Japanese subcontracting is complex and evolutionary, a result of the interplay of historical events and human agents. Consequently, no single theory —whether dualism, flexible specialization, transaction cost economics, or cultural specificity (which we discuss later) — is sufficient to explain it. In this paper, we detail the evolution of Japanese industrial sourcing and view subcontracting as a movement toward collaborative manufacturing based on problem-solving principles.1 We argue that the evolution is best explained by political, economic, technological, and strategic factors, not by a single theory. Today, the major advantages of Japanese subcontracting are the economic benefits of interfirm problem solving that ensures the continuous production of high-quality, low-cost products.

Evolutionary Development

Japanese subcontracting after the Meiji Restoration can be divided into four periods: early, wartime, post-war, and modern. Each period is distinct, which makes generalizing hazardous. Taken together, however, they are evidence of the importance of political, economic, technological, and strategic factors in the evolutionary development of subcontracting.

Dualist System Emerges

In the nineteenth century, labor immobility in Japan resulted in substantial regional wage differences. By the turn of the twentieth century, however, such differences had narrowed, and workers frequently moved from one employer to another seeking better working conditions. There is little evidence of company commitments to lifetime employment at this time.

We can trace the emergence of segmented labor markets in Japan to a period just after World War I, when the labor market divided into two areas: (1) large firms, especially those in the heavy manufacturing industries, and (2) the rest of the economy. This stabilization resulted from a deliberate management strategy. Business expansion during the wartime economic boom made it imperative that, along with extensive investments in facilities, labor be more finely differentiated. Although the navy arsenals and Mit-subishi Shipbuilding had already adopted this practice around the time of the Russo-Japanese War (1904-1905), it was not common throughout the economy until World War I.

Along with the division of labor came firm-specific skills. How jobs were classified and tasks defined differed substantially from one firm to another, so workers had to be trained according to new requirements. Of course, it would make no sense to train workers if they left for better opportunities. Worker education, therefore, needed to coexist with a more stable labor situation.

A powerful way to stabilize labor was to link length of service and incremental pay and also introduce retirement pay schemes. As a result, wage differentials in the various segments of the economy expanded, large firms began to hire temporary workers extensively, and a dualist system emerged.

Government Mandates

The roots of Japan’s subcontracting are found in its industrial reorganization during World War II. Several factors were responsible for the development of subcontracting, including a burgeoning demand for munitions, the availability of cheap labor, changes in infrastructure and technology, and politics. Demand continued to grow along with Japan’s expansionist involvement, and in 1937, when military requirements skyrocketed, output in the machinery industry followed suit, expanding from ¥1,609 million in 1936 to ¥4,359 million by 1938.

Japanese manufacturers turned to subcontracting, which had become feasible on a large scale due to improvements in the industrial infrastructure and technology. The expansion of national transportation and communication systems made small firms’ output more marketable. Standardized manufacturing processes and mass production methods gave rise to simple machining and drilling jobs for subcontractors. Electrical power and increased use of electrical or gasoline-powered machines in small firms moved them away from traditional craft production. All these changes permitted even the smallest family operation to be integrated into the standard routines of large, advanced firms.

Between 1940 and 1945, in a struggle to establish an independent economy, the Japanese government, through the Ministry of Commerce and Industry (MCI),2 intervened to rationalize subcontracting systems. The ministry mandated that companies should promote specialists, divide production by specialty, and phase out production of overlapping components by major contractors. It required subcontractors to maintain dedicated relationships with designated “parent” contractors and established subcontractors’ organizations (kyoryokukai) controlled by their prime contractors. In 1943, the government created enterprise groups (kigyo shudan).

In the aircraft industry, the government carefully mapped out machine tool production. In 1944, the Ministry of Munitions designated hundreds of firms that it intended to control directly and ordered them to manufacture aircraft components and other armaments. Shortly thereafter, the government rationalized vertical enterprise channels (kigyo keiretsu), imposing a vertical hierarchy between prime contractors and subcontractors.

MCI, however, did not implement its plans well. Companies did not like the rules, and there was little coordination. Sub-subcontracting continued, and dedicated subcontractors often switched or served multiple customers. Although government aims proved unrealizable, the rules eventually had a major impact on subcontracting in Japan. The mandates initiated a considerable transfer of technology from large companies to small ones and established a pyramidal structure in subcontracting relations.

