1. INTRODUCTION
Globalization has posed complex and completely new challenges to the whole of society since competition is not solely between business enterprises, but also between economic systems.

From the Corporate dimension, Globalization means the total re-shaping of the way in which firms design, manufacture, market, distribute and innovate their products, as well as the way they decide on why/how/when and where to re-locate industrial facilities. This involves transforming their organizational structures, their ownership, their roles and their very own nature. As a result, an elaborate international network of suppliers, assemblers, designers and supporting 'service' activities has already been developed (The Economist, 1998). Global manufacturing can be defined as 'the process of moving from an independently managed business serving local markets to networks of business serving the businesses' chosen markets in a coordinated and optimized way "(Shi and Gregory, 1998: 212). Global Companies therefore are blends of units for product lines, country subsidiaries and business functions which dynamically move among and between different national settings (Galbraith, 2000).

Globalization has brought massive implications for the whole production chain - core firms, partner firms, specialized suppliers, contractors and service providers - creating virtual networks of value creation (Berger et.al., 1999). As a result, new partnership strategies, new external relations among companies (contracts, alliances, equity purchases, and outright ownership) and supporting policies are emerging to integrate producers, suppliers and customers within and among different economies (Galbraith, 1995). Individual companies have been redefining their core competencies and negotiating their strategic role in order to build capabilities across the whole? production chain (Bartmess and Cerny, 1993; Ferdows, 1997).

In the electronics industry, this process has quickly evolved towards what Sturgeon (2002) called Modular Production Networks (MPN). Under MPN arrangements, the whole industrial chain is divided into two parts. One part, the customers (usually referred to as OEMs - original equipment manufacturers), are responsible for design, marketing, and sales. The second part, the manufacturing and logistics services, is performed by contract manufacturers (CMs).

Since depending on institutional factors such as income distribution patterns, rules, norms and dominant ideologies, different combinations of a company's strategy, as well as its production models and organizational architectures, need to be coherent in order to support firm-level competitiveness (Guillen, 1994; Guzman, 2000; Boyer and Freyssenet, 2000), how do specific contextual conditions impact on the process of organizational and production re-structuring? And how do firms strategies, organizational architectures and production best practices are adapted to existing (and non controllable) contextual conditions?

While all those questions are still waiting for answers, this paper attempts to contribute to understand some key macro-micro connections of the adaptation process of a Performance Measurement System, Manufacturing Cells and Matrix Organizational structure at a Contract Manufacturing factory located in Brazil.

Two complementary theoretical frameworks were applied:

• A comparative Business System approach (Lane, 1992, Maurice et.al. 1986; Streeck, 1995; Whitley, 2000) that enables a multi-dimensional examination of macro- and meso- contextual factors such as employment practices, work systems, and types of dominant firm that influence both production and organizational practices (Guillen, 1994; Lindberg, Voss, Blackmon, 1998).

• Network Organizational? theory (Nohria, 1992; Nadler and Tushman, 1997; Bartlett and Ghoshal, 1989; Nohria and Ghoshal, 1997; Belanger et.al., 1999; Galbraith, 2000; Hedlund, 1994; Pettigrew and Fenton, 1996) allows an improved understanding of the grounds on which organizational architectures, organizational processes and management practices are deployed inside and beyond national frontiers.

2. MODULAR PRODUCTION NETWORKS (MPN)
MPN can be interpreted as part of a major trend that has two dimensions. The first is further economic concentration by OEMs and CM businesses, which reinforces their capacity to innovate, increase sales and improve quality/productivity (Sturgeon, 2002). The second dimension is a trend towards the externalisation of fixed costs, in terms of both business risks and manufacturing costs (Appay, 1998). In this context, lower manufacturing costs might result also from lower wages, workforce size, health and safety costs and human resource management costs.

