The New Dynamics of Global Manufacturing Site Location

The past twenty years have seen the development of a global marketplace in almost all major industries. Since 1962, worldwide exports have increased from 12 percent to more than 30 percent of world GNP,1 totaling $3.5 trillion in 1992.2 If one considers the potential exposure to import penetration, more than 70 percent of goods now operate in an international marketplace.3 Every organization must now formulate strategies within a global context.

Global competition affects a firm’s manufacturing strategy by dramatically increasing the complexity of decision making. Worldwide markets can be served in many ways; for example, by export, local assembly, or fully integrated production. Underpinning these factors is the optimal configuration of the organization’s production resources. Location is an important part of that picture, but one that is usually given only limited attention. Decisions are often based purely on quantitative analyses that trade off transport costs, scale economies, and other cost-based variables. This practice, however, can lead to suboptimal results, as decision makers tend to focus only on factors that are easily quantifiable. Important qualitative issues are frequently neglected or used only to temper results. These factors are often central to supporting or creating a competitive advantage. For example, location dictates the level of knowledge embedded in the workforce; as such, it can affect the ability of firms to implement skill-based process technologies, or it can limit the effectiveness of quality programs.

Another disadvantage of strictly cost-based methods is that they tend to focus on factor cost advantages, which are all too often transitory. Government regulations, tax systems, and exchange rates can quickly change. Strategies based on such parameters may eventually be rendered obsolete by the very factors that first created advantage.

When formulating a site location strategy, companies should therefore emphasize the qualitative factors required to ensure that it supports the business strategy. Only after establishing a set of desirable location options should companies refine choices using cost-based algorithms.

In this paper, we examine how recent macroeconomic-and business-level trends have affected site location decisions. We describe how the dynamics in production systems, technologies, and management philosophies have changed location requirements. Finally, we propose a new framework for assisting in site location decisions and a model of the future global manufacturing firm.

Macroeconomic-Level Trends and Implications

The presence of large overseas markets suggests that there are benefits of scope for firms that sell globally. However, trends in trade and investment patterns indicate that the most effective way to serve global markets is with a regional approach, which reduces cost, provides better customer feedback, and minimizes risk resulting from exchange rate fluctuations and other political factors.

Overseas Markets

The development of large and sophisticated overseas markets dictates a global presence for leading manufacturers.

The global company cannot ignore the large potential markets that have emerged overseas. Many countries once regarded as less developed or developing now constitute large markets for leading-edge, quality products. U.S. companies no longer possess the world’s largest, most sophisticated market. The collective income of Europe has now surpassed that of North America.4 Japanese consumers have the largest per capita income of any large industrialized nation.5 In addition, the fastest growing economies are not the existing industrial giants but are in the emerging and developing markets and, in particular, those areas of Southeast Asia that have historically been the base for low-cost exporters.

As global firms expand their presence to these markets, the economies of scope they can generate will provide a significant source of advantage against local players. Global giants can sink large sums into product development, which are recouped across multiple markets. They have deep pockets to fight local market share wars. For the national player with no niche to hide in, the choice can be stark. Get global, get eaten, or get out.

Firms cannot compete in these global markets, however, by pushing mature products to overseas subsidiaries. Vernon’s “product cycle” approach to overseas markets became antiquated long ago.6 Product innovations introduced in the United States will be rapidly copied, improved on, and introduced to other markets within months if the producer doesn’t quickly exploit worldwide opportunities.

The development of such sophisticated markets means that lead users are no longer in one place but have to be sought out in the most demanding market (for example, in consumer electronics, Japan; in sophisticated computers, the United States). Locating production within such markets will facilitate quicker customer feedback, giving the firm a product development edge and enabling it to capture spillover benefits from the local industry base. It also allows tailoring to local tastes.

Trade and Investment Patterns

Increasing levels of nontariff barriers (NTBs) are forcing firms to localize production resources.