Expanded Subcontracting

After World War II, demand for munitions abruptly ceased, and investments in new territories were lost. There was a huge influx of demobilized soldiers and an enormous scarcity of basic materials, including food and clothing. The nation’s manufacturing capacity was temporarily paralyzed.

In 1945, the Supreme Commander for the Allied Powers, General Douglas MacArthur, ordered the Japanese government to promote the production of basic commodities, which created a shift to nonmilitary manufacturing. The first Japanese labor union laws were passed, guaranteeing most workers the right to organize, collectively bargain, and strike — rights further protected by the constitution.

Under such protection, unions flourished, evolving into firm-specific unions with both blue-collar and white-collar members. “Us” became those people within the firm, rather than the working class. “Them” referred to competitors or, more generally, those outside the firm. By structuring labor interests with a subtle interplay of competition and cooperation, enterprise unions had an effect on the subsequent development of Japanese industrial relations.

Because unions were so well protected, management could no longer arbitrarily fire employees. Given this constraint, the logic of using more flexible labor resources outside the firm as a buffer was compelling. Unions resisted this trend, but the 1949 Dodge Plan put them in a very weak bargaining position. Joseph Dodge, financial adviser to MacArthur, recommended stringent, deflationary policies to abolish subsidies to Japan, which paralyzed many firms’ cash flow. Between 1949 and 1950, there were more than 1,100 bankruptcies and more than 510,000 redundancies.

Management achievements during this period produced a new equilibrium of power, with wage differentials widening and the number of union members declining, particularly in the small manufacturing sector. In the immediate postwar period, wages in small companies had been as good as or better than in the large firms, due to small firms’ maneuverability in the black market. By 1948, compensation in large firms had surpassed that in smaller firms.

The effects of the Dodge Plan, however, were soon superseded by the Korean War (1950-1953). Munitions demand for United Nations forces skyrocketed, and Japanese industry, close to dying, experienced an immediate rejuvenation. The iron and steel, heavy machinery, and electric power industries benefited. For example, from the mid-1950s, electrical manufacturers began to set up full-scale mass production facilities and expand their subcontracting base.

As Japanese subcontracting grew, suppliers were squeezed by lower prices and deferred payments. Prime contractors adopted multiple sourcing strategies, strengthened their bargaining power over price, and increasingly used temporary workers in a clear manifestation of a dualist strategy. To cope, first-tier suppliers segmented their jobs and assigned their simple, messier, more labor-intensive jobs to second-tier suppliers, earning a small profit. As a result, a fine-tuned division of labor among clusters of interrelated manufacturing units was established.

Politics, too, played a significant role in stabilizing the small-firm sector. Because managers and workers in small firms were a sizable source of votes, it was in the state’s interest to secure their support. Those efforts helped create a favorable environment for the transformation of Japanese subcontracting.

The Small and Medium Enterprise Agency (SMEA), an independent organization within MCI, was responsible for three important policies in the 1950s: legislation prohibiting the discriminatory treatment of subcontractors, the promotion of cooperatives (kyodo kumiai), and the establishment of financing institutions for small businesses.

  • In 1953, the Fair Trade Commission (FTC) announced twelve types of unfair trade practices.
  • In 1949, a new law allowed small firms to collectively purchase new machinery and technologies, facilitated borrowing by aggregating individual risks, and provided a way to administer government policies. The new cooperatives helped small businesses, hard-pressed for money under the Dodge Plan austerity, to get loans otherwise unavailable. By 1960, such groups were widely called “borrowing money” cooperatives.
  • The government also established three financial institutions specializing in small-business financing in the 1950s: the Central Bank for Commercial and Industrial Cooperatives, the People’s Finance Corporation, and the Small Business Finance Corporation. Furthermore, it introduced a Credit Guarantee Association to underwrite small firms seeking credit from ordinary financial institutions. Finally, it set up the Small Business Credit Insurance Corporation, an agency to automatically reinsure the debts of the association.

Overall, this legislation made it easier for small firms to obtain the capital they needed to grow. Given a viable small-firm sector coupled with continued labor difficulties, large firms continued to expand their subcontracting.


