Transferencia y cambio de sistema de trabajo a través de las fronteras: una empresa danesa va a Brasil

Based on qualitative research, this paper explores the transference and change of work systems across national borders in a Danish Multinational Corporation (MNC). A qualitative longitudinal study was made of two factories belonging to the same MNC and the processes of transforming the work system were analyzed. The paper presents a case that has not been subject to a great deal of research: the problems faced by a Scandinavian company when internationalizing into a quite distinct business system. The main object of analysis was the negotiation of the changes in the work systems at the two sites. The host country institutions and worker/manager traditions are shown to have a strong influence on the way  the work system was translated and negotiated. The organizational actors had an understanding of quite different national institutions and made it difficult to transfer the work system from Denmark to Brazil.


Resumen
Este artículo explora la transferencia y el cambio de sistemas de trabajo transnacional en una Empresa Multi-Nacional (EMN) de Dinamarca a una filial en Brasil. Es un caso no muy estudiado de una empresa escandinava que se internacionaliza en un sistema marcadamente diferente. El estudio longitudinal recoge el caso de dos factorías pertenecientes a la misma EMN en donde se estudiaron los procesos de transformación del sistema de trabajo. Se analizan las negociaciones en torno a los cambios en el sistema de organización del trabajo de los dos centros de producción. Las instituciones y las tradiciones de las relaciones entre empleado y directivo del país receptor tienen una fuerte influencia en la forma en que el sistema de trabajo fue adoptado y negociado. La diferente interpretación de las instituciones nacionales por parte de los actores organizativos hizo difícil la implantación del sistema de trabajo de Dinamarca a Brasil.

Introduction
It has been suggested that multinational corporations (MNCs) are becoming increasingly footloose and less country-bound, and at the same time more organisationally integrated than their parents, the international companies. MNCs are said to have substantial direct investments in foreign countries and actively manage these operations as integral parts of the company, strategically and organisationally, thus becoming transnational corporations (Bartllet and Goshal, 1989). As a result, the strategies of firms in one market depend on those in others, and firm-specific advantages, developed in the domestic economy, can be extended to other economies (Whitley, 2001). This integration and interdependence of economic activities creates a need for competencies and practices to be continuously transferred from one business environment to another. Considering the ferocity of international competition, it is only to be expected that employers adopt the most efficient techniques of production and use them in the most efficient way, regardless of national institutions. In factories producing the same product for the same markets with similar technology, work should be organised in the same way. If this were to be the case, firms should present similar structures and forms of control. However, as Whitley (1992: 277) puts it, 'the extent to which a transfer of business system characteristics occurs does, of course, depend on the openness of dominant institutions in those economies to novel forms of economic organization and the relative strength of business systems in them. ' Little attention has been paid to how new organizational practices are 'translated' and negotiated by different organizational actors when organizations cross national borders (Geppert and Williams, 2003). This article aims to contribute to fill this gap by examining the process of transferring the work system in a Danish multinational from the headquarters to a Brazilian subsidiary, and focusing on how the work system changes when it crosses national borders. By exploring the development of MNCs in Brazil, this research contributes to the debate on transference and change in work systems across borders. It focuses on a Scandinavian corporation operating in Brazil. Special attention is paid to the changing dynamics of the original work system that was transferred by the local agency within the framework of the host country' s traditions and institutions. When discussing the successful transference of work systems across national borders, the literature rarely looks at the social processes that are set in motion when attempts are made to transfer a new organizational model or work system from one environment to another. The article explores how different work and ocupational groups organise themselves to affect, resist or change the fundamental characteristics of work systems. RIO, Nº 4, 2010 Transference and change of work system across borders. a Danish company goes to Brazil The article is organized as follows. In the second section, it looks at MNCs and the transference of work systems across national borders from the perspective of the two institutional environments involved. In the third section, the methodology of the research is presented. The case study presents how work was organized in the two factories and compares the two work systems. It is shown that the work system radically changes in Brazil. By dividing the empirical material in this way, the negotiation processes are shown to affect organizational interdependencies. In the forth section the findings are presented. Finally, section fifth discusses the findings, and section sixth presents the conclusions.

