An Overview of the Japanese Distribution System
The Case of the Automobile Industry
Author: Sayuri Shirai1
  • 1 0000000404811396 Monetary Fund

This paper surveys the recent literature on the Japanese distribution system to consider two propositions: first, that the system is inefficient, and second that prices of imported products tend to be higher in Japan than in other markets. Most of the literature demonstrates that the system is efficient. However, the efficiency has not necessarily resulted in high social welfare as consumers have had limited access to various product lines or paid high prices for some products. This paper examines the distribution system in the automobile industry to promote understanding about the impacts of the system on price differentials.


This paper surveys the recent literature on the Japanese distribution system to consider two propositions: first, that the system is inefficient, and second that prices of imported products tend to be higher in Japan than in other markets. Most of the literature demonstrates that the system is efficient. However, the efficiency has not necessarily resulted in high social welfare as consumers have had limited access to various product lines or paid high prices for some products. This paper examines the distribution system in the automobile industry to promote understanding about the impacts of the system on price differentials.

I. Introduction

Bilateral trade conflicts between the United States and Japan have emerged since the late 1970s. The trade conflicts, which began in the textile industry, lessened when Japan experienced a large-scale current account deficit after the two major oil crises. Then, the conflicts appeared again in the beginning of the 1980s when Japan experienced considerable net exports in the automobile industry. The conflicts expanded when the United States experienced a deteriorating current account deficit while Japan enjoyed a swelling current account surplus in the early 1980s. The current account surplus of Japan with the United States increased between 1985 and 1987 even when the yen appreciated sharply. The surplus has since declined and after 1990, finally dropped below the 1985 surplus level.

Among the factors that helped reduce the surplus was an increase in Imports by Japan. Petri (1990) showed that since 1987, Japan’s trade imports Increased 10 to 20 percent ahead of the historically estimated import functions Corker (1989) and since 1985, the sensitivity of Imports to the relative imported goods prices increased significantly. Baldwin (1988) explained such an unusual increase in Japan’s imports by demonstrating that the prolonged and large appreciation of the yen since 1985 encouraged foreign firms to establish associated new sales - distribution networks, which increased Japan’s imports of their products to a large extent.

The net reduction in Japan’s trade surplus since 1985, however, seems relatively small considering the large appreciation of the yen. The current account surplus declined by only 3.9 percent between 1985 and 1990 while the yen appreciated nearly 100 percent. Some argue that the unique trade structure of Japan, characterized by the small scale of intra-industry trade (IIT) and substantial net exports of manufactured goods, may have affected the sluggish Increase in imports by Japan.

Regarding the small scale of IIT, the Economic Planning Agency (EPA) estimated the IIT indices of some developed countries between 1978 and 1988, as shown in Table 1. The IIT index of Japan was smaller relative to that of other countries although the index increased in recent years. For example, the IIT index of the year 1988 was 18 for Japan, 44 for the United States, 53 for the United Kingdom, 48 for the former West Germany, 50 for France, and 48 for Italy. In addition, Japan’s IIT was particularly small in the machinery and transportation equipment industries which accounted for 47 percent of Japan’s total trade value in 1988. The EPA concluded that these industries achieved considerable net exports as a result of achieving competitiveness and fuel-efficiency after the oil crisis. Petri found a similar result in that Japan’s IIT was inversely related to competitiveness, which differed from other countries where strong industries developed close trade ties with their counterparts. Meanwhile, Lincoln (1990) considered that the IIT indices of Japan in the early 1970s declined because the Japanese market became more insular over time. However, it is difficult to follow this view as the IIT indices showed considerable fluctuations over the period and the Japanese market could not respond to the fluctuations so frequently. EPA (1991) and Saxonhouse (1989) attributed the low IIT index of Japan to an increase in inter-industry trade after the oil crisis and the resulting specialization in the machinery industry.

Table 1.