Japanese subcontracting, as we know it today, is a product of the high-growth economy that began in the early 1960s. Since then, Japan’s domestic market has had rapid, continuous growth, heavy competition, and product proliferation.3 A strong upward trend in real disposable income has allowed continued dissemination of consumer products. “Congested” competition among many players, each constantly replacing product lines, has occurred in many Japanese industries, especially electronics and motor vehicles.

For example, in 1967, Toyota produced 670,000 motor vehicles on 33 different platforms. In comparison, in 1965, General Motors’s Chevrolet division produced 1.65 million Impala, Bel Air, and Biscayne models, all on the same platform. Similarly, the engine, transmission, and various features and colors of a Toyota Corolla jumped from 72 in 1966 to 5,200 in 1987. In short, increased product variety coupled with shorter product cycles has made manufacturing considerably more complex.

The fact that subcontracting continues to play an important part in the Japanese economy is itself revealing. As demand surged in the 1960s, labor became scarce, and the wage differential between the large and small company sectors of the Japanese economy narrowed (see Figure 1). Yet, even without the attraction of low wages, there was no decline in the small-firm sector. In fact, the share of added value in the small-firm sector remained stable at about one-third the national aggregate from the 1960s through the 1980s. Meanwhile, the number and proportion of small firms that subcontracted steadily increased in Japanese manufacturing industries (see Table 1). Moreover, the percentage of temporary workers noticeably decreased from 7.7 percent to 4.5 percent between 1959 and 1965. In short, subcontracting continued but less along dualist lines.

Understanding the context here is essential. By the late 1950s and early 1960s, large Japanese producers faced difficult challenges in manufacturing complexity, and small firms were constrained by technical and financial resources. To meet these challenges (and plan for impending liberalization of foreign investment regulations in Japan), large manufacturers began to invest heavily in their subcontractors. In the process, Japanese producers developed a distinctive sourcing strategy to delegate work to major subcontractors.

In general, major firms concentrated their resources on strategically important products and processes, while they delegated the assembly of more mature products, small batch items, and the manufacture of subsystems to subcontractors. Often, subcontractors, originally responsible only for discrete treatments, were strategically converted over time into subsystems manufacturers and contract assemblers. Outsourcing relieved assemblers from increasingly complex operational and administrative tasks. Moreover, many practices that we now associate with modern Japanese subcontracting (clustered control, long-term contracts, target pricing, collaborative engineering, black-box design, and so on) developed during this period.

Strategic Industrial Sourcing

Modern Japanese subcontracting relies on clustered control, in which firms at the top buy complete assemblies and systems components from a concentrated base of first-tier subcontractors, which in turn buy specialized parts from a group of second-tier subcontractors, which buy discrete parts and labor from third-tier subcontractors, and so on. Distinctive practices and routines have developed around this system of clustered control and joint problem solving. Taken together, they have moved Japanese subcontracting from a system of exploitation and “win-lose” decisions to collaborative manufacturing.

· Target Costing.

Target costing of new product development may be the most critical practice. By using the target cost (or cost planning) method, Japanese automakers have attempted to lower costs of new products at the design stage. The method uses “market price minus” rather than “cost plus” principles. First, an automaker determines the sale price of a new car model. Second, it decomposes the price into desired profit and costs. It then breaks down costs to evaluate and price every part. Throughout this process, the company encourages proposals from suppliers because of their intimate knowledge of the part. A process of continuous innovation, it often results in annual (or even semiannual) decreases in component prices. When the automaker and supplier jointly evaluate various possibilities in view of customer needs and desires, they can reduce the combined costs of the parts without degrading specified part quality. (For a view of the evolution of joint price determination in Japanese manufacturing, see Table 2).

· Value Analysis.

Value analysis (VA), a technique to decompose increasingly complex cost structures so that cost-sensitive elements can be identified item by item is essential to Japanese target pricing. Companies use the technique at both the prime contractor and subcontractor levels. Customers often request detailed cost data from subcontractors in order to look at possible cost reduction at the source.

The Japanese have used such joint problem-solving exercises to both parties’ mutual gain. This has been possible, however, only with the development of reasonable profit-sharing rules between purchasers and vendors. If, for example, the agreed-on price for an instrument cluster was 120 points for the first car-model year, and the cost was reduced to 110 points (target price for the second year) through “joint” problem-solving efforts, the assembler would pay the supplier 115 points, thus sharing the profit evenly. If, however, the cost was further reduced during that period, say, to 108 points, the balance would go to the supplier. By these rules, an assembler would not ask for a price that was less than the second-year target price. A decrease in price, however, would be expected throughout the life cycle of the car model, with price rectification as often as every six months.4

· Bilateral Design.