In economic terms, MPN arrangements can be viewed as the combination of two ideas:

• The further exploration of economies of scale by CMs (Chandler, 1990); and [problems : increased scope and increased value added, + complexity , influence of domestic demand trend; competition among plants of same CM]

• An increased focus on core competencies by OEMs (Hamel and Prahalad, 1994).

The concentration of manufacturing activities by the CM requires the development of volume flexibility in order to meet sudden changes in clients' demands. By out-sourcing manufacturing activities, OEMs transform manufacturing costs from fixed costs to variable costs. This also requires that CMs develop the capacity for mix flexibility in order to meet the variety of orders from customers (Sturgeon, 2002).

From the production engineering perspective, Zarifian (1999) has pointed out that CMs perform manufacturing activities under the logic of industrial services. CM arrangements, then, not only have features of both industrial and service sectors, but also follow different logics which are hard to combine. While the industrial logic follows mass production/low cost/standardised product/rigid technological processes and hierarchical employee-employer relationships, the service logic pursues small and medium volumes/customised production/flexible technological processes and co-operative type employee-employer relationships. In other words, CMs need to develop service type capacities and competencies, such as mix, volume and new product flexibility (c.f. Suares, Cusumano and Hine, 1995), in order to fulfil customers' product differentiation needs, while at the same time carrying out manufacturing activities with the same continuity/stability and low costs as mass production processes.

From the organizational perspective, the efficiency of the whole production chain depends to a significant extent on the quality of co-operative relations developed by different organizational actors. Crucially, it is important to remember that OEMs and CMs perform three different tasks separately, namely, product design, manufacturing and marketing/sales. Despite the significant differences in the type and nature of tasks performed by CMs and OEMs, success seems to depend on the convergence of work methods, organizational structures and organizational goals. Thus, the co-ordination of the MPN becomes a key issue, as design and manufacturing activities are concentrated but performed by different organizations (in different locations) and, at the same time, marketing and sales are de-concentrated (Zarifian, 1999). Matrix structures, accompanied by lateral integration mechanisms, such as mutual adjustment, formal groups and integrator roles and/or departments, need to be purposely designed in order truly to differentiate and integrate organizational units inside the whole MPN (c.f. Galbraith, 1995; Nohria and Ghoshal, 1997).

3. THE CASE STUDY
CM-BR^[1], a branch of CM-HQ, has approximately 250 personnel (150 direct production workers and 100 indirect) and five competitors in Brazil. CM-BR manufactures finished products such as printers, personal computers, ATMs (automatic teller machines) and components such as PCBs, and provides industrial services such as repairing and testing^[2].

CM-BR operations
The manufacturing of CM-BR products involves two basic operations: assembling micro-electronic components on a board (PCBA) and assembling the PCB with other hardware components (box building (BB))^[3]. Due to the wide variety of products, the number of heterogeneous models per product and the small size of batches, logistical and organisational complexity at CM-BR is higher than some other overseas CM-HQ plants, such as those located in China and Mexico.

Structural changes towards the implementation of a series of industrial best practices that are already being applied world-wide are occurring at CM-BR. Manufacturing cells, Drive by Wire (performance measurement system),? and matrix organisational architecture are some of the best practices which CM-BR need to meet in order to be aligned to CM-HQ. Nevertheless, the implementation process of the above-mentioned best practices seems to be constrained by a combination of factory-level issues and contextual forces.

Contextual conditions shaping Operations
Four contextual conditions shape the implementation of industrial best practices.

1. The current tariff system combined with informatics law requires that more than 80% of total costs be from imported components. As a result, the general operations manager spends approximately 70% of his or her time trying to discover materials/component configurations that minimise custom tariffs, and only 30 % of his or her time in manufacturing-related matters.
   