Intra-Industry Trade Indices

article image
Source: Economic Planning Agency, Economic Institute, “Keizai Bunseki” vol. 125, July 1991, pp. 14-18.



i is the country

j is the trade counterpart

k is the industry

As for the substantial net exports of manufactured goods, the Japan Economic Institute JEI, 1991) pointed out that exports of manufactured goods from Japan to the United States were much larger than those from the United States to Japan. Noland (1993) also showed that Japan’s manufactured import share in domestic consumption in 1990 was 5.9 percent while the share was 15.3 percent for the United States, 15.4 percent for the former West Germany and 17.7 percent for the United Kingdom. He further demonstrated that the foreign-owned firms’ share of domestic sales in 1986 was 1 percent for Japan, 10 percent for the United States, 20 percent for the United Kingdom. Noland stressed that Japan’s import penetration ratio for manufactured goods remained essentially flat, never having risen above 6 percent since 1975. Lawrence (1987) showed that unusual trade barriers of Japan and specific preferences of Japanese consumers reduced Japan’s imports of manufactured goods by an estimated 40 percent. Bhagwati (1991), on the other hand, emphasized that careful micro-level studies were necessary to provide more natural and less adversarial explanations for Japan’s low imports of manufactured goods (e.g., preference of Japanese consumers for small refrigerators because of small apartments and houses).

The unique trade structure of Japan might be closely related to price factors. 1/ Some studies (for example, JEI, 1990) showed that prices of many goods and services were higher in Japan when compared to those in the United States and some other developed countries. Furthermore, the price report by the EPA showed that some imported products were more expensive in Japan than in other countries. Ohno (1989) demonstrated that American exporters had higher pass-through coefficients than Japanese exporters so that the movement of the dollar would be almost completely passed through to foreign prices. His study suggested that the appreciation of the yen against the dollar reduced the yen-valued prices of imported products from the United States. However, the retail prices of some imported products did not decline much despite the large appreciation of the yen, while the CIF prices did. Most of these products were brand name goods produced by European manufacturers and automobiles produced by both American and European manufacturers. Foreign manufacturers of these products generally depended on importers that distributed imported products with exclusive sales rights (sole representative importers). Japanese manufacturers of similar products distributed their products through their exclusive dealer networks. These studies reinforced the view that the Japanese distribution system limited foreign firms’ access to the Japanese market by increasing set-up costs, or that it lowered the import market share by enlarging price differences between Japanese and imported products. Such a view made a substantial impact on the Structural Impediments Initiative (SII) talks that began at the end of the 1980s.

The purpose of SII talks was to narrow the trade gaps between Japan and the United States by removing the impediments to trade. The talks viewed the Japanese distribution system as one of the structural impediments that created the trade imbalance. The United States Trade Representative’s Office (USTR) pointed out the following issues as the factors that may have limited imports by Japan and may have created price differences between Japanese and imported products. First, the Japanese distribution system Is complicated and inefficient because there are a large number of small-sized retailers supported by a multi-layered wholesale structure. Bhagwati (1991) argued that the complexity of the system may have created the perception that the Japanese market was closed, which may have prevented attempts to enter, or may have led to hasty withdrawals by foreign firms after a short period. Foreign manufacturers generally face cultural unfamiliarity, language barriers and other unfamiliar business practices whether or not the targeting market is Japan. Thus, the issue about whether the overall Japanese distribution system is closed should be approached through careful investigation and the accumulated evidence of case studies of the system. For this reason, case studies are important since each industry has a different distribution system that depends on the nature of competition, the degree of production concentration, product attributes and industrial history.