As the practice of bilateral design has spread, supplier-driven innovations have become increasingly significant. The modularization of wire harnesses is one example. In a typical passenger car, the electrical wires for the dashboard alone doubled from 70 in 1973 to 140 in 1985. For a luxury model, there were approximately 1,500 to 1,600 electrical wires per vehicle in 1988, which increased the weight, bulk, and manufacturing complexity of the wire harness assemblies and potentially compromised their reliability.

Seeing the problem as a possible profit opportunity, several Japanese wire harness suppliers proposed dividing the one unit into blocks and connecting them by small junction boxes in plastic housings. They also suggested incorporating microchips and microprocessors into the junction boxes, resulting in a new “intelligent” wire harness system to replace the conventional, bulky wire assemblies. This new modular design radically reduced cost, weight, bulk, and manufacturing complexity, while substantially enhancing reliability. More important, it helped accommodate frequent design changes throughout the model cycle of a given car.

The result — a substantial cost reduction and an increased capability for design changes — was not achieved by attention to scale economies or downstream price negotiations, but rather from suppliers’ proposals. By gaining advanced expertise in modular design, Japanese suppliers also benefited. National statistics on the prevalence of subcontractor-developed technologies reinforce this point. In 1983, one-third of the subcontractors surveyed provided their own techniques to customers (see Table 3).

Another indicator of the extent of bilateral design is the prevalence of black-box design, in which the customer provides basic ideas and size and performance specifications for a component, and the supplier uses its expertise to attend to the design details. In sharp contrast to practice in the United States, where, in 1987, assemblers still controlled three-fourths of component design details, Japanese auto component development relies on black-box design.5

As an aid to bilateral problem solving, Japanese suppliers regularly assign engineering groups to customer facilities. Often they send several engineers who become part of a project team two to three years before assembly production. Approximately one year before mass production, production managers from the assembler and suppliers run pilot productions, frequently along with second-tier subcontractors. They decide on final layouts and process technologies approximately six months before launching a new car model. As a result, there are generally few design changes after launch. In contrast, design changes in the United States tend to skyrocket immediately after launch.

In Japan, after launch, resident design engineers usually return home, while resident production engineers stay to observe components’ assembly into cars and suggest changes. These production engineers are, in a sense, customer-specific intelligence agents who do preventive maintenance and problem solving. They gather information about their customers’ long-term product strategies. The feedback they provide to the home supplier is divided into three categories: what to do immediately, what to adopt in the next model change, and what to discard. The engineers also relay the suppliers’ comments to the customer.

· Subcontractor Evaluation.

While Japanese companies tend to share more design work with their suppliers than companies in the United States do, they do not do so blindly. The prime contractor continually evaluates subcontractors’ performance on product quality, price, delivery, engineering, management competence, and long-term viability. In general, the contractor gives subcontractors grades and detailed scores that indicate weak areas for improvement. It gives those with better grades more responsibilities and long-term commitments and those with poor grades a chance to do better. If they fail to improve after several checks, however, poor performers may be either discharged or forced to become lower-tier subcontractors. The move toward clustered control is thus aided by the application of systematic performance measurement.

Subcontractor grading is closely linked with quality assurance. Those companies with top grades are designated “self-certified,” meaning that they are wholly responsible for achieving a zero-defect rate and that their customers will no longer routinely inspect the parts they deliver. In this way, Japanese companies have delegated not only assembly and design tasks to the subcontractors but also testing and quality functions.6

· Purchasing Agents’ Role.

Widespread evaluation of suppliers coupled with suppliers’ increasingly sophisticated design skills have substantially changed the purchasing agents’ role. No longer are they merely downstream price negotiators but evaluators of subcontractor performance, interfirm coordinators, and teachers. At Toyota, all purchasing agents acquire sufficient technical knowledge to evaluate subcontractors’ technical competence and teach them the principles of the Toyota production system.

