicsCo1 and ElectronicsCo2, which had overseas factories or refineries, all commented that they had not expected local unions and employees in some host countries to have been so difficult to deal with. As a result of the difficulties, IIR had become their one big concern and they were now cautious about choosing the overseas investment destinations. Both SteelCo and MinmetalsCo were interested in setting up factories in India, but had abandoned their interest when they learnt that Indian unions could and did organize strikes, quite frequently. Minmetal-sCo's African copper refineries had experienced strikes since 2001, as had SteelCo's copper refinery in South America. The HR managers from these two companies commented that they had experienced great difficulties in settling these disputes. The two companies had considered withdrawing investment from the two particular nations if the host unions did not accept their offers. In the words of the HR Manager of SteelCo, 'Our offers were all on the table. If these (offers) still couldn't satisfy them (the host unions and local employees) we had to sell off our refinery'. The Executive Managers in BankCo and PlantCo had also been troubled with labor disputes and management revue, volume 18, issue 4, 2007 419 were ready to review their choices of host nations and be more careful when dealing with labor issues in the future. The 2001 interviews revealed that general and HR managers at the HQs allowed subsidiary managers substantial autonomy to manage IIR: 'We do not dictate subsidiaries how to deal with trade unions and labour issues. We have not heard of a problem about IIR so far'. All overseas executive managers emphasized that they hoped HQs could help them with the IIR issues but did not think HQ understood what was happening overseas. However, the 2006 interviews revealed that there had been a move towards centralisation of decision making on IIR during the study period. All the case companies, except InstrumentTradeCo, OceanShippingCo and PlantCo, had requested subsidiaries to report any new development in dealing with local unions, particularly with collective labour contracts. The managers at the HQs of five of the firms BankCo, MinmetalsCo, SteelCo, ElectronicsCo1 and ElectronicsCo2 were directly involved in negotiations with host unions over wages and labour dispute resolution. When asked whether the importance of overseas subsidiaries affected HQs involvement in decision-making on IIR issues, interviewees typically replied, 'It could be, but all overseas subsidiaries are important even when they are small or not making profit. We pay equal attention to all subsidiaries.'

Union recognition and collective bargaining
None of the managers in the 2001 interviews were opposed to unionism in their overseas subsidiaries. A typical response from the general and HR managers was, 'We are not against host unions' recruitment campaigns as we are in our home operations'. The executive manager at BankCo said, 'We observe other MNCs and local firms and do the same as they do'. Similar statements were made by the other nine executives. In the 2001 interviews most managers commented that host country unions did not seem to be a bad thing and they did not have a problem with them. Unionization rates were generally high in the case Chinese MNCs' overseas subsidiaries: 185 out of 293 subsidiaries (63 per cent) were unionised in September 2001.
The senior managers' perceptions of host country trade unions changed during the study period. In 2006, the HR managers at ElectronicsCo1, BankCo, Minmetal-sCo, ElectronicsCo2, SteelCo and PlantCo saw host country unions as troublemakers, and the HR mangers in other firms regarded them as potential troublemakers. Even so, although the 2006 interviews showed that the managers had come to realise how different host country trade unions were from Chinese trade unions, they were not willing to take an anti-union stance. They felt 'obliged to recognise unions [and our] sentiment towards home unions may play some role in that'. Nevertheless, the 2006 interviewees were not against unionism but had become uncooperative with host unions' recruitment campaigns. Unionization rates had dropped from 63 per cent to 59 per cent between 2001 and 2005, during which period 33 new subsidiaries had been established, 12 in manufacturing, of which nine had recognized trade unions. (As noted above, HealthCo was not included in the 2006 study and SteelCo was not included in 2001). Two circumstances had not changed during the study period: parentcountry nationals (PCNs) working overseas could only join Chinese workers' con-gresses under the leadership of the ACFTU, and host-country nationals (HCNs) were free to join local unions.