Second, it was pointed out that the “sales-distribution Keiretsu” system promotes collusive pricing behavior among Japanese manufacturers. The system is characterized by exclusive dealership arrangements, territorial restrictions, complicated rebate systems and resale price maintenance systems. Such features are frequently observed in the automobile, home electronics and cosmetics/medical industries. Since the late 1970s, the Japan Fair Trade Commission (JFTC) has examined whether the system violates the Anti-monopoly Act [EPA (1986) and JFTC (1981, 1986, 1990)]. Meanwhile, the development of the industrial organization theory and microeconomics in recent years provides theoretical foundations to support the system by emphasizing efficiency advantages [Ito and Matsui (1987), Ito (1989), Maruyama, et al. (1989) and Shibayama and Kiji (1990)]. However, most of these studies emphasize the advantages in a closed-economy framework. Thus, their arguments are not always relevant to the important issues associated with international trade. The efficiency advantages created by the system might have improved domestic producers’ surplus but may have sacrificed domestic consumers’ surplus when the system provided limited product variety and set higher prices on imported products.

Third, it was argued that the sole representative import system (So-yunyu-dairitensei) allows importers to set up high margins, which results in high retail prices of imported products. In this system, foreign manufacturers generally provide exclusive sales rights to importers. The system is also criticized as creating retail price rigidity to the movement of exchange rates as importers attempt to maintain the “luxury” brand image of imported products by preventing price reductions. 1/ The imported products whose prices are more expensive than those of Japanese products are consumer goods which are mostly sold by the sole representative importers. However, margins can be high partly because these importers, as late-comers to the market, have to pay large initial entry costs [Kida and Kiji (1990), and Reilly (1992)]. On the other hand, an increase in the number of parallel importers, who bypass the sole representative importers, suggests that rents earned by the latter exist. Thus, the further increase in parallel trade will probably enhance intra-brand competition among importers and induce the sole representative importers to lower prices.

Section II focuses on efficiency arguments regarding the Japanese distribution system. Section III analyses the Japanese distribution system in the automobile industry. In that industry, business practices such as exclusive dealership arrangements, exclusive territories, retail price maintenance and rebate systems are commonplace. These practices are also observed in other countries such as the United States, the United Kingdom, and Germany although the degree of complexity varies among countries (Shibayama and Kiji, 1990). Foreign automobile manufacturers typically depended on sole representative importers to distribute their products. The sole representative import system and the “sales - distribution Keiretsu” system were closely related in the automobile industry and therefore, this paper discusses them together.

II. The Japanese Distribution System: The Macro Approach

1. Japanese distribution system and efficiency

The Japanese distribution system can be characterized as having a large number of small-sized retailers with a multi-layered wholesale structure. The JEI (1987) and Maruyama, et al. (1989) found the following characteristics, as shown in Table 2. The numbers of persons engaged per wholesale establishment and per retail establishment were smaller in Japan than in other developed countries, except Italy. The sales floor space per retail store was smallest in Japan among the countries whose data were available. Japan’s density of retail stores, measured by the number of stores per one thousand population, was the second highest after Italy. The multi-layered wholesale structure of Japan, measured by the ratio of wholesale annual sales over retail annual sales, was highest among the countries considered. The wholesale/retail sales ratio was high because of the larger sales value per wholesaler compared with the sales value per retailer.

Table 2.

Size and Wholesale/Retail Sales Ratio

article image
Source: Maruyama, et al., 1989, pp. 37, 48 and Economic Planning Agency, 1991, pp. 7, 22.

W/R refers to wholesale trade sales divided by retail trade sales.

The datum is based on enterprises statistics.

The sales data are based on 1978 data.

The sales data are based on 1984 data.

The 1982 report by the USTR concluded that the Japanese distribution system is inefficient and complicated. However, the system, although complicated, can be justified in a closed-economy setting if it attains efficiency in distribution. Thus, there is need to test whether the Japanese distribution system is efficient. Various studies suggest at least four ways to estimate efficiency. The first measure is to estimate annual sales both per establishment and per worker, as shown in Table 3. The table, using both measures, reports that the Japanese wholesale industry is more efficient than that of other countries. The Japanese retail industry, on the other hand, is less efficient than that of other countries, based on the measure of sales per retail establishment. However, the Japanese retail industry is as efficient as that of other countries, with the measure of sales per worker [Maruyama, et al. (1989), and EPA (1991)]. The performance of the Japanese wholesale industry can be attributed to large general trading firms (Sogososya). Sales per worker of these trading firms were $2,857,800, while those of wholesalers of equivalent size in the United States were only $408,900 in 1982 (EPA 1991).