Although implementing all these practices has dramatically transformed Japanese subcontracting, many subcontractors have welcomed the new system, with its more stable relations, greater opportunities for technical learning, and improved growth prospects. In 1987, approximately 60 percent of all subcontractors in the Japanese electronics, transportation equipment, and machinery industries were involved in joint design with their customers — ready proof of a radically new model of subcontracting.

Theories on the Evolution of Subcontracting

Several theories have attempted to explain the evolution and growth of Japanese subcontracting, including liberal and radical dualism, flexible specialization, transaction cost economics, and cultural specificity. As we review the theories, we find that all have critical shortcomings and, because the historical origins, economic rationale, and organizational functions of subcontracting differ so dramatically over time and place, no single theory may be sufficient to explain it fully.


The core of dualist theory is that society may be segmented into separate and unequal parts. Dualists see inequality between internal and external labor markets, primary and secondary industry sectors, and core and peripheral economies. Economic agents —whether workers or firms — located in different segments of the economy are treated unequally, regardless of their “objective” worth. Neither workers nor small firms in the secondary sector receive in payments what they should. Screening effectively discriminates outsiders from insiders. Internal workers are promoted according to clearly defined administrative rules for mobility. Small firms on the periphery are discriminated against in terms of access to technologies, capital, and human resources. When the economy is booming, the periphery firms are used extensively, but when it contracts, they are the first to be forced out.

Berger and Piore have proposed a dynamic theory of dualism in which the strategic use of subcontracting, by shifting many productive processes and therefore risks to the secondary sector, would help large firms survive uncertainty and change.7 They cite examples of companies in France and Italy that pursued such a strategy in response to labor inflexibility and unrest in the primary sector. The liberal school sees the flexible use of subcontracting as a response to uncertainty.

In contrast, Edwards has proposed that differentiated labor markets result from employers’ attempts, in a capitalistic economy, to break up the working class.8 From this perspective, fragmentation of the working class through the creation of artificial inequalities and gradation between different factions is essential to bourgeois dominance. The radical school emphasizes a class-conflict aspect of dualist strategy from a Marxist perspective.

Both theories lack an analysis of the properties of products, markets, and producer strategies that constrict the use of subcontracting. Technological requirements, for example, may even differ for the same product, depending on the area of processing. Moreover, some dualist literature uncritically assumes that opportunities to exploit advanced technology are unequivocally limited to the primary sector.9 Furthermore, dualism often implies a retreat of the small-firm sector that parallels larger firms’ domination. As we have seen earlier, however, evidence rejects this notion.

Neither subcontracting nor dualism were essential features of the Japanese economy at the beginning of the twentieth century. The formation of internal labor markets began only around World War I, and not until World War II is there evidence of extensive subcontracting in Japan. Dualism, however, does help to explain events from the immediate postwar period through the late 1950s. After the Korean War, the war-shattered institutions of subcontracting were revived in a new guise. At the same time, the harsher side of postwar subcontracting also became apparent in the increased use of temporary workers, unilateral price reductions by prime contractors, and pay delays. Other than some embryonic efforts in the auto industry, there were few signs of collaborative manufacturing among Japanese firms. This situation was only temporary, however, as government intervention helped to promote small businesses and stabilize the Japanese subcontracting system. Combined with high-growth beginning in the 1960s, these factors led to a transformation of subcontracting relations in Japan.

As a traditional argument explaining Japanese prosperity, dualism posits the sacrifice of many small subcontractors. Small firms are characterized as “sweatshops” with cheap labor and labor-intensive technologies. If this claim were correct, the competitiveness of Japanese industry might have declined when inter-scale wage differentials substantially narrowed in the mid-1960s to international levels (see Figure 1). If Japanese subcontractors were only an exploited pool of cheap labor, they might have withered away, as the dualists predicted.

Empirical evidence rejects these contentions. Despite a narrowing wage differential (see Figure 1) and a stabilized union density, small firms continue to keep pace with larger firms. During the past three decades, small Japanese firms have consistently contributed one-third of the value added in the Japanese economy. Their share of the total working population has been three-fourths. Moreover, the number of small manufacturing firms that subcontract has steadily risen (see Table 1).

Small firms in Japan not only have grown in parallel but also have connected with large firms via subcontracting. As we have already indicated, producers’ distinctive strategy to manage the demands of increasing product proliferation and manufacturing complexity has led to increased outsourcing and subcontractor development in Japan. Over time, many subcontractors have been converted from single discrete-process specialists to contract assemblers and components manufacturers. This arrangement has allowed primary producers to focus on strategic activities while providing subcontractors with new opportunities to learn and develop technical expertise (see Table 4).