In 2001, of the case MNCs only ElectronicsCo1 had signed collective agreements (at its North American factories) on wages. Wages and working hours in the other case MNCs had been determined unilaterally by their managements in compliance with host country labour regulations and where unionized with the host unions' consent. By June 2006, ElectronicsCo1, BankCo, MinmetalsCo, ElectronicsCo2, SteelCo and PlantCo had all signed collective contracts on wages in their overseas manufactories. The contracts in resource-based MinmetalsCo and SteelCo were mainly between the subcontractors and trade unions, and modified twice between 2004 and 2006. After the interviews in June 2006 the researcher was informed that the collective contract in SteelCo's South American copper refinery had been modified for the third time by increasing wages following industrial action in August 2006. The 2006 interviews indicated that successful collective bargaining had played a pivotal role in maintaining sound industrial relations in the subsidiaries of the case Chinese MNCs, but a failure to reach collective agreements often became a cause of labour disputes.

Labour relations and labour dispute resolution
The 2001 interviews revealed that the majority of the Chinese MNCs had maintained cordial IR in their overseas operations. Only one individual labour dispute was reported, at BankCo's London subsidiary. A HCN middle manager alleged that she had been deprived of a deserved promotion and had brought the case to an employment tribunal in 2001. Management however claimed that she did not deserve promotion. The rare occurrence of labour disputes was attributed by the managers to effective communication and social contacts with host trade unions, and their compliance with local labour regulations, including minimum wages, paid holidays, and diversity management. The following statements by the executive managers were typical: There has been no big problem so far. We have treated HCNs well. We have complied with local employment regulations, such as working hours, minimum wages, health and safety, employers' contributions to superannuation and paid holidays. When there were isolated cases of individual complaints we dealt with them very carefully. We respect unions' demands and communicate with them well. We have been very serious about HCN employees' complaints and suggestions. We have acted on them without delay because we do not want to be accused of violating workers' rights in other countries. We have made efforts to communicate well with HCN employees and unions.
Chinese MNCs usually paid HCNs at rates higher than the local market average in developed nations in order to compensate for a lack of training and management development opportunities (Shen/Darby 2006). Competitive wages became a major source of attraction of Chinese MNCs to HCNs in those countries. Even though Chinese MNCs were weak in training provision, management development and employment participation and involvement (Shen/Edwards 2006), these soft aspects rarely became the cause of labour disputes.
However, the interviewees in 2001 generally admitted that they did not know how to handle it if there were to be industrial action place in their overseas operations. As the HR Manager of BankCo speculated: What we can do, more likely, is to compromise by meeting the demands of employees or unions, and conforming to local and international labour standards to avoid any trouble to happen. Otherwise, our alternative choice is to leave the trouble land.
The Executive Manager of HealthCo echoed this with: 'We won't make big mistakes if we comply with local labour regulations'. These views were iterated by the majority of the managers interviewed in 2001, reflecting the manner in which Chinese MNCs dealt with labour disputes.
Since 2001, there have been eight collective labour disputes, involving Electro-nicsCo1, ElectronicsCo2, BankCo, MinmetalsCo, SteelCo and PlantCo, the manufacturing, resource refining and retail banking MNCs, over wages, working conditions and dismissals. While negotiations between host unions and enterprise management were smooth in BankCo and PlantCo, host unions did threaten to strike in the two electronics firms, which eventually met the union demands. On the other hand, negotiations did not go so well at MinmetalsCo and SteelCo, where both paid above average wages but which were lowered by contractors' deductions. Consequently large scale, violent strikes occurred at MinmetalsCo's African subsidiaries in 2004 and 2005. In 2006, SteelCo's subsidiaries in South America experienced two strikes over pay and dismissals.
The disputes at SteelCo and MinmetalsCo went to local labour courts, where they were declared unlawful, and the companies required to pay minimal increases in wages. Both SteelCo and MinmetalsCo intended to sell off the refineries where disputes occurred. The HR Manager at SteelCo lamented: 'It was really out of our control. We paid HCNs above the local standard. But, they always expected too much'. His lament was shared by the HR Manager at MinmetalsCo, and the HR managers in other firms speculated that the potential for aggressive labour disputation would result in their shifting their investments elsewhere.