An institutional perspective on the study of MNCs
Institutional arrangements embody collective action that constrains, guides, and liberates individual action. National institutions play an important role in influencing the development of technology, the use of labour, and the development of organizational competencies and capabilities (Fligstein, 1996). Thus, an understanding of the social and political environments within which industries and organizations flourish and perish has been a central theme in scholarly works on country differences leading to industrial development and contrasting ways of controlling and coordinating economic relations.
When attempts are made to transplant the organisation of work from one country to another, national patterns may clash, and it may be no easy matter to reconcile different traditions and interpretations of the division of work and the relations between different work groups. These traditions may also be in a process of change for other reasons (for example, economic and democratic changes in macro level institutions). This means that quite important methodological questions must be asked about how the various factors that affect a work system should be evaluated. Elger and Smith (1998) argue that multinational companies are important for transmitting innovations in the organisation of production and the regulation of labour. However, as the home country model is also embedded in the home institutions, the national recipe cannot easily be reproduced; it is always evolving, and will therefore vary from one site to another, even in the same host country.
Muller et alii (2001) argue that the more institutionally different the country of origin is from the host country, the easier it is to identify a host country effect on a MNC. An important expectation is that MNCs are under more pressure to comply in more tightly regulated business systems than in weaker institutional environments. Thus, multinational corporations in a weak business environment such as the Brazil-ian would have fewer difficulties in transferring their ways of organizing work from home. On the other hand, as Whitley (2001) argues, MNCs from coordinated market economies (CME) would have more difficulties in transferring their capabilities across borders than MNCs from liberal market economies. «Cooperative hierarchies are less able to extend their prevalent ways of managing activities to subsidiaries in different kinds of environment that are opportunistic and isolated hierarchies, unless they can transfer key features of their domestic home business system» (ibid: 50).
Instead of looking on multinational corporations as unified rational actors, we should attempt to see them in a new perspective. Kristensen and Zeitlin (2005) complicate the picture by raising larger questions about how multinationals and the globalisation of economic activities should be conceived. Subsidiaries should be seen as sites where organisational knowledge is translated and negotiated, and exactly how these translations and negotiations take place should be studied: what is negotiated and how?

The Danish business and work systems
It has been argued that the Danish work system combines considerable fluidity in the external labour market with a reliance on highly skilled workers, who exercise a high level of discretion over work performance. Historically, flexibility has been continually pursued by skilled workers in their working life. Flexibility is closely linked to the possibility of permanently upgrading their skills (Rocha, 2003). This permanent upgrading of skills has two consequences among different work groups: one is the growing organizational flexibility of these groups, and the other is the blurring of boundaries between them. Gjerding (1996) in a survey of 1900 privately owned firms in the manufacturing and service sectors in Denmark showed that in 50,3% the daily planning is executed by workers; that supervisors and middle managers carry out 38.5% of these tasks; that top managers are responsible for 10.4% of daily planning; and that a total of 39.5% of the firms affirmed that workers are responsible for organizing their work. Lorenz and Valeyre (2005:435) found that Denmark, Holland and Sweden presented the highest percentage of organizations characterised as learning forms. No less than 60% of Danish employees work in learning organizations. 7 Danish trade unions are strongly involved in the governance of vocational and further training institutions (Kristensen and Rocha, 2007), which need to be co-ordinated locally by networks among the convenors and shop stewards of different firms, who in this way become important gatekeepers of the flexicurity 8 model.

The host country effect: the Brazilian business and work systems
During most of the twenty century, the principles of Taylorism in Brazil dominated Brazilian industry. The implementation of Taylorism was largely the result of the Brazilian social configuration, in which political power was placed in the hands of economic and military groups who controlled the country during the many years of dictatorship. Taylorism was adopted by these groups in an attempt to reduce workplace struggles, to minimize the control the skilled workers had of the production process and to transfer this control to engineers.