While workers in small firms do not earn as much as those in large firms, there is an institutional outlet for workers in small firms. In Japan, a man typically starts working for a medium-sized manufacturing firm after leaving high school. He stays there until he accumulates enough knowledge and money to become independent. Then, depending on his ambition and other contingencies, he opens his own subcontracting firm.

While, in most accounts, the livelihood of subcontractors is generally depicted as “miserable,” Nishiguchi’s research belies such stereotypes.10 Many subcontractors in the study stood out as highly motivated individuals content with their lives and circumstances. As one subcontracting entrepreneur said, “Workers in large firms may have security, but no great money. In my business, I have less security but a lot of money. The more you work, the more you earn. The more clever you are, the more business you get. There’s no limit.”11

Flexible Specialization

While we could also consider flexible specialization a form of dualism, its reasoning and implications are so different that it is best to discuss it separately. The main proponents of flexible specialization, Piore and Sabel, argue that the mass production technology and existing regulatory institutions based on mass production principles are crumbling because of their diminishing adaptability to emerging externalities.12 If industrial society is to prosper again, they say, it must instead adopt institutions and production methods based on the principles of flexible specialization, a derivative of craft production that lost to the mass production model in the nineteenth century.

Piore and Sabel argue that there are two alternative strategies. The first builds on mass production technology but is inherently inflexible in its use of dedicated machines and semiskilled labor to produce standardized goods. The second is to adopt flexible specialization technology, which allows skilled workers to use general-purpose machinery to turn out a wide, constantly changing assortment of goods for shifting markets. As a result of the diffusion of manufacturing flexibility in the small-firm sector, they see an impending reversal of roles between the hitherto dominant and the subordinate sectors. Their model stands the regnant paradigm of production on its head. Dominant sectors of the established system will be subordinated; subordinate sectors will become dominant.

Piore and Sabel have identified the importance of interactions among product markets, technology, and organizations, with technology as a key variable for success or failure. Their argument for flexible specialization, however, suffers as a result of its rather dichotomous division and reversal of roles between large-and small-firm sectors and its confrontational positioning of the two technological paradigms.

In particular, Japanese subcontracting since 1960 does not seem to fit the framework of flexible specialization. Given that the value added by small firms in the Japanese economy has stayed roughly constant during that time, it is hard to see any sector reversal. Moreover, the success of lean Japanese manufacturers and their supplier hierarchies seems to indicate more an accommodation than a confrontation between flexible specialization and mass production. Cars and electrical appliances are still produced en masse under the lean production system. At the same time, flexible manufacturing by lean Japanese suppliers has improved the productivity of the whole system,13 with the increasing coordination between the two sectors refuting any notion of separate sectors (see Table 1).

To their credit, the importance that Piore and Sabel have attached to technology has made us rethink the concept of flexibility. Whereas the liberal school of dualism argues that flexibility is primarily the ability to hire and fire at will or change work rules, Piore and Sabel see flexibility as the supplier’s ability to accommodate increasing product mix, design changes, and quick delivery without prohibitive costs. They further note that such flexibility is not automatically produced by contracting out. Because the traditional dualist approach does not distinguish the benefits of exploiting subcontractors (as, say, buffers) from their flexible manufacturing capabilities, the benefits of Japanese subcontracting go far beyond what liberal dualists understand to be the benefit of “flexibility.”

Until now, differences in understanding subcontractor flexibility have been masked by a remarkable lack of data on the performance of component suppliers (which contrasts strongly with what is available on assemblers in the automobile or electronics industries). If we understand flexibility as the ability to get the right goods to the proper place at the desired time at minimal cost, however, from Nishiguchi’s data, we can see that substantial differences exist in the flexibility of auto component producers in the United States, Japan, and Europe (see Table 5).

Transaction Cost Economics

Williamson’s discussions14 of transaction cost economics constantly refer to bounded rationality and the assumption that human nature tends toward opportunism.15 The pairing of bounded rationality with small numbers exchange (e.g., a single buyer versus a single seller) and opportunism with uncertainty gives rise to what Williamson calls “information impacted-ness,” which causes exchange difficulties (i.e., trading becomes difficult based on market mechanisms), possibly even market failure. One recourse to coping with these market costs is the establishment of organizational hierarchies.