NTBs are trading requirements that circumvent the General Agreement on Tariffs and Trade (GATT) legislation. By placing restrictions on local content, sales volumes, or market share, NTBs interfere with a firm’s ability to serve a market through an export-based strategy. The 1980s saw a trend of increased protectionism in developed economies through such mechanisms. During this decade, the percentage of U.S. imports subject to NTBs doubled to 25 percent.7 For major industrialized countries, over 22 percent of imports are affected.8 More important, such practices are increasingly directed at imports from less developed, low labor-cost countries. The European Community (EC), for example, restricts only 13 percent of imports from advanced economies, but almost 25 percent of imports from less developed countries.9

Particularly in Europe, there has been an increase in restrictive rules of origin and local content requirements, forcing firms to localize more parts of their manufacturing value chain. Products sourced within the EC qualify for national origin status and its compensating benefits. Voluntary Export Restraints (VERs) have also found favor as a trade-restrictive tool. The value of trade affected by VERs increased by 60 percent during the 1980s, representing half the growth in government intervention. Currently some 250 agreements are in force, most directed at Japan and other Asian companies.10

These policies penalize an export-based strategy; they force a shift toward foreign investment and therefore promote a decentralized manufacturing strategy. For companies that have established facilities within a region, these policies shape the nature of the operation, expanding the scope of local activities. More components are sourced from local suppliers, or the firm increases the local value-added within its manufacturing process. This trend is illustrated by a recent study of Japanese companies operating in Europe: more than 40 percent had increased local content during the preceding twelve months.11

Regionalization of trading economies is increasing the benefits to decentralized manufacturing structures.

The past decade has seen a shift from an economic world dominated by the United States to a world with three relatively equal economies: the United States, Europe, and Japan. These economic powers and others are now moving toward a system of regionalized trading blocs, each characterized by relatively free internal movement of goods and production resources, common standards, and coordinated macroeconomic policies.

The EC constitutes one focus of regional trade; it encompasses twelve member nations and special arrangements with most other European states. The United States, Canada, and Mexico, through the North American Free Trade Agreement (NAFTA), are also linking their economies. The Pacific Basin, centered principally in Japan, but including the newly industrialized countries (NICs) and Association of South-East Asian nations (ASEAN), although not technically a free trade area, has become the fastest growing and foremost trading region of the world. Finally, no fewer than four free-trade blocs have been created or proposed in Latin America.12

This shift toward regionalization can be detected from the trends in trade and investment flows. The levels of intraregional trade have been growing faster than interregional trade. For example, the percentage of total EC exports flowing between member countries increased from 53 percent to 61 percent between 1975 and 1990.13 Similar patterns occurred in the Americas and in the Pacific Rim.

The evolution of a world trade system based on regional blocs creates incentives for firms to follow direct investment strategies that give them a manufacturing presence in each region of significant demand and unrestricted trade. The more formally managed nature of trade between blocs will mean that firms using export-based strategies will face additional administrative hurdles and potentially damaging regulatory barriers.

Risk Management and Exchange Rates

Exchange rates and other aspects of risk are forcing firms to be flexible in terms of capacity and locations and to view their global networks in a coherent way.

Exposure to risk becomes critical as firms develop global networks with multiple facilities serving many markets. If a company sells in a particular market, not producing in that market exposes it to a risk of currency depreciation, thus lowering revenues. Conversely, having a large production site in a country exposes the firm to the risk of currency appreciation. In addition, the firm may face a range of political risks from political instability to increased barriers to trade.

Clearly, the trend to global operations has heightened the impact of these economic risks. From a financial and operational point of view, increased flexibility reduces risk and can even reduce average costs.14 From a production viewpoint, such flexibility can be gained by having a number of plants serving demand, with the ability to vary plant loadings according to exchange rate trends.

While it is clear that a firm does not want to shift production cavalierly around the world, chasing every fluctuation in exchange rate, moving marginal production can be useful. That is, a firm can set up extra capacity so that it can execute relatively small volume shifts easily. In fact, evidence indicates that companies value such flexibility. In a recent survey, 63 percent of foreign exchange managers cited having locations “to increase flexibility by shifting plant loading when exchange rates changed” as a factor in international siting.15 In general, a company needs to examine the complexities of multiple markets, sources, and production stages through sophisticated modeling approaches that consider the entire supply chain.

Business-Level Trends and Implications

The macroeconomic trends discussed above suggest that a sound strategy for the global firm is to have a manufacturing presence in each region of significant demand. Business-level trends suggest that, while declining economies of scale facilitate this approach, the need for highly skilled labor and developed communication, transportation, and institutional infrastructures has increased. Organizational trends suggest that there are benefits to locating close to major customer markets.

Changes in Production Systems

The emergence of manufacturing technologies and methodologies such as flexible manufacturing systems, just-in-time manufacturing, and total quality management have reduced scale, increased the importance of worker education and skill, and placed demands on local infrastructure.