Brazilian industry was not so successful on the international market, but the protected economy for industrialised products was a defence against international competition. The state, which regulated the internal market, did not allow for powerful competition among national producers. The industry was very weak in terms of quality and productivity, but able to survive in this low-competition environment. The result was a dynamic of little competition, low quality and state-controlled prices. The state controlled access to the internal market through high import tariffs and regulations, thus reducing competition. During the 90s, the manufacturing sector had to deal with an intense and extensive agenda of trade liberalisation.
Since the liberalization of the national market, Brazil has gone through a intensive process of productive restructuring. Many leading companies have adopted new organizational and technological templates which have transformed the industrial landscape. Changes have been made to the management patterns of the labor force in Brazilian industry, and in some companies the just-in-time system, a production philosophy based on team work in total quality programs, has been introduced. Participatory methods encouraging workers to get involved in companies' plans, which were slowly initiated at the beginning of the 1990s (Gitahy and Rabelo, 1991), have become quite widespread (Alves, 2000). It has also been pointed out that MNCs have led their subsidiaries in Brazil to adopt new organizational and technological patterns that were, to a certain extent, inspired by the Japanese model (Hirata, 1993). MNCs led the transformation process in different regions and industries, and sought new organizational and technological patterns that were, to a certain extent, inspired by Toyotism and by flexible forms.
But in spite of the democratic changes (Alves, 2000), the trade union influence in the work place remains quite weak although some inroads have been made. There is no established institution for workplace representation, and the level of trade union influence at firm organizational issues is extremely low. How work is organized and divided up, the boundaries between different work groups, are issues that are of no real importance in the trade unions' agenda.

Methodology
The intention of the research was to investigate whether work systems in multinational corporations change when they cross institutional borders and, if so, how. With this goal the most appropriate research design was considered to be comparative research. The example here is an in-depth longitudinal case study that spans a period of five years. The case-study methodology is useful in developing a processual and holistic approach (Elias, 1978) which fits in well with neo-institutional theory. The research presents two institutional environments that differ enormously and may shed light on the interesting social processes that occur in multinational companies. Case studies make it possible to explore how the institutional contexts in which MNCs are embedded impact on the way they are controlled and coordinated (Edwards et alii, 2007).
In order to analyse the work system at the two sites, the most important characteristics in a work system will be addressed (Whitley, 1997): how tasks are organised, performed and controlled -job fragmentation and managerial control of work; workplace relations between social groups; employment practices and policies-the extent to which employers are committed to retaining their workforce; the development of skills; the extent to which the different work groups are separated on the basis of their backgrounds; the basis for reward allocation and differentiation; the development of the employees' careers and their levels of commitment to company objectives.
As the research addressed the transference of the Danish work system from headquarters to a subsidiary in Brazil, the first step is to characterise the Danish work system in a mass production unit, since much of the research conducted in Denmark relates to work systems in small and middle sized companies with batch production (Kristensen and Zeitlin, 2005).

RIO, Nº 4, 2010
Transference and change of work system across borders. a Danish company goes to Brazil

Danfoss-The Danish Multinational
The Danish thermostat factory in Denmark (DDK) is located in the southeast of Jutland. The factory is part of the compressor and thermostat division of one of the largest Danish multinational companies and assembles thermostats for refrigerators and freezers. The Danish subsidiary in Brazil (DBR) is located in an industrial district of Campinas. The same products and product variations are assembled in both factories.
In spite of the fact that most of the machines are the same, or at least perform the same function, two automatic machines that assemble pieces in the Danish factory are not encountered in the Brazilian factory. These machines are substituted by manual operations. The industrial engineering, the design of the product and the changes in more expensive or core technology are the responsibility of headquarters.