This dichotomous formalization, however, has been criticized for its failure to explain the continuation of interfirm contracting. To explain such contracting, Williamson turns to the notion of asset specificity or the degree to which durable investments are undertaken in support of particular transactions (and thus have no alternative uses). Williamson distinguishes four types of asset specificity: site, physical, human, and dedicated. Although transaction-specific assets are assumed to be at risk by unilateral long-term trading, a reciprocal long-term exchange agreement supported by both parties’ separate but concurrent investments provides a mutual safeguard against various trading risks. Such risks include the buyer’s premature termination of the contract and possible supplier exploitation of the bond required from and posted by the buyer.

To test Williamson’s assertions, in the 1980s, Nishiguchi surveyed electronics subcontractors in Japan and the United Kingdom. Even within the same industry, he found differences in subcontractor organization between British and Japanese samples. Customer concentration was considerably higher in the Japanese sample, and contract assembly was far more common. While site specificity was broadly equivalent, Japanese relationships varied by tier —Japanese second-tier firms were tightly coupled to their first-tier buyers, while first-tier firms were often quite removed from their prime contractors. Overall, Nishiguchi found greater asset specificity among Japanese subcontractors than among British subcontractors (see Table 6).

Clearly, the asset-specific character of such out-sourcing contributed to the stabilization of contractual relationships. When we interpret the results of Nishiguchi’s research, however, we do not see asset specificity as the cause but rather the consequence of strategy. Williamson views asset specificity as causally preceding obligational contracting or, at the very least, as being determined simultaneously (a traditional, ahistorical view of the world shared by many economists). Surely, asset specificity is important, because it enables a virtuous cycle of increasing cooperation. Our historical examination, however, clearly demonstrates that it is the producer’s strategy that is responsible for the genesis of obligational contracting (and once established, also has had a perpetuating effect). In this, our research supports Chandler’s notion that structure follows strategy.16

Cultural Specificity

Dore’s contention of cultural specificity and Williamson’s transaction cost economics begin poles apart, yet as a natural outcome of their reasoning, they both envision similar social structures.17

Dore offers a cultural explanation for Japanese subcontracting, arguing that relational (obligational) contracting is reliant on the benevolence, loyalty, and goodwill found in Japanese culture.18 As a result of these cultural characteristics, transaction costs for large Japanese firms may well be lower than elsewhere, and opportunism may be less a danger because of the explicit encouragement and actual prevalence in the Japanese economy of what we might call moralized trading relationships of mutual goodwill. According to Dore, the cultural disposition of the Japanese includes four significant elements: collective risk sharing, dutifulness, friendliness, and nonalloca-tive economic efficiency (the last being a kind of efficiency nonlinear with price signals). The advantages of relational subcontracting include a facilitation of investment, a broader flow of information, and an emphasis on product quality.

Given such a characterization, we might ask how uniquely Japanese it is. Dore argues that although the Japanese are believed to have an unusual preference for relational contracting, they are not uniquely susceptible to it. Moreover, Dore sees such contracting as a phenomenon of subjective affluence in which stable, friendly relationships become a goal in their own right, and the assurance of quality necessitates trust.

Questions that lead to empirical tests of Dore’s theory include: Has the institution of subcontracting always been an essential feature of the Japanese economy? What would be the consequence of moving the current system to an entirely different cultural environment? Would the Japanese system wither away, or could it succeed despite the foreign environment’s apparent lack of the cultural values claimed to be the determinants of Japanese success?

In answer to the first question, our examination of the history of Japanese subcontracting clearly indicates that subcontracting has not always been a prominent part of the Japanese social system and, in fact, hardly existed before World War II. Moreover, from the immediate postwar period through the late 1950s, it seems that Japanese supplier relations were as contentious and antagonistic as those in the West. In short, after we examine the record of subcontracting in Japan, it appears that no timeless national predisposition can account for the various forms of subcontracting.