The approach to IIR and the determinants
The data show that the IIR policies and practices in the case Chinese MNCs were a combination of both the home and host IR systems an integrative approach, mainly the result of the interaction of the country of origin effect and the host country effect. In particular, the senior management's attitudes toward host unions, including host union recognition and unions' involvement in settling employment terms, were largely influenced by the home IR system (the country of origin effect). Chinese MNCs were generally not against unionism but unwilling to involve host unions in setting labour terms in their overseas subsidiaries. In addition, labour standards were to a certain extent influenced by the low labour standards in China. As a result, the overall labour standards in Chinese MNCs had not been better than those of their competitors in the same host countries. It becomes evident that Chinese MNCs were not readily accepting global social responsibilities. IIR had not been a major factor affecting the decisions of the HQs of Chinese MNCs on overseas investment and, until a few years ago, this remained the responsibility of the managers in the subsidiaries, a result of organ-izational inertia and, as IR and trade unions in their domestic operations do not have strategic implications, e.g. setting labour terms above minima the country of origin effect.
Collective bargaining and labour dispute settlement were mainly host-based. Although the case Chinese MNCs were not willing to bargain with unions at home, the majority of them did overseas. Political and administrative mechanisms play an important role in labour dispute settlement in China (Shen 2007), but overseas labour arbitration and litigation were the main means. These host-based IIR practices had resulted from observation of local practices, compliance with host IR legislation, labour traditions, and political compromise with the host authorities. While the growing incidence of labour disputes was commensurate with the rapid expansion of Chinese MNCs, it also resulted from the localised IR policies. Chinese MNCs tended to localize employment conditions, particularly wages, working conditions, and labour disputes settlement. Localization had led to higher wages and better working conditions in developed countries but lower labour standards in developing economies. Such an IIR approach had caused problems in South America and Africa, where there was an expectation of MNCs providing better wages and working conditions.
If they were to compete successfully, the case Chinese MNCs were unable to deviate from local IR systems, especially in developed economies (Gunnigle et al. 2003). Many of the managers of the case Chinese MNCs said that they did not have the power that some world market leaders had to impose their preferred IR systems on overseas subsidiaries. This was partly attributed to the case Chinese MNCs' heavy reliance on host markets and their inability to transfer knowledge and technology to host enterprises. Of the case Chinese MNCs, the resource-based were in the weakest position because they could not easily move elsewhere. The two electronics MNCs. BankCo, AirlineCo and OceanShippingCo also did not have much room to manoeuvre as they were highly reliant on international markets. Interviewees thought that Chinese MNCs in general were not strong on knowledge and technology transfer as they put it, 'our advantage is low price not technology', a competitive advantage not an IR bargaining one.
The data show that collective labour disputes took place mainly in manufacturing and resource refineries where there was a high density of blue collar workers. Because labour standards in the case Chinese MNCs were partly influenced by the home standards, blue collar workers were more likely to be employed under conditions from which disputes arose.
This longitudinal study reveals considerable changes in senior management's perception of IIR significance, host unions and labour relations between 2001 and 2006. Some of these changes are attributable to the managers' growing international experience. Most Chinese MNCs started overseas operations after China adopted the 'Open Door' policy in the late 1970s. The knowledge of IR of many managers in the early stage of their international businesses was confined to Chinese IR. Throughout the 1990s, Chinese trade unions remained an agency of the CCP and a part of enterprise management. Although labour disputes in China increased during the 1990s, employers were not seriously concerned as the advantage remained with them. However, the 2006 interviews indicated that in the first five years of the 21 st century, labour disputes management revue, volume 18, issue 4, 2007 423 in overseas operations had strategy implications for Chinese MNCs, e.g. on the choice of outward investment sites. As international experience is not only related to years of operations but also to modes of operation (Welch 1994): a manufacturing MNC is likely to have more international experience in dealing with labour relations than a trading MNC. The perceptions of managers in ElectronicsCo1, ElectronicsCo2, BankCo, MinmetalsCo, SteelCo and PlantCo, six case MNCs that all experienced collective labour actions in their overseas operations, had changed significantly. These were manufacturing, resource refineries and retail banking MNCs that employed a large number of HCNs, many of whom were blue collar workers.