2. The transfer of best practice from CM-HQ to CM-BR calls for important adaptations of the concepts, since the predominant production system in main overseas plants is different >from the Brazilian operation. Typical CM-HQ overseas plants (e.g. in China and Mexico plants) are high volume low mix. They perform mainly PCB activities, some BB (assembly) activities, and produce low cost products. A few CM-HQ overseas plants (e.g. in West Europe and USA) are low volume high mix, but they manufacture high value-added products. On the other hand, the Brazilian plant is low volume high mix and manufactures low cost products. Predominant operations involve both PCB and box building activities. Therefore, while CM-HQ have a great deal of experience in handling product variety, small batch production and ongoing engineering changes, this experience is not being successfully transferred to CM-BR. The reasons for this are (a) they have a different production system (i.e. high volume low mix); (b) they manufacture different products (i.e. very high value-added); and (c) current low volumes and the recency of the CM-BR acquisition make it a rather low priority for attention by CM-HQ management. As a result, best practice is selectively implemented to varying extents.
   
3. While USA and Asian plants are relatively close to their main suppliers, the Brazilian operation is a great distance from its main suppliers of components. The geographical location significantly affects purchasing because of transport time and the level of service provided to the purchasing plant. Plants located near component suppliers can easily ask for changes in their orders since both transport time (from supplier to plant) and transport cost are relatively low. Additionally, because the main CM-HQ overseas plants are high volume they have greater purchasing power. This is not the case with CM-BR. Time and costs of transport are high and this limits delivery flexibility. As well, because CM-BR is a small customer it has low purchasing power.
   
4. By definition, competitiveness in the CM industry comes partly from lower wages, employment numbers and health and safety costs. However, while this arrangement results in lower production costs, it also brings important limitations for competitiveness building at the plant level. Low pay for production workers restricts the much-needed improvement in the company knowledge base as well as the commitment of production line personnel. These outcomes reflect the lack of a specific labour policy to stabilise manpower at CM-BR.

Competence acquisition, is also limited at CM-BR. Accepting a high rate of turnover is a mechanism developed by the industry (and allowed by labour laws) in order to get qualified personnel without raising wages. The idea is to develop the hire/train/fire/hire cycle, keeping wages at the same level. This cycle is intensively repeated to stabilise the supply of skilled personnel. While this mechanism is useful to keep wages low, it constrains competence accumulation. Barriers to competence acquisition are related to company policies (personnel and training), labour laws and the local labour market situation. All those factors hinder the implementation of industrial best practices that demand cooperative relations, stable and skilled manpower and fine coordination oof operations inside and between plants.

4. CONCLUSIONS
This article has attempted to shed some light on the key management issues that affect competitiveness building of CM operations in an emerging nation such as Brazil. Despite the fact that branch operations are framed by the philosophy of headquarters, CM-BR possesses leeway to customise CM-HQ's manufacturing concepts and practices. Both contextual and factory-level factors at CM-BR limit its full /adoption/realisation of CM-HQ's concepts.

Non-controllable contextual factors are related to the wide range of small customers (ordering small volumes), the location of main component suppliers (overseas), and low wages and labour laws which negatively impact on training and knowledge acquisition practices. 'Controllable' factory-level factors seem to explain the slow implementation of matrix organisational architecture, manufacturing cells, teamwork practices and the limited application of the DBW system.

Building manufacturing competencies in developing nations seem to be an ongoing task that not only involves knowledge of new production methods, but also knowledge of methods of implementation that take into consideration both the context and factory-level limitations. Both formal/explicit/hard and informal/tacit/soft issues are related to contextual and factory-level forces. One of the challenges for practitioners and academics, therefore, is to learn to correctly interpret particular contextual factors that may undermine the implementation of advanced production techniques/concepts.