Similarly, with the economic and operational benefits of various institutional innovations of Japanese subcontracting being increasingly recognized outside Japan, a trend has emerged among many Western producers to follow established Japanese practices. This foreign imitation, combined with the success of Japanese transplants abroad, has diffused Japanese subcontracting practices across national boundaries and demonstrated their cross-cultural applicability. What British subcontractors, for example, share with Japanese customers are not common national characteristics but an appreciation that high quality and low costs are potent strategic weapons and that joint efforts are more effective than mutual recriminations.19 The institutions that have given rise to the modern Japanese subcontracting system are advantageous and imitable. What we need is to identify particular methods that have led to success, rather than an appeal to some vague concept of national culture.


Since the 1960s, Japanese producers have achieved notable growth, not by unilaterally exploiting subcontractors but by strategically creating, and benefiting from, distinctive institutional arrangements. As the product of a complex historical interaction among market demand, politics, technology, and producer strategy, the evolution of Japanese subcontracting cannot be explained by a single theory. Subcontracting, as it now exists in Japan, consists of a series of collaborative relationships based on problem-solving principles for the manufacture of high-quality, low-cost goods; its essence can be applied across national boundaries.


1. This account of the early history of Japanese subcontracting comes from traditional secondary sources. To analyze more recent events, Nishiguchi conducted 1,035 field interviews between 1983 and 1989, involving 813 people at 394 organizations in 14 countries. The core of his research is discussions using a semistructured interview format with 251 interviewees in 71 prime contracting firms and 320 interviewees in 172 supply and subcontracting firms for the automotive and electronics industries in 10 countries. In addition, in 1986, Nishiguchi distributed a mail questionnaire about the reasons for subcontracting as perceived by 20 large automotive and 20 large electronics firms in Japan. A fuller exposition of this work may be found in:

T. Nishiguchi, Strategic Industrial Sourcing (New York: Oxford University Press, 1994).

2. MCI later became the Ministry of Munitions.

3. In the past few years, economic growth has slowed, but competition continues to be fierce and product variety remains great.

4. Many companies outside Japan now have profit-sharing rules. Interestingly, in Japan, the split is generally 1:1; in the case of Ford Motor Company, the split is 2:1, with Ford taking the lion’s share. See: Automotive News, 1 August 1994, p. 1 and p. 36.

Additionally, the tendency of Japanese purchasers to not ask for a price lower than that expected in the second year is a detail often overlooked by non-Japanese incentive systems.

5. See, for example:

K.B. Clark, T. Fujimoto, and F. Dubinskas, “Product Development in the U.S., Japanese, and European Auto Companies” (Boston: Harvard Business School, collection of viewgraphs, 1987).

6. With the benefits of subcontractor grading and self-certification becoming increasingly apparent, automakers around the globe began to establish such programs in the 1980s. Ford Motor Company’s Q1 program and General Motors’ Supplier Assessment Program (SAP) are prominent examples.

7. S. Berger and M. Piore, Dualism and Discontinuity in Industrial Societies (Cambridge: Cambridge University Press, 1980).

8. R. Edwards, Contested Terrain (New York: Basic Books, 1979).

9. Ibid., p. 77.

10. Nishiguchi (1994).

11. Ibid., p. 187.

12. M. Piore and C. Sabel, The Second Industrial Divide: Possibilities for Prosperity (New York: Basic Books, 1984).

13. J. Womack, D.T. Jones, and D. Roos, The Machine That Changed the World (New York: Rawson Associates, 1990).

14. O. Williamson, Markets and Hierarchies: Analysis and Antitrust Implications (New York: Free Press, 1975); and

O. Williamson, The Economic Institutions of Capitalism (New York: Free Press, 1985).

15. Bounded rationality refers to the human mind’s limited capacity for formulating and solving complex real-world problems. Opportunism here is defined as the fulfillment of one’s own interests by means of guile, including lying, stealing, cheating, and incomplete or distorted disclosure of information.

16. A. Chandler, Strategy and Structure (Cambridge, Massachusetts: MIT Press, 1962).

17. R. Dore, Taking Japan Seriously (Stanford, California: Stanford University Press, 1987).

18. Ibid., p. 170.

Dore defined benevolence here as “something shown in relations between unequals, by superior to inferior, the reciprocal which is usually called loyalty” and goodwill as “the sentiments of friendship and the sense of diffused personal obligation that accrue between individuals engaged in recurring contractual economic exchange.”

19. For more details, see:

Nishiguchi (1994), pp. 181–185.