Conclusions
This study develops a framework of the integrative approach to IIR and tests the framework with Chinese MNCs. It shows that the Chinese MNCs adopted an IIR approach that integrated home and host IR systems and gave effect to each. The empirical evidence shows that there had been considerable changes in the IIR policies and practices of the case Chinese MNCs during the study period: a centralisation of decision-making on IIR; a decrease in unionisation rates; an increase in collective bargaining; an increase in labour disputation. These changes, however, did not change the integrative nature of the MNCs' approaches to IIR. The findings of this study, hence, contribute to the international business literature by revealing that MNCs tend neither to adopt wholly home-based nor host-based IR systems, as suggested by Ferner (1997), Gunnigle et al. (2003), Royle (2002) and Tuselmann et al. (2002). Instead, they are likely to adopt an integrative approach that combines both the home and host IR systems.
The degree to which IIR is home-based or host-based is determined by the country of origin effect, the host country effect and a range of firm-specific factors. The country of origin effect was illustrated by senior management's attitudes toward unionism and collective bargaining. The host country effect was by political pressure from local authorities, host legislation and regulation, and IR traditions. The interaction of the home and host effects was typically found in the labor standards in the Chinese MNCs' overseas subsidiaries. On the one hand, the Chinese MNCs tended to localize labor standards, e.g. high wages and good working conditions in developed countries, and low wages and poor working conditions in developing countries. On the other hand, labor standards in subsidiaries were influenced by the home country's low level labor standards. This, and the fact that labor standards in the case Chinese MNCs were generally lower than their western counterparts, explains why most collective labor disputes occurred in the developing countries. Even so, disputation was lower than in the home country (Shen 2007).
The ability of the case Chinese MNCs to alter IR systems in host countries was a function of their bargaining power, which was in turn affected by their importance to and influence on host economies, to their ability to transfer knowledge and technology to host enterprises, and to the degree of their reliance on host markets. The Chinese MNCs were smaller and less influential than their Western counterparts, and less advanced in the use of knowledge and technology. Most of them were also heavily reliant on international markets. Consequently, their bargaining power was weak, and they usually had to comply fully with host IR regulations and traditions. The incidence of labour disputes was related to type of industry, each with a different level of labour density and labour conditions. The changes in the MNCs' senior management's perception of IIR significance, host unions, and labor relations over the study period were attributable to the managers' growing international experience.
IIR had not been a major factor affecting the decisions of headquarters of the case Chinese MNCs on overseas investment and until a few years ago it remained the prerogative of subsidiary managers. There had been a centralization of decisionmaking on IIR over the study period. For example, the HQs started exercising more control over IR issues, particularly collective labor contracts and labor dispute resolution. However, given undetected changes in management's perceived importance of overseas subsidiaries this study has been unable to confirm whether the importance of overseas subsidiaries determines the degree of HQs' involvement in IIR. The underlying cause of such centralisation was the growth in the number of labor disputes that gave rise to the significance of IIR matters at HQs. This study has significant implications for managers in Chinese MNCs as well as MNCs of other nationalities. First, although unwillingly, the case Chinese MNCs adopted host-based unionism, collective bargaining and labor dispute resolution practices. It appears that the country of origin effect on these IIR practices was normally overridden by the host country effect if the former is anti-localization. Therefore, MNCs should be prepared to formulate IR policies that comply with host country legislation and traditions, especially if they do not have strong bargaining power. Second, it is inappropriate for MNCs to localize labor standards as such practices are unlikely to meet global social responsibilities by their maintaining low labor standards in developing countries. Chinese MNCs are expected to grow rapidly in the future. Developing IIR policies for their current, interests in Africa and South America that fulfill global social responsibility is necessary for them to become sustainable and competitive global players.