A flexible manufacturing system (FMS) integrates computer-controlled tools and material handling systems with a centralized monitoring and scheduling function. Such systems are most efficient when numerous different parts need to be manufactured in relatively small batches. They offer significant advantages over other manufacturing methods when the nature of product demand requires differentiation. Penetration of FMSs has been rapid. In Europe, the established base has grown at around 33 percent annually.16 In the United States and Japan, diffusion has been even faster; the number of installed FMSs appears to be doubling every two years.17

Two dynamics explain the increasing attractiveness of FMSs. First, product life cycles are rapidly declining, and customers increasingly prefer customized rather than generic products. The implication is that, over time, manufacturers will be forced to produce greater product variety within shorter lead times. Second, technical advances in FMSs are making them increasingly attractive to low-volume producers and more competitive with hard automation for high-volume production.

Just-in-time (JIT) manufacturing is a demand-pull production system first adopted by Japanese manufacturers and now used in many industries and countries. Based on a daily demand schedule, parts are pulled through each of the manufacturing steps, each process producing only to the demand of the succeeding process. In essence, the production process is better synchronized with customer demand, wasteful in-process inventories are avoided, and cycle times are subsequently reduced. One of the many benefits of JIT manufacturing is that these advantages can be realized with very little investment — and with wider applicability to a more diversified industry base.

The automotive and electronics industries were the torchbearers of non-Japanese JIT implementation. In 1986, a survey established that 71 percent of automotive products manufacturers had already instituted JIT, or at least a pilot program, and that 87.5 percent intended to do so within a year.18 Similarly, in 1988, a survey of top electronics manufacturers showed that 71 percent used JIT to some degree.19 This trend is not limited to the auto and electronics industries. A 1990 survey of 260 high-tech firms found that one-half had established JIT programs.20 JIT is also applicable to smaller businesses, suggesting that scale is not necessarily a prerequisite.

Total quality management (TQM) is a philosophy that changes the nature of an organization. Its principles differ from classic quality control in that it takes a proactive rather than a reactive approach to improving quality. At the heart of TQM methodology is the concept of continuous improvement. By constantly looping through the “plan, do, check, act” steps, improvements can be made to a process. Additionally, heavy emphasis is placed on understanding and incorporating customer requirements into daily job routines at every level. In 1989, roughly 26 percent of Japanese firms had adopted TQM.21 Similarly, a 1990 survey of senior managers in 260 U.S. high-tech firms indicated that 22 percent had TQM programs in all manufacturing areas.22

All three approaches tend to change the optimal configuration of production resources: they reduce the benefits to scale; they require well-educated, highly skilled workers; and they depend on sophisticated, well-maintained local infrastructures.

Reductions in Scale.

Studies of FMSs show dramatic increases in equipment utilization, thereby decreasing plant scale.23 In one study, a comparison of a U.S. and a Japanese automotive component supplier showed that the Japanese producer manufactured a similar product in a factory with one-third the scale and three times the variety at half the cost of the U.S. producer.24 In the days of U.S. supremacy in car production, a conventional automation system required annual volumes of a million units, relatively little year-to-year variation in demand, and a market of at least three years duration.25 FMS scale, on the other hand, corresponds to annual sales of roughly 24,000 units.26

Although FMSs are the most prominent hard technology for reducing scale, the general trend is toward a range of scale-reducing technologies and methodologies, which allow smaller facilities to be constructed without the cost penalty of operation at lower volumes. However, technology by itself does not provide the most significant advantages. For automation to be fully leveraged, it must be accompanied by improvements in production methodologies such as JIT and TQM.27

A survey of firms that have implemented JIT production shows 33 percent to 40 percent reductions in setup time, scrap, and downtime.28 Combined, these improvements increased utilization rates 26 percent and reduced the amount of workspace by 34 percent. TQM activities share similar results. When NEC IC Microcomputer Systems (NIMS) introduced TQM, productivity increased between 2.4 and 2.9 times every two years.29

A study quantified the benefits gained from production automation with simultaneous implementation of JIT and TQM.30 Focusing on mature industries, the study showed that savings, depending on the country in which they were implemented, ranged from approximately 10 percent to 20 percent. Interestingly, automation alone accounted for only 2 percent to 4 percent of overall savings, while improved manufacturing methodologies like JIT and TQM had created most of the improvement.

Through the productivity increases common to these new management philosophies and improved production techniques, a plant is able to produce to the same demand with less overhead. Plant scale is thus reduced, which implies that smaller, more focused plants can satisfy given levels of demand. In addition, overhead reductions decrease labor costs as a percentage of overall product costs. The cost structure of products is therefore primarily determined by material cost, equipment depreciation, capital charges, and overhead support costs. Many of these costs are not location specific, but rather are embedded in the product through both the technologies used within it and the processes used to manufacture it. Low-cost labor therefore becomes less of a consideration in location decisions.