Table 3.

Sales Per Worker and Sales Per Establishment

article image
Source: Economic Planning Agency, 1991, p. 7.

The data are based on enterprises statistics.

As the second measure to estimate distributors’ efficiency, value added per worker can be used (Maruyama, et al. 1989). The value added measure or the net of input cost measure is useful because it eliminates the double counting problem that occurs in the process of measuring sales values in a multi-layered wholesale structure. Japan’s ratio of value added per worker in the distribution sector to that in the total industry was higher than the ratio in the United States and Germany. For example, the ratio was 0.76 for Japan, 0.7 for the United States, 0.68 for the former West Germany, 0.58 for the United Kingdom, 0.82 for France and 0.9 for Italy. Furthermore, with the ratio of value added per worker in the distribution sector to that in the manufacturing sector, Japan’s ratio was not very different from that of other countries. For example, the ratio was 0.64 for Japan, 0.63 for the United States, 0.71 for the former West Germany, 0.61 for the United Kingdom, 0.85 for France and 0.88 for Italy.

As the third measure of efficiency, Maruyama, et al. (1989) suggest the ratio of inventories to sales. With 1985 data, Japan’s ratios, both in the wholesale and retail industries, were smaller than those of other countries. For example, the ratio in the wholesale industry was 0.044 for Japan, 0.11 for the United States, 0.073 for the former West Germany, 0.081 for the United Kingdom, and 0.099 for France. The ratio in the retail industry was 0.097 for Japan, 0.126 for the United States, 0.123 for the former West Germany, 0.116 for the United Kingdom, and 0.121 for France. These figures suggest that the inventory turnover rates, in the retail and wholesale industries, were higher in Japan than in other countries. Furthermore, Maruyama, et al. introduce an inventory ratio that combined both sales levels, estimating the sum of wholesale and retail inventories divided by retail sales. The inventory ratio of Japan was similar to that of other countries. The ratio was 0.248 for Japan, 0.232 for the United States, 0.266 for the former West Germany, 0.286 for the United Kingdom, and 0.237 for France. Nishimura (1990) rejected the conclusion drawn by the Ministry of International Trade and Industry (MITI, 1988) that Japan’s commercial margins were lower than those of the United States by about 10 percent and demonstrated that in fact these margins exceeded those of the United States by 2.9 percent.

The fourth measure of efficiency is to estimate distribution margins. Maruyama, et al. (1989) estimated the ratio of gross margins to sales, defined as the difference between sales and costs of goods sold, divided by sales. The cost of goods sold included costs of purchases as well as the difference between beginning-period inventory and end-of-period inventory. The ratio in the wholesale industry in 1986 was 11.2 for Japan, 19.4 for the United States, 12.6 for the former West Germany, 13.4 for the United Kingdom and 21.8 for France. The ratio in the retail industry in the same year was 27.1 for Japan, 31 for the United States, 34.2 for the former West Germany, 27.6 for the United Kingdom and 29.6 for France. Japan’s ratios in both industries were the lowest among the countries considered. However, when the ratio of the sum of wholesale and retail distribution margins to retail sales was measured, Japan’s ratio was higher than that of other countries except the former West Germany. 1/ The ratio was 57.6 for Japan, 49.7 for the United States, 58.9 for the former West Germany, 55.6 for the United Kingdom and 55.3 for France.