The initial intention was to transfer the organization of work from Denmark to Brazil. A former production manager at DDK was sent to Brazil to act as production manager, and skilled and semi-skilled workers from Denmark were also sent to teach the Brazilian workers how to perform the operations and repair the machines.

Danfoss in Denmark-Negotiating flexibility
In order to cope with the challenges of globalisation, production flexibility has been quite an important strategy for the company. In recent years, both managers and workers have been paying particular attention to increasing the flexibility of the workforce, in an attempt to increase productivity (from a managerial perspective) and to preserve jobs at home and improve employability (from the work force perspective). Considerable effort has been put into increasing the flexibility of workers' qualifications and skills because if the intersection between skilled workers and semi-skilled workers can be expanded, the factory can adapt more easily to increasing or decreasing demand.
Skilled workers develop their skills in different ways. They learn by doing and exexchanging knowledge with other skilled workers on the work floor. They complement this knowledge with evening classes. They can also learn by moving within the company from one factory to another, this being able to experience and learn from different jobs and related tasks. In the company the same kind of career path is pursued by most engineers and technicians, who move from factory to factory without changing their employer.
Semi-skilled workers upgrade their skills in a less structured way. Of particular importance is their own personal interest in moving from jobs involving simple tasks to jobs involving more complex machines. The social process of competing for skills within the group of unskilled and semi-skilled workers was not the same as in the group of skilled workers. In recent years, unskilled workers who are interested in moving up in the qualification hierarchy and upgrading their skills to compete with those of skilled workers have found new opportunities and receive support and interest from the management for their endeavour. In these cases, the process of searching for qualifications has made semi-skilled workers more flexible and blurred the boundaries between the various groups in the organisational sphere. This means that skilled and semi-skilled workers can perform some of the same tasks, even though they do not share common skills, identity or organisational status. Set-up tasks, which for many years were allocated to skilled workers, are now allocated to unskilled and semi-skilled workers. Skilled workers now focus on learning and mastering the new technologies and maintenance tasks related to the new machines on the shop floor.
The role played by the shop steward in the organisational arena has been of central importance to the move towards a more flexible organization. These changes in the organization and division of work in the factory have been characterised as a long-term process. The already multi-skilled craft workers have taken on greater responsibilities. They now have responsibility for all plant maintenance and have also started to participate more actively in industrial engineering matters. Thus, rather than finding their roles eroded due to operatives taking an increasing role in the set-up of the machines, craft workers have been entrusted with increased responsibility.
The pattern of change tends towards spreading qualification to unskilled workers through continuous education so that they can upgrade their skills. Unskilled workers can now achieve semi-skilled status, and setters can move into the area of maintenance work. Instead of creating a dual structure that sets central and peripheral workers against each other, this process of change leads to an organisational structure that incorporates the two categories without stamping them with stereotyped marks of skilled versus unskilled, central versus peripheral. The differences in the status and employment between skilled and semi-skilled workers are still and will be considerable, but the intersection area of the tasks and the jobs is growing.

The Danish Multinational in Brazil (DBR): flexible Taylorism
DBR was organised in a few small sections some of which were made up of unskilled workers. The flexibility within and between these teams was considerable. Workers could occupy a variety of workstations within their sections, and also move to others. The exact position they should occupy was not regulated, there was no agreement, and the internal movements of workers were largely ad hoc. It was the supervisor' s responsibility to indicate where a worker was needed. The unskilled workers were assigned to a specific work area to a greater extent, but this did not signify that they could not work in other sections. Most of the tasks were very simple and could easily be learned, and workers could be placed at different workstations without major problems. The «stick and carrot» was the number of thermostats assembled during the day, and at the moment of the research the goal was 4000 pieces, a goal that had recently been achieved.