These benefits do not come without concessions. All these methodologies rely heavily on workforce quality and skill. Thus, while the total number of production employees may decrease, the remaining workforce must be highly qualified.

Increases in Worker Education and Skill Levels.

The benefits of FMSs can be obtained only through a dramatic change in the nature of the workforce. FMSs are highly automated and thus reduce direct labor, relying instead on qualified engineers. In a typical system, engineers outnumber production workers three to one.31 In considering a location for an FMS or other highly automated manufacturing facility, the company must therefore consider the local engineering labor pool.

The success of FMS, JIT, or TQM also depends on the quality of direct labor. All employees must be highly flexible and multiskilled. For FMS, an ability to understand complex machinery and computers is essential. Successful JIT manufacturing requires that employees perform preventive maintenance, repairs, and complex planning activities. TQM also requires a highly skilled workforce, since the improvement tools make extensive use of mathematics and statistics. Softer skills such as team dynamics and proactive problem-solving techniques are also important.

Some companies have assumed responsibility for educating employees through internal training programs. Motorola, a company committed to TQM and JIT, spends approximately $60 million annually for internal training. As a percentage of payroll, this exceeds that of the average company by a factor of two.32 Nevertheless, Motorola recognizes that the workforce must have a knowledge base to enhance the internal training program’s effectiveness. All new employees are required to pass entrance tests and must have at least a high school diploma.33

While these methodologies may emphasize different skills, employees must be more highly skilled than the factory worker of a generation ago. When considering potential sites, firms must regard local employee skills as a key decision variable. Plant locations should be supported by an educational infrastructure that provides workers with basic science, math, and communication skills. Preferably they should also have some exposure to modern technology and manufacturing practices. Such factors suggest that progressive manufacturing practices will be more effective in developed regions, close to relatively sophisticated markets, rather than in less developed, low labor-cost regions.

The importance of labor skill and effective management philosophies has been corroborated in a recent empirical study comparing select mature industries in less developed countries (LDCs) to those in NICs.34 The authors demonstrated that the overall costs saved from locating in NICs exceeded the labor cost savings from locating in LDCs. In effect, the authors showed that NICs are more competitive than LDCs even though they have higher factor costs.

Trends in foreign direct investment parallel the conclusions of this study. In the 1970s, one-third of foreign direct investment went to less developed countries. In the 1980s, this proportion declined substantially; less than one-fifth of foreign direct investment went to less developed countries. Instead, investment flowed between the sophisticated markets of the EC, the United States, and Japan. The proportion of direct investment flowing between these regions increased from 30 percent to 39 percent during the 1980s.35

Importance of Institutional and Transportation Infrastructures.

The adoption of JIT policies increases a manufacturer’s dependence on its supplier network and support services. Complete production shutdowns can result if raw materials, components, or assemblies are delivered late. Therefore, a reliable institutional infrastructure is critical. In major industries like automobile production, it is possible to develop such an infrastructure, as Japanese automobile transplants in the United States demonstrated when they attracted component suppliers to the “transplant corridor” from Ontario to Tennessee. However, most companies possess neither the scale nor the leverage to create their own supplier network where they choose to locate. It is generally easier to choose a manufacturing location where an infrastructure already exists. Whether existing or attracted after the fact, such infrastructure is critical.

These considerations have influenced Motorola’s manufacturing site location decisions in the cellular telephone industry. Currently, 80 percent to 90 percent of Motorola’s vendors deliver directly to the line. In considering the location of manufacturing facilities, the presence of a sophisticated supplier infrastructure was of primary importance. Not all global manufacturing locations meet this criterion. A previous attempt to take advantage of cost-based manufacturing, through location in Puerto Rico, was unsuccessful.36 When faced with the decision of how to serve the European market, Motorola chose to produce cellular phones in Easter Inch, Scotland. Located in the “Silicon Glen” region of Scotland, vendor proximity and a well-educated labor pool were readily available.37

That JIT manufacturing requires constant contact with suppliers also places demand on the local infrastructure. Companies frequently request suppliers to make deliveries several times a day. This consideration tends to favor locating JIT-based manufacturing facilities in industrialized regions with developed transport and communications systems.

Organizational Philosophies

The need for organizational learning has increased the benefits of being close to all major, sophisticated markets.