REFERENCES

• Appay, B. (1998), "Economic concentration and the externalisation of labour", Economic and Industrial Democracy, Vol 19 No 1, pp.161-84.
• Bartlett, C. and Ghoshal, S. (1990), "Matrix management: not a structure, a frame of mind", Harvard Business Review, July/August, pp.138-45.
• Bartmess A Cerny K. (1993) Building competitive advantage through a Global Network of Capabilities, California Management Review, Winter
• Belanger J., Berggren C., Bjorkman T., Kohler C. (1999) Being local Worldwide, Cornell University Press, Ithaca
• Berger S., Sturgeon T., Kurz C., Voskamp U., Wittke V. (1999) Globalization, Value Networks and National Models, IPC Globalization Meeting, MIT, October 8
• Chandler, A.D. Jr. (1990), Scale and Scope: The Dynamics of Industrial Capitalism, Harvard University Press, Cambridge.
• The Economist (1998) Manufacturing - The world as a single machine, June 20th
• Ferdows K. (1997) Making the most of foreign factories, Harvard Business Review, march-April
• Freyssenet M. (1998) Intersecting trajectories and Model Changes, IN Freyssenet M., Mair A, Shimizu K., Volpato G. (Eds.) One best way ? trajectories and industrial models of the world's automobile producers, Oxford University PressNew York
• Galbraith, J. (1995), Designing Organisations, Jossey-Bass publishers, San Francisco.
• Galbraith, J. (2000), Designing the Global Corporation,? Jossey-Bass publishers, San Francisco.
• Guillen, M. (1994), Models of Management, Chicago, The Chicago University Press, Chicago.
• Guzman, G. (2000) New Production Systems in Transition - Implications for Brazilian Industry, The Brazilian Journal of Political Economy, Oct.-Dez.
• Hamel, G. and Prahalad, C. (1994), Competing for the future, Harvard Business Review Press, Boston, Masachusetts.
• Khurana, A. (1999), "Managing complex production processes", Sloan Management Review, Vol 40 No 2, 47-55.
• Lindberg, P., Voss, C. and Blackmon, K. (1998), International Manufacturing Strategies - Introduction, Kluwer Academic Publishers, Dordrecht
• Nadler D., Tushman M. (1997) Competing by Design, Oxford University Press, New York
• Nohria, N. (1992) Network and Organisations: Structure, Form and Action, Harvard Business School Press, Boston, 67-82.
• Nohria N., Ghoshal S. (1997) The differentiated Network, Jossey-Bass publishers, San Francisco.
• Shi Y., Gregory M. (1998) International manufacturing networks, Journal of Operations Management, Vol. 16, pp. 195-214
• Sturgeon T. (2002) Modular Production Networks: A New Model of Industrial Organization, Industrial and Corporate Change, 11(3)
• Suarez, F., Cusumano, M. and Hine, C. (1995), "An empirical study of flexibility in manufacturing", Sloan Management Review, Fall, pp.25-32.
• Whitley R. (2000) Divergent capitalisms - The social structuring and change of Business systems, Oxford University Press, New York
• Zarifian, P. (1999), El modelo de Competencia y los Sistemas Productivos, Papeles de la oficina tecnica, OIT, Cinterfor, Montevideo.

[1] In keeping with confidentiality requirements, the company operating in Brazil will be referred to as CM-BR and its headquarters (in the USA) as CM-HQ.

[2] CM-BR supplies three types of services. The first is turnkey, in which CM purchases or procures components, in addition to PCB mounting, testing and final assembly activities. Although this mode increases risks (e.g. CM is responsible for capital outlays for components and is under pressure to optimise inventories) it also assures higher margins. The second is where the OEM supplies all raw materials and CMs perform only the manufacturing and test activities (known as 'consignment'). The third is partial turnkey, in which a portion of the components is supplied by the OEM and a portion is supplied (through procurement, purchasing and logistics) by CM (Porto, 1999).

[3] PCBA is done either by machine, using highly automated/high speed surface-mount technology (SMT), or by hand, using labour-intensive placement through-hole (PTH) technology. At CM-BR, PCBA operations are hybrid, as almost all PCBs use both SMT and PTH lines. This means that human-related factors are important for plant competitiveness. CM-BR has currently one SMT line, three PHT lines and five BB production lines.