Nishimura (1991) proposed a new measure considering buyers’ perspectives rather than those of distributors. He defined the distribution-margin ratio as the ratio of distribution margins to purchase prices. Distribution margins included wholesale margins, retail margins and transportation costs. The purchase prices consisted of producers’ shipping prices, wholesale margins, retail margins and transportation costs. With this measure, he concluded that distribution margins, particularly those of consumer goods, increased rapidly between 1965 and 1985 and exceeded those of the United States in 1990. His finding suggested that the Japanese distribution system particularly for consumer goods was relatively inefficient. Furthermore, he found that an increase in distribution margins particularly for consumer goods was the result of an increase in retail margins. The retail margins increased from 15.2 percent in 1965 to 26.8 percent in 1985 while the wholesale margins decreased from 9.2 percent to 8.3 percent and the transportation margins remained stable at around 1.5 to 2 percent. He argued that the poor performance in the retail industry was due to an increase in real wages and a failure to improve productivity. The labor productivity growth rates, in the wholesale and retail industries between 1973 and 1986, were only 1.92 percent and 1.11 percent, respectively. In comparison with the growth rate of the manufacturing labor productivity at 4.06 percent, these industries were not successful in improving productivity.

Kida and Kiji (1990) found that retail margins increased between 1967 and 1986 while wholesale margins were stable at around 11 percent. Furthermore, they decomposed consumer goods into “convenience” goods such as foods and beverages and “nonconvenience” goods such as clothes and vehicles. Then, they found that “nonconvenience” goods, which had higher wholesale margin rates, had a smaller number of wholesale tiers than “convenience” goods. Thus, there was a negative correlation between wholesale distribution margins and the number of tiers. Their findings suggested that the wholesale multi-layered structure of Japan did not necessarily reflect inefficiency. Nishimura hypothesized that the multi-layered structure realized a division of labor and may have increased efficiency by reducing prices of distribution services. Furthermore, some theoretical studies also explained the existence of many small retailers by consumer’s buying behavior, storage costs, ordering costs, distances from a residence to a nearby store and transportation costs (Flath (1986) and Maruyama, et al. (1989)). 1/ This theoretical work suggested that the Japanese distribution system emerged in response to those variables that were specific to the Japanese situation and hence, achieved efficiency.

Finally, we need to mention the Large Retail Store Law (Daiten-ho), which made it difficult for large-sized retailers to establish their stores. The law was originally enacted as the Department Store Law in 1923 to protect small- to medium-sized retailers from the expanding department stores by introducing an approval system for establishing or expanding stores. The law also required large-sized retailers to close at 7:00 p.m. between April and October and 6:00 p.m. between November and March, and obliged them to close more than three days per year. 1/ The law was abolished by the General Headquarters in 1947 but was reintroduced in 1956 as a result of active efforts made by small- to medium-sized stores to regulate the rapidly expanding department stores. In 1974, the law then became the Large Retail Store Law to regulate large super markets, discount stores and other large chain stores by requiring applicants to submit a report for construction plans. The MITI initially attempted to change the previous law by only requiring applicants to report the plans for establishing or expanding stores rather than waiting for approval. Due to strong protests, the MITI compromised by introducing a reporting system that was conditional on prior appraisals. The law was revised in 1979 and frequently revised in the 1980s to reinforce regulations and prolong the period spent before opening stores. One example was Ito Yokado which spent 10 years to open a new store in Shizuoka City. This law was finally modified in 1991 as a result of pressures from the United States. The new revision abolished the “Shochokyo” or the committee for “adjusting” retail activities in the community, which was organized under the Chamber of Commerce. The committee discussed floor space, opening date, store hours and the number of days closed per year after applicants submitted the construction plan to the governor of the prefecture. With the deregulation movement, large-sized retailers could provide consumers a wider variety of products including imported products at discount prices by realizing scale economies in distribution.

To summarize, the existing studies showed that the Japanese distribution system, characterized by a large number of small-sized retailers with a multi-layered wholesale structure, was as efficient as that of other developed countries. The efficiency measures used to derive this conclusion were sales per worker and sales per establishments, value added sales per worker, inventory ratios, and distribution margins. Some studies showed that retail distribution margins of Japan increased rapidly particularly for consumer goods, due to an increase in operation costs. However, the margins exceeded those of the United States only recently and by a small amount. Thus, most of the studies rejected the view that the Japanese distribution system was substantially inefficient. Furthermore, the multi-layered wholesale structure was shown to be negatively correlated to wholesale distribution margins.