During the 1980s, U.S. corporations awoke to the fact that technical superiority alone does not provide a sustainable source of advantage. Management approaches can have even more dramatic impact on performance, as the Japanese automakers demonstrated. Product innovations did not initially drive their success, but rather a new managerial philosophy that overcame scale disadvantages by reducing time to market, improving quality, reducing inventories, and increasing productivity.38

In the 1990s, however, both product and management advantages are no longer sustainable. For example, technical positions can easily be copied and even built upon, as demonstrated by PC clone manufacturers like Dell. As globalization increases, firms will increasingly find that static advantages based on traditional competitive positions are no longer enough. Those firms that are capable of learning and disseminating knowledge faster than their competitors will achieve superior performance.

A proponent of this view, Ray Stata, CEO of Analog Devices, argues that “the rate at which individuals and organizations learn may become the only sustainable competitive advantage, especially in knowledge-intensive industries.”39 Research in the shoe, bike, PCB, and steel industries supports the view that the key to establishing manufacturing excellence is to create a learning organization capable, for example, of assimilating new production methodologies.40 Companies that succeed not only will harbor the cost advantages noted earlier but, since knowledge is cumulative, will also be better poised for further incremental improvement.

Pivotal in establishing a learning organization is an awareness of global developments, suggesting that international information networks or linkages are a key factor in organizational and technical learning. The importance of creating “global feelers” is often underestimated. It could be argued, for example, that if U.S. firms had had a stronger presence in the Japanese market, the value of JIT and TQM philosophies would have been recognized and adopted sooner.

The importance of building a learning organization suggests that decentralization can benefit the firm. Such companies will have a network of feelers that can recognize new technological, market, and management trends. Although disseminating this knowledge through a far-flung, decentralized organization is difficult, firms that have experienced such difficulties and reembraced more centralized control are missing an opportunity. Those that harness the advantages of decentralization to create a dynamic learning organization will possess a formidable and sustainable source of advantage.

Toward a New Framework

Traditional approaches to production location no longer apply. Large, centralized manufacturing facilities in low-cost countries with poorly skilled workers are not sustainable. The trends we have discussed lead to a more decentralized manufacturing structure with smaller, lower-scale plants serving demand in regional markets. Location will depend increasingly on educational and institutional infrastructure.

These trends must be set in the context of the overall environment, with both internal and external constraints. Internally, the corporate strategy, the availability of capital, the characteristics of the existing site network, and the company’s growth strategy will influence its decisions. Externally, the location strategies of competitors will also have impact.

These factors can be synthesized into a new framework for setting location policy. We propose a four-phase procedure to aid decision makers.

Phase 1

Establish the critical success factors of the business, the degree of global orientation necessary, and the required manufacturing support role.

The first step in formulating a site location strategy is to examine how the firm competes. Managers should assess the degree to which the business strategy requires leading performance in each of four areas: cost, quality, innovation (time-to-market), and flexibility. The differences in emphasis between each of these factors will have significant impact on aspects of the location strategy.

For businesses where cost leadership is the sole driver of competition, location is important only as a driver for reducing transportation, labor, and inventory costs. For markets with a stronger emphasis on other factors, however, location has an expanded role: leadership in quality places higher demands on both the workforce and the suppliers. Competing through innovation and time-to-market requires customer proximity, a coordinated design-manufacturing link, and, potentially, local development resources. Achieving best-in-class flexibility may again require customer proximity and, potentially, a different set of production techniques and skills. Such requirements must be fully understood at the outset to help define many features evaluated in subsequent phases.

The second step is to forecast the likely evolution of the industry in terms of global market requirements. That is, can companies survive by serving only localized markets, or will they eventually submit to larger, financially strong companies, operating globally? In addressing this question, it is useful to examine three outcomes for an industry’s evolution: scale-driven consolidation, scope-driven consolidation, and niche/local market-driven fragmentation. The factors that determine this evolution are the underlying basis of competition, strategies of competitors, and the industry cost structure.

To develop a forecast of evolution, managers need to evaluate the drivers most likely to affect the industry’s development, and to what degree. For example, the presence of high economies of scale is often a key factor in the need to leverage greater product volumes. This tends to take effect at the plant level, suggesting ever-bigger plants serving multiple markets. Economies of scope also drive consolidation, but in a different manner; these are gained through scale in market access or capital resources. Large pharmaceutical companies, for example, have escalated research expenditures to levels that smaller companies cannot sustain. Those levels can be achieved only by serving many markets, across which investments are recouped.

The third step is a review of the internal constraints that may limit a firm’s ability to implement the optimal location strategy — primarily, the availability of investment capital and managerial resources. These limitations should be fully understood and defined at the outset. For example, while industry dynamics may suggest that global scale is a necessity for success, financial and managerial limitations may restrict the options available for expansion. In such cases, firms may be better served by seeking alliances to increase product volumes.