The intention of the Danish company was that workers in Brazil should be able to perform the tasks at as many workstations in the factory as possible. When the production in the assembly line needed to be intensified, workers were moved onto the line. When the urgency was over, they could be re-distributed to other workstations. Parts of the workforce, then, were moving most of the time, practically accompanying the parts being assembled. They could be placed at the end as well as at the beginning of the assembly operations. This flexibility was closely linked to the sharp division of work in a Taylorist organisation of work. The fact that the tasks could be easily learned helped to build a flexible work force using the essentially Taylorist way of organising workstations. It was important to this movement that the workers could move from one workstation to another and quickly learn how to perform the tasks there.
In spite of the fact that the management aimed to create a flexible workforce, which could occupy a variety of workstations, the unskilled workers had developed a hierarchy and status system that seemed to contradict the «official» work system. For instance, female workers identified the automatic machines by the person working at them; they gave the machines the name of the workers. The supervisor decided where people should work, and who should learn to operate new machines, and he also took care of the production flow in order to keep it running and avoid bottlenecks.
The Danish supervisor is above all a technical position. At DBR the skilled workers did not show much respect for the supervisor' s technical competence: rather, his function was largely that of a bully. This was in contradiction with the fact that the Danish manager was trying to introduce a flat organisational structure, in which the authoritarian role of the supervisor would be minimal. The result was that the supervisor at DBR was in an ambiguous position and had a difficult task. On the one hand, he had to exercise authority and co-ordinate the production process. But on the other, the Danish manager made it clear that he did not like supervisors in the organisational structure, and that he did not intend to maintain this position in the factory. The supervisor tried to impose his authority, sometimes threatening people with the sack.
The power to hire and fire was the most important constituent of his authority. Good personal relations with supervisors and bosses are fundamental in an environment where authority is concentrated, knowledge is not dispersed, and human resources are seen as abundant and redundant. At DDK, authority and the power to hire and fire personnel are not connected.
One area of conflict among blue collar workers was the learning process. Some of the unskilled workers declared that the process of learning to operate some machines was complicated by the fear of being fired. They were not interested in teaching the whole operation or passing on tacit knowledge. Sometimes they would teach their colleagues the worst way to perform a task. According to them, this was a way of maintaining some kind of job security. If very few people knew how to operate a machine, then there would be less risk of getting fired. Workers at DDK, however, had a different strategy. They interpreted their flexibility as a way of securing their jobs. Skilled workers at DDK who resisted upgrading the skills of unskilled workers faced problems within their own work group. At DBR, once again, there was no opportunity for discussing these problems: negotiation related to the learning process was minimal. There was an explicit conflict among workers which led to a non-negotiated division of labour, a combination of imposition and continuous resistance in which it was difficult for co-operation between the different work groups to emerge.
There were two young maintenance workers in the factory who were responsible for solving maintenance problems on all machines in the factory, and the area of their work was less limited than the area of skilled workers at DDK. This was a relatively surprising observation, however, because skilled workers in Brazil did not overlap with unskilled workers or technical personal. These two workers did not split the factory into specific areas, and they tried to help each other in all situations. However, they were not requested to perform direct production tasks on the assembly lines or at the automatic machines, or to fill in for absent workers. To perform the tasks of semiskilled or unskilled workers was an idea they resisted and vehemently rejected.

Comparing the work systems
What are the similarities between DDK and DBR as far as the organisation of work is concerned? Very few, if any. I have shown how radically the organization of work changes when institutional divides are crossed in spite of the continuous attempts to transfer the Danish model, not only by having an expatriate in charge of the transference but by sending Danish skilled and unskilled workers to Brazil. The way in which the different work groups relate to each other at DBR does not resemble the relations RIO, Nº 4, 2010 Transference and change of work system across borders. a Danish company goes to Brazil at DDK. The forms of competition and co-operation between and within work groups do not have much in common.