2. Japanese distribution system and international trade

The studies discussed in the previous subsection did not focus on the relationship between the Japanese distribution system and international trade. The efficiency of the distribution system measured by the macro-oriented methods did not necessarily result in higher social welfare. The distribution margins data included all goods and services that were distributed through the wholesale and retail sectors. Thus, the margins were the sum of distribution margins for tradable goods and nontradable goods. The efficiency results derived with the macro-oriented measures could be consistent with inefficiency in resource allocation or high distribution margins for some tradable goods. (Sazanami, Urata and Kawai (1993a, 1993b)) demonstrated that an elimination of the gaps of imported products between domestic producers’ prices and import prices on a c.i.f basis would have increased social welfare of Japan by ¥4,200 billion or by 1.1 percent of GNP. 1/ To analyze the impact of the Japanese distribution system on the import market, we need to find out if there existed large differences in margins and resulting price gaps between domestic and imported products.

Before the Plaza Accord in 1985, Japan’s purchasing power parity was ¥218, which was not much different from the exchange rate of ¥239 per U.S. dollar. In 1988, however, the purchasing power parity indicated ¥201, while the average exchange rate was ¥128 per U.S. dollar. Thus, prices in Japan were 57 percent higher than those of the United States (JEI 1990). We should discuss two issues regarding Japan’s price differences separately: (1) the high overall level of prices; and (2) the high prices of imported goods.

The 1989 price report by the EPA and the 1989 survey by Keizai Doyukai focused on the high overall level of prices. These reports suggested that many products were more expensive in Japan than in New York and other cities. In the previous subsection, we concluded that the Japanese distribution system was as efficient as that of other countries because the margins, for example, were similar among countries. Thus, other factors affected the overall high prices in Japan. Kida and Kiji suggested that an increase in operating costs and a decrease in operating profits, especially in the Japanese retail industry, may have affected the high overall prices of “nonconvenience” goods as compared to those of “convenience” goods. They pointed out that operating costs increased due to an increase in wages, rents and the purchase of information-related facilities. Alternatively, the overall high prices existed because cartels were granted exemption from the Anti-Monopoly Act or firms were protected by regulation in some industries. 1/ In some industries, manufacturers openly discussed retail prices and set uniform prices. Such a practice called resale price maintenance tended to promote collusive price - setting behavior among manufacturers, which resulted in higher prices at the expense of the consumers’ surplus. The practice also created price rigidity at the retail stage and reduces inter-brand and intra-brand price competition (Economist, 1991). Those industries included barber shops, newspapers, books, casualty-insurance and some cosmetics. In recent years, the JFTC began to investigate the impact of the practice on the overall high prices and provided some restrictions on the practice. 2/

As for the high prices of imported goods, the 1989 price survey, performed by the Tokyo Metropolitan government, showed that 14 out of 20 foreign brand name products were priced higher in Tokyo than in other major cities. Also, the survey reported that Japan’s retail prices of the imported products were 3 to 7 times higher than their landed prices. The survey suggested that large distribution margins existed for imported products. The 1990 price survey by the JFTC also found that 6 imported products, such as instant coffee, spaghetti, men’s suits, bags, passenger cars, and china, were more expensive in Tokyo than in other cities. Furthermore, the price survey, jointly performed by the Department of Commerce (DOC) and the MITI, suggested that 31 out of 35 American-made products and 22 out of 23 European-made products were more expensive in Japan than abroad. The recent report by MITI (1993) demonstrated that some brand-name products made in Europe and the United States were still 60 percent more expensive in Japan.