The nature of a firm’s growth strategy may also have substantial impact on location decisions. Building a coherent network of plants serving local markets can take more than ten years. Any company that generates growth primarily through acquisition is likely to become a significantly different organization over such time frames. Location is unlikely to be a strategic priority for firms that rely on frequent asset trading to increase shareholder value.

The final step in this phase is to synthesize the impact of all these factors on the overall manufacturing strategy. The business strategy, for example, will dictate to a great extent the type of production processes and technologies to employ: at one extreme, cost-sensitive markets may require large transfer line systems; at the other, locally responsive markets may require more flexible operations, located within markets, and with the ability to customize products for individual customers.

The output of phase one is a comprehensive assessment of the company’s basis of competition and the manufacturing capabilities for supporting it. These factors will be the primary determinants of whether a global manufacturing network is appropriate, and to what degree such a network can generate and sustain competitive advantage.

Phase 2

Assess options for regional manufacturing configuration, considering market access, risk management, customer demand characteristics, and the impact of production technologies on plant scale.

The first step is to assess political and market access requirements. To avoid tariffs and other nontariff trading restrictions, firms should locate within trading blocks. They should understand the nature and composition of the blocks, as well as their likely development. In addition, market requirements for exempt status, such as local content stipulations, should be well defined. Firms should assess other political issues, such as incentives for local investment in facilities or contract offset agreements, for an overview of the political imperatives. Obviously, the level of long-term political stability is also critical.

The second step is to assess the degree of risk inherent in serving the regional market. Assessing risk and the options for managing it encompasses both the forecasting of exchange rate movements and development of scenarios on product sourcing. Rapid fluctuations in rates can be hedged in the short term with financial instruments, but the value of production flexibility should be evaluated for longer-term changes, within and between given regions of demand.41 The value of flexible production capacity will influence the assessment of optimal site numbers and their specific locations.

Step three defines the characteristics of regional demand: the level of homogeneity in customer requirements across a region, the size of demand in each country, and the forecast for future development. The objective is to elaborate on the firm’s ability to serve demand from one point within a region versus the benefits of multiple sources. For example, in industries with a premium for local differentiation, markets may be better served by multiple plants, each with a high degree of production flexibility. When combined with the basis of competition, such characteristics will define priority areas of demand, potential plant numbers, and those areas most likely to be considered potential location choices.

The final step is to establish the effect of production technologies on plant scale in order to define a range of production parameters, such as the potential number of sites in each region, the ranges of product volumes, and the capacities for each stage of production. As we discussed earlier, developments in production process technologies have significantly reduced the benefits to scale at the plant level, thus enabling smaller plants to serve markets effectively. For a company competing on, say, innovation and time-to-market, maximizing the number of plants subject to minimum scale may be appropriate. Companies adopting such a dispersed strategy, however, must weigh the costs of complexity against forecast market benefits.

All these factors must be set in the context of competitors’ strategies. Many competitors will likely be attempting to establish a global presence simultaneously. First movers often have advantages, for example, in tying up local joint-venture partners, or establishing a “blue chip” distribution network. In fast-growing regions, such dynamics can lead to preemptive capacity expansion, committing investments to markets with less volume than appears economically viable.

Phase 3

Define a set of potential sites, primarily based on infrastructure, that adequately supports the business and manufacturing strategies.

The first step in this phase is to assess the production methodologies for each location. This assessment will dictate demands on the workforce, the supplier base needed, and the requirements of transport/communication networks. As discussed, new techniques such as JIT, FMS, and TQM give more responsibility to line workers than traditional systems do. They also make more stringent demands on local suppliers and physical infrastructure.

The components of site infrastructure should be separated into “hard” and “soft” requirements. Hard or tangible factors relate to the physical availability of supplier, communication, and transport systems. These factors are influenced by the nature of the manufacturing system and the degree of vertical integration. Softer factors relate to organizational and educational infrastructure, for example, inherent workforce education levels, or suppliers with specific technical know-how. These factors, often overlooked in traditional location algorithms, are among the most important for future sources of competitive advantage.

The second step is to examine, on a pro forma basis, the levels of infrastructural development for each broad location option defined in previous phases. For example, the need for trained technical staff might suggest a location near specific institutions or science parks. On this basis, the range of potential sites can be narrowed to specific areas in a region. In practice, there will be many other such factors to integrate in the analysis. The resulting options are those that best meet a range of infrastructural parameters, yet do not fail to meet any that are critical.