Although the similarities between the two factories are few, there is one important similarity that must not be forgotten: the struggle of the unskilled workers to gain more space in the factory. At DDK a long-term process had resulted in greater opportunities for the unskilled workers to earn promotion and upgrade their skills. In Brazil, the unskilled workers were also involved in fighting for recognition, and resisting the exploitation as they had experienced previously. Despite having won some battles, they were still a long way from achieving the conditions in DDK. DDK had a special scheme to support the mobility of internal workers. Skilled workers could move from one factory to another, gaining experience from different production sites, technologies and products. At least three different factors were involved in creating a system for continuously upgrading skills: the opportunities for internal mobility, the institutional interest in improving skills, and the reputation-based system of authority derived from technical competence.
This link between mobility and skills development was not found in Brazil. Workers were mobile because they were fired during downsizings in different companies to cope with economic instability. Beyond that because the high rates of turnover represent a strategy of control based on punishment. At DDK, the solutions to problems were negotiated collectively. The system left little room for supervisors to play a centralised authoritarian role on the shop floor. They had no direct control over blue collar workers, because the strong and important presence of the shop steward who is in charge of negotiation on the behalf of the workers. At DBR, the supervisor' s function was a disciplinary one, and he acted as the eyes and the whip of the owner on the shop floor.
The figures compare the two work systems. Table 1 outlines the characteristics of the work system described above, and figure 1 illustrates the boundaries between the different work groups in the factories. The arrows in figure 1 indicate the paths individuals from different work groups may take to construct their careers. In Brazil, the opportunities for career development are quite restricted: engineers are expected to become managers, but the same cannot be said of the other groups. Semi-skilled and unskilled workers have few opportunities to move up in the hierarchy. At DDK, some skilled workers had decided to be employed as semi-skilled or even unskilled workers, as a way of reducing the stress of the continuous upgrading of skills expected of their original work group.

Discussion
At DDK, in striking contrast with the Brazilian factory, the shop steward plays an important role in regulating the games between the managers and the blue-collar workers. At DDK the shop steward was responsible for negotiating a wide range of agreements between workers and the management, between the workers and the vocational institutions, and with other shop stewards representing other work groups. She pursued grievances and other worksite-specific issues, working to guarantee that workers could have their demands negotiated and attended to, and reaching agreements not only with management, but also among workers. By playing a role in the negotiations between managers and workers, the shop steward could reduce the direct confrontations between them and organise the negotiations on both individual and collective issues. She needed to negotiate all changes in the organisation of work with the management, not because it was prescribed by law (on the contrary, Danish law gives sovereignty to the management to choose the management model), but because there is an assumption among all the actors involved that negotiation is necessary to bring about change and to solve problems. The Danish system, then, is continually being built on the basis of negotiation and agreements among the different groups. This is coherent with the arguments of Kristensen (2003) about the role played by the shop steward in Danish firms: managers have the right to manage the factory but they seldom have the knowledge that makes this possible without the shop stewards. At DDK unskilled and semi-skilled workers have colluded to gain recognition and power; they have entered into the dynamic of skills development. They have been quite articulate in challenging the domination of skilled workers in this game. The ongoing system of negotiation in the factory includes all the work groups, and it is through agreements and representations that these groups create partnerships between managers and workers. The shop steward at DDK not only has to try to find solutions to workers' problems: she also has to negotiate the problems faced by the management. The shop steward is therefore co-responsible for the continuous improvement of the organization, not only to protect workers from any downgrading in work conditions, but to reach agreements by which the organization can support the process of skills upgrading and higher productivity. When the factory faces market problems, the shop steward needs to negotiate the possible solutions with the workers, which may involve sending workers to courses, or even negotiating a list of workers to be dismissed.