Cline (1990) argued that high prices of Japan’s Imported products were the consequence of foreign firms’ pricing strategy to sell their products at higher prices rather than achieving a large sales volume with lower prices. However, Lawrence (1991) demonstrated that the weighted average of the unit value of the sampled consumer goods when sold to Japan was only 17 percent higher than the sample sold to Germany, but the retail price ratio weighted by export values of the samples was 99 percent higher in Japan than in the United States. He suggested that the high prices of Japan’s imported products came from distributors’ pricing behavior rather than foreign manufacturers’ pricing behavior in the Japanese distribution system. Thus, he concluded that the efforts to reduce distribution margins would lower the prices of imported products and increase import volumes without hurting the foreign manufacturers. Petri (1990) estimated that a 1 percent increase in distribution margins would reduce an import market share by 0.16 percent. He. suggested that since the distribution margins in the manufacturing sector were around 20 percent, the import demand would increase by 3.2 percent if the products were sold directly.

The high prices of imported products were observed mainly for brand name products produced by European manufacturers and automobiles produced by European and American manufacturers. Foreign manufacturers of these products typically distributed their products through the sole representative importers. In Japan, 70 percent of importers obtained exclusive sales rights from foreign manufacturers. 1/ Their main sales strategy was to create the “luxury” brand image of imported products to set high distribution margins, rather than to lower retail prices to sell a large number of products. The high markups set by these importers, thus, may have enlarged the price differences between Japanese and imported products. According to the 1985 consumer survey by the JFTC, 48 percent of consumers pointed out, as the reason for not buying brand name imported products, their higher prices compared with those of domestic products. Also, the survey reported that 40 percent of consumers would purchase imported products if their prices declined.

The number of the parallel importers of brand name products has increased in recent years. The increase probably limited other importers from setting high markups and eventually, lowered prices of imported products to a certain degree (JFTC 1987). The parallel importers supplied imported products without passing though the sole representative importers. 2/ They directly acquired imported products through three sources: from distributors in the foreign manufacturers’ home country, from distributors in a third country, and from foreign manufacturers’ factories in a third country. They distributed imported products at lower prices than the sole representative importers, supplied products that were not handled by other importers and distributed products faster than other importers. However, the parallel importers usually had a limited number of sales outlets and insufficient after-purchase service facilities. Therefore, organizing parallel importers’ associations was one way to achieve a stable supply of parts and products, to maintain stable prices, and to expand the number of parallel importers.

According to the 1985 survey by JFTC, 73 percent of the sole representative importers were Japanese distributors while the rest were Japanese manufacturers. Among the latter-type of importers, 68 percent of manufacturers distributed imported products which were similar to those of their own products, and 48 percent had exclusive sales rights. The main purpose for Japanese manufacturers to sign such sales contracts was to increase product variety. Thus, Japanese manufacturers were likely to sell differentiated products rather than directly-competing foreign products. Furthermore, the survey suggested that 62 percent of Japanese manufacturers charged the same or higher sales prices on imported products compared to those of their own products. Such a system may have created the overall high prices of tradable products as well as the price differences between Japanese and imported products. While large manufacturers accounted for 19.4 percent of total distributors, the impact of these manufacturers’ pricing behavior on imports could be large in some industries (JEI 1990).

Furthermore, Japanese manufacturers operated under the “sales -distribution Keiretsu” system in some industries. Under the “sales-distribution Keiretsu” system, Japanese manufacturers controlled dealers with exclusive dealership arrangements, exclusive territories and complicated rebate systems. The “sales-distribution Keiretsu” system reflected market power of oligopolistic Japanese manufacturers. Maruyama, et al. (1989) emphasized the role of the system as a device to deal with market failure. The system provided dealer incentives to promote adequate services, to transmit reliable and stable demand information to consumers and manufacturers, and to lower transaction costs. It also provided a risk-sharing function between manufacturers and distributors, and realized higher joint profits between them by eliminating double markups. Although the system had efficiency advantages, the system may have created higher overall prices by reducing intra-brand price competition (JEI 1987). In addition, such a system may have promoted collusive pricing behavior among domestic incumbent firms by increasing entry costs or limiting newly entering firms’ access to the market.