The output of this phase is a detailed evaluation of specific sites that have sufficient infrastructure to maintain the firm’s basis of competition.

Phase 4

Rank the most cost-effective solutions, using a quantitative analysis of remaining location options, and define the manner of operation.

After defining a range of sites with suitable infrastructure, as well as the potential ranges of activity for each site, the company should establish the size of the facilities, the specific locations, the product flow from source to market, and the sourcing and location of specific products and processes. Computer-based models should be used to detail the manufacturing network. The company should decide which facilities to use and which production and distribution activities for each time period to handle at that location. Specifically, it should define a decision level for each combination of plant, market, and production stage for a given product, as well as a distribution path. For a given scenario of exchange rates and facility options, a numerical algorithm would identify activity levels. The company should then iteratively evaluate currency scenarios and major strategy options. While the details of such an approach are beyond this paper’s scope, an example can give a sense of the nature of the decision variables.

A pharmaceutical manufacturer had already selected sites, but it needed to determine product sourcing and material flow. It had a mixture of plants in the United States, Europe, Asia, and Latin America and made syrups all over the world, pills and capsules in two U.S. plants, Puerto Rico, and Ireland, and animal-based products at a different U.S. plant. The plants were each a jumble of different products and processes, and detailed costs and levels of scale were little understood. While scale was important for some products, others could be made efficiently in different markets. A decision support system, based on the classification of products into twenty-five groups, showed the ideal locations and volumes for each of the products and processes. As a result, the U.S. plants were better focused on different types of products and process steps for some products (the latter when different scale and technology needs characterized the process steps for a given set of products). The two tax havens of Puerto Rico and Ireland were focused on their most appropriate products.

Toward a Model of the Global Manufacturing Company

Three basic principles will direct global decision makers’ thinking as they reconsider how to make manufacturing location decisions: increasingly, companies will serve all major markets; manufacturing will be based on regional presence; and infrastructure will be more important than pure “cost” factors.

Establishing Global Product Volumes

Global companies of the future will serve all major markets, leveraging economies of scope to undermine the competitive strength of domestic players. These dynamics are being brought about through the evolution of fixed costs as the primary determinant of a firm’s cost structure. Although manufacturing scale is decreasing, other effects are playing out in areas such as technology and marketing.

In high-tech products, scale in technology has become more relevant than scale in manufacturing. Investment levels have risen in the drive for rapid innovation, so industry players have sought more markets across which to recoup R&D expenditures. In pharmaceuticals, this evolution has already occurred. Industry giants such as Merck have raised the stakes through massive R&D investment. During a short time, the industry underwent substantial transformation through mergers, acquisitions, and divestments. Now there are few, if any, purely national players.

In consumer products, financial scale is playing a transformative role. Global competitors use financial scale to their advantage by managing cash flows between different product segments or national markets. A giant like Procter & Gamble can gain entry into a local market by undercutting domestic competitors and investing heavily in promotion. Local players cannot compete with P&G’s deep pockets without significant profit penalty.

Establishing global product volumes does not mean that products have to be homogeneous across markets. Scale in technology can be spread across locally tailored products, differentiated products, or, indeed, different product families. Honda, for example, has unique capabilities in the design and manufacture of engines. It uses this to great advantage across a line of businesses, from automobiles to power boats to lawn mowers. Honda’s competitive advantage now lies in fast-cycle innovation and in the scope of its operations. In essence, economies of scope rather than scale are now the driving force in global competition.

Establishing a Decentralized Manufacturing Network

The global company of the future will establish a manufacturing presence in each region of significant market demand. These facilities will be smaller and more flexible. Trends in trade and investment incentives are exerting powerful pressures in this direction, and new production methods and management techniques facilitate it through reduced economies of scale and increased flexibility. Regional manufacturing plants will benefit from rapid customer feedback on product performance and will help efforts to build a learning organization through exposure to multiple markets.

Once a manufacturing presence is established in a market, the number of sites for serving the region becomes a function of the classical trade-off between scale benefits and transport costs. Where scale benefits are reduced significantly, there is potential for multiple plants to serve a single region. Currently, the major world markets are North America, Europe, and Asia-Pacific. Competing in these regions is essential for the global player. As developing economies grow to significant size, however, other regions of critical scale will emerge.

Emphasizing Infrastructure over Cost

Increasingly, the location of manufacturing sites will be based less on the classic measure of labor cost and more on manufacturing infrastructure. Companies will locate facilities closer to final markets, where the workforce has the necessary skill and knowledge base.