In Brazil, trade union participation in shop floor politics is an emerging phenomenon; there is no tradition and the right to organize on the shop floor level is quite new. Trade unions still mainly focus on playing a role in central bargaining over wages. Most workplaces -normally the majority-have had no plant representative to pur-sue grievances. In the late 1970s, de facto shop stewards had begun to emerge in São Paulo. They were named factory commissaries (comissão de fábrica), but they had no legal standing. Only after democratization and the new civil government in 1985 did new arrangements and new labour laws enter into discussion. A new labour law made strikes legal and strengthened trade unions. Making strikes easier shifted the focus from police repression (or the threat thereof ) to the real issues at stake between employers and unions, and new ground for bargaining was created. But salary issues are still the main focus of attention.
At DBR, individual problems could easily turn into conflict. Production problems were resolved in an ad hoc way: the supervisor distributed tasks, and the workers accepted or resisted with no collective negotiation. The absence of a shop steward implied that workers needed to seek individual solutions to their problems. The absence of spaces and formal processes for conflict resolution meant that informal and more personal ways needed to be constructed, and people with no mediation function tended to try to resolve conflicts informally. These arrangements are hardly sufficient to substitute formal and legitimised channels for conflict resolution in which the different parts involved can negotiate by recognizing each other' s different roles and rights, thus helping to prevent conflicts from continually emerging from mutual misunderstandings. Partnership or other forms of collaboration between the different organizational groups could not emerge. In these informal systems of negotiation at DBR, partnership arrangements were difficult to generate, representation was impossible, and war among fiefdoms was commonplace.
At DDK, the management and the blue collar workers were aware of the impossibility of achieving organisational goals or changes, or any other modification in the way the work was divided and organised, by means of a top-down decision making process. Negotiation was necessary if organizational co-operation was to be constructed and a continuous improvement process implemented. It was not just a matter of discourse and new management fashions: both groups agreed that pressure from the top had never worked in the factory.
At DBR, a group that the blue collar workers called the Mafia played a game in which the rules were still being constructed. The intention seemed to be to establish an informal agreement between the unskilled workers and the management about the number of pieces to be produced daily. The Mafia was inclined to resist high levels of productivity, which would be costly in terms of controlling the pace of work and also in terms of physical strength. If a system of continuous improvement could be achieved, it would jeopardise their control over the pace of work. The group consciously played against all those workers who were able to demonstrate that a higher level of production could be achieved.
It would be easy to get a bad impression of the Mafia, one group of workers devoted to oppressing others. A more contextual interpretation would show the group to be defending the organisation of work against the emergence of all the things that they had experienced before: powerful supervisors, opportunistic behaviour among workers, sexual abuse and firing as a disciplinary tool to name just a few. The group can also be regarded as the only organised force among blue collar workers who could keep destructive competition among workers away from the shop floor. They were fighting to determine what could be considered as a fair day' s work. The workers they fought against represented the kind of opportunistic behaviour that they would like to avoid. The same could be said about their struggle against the supervisor. An interesting aspect of this group was that in spite of having been described as a Mafia, their leader was also nominated as the most qualified person to substitute the supervisor. They fought against the emergence of a hierarchical organisation in which the supervisor had a great deal of arbitrary personal power and was able to make crucial decisions about the division of work without taking into account the negative consequences to the workers' health and well being. If the Mafia could control the speed of improvement, they would give the management and its targets less control over the workers. They played the game with the supervisor, and subverted his orders. And by so doing they induced a greater level of uncertainty and therefore fluctuation in the level of output. The game was being played with the aim of making one of them a supervisor.
The game can also be interpreted as a way of fighting against the unequal distribution of income. Why should workers feel compelled to work harder and be more affected by work related diseases when the distribution added value remained the same? Recognising unskilled workers as a group, and communicating the idea that they should be treated equally, would be constitutive for making fair rules of the game. Imagine the economic situation of these unskilled workers: the struggle of a single mother or a person who is an important source of income for a whole family to achieve levels of high productivity can jeopardise health and put entire families ate risk. 9 On the other hand, some workers think that the best way to keep their jobs is to work as hard as possible and as efficiently as possible. Conflicting strategies would not be a problem, or at least not as big as they once were, if the distinct strategies and perspectives could be discussed in an institutionalized way. If, for example, a shop steward in the factory could discuss these different points of view and negotiate the aspects of the organisation of work with workers and management. However, at DBR there was no institutionalized way of discussing the organizational issues, for achieving common goals and solving problems. The informal organization was built on mutual misunderstandings and destructive power games.