The labor requirements of new systems and techniques are driving the need for a better educated direct-labor workforce. FMS, JIT, and TQM systems place greater importance on the flexibility of the workers and their ability to operate under growing autonomy. The increasing sophistication of product and process technologies has also increased skill requirements.

As new production systems have reduced the amount of direct labor in product cost, capital has become proportionately more important. Capital costs are essentially the same the world over, and therefore increasing the capital intensity of the production process decreases the sensitivity to site location. Companies continuing to focus on direct labor cost savings may find transitory advantages, but eventually, as has happened in Korea, cost pressures will wipe out such advantages. Given these factors, locating facilities closer to final markets, with less emphasis on labor cost, is a natural progression.

The final factor in site infrastructure as the key determinant of production location is the need for a sophisticated vendor network. JIT systems require a supplier base that is capable, reliable, and physically close (around two hours or 100 miles). Such requirements dictate that existing industrialized regions will increasingly attract the investment of global manufacturing firms. Deploying manufacturing resources in larger, sophisticated regions will allow leverage of existing suppliers’ capabilities. This will aid entry strategies that rely on significant outsourcing during initial production while higher value-added capabilities are developed.

Some Caveats

Although the trends affecting production location are pushing all firms toward globally dispersed manufacturing, their effects will not be felt equally by all industries, and not in any one industry at all stages.

Young Industries.

Industries still in the early stages of their life cycle will tend to adopt more centralized manufacturing networks than industries in more mature stages because of the rate of change in product and process technology and the nature of product demand.

Early on, the technology involved in a product fluctuates as firms are still making design innovations. At the same time, the process technology for manufacturing evolves through a flexible, general purpose form toward a more efficient, capital-intensive process.42 The control of this evolutionary process provides many challenges to the firm, such as maintaining a strong link between manufacturing and development. The firm’s manufacturing systems must be flexible enough to incorporate the changes, and the workforce must be able to understand and adapt to different technologies.

In industries characterized by this fluidity, managing a dispersed manufacturing network is particularly complex. As each innovation occurs, it has to be rolled out coherently. Possible major changes in product and process technology could render existing investments inefficient or obsolete. In such an environment, therefore, there is a tendency to limit the amount of decentralized manufacturing to a level below that expected in the long term.

Although these dynamics also occur in life cycles at the individual product level, the manufacturing location decision is usually concerned only with the dynamics at the industry level. In a mature industry, even new products should have relatively stable technologies, and the product rollout decision is driven by the need to access all global markets quickly. Thus a decentralized strategy makes sense.

Products with High Unskilled Labor Content.

For products with unavoidably high levels of direct labor, the tendency will be to choose lower-cost locations near final markets rather than the final market itself. In this way, advantages of proximity are not diminished significantly. In Europe, for example, assembly operations would conceivably be located in Portugal, rather than in a more developed country.

A high labor content, however, should not immediately be viewed as sufficient grounds for a low-cost country strategy. New production technologies are getting cheaper and provide many other advantages in terms of flexibility, quality, and consistency. Successful companies will drive changes in production technologies, rather than tacitly accepting the existing way of doing business.

Production Systems with High Minimum Efficient Scale.

Industries with large scale-intensive elements will tend to consolidate these operations and, potentially, split the supply chain. Where these operations constitute the major part of the manufacturing process, large, scale-driven plants may still exist. In the semiconductor industry, for example, wafer fabrication is tremendously scale intensive because of the immense investments in equipment — between $500 million and $1 billion. Most manufacturers, therefore, consolidate all such processing to a few major facilities, while relying on local assembly and tests to satisfy local content regulations.

Although a significant minimum efficient scale in a process makes a strong case for the breakup of the value chain, firms should be careful not to take these economics as given. A thorough investigation of all production methods is needed. For example, the advent of FMSs substantially changed the economics of transfer line systems. Decades ago, U.S. automobile firms setting their facilities strategy developed large-scale myopia. The Japanese have since shown that economies of scale are not static; they tend to move in one direction — down.

Conclusion

In this paper, we have examined recent trends in global competition that affect the location of international manufacturing sites. Advances in technology, changes in management philosophies, and shifting market requirements are the new dynamics that shape a firm’s facility network. These trends suggest that global corporations of the future will move to a manufacturing network of decentralized plants, based in large, sophisticated regional markets. Specific locations will be based more on local infrastructure, such as workforce capabilities, than on cost-based considerations. Each plant will, in general, be smaller than current ones, yet have significantly more flexibility to produce multiple products.