At DDK, one convenor represents all employees at the board level. He/she plays the role of negotiator, a person who has access to different organizational layers, who is able to negotiate from the shop floor level to the board of directors. It is a problematic and difficult role, since the negotiator often represents the blue collar workers as their convenor or shop steward, but by becoming a negotiator he/she must be able to play the game as a manager, and on many occasions convince workers that to follow a specific managerial strategy is necessary. However, this must be done without colleagues losing confidence in the work of the representative. A negotiator is usually defined as someone who presents high goals, makes high initial demands, and resists concessions. Negotiators in Danish organizations play a different kind of role; they are responsible for negotiating at different levels and at the same time are also able to position themselves on opposite sides in order to manage organizational interdependency. What is generally being negotiated is how the organization can survive, which kinds of concessions the various organizational actors need to make, and how the fruits can be equally distributed. In Brazil, it seems that there is still a long way to go before agreements will be negotiated: there is no tradition for negotiation between and within work groups, and the question of authority seems unresolved (Rocha, 2003). There is some confusion about who has authority to negotiate and what can be negotiated. The question of who has the authority in Brazilian organizations, the technical expertise or the ownerships rights, is still unresolved. And this impacts strongly on the way the Brazilian white and blue collar workers build their strategies and are able to co-operate and produce win-win results. The Brazilian vocational system was designed to upgrade the skills of just a small fraction of the workforce (Rocha, 2003). At DBR the few opportunities to upgrade skills were interpreted as opportunities for job security that could not be shared by all workers. The opportunities for improving and learning in the Danish factory were largely supported by partnership agreements between managers and shop stewards; the collective solution was creating new ways of coping with the challenges of globalization where the salaries are among the highest in the world. These negotiations facilitate the renewal of company operations and the upgrading of products and processes to match the challenges of global competition.
In the Danish case, the parties were able to reach a compromise, and open negotiation was an important mechanism for achieving and supporting change; the norms were continually being worked out. The greater flexibility and the possibility of tailoring solutions to different needs of the groups involved resulted in a wider variety of solutions.
At DBR, workers and managers were unable to create a forum for open negotiation. Brazilian workers avoided an open process of negotiation and relied on more informal and conflictual ways of dealing with their disagreements. Their past experiences showed that they had nothing to gain from each other, and the discursive strategies on empowerment could be just one more attempt by the management to dominate and control how work was divided and organised.

Conclusions
The case studies show that the host country institutions and worker/manager traditions had a strong influence on the way the organization of work was translated and negotiated. Traditions were too strong and hindered the transference of the work system from Denmark to Brazil. The result could have been different had the company isolated itself from the host country traditions and institutions by internalizing the human resources management and changing the salary differentials to lower levels as some companies have done to reduce the abyssal gap between managers' and rank and file salaries. In the host country there were no institutions like those in the home country that supported a flexible workforce.
In the case presented, the home and host country effects both create difficulties for the company. The flexible work system of the Danish site is highly dependent on the continuous upgrading of skill among its employees; this highly flexible labour force is the result of a well managed welfare state which supports this social dynamic. Danish workers are enabled to continually upgrade their skills by welfare institution, not by their employers, so even the largest Danish manufacturing companies have no internalized system for improving and upgrading the skills of its labour force. The dependence of Danish firms on their institutional environment seems to become a competitive disadvantage in the internationalization process.
Future empirical research into multinational corporations would be greatly improved by recognizing that the transference of work systems between sites of multinational corporations needs to be seen as a process of organization and negotiation. And that the negotiation process unfolds in quite different ways in different locations, which results in transformations and adaptations of the negotiated order.