Meanwhile, the problems seen under the “sales - distribution Keiretsu” system seem set to lessen gradually in the future. In the home electronics industry, for example, the number of large general merchandise retailers has been increasing recently. 1/ The general merchandise retailers have been distributing products at discount prices by achieving economies of scale in distribution and maintaining high profits by obtaining volume rebates from many manufacturers. Such a movement is likely to speed up as a result of a weakening in the Large Retail Store Law. In the cosmetics industry, fewer retailers use exclusive dealership arrangements. 2/ Dealers’ sales strategy has recently shifted to attract customers with a wider variety of products in addition to a provision of sufficient information about attributes of particular manufacturers’ products. As consumers’ tastes have become more diversified, dealers have more incentives to sell several manufacturers’ products, including imports.

To summarize, the higher prices of imported products compared to those of similar Japanese products were partly attributed to high markups charged by sole representative importers. These products were mainly expensive brand name products and automobiles. Some of these products are likely to be less expensive in the future due to an increase in the number of parallel importers. However, the prices of the products such as automobiles, which need sufficient after-purchase repair and maintenance services, will not decline as quickly as those of other products. The limited after-purchase service facilities and reliable sources to purchase parts will deter the expansion of parallel importers.

III. Japanese Distribution System in the Automobile Industry

1. Distribution channels

The automobile industry in Japan has maintained its leading position in world production since 1980. The output volume produced in Japan Increased by 3.5 percent per year to reach approximately 13.5 million units in 1990. The total output volume produced domestically and abroad by Japanese manufacturers accounted for about 28 percent of world production volume in 1990. There were 11 automobile manufacturers in Japan. Among the 11 manufacturers, 9 manufacturers produced passenger cars and 2 manufactured trucks. In the domestic market, Toyota, Nissan and Mazda were the three major manufacturers. They had market shares of 41.9 percent, 23.5 percent and 8.4 percent, respectively for the vehicles including passenger cars, trucks and buses in 1990. Honda took the third position instead of Mazda once mini-vehicles were included (Table 4). Among all manufacturers in the world, Toyota was the second biggest manufacturer next to General Motors, producing 4.2 million vehicles in 1990.

Table 4.

Distribution Channels of the Japanese Manufacturers in 1990

article image
Sources: Toyota Motor Corp., “The Automobile Industry: Toyota and Japan,” 1991 Edition: Japan Automobile Manufacturers Association (JAMA), “Automotive Distribution In Japan, June 1990, p. 3.

The data for used-car outlets are for 1989.


(1): The data include all vehicles, except mini-vehicles.

(2): The data include all vehicles, including passenger cars, buses, trucks, and mini-vehicles.

Domestically, Japanese manufacturers distributed their products through their dealers under the “sales-distribution Keiretsu” system, as shown in Table 4. Japanese manufacturers distributed several different types of automobiles through each distribution channel. The door-to-door sales method by sales persons was their major marketing strategy. Thus, the number of sales personnel is one of the most important measures in evaluating each manufacturer’s sales strength. Toyota had the largest sales personnel with 113,467 employees in its dealer network in 1990. Nissan had the second largest personnel with 72,695 employees in the same year. The number of vehicles sold domestically in 1989 were 2.3 million units for Toyota and 1.3 million units for Nissan. These data suggested that the number of sales personnel had a substantial impact on sales performance.

From the early period, Japanese manufacturers attempted to maintain exclusive dealership arrangements with their dealers. Dealers distributed only particular manufacturers’ products and usually furthered the interests of their manufacturers by providing efficient services to customers and investing in advertisement. As dealers belonged to particular manufacturers’ organizations, they frequently exchanged information about product inventories, parts inventories, market trends, and used car handling between dealers. In return for this loyalty, Japanese manufacturers provided a sufficient product variety and model changes, trained sales personnel with adequate information and technical advice, and granted financial support and rewards with rebates and discount prices. Thus, these dealers rarely supplied other manufacturers’ automobiles, including imported ones.

Foreign manufacturers generally distributed their automobiles through importers. There were at least three kinds of importers. The first were sole representative importers. The second were Japanese manufacturers which played the role of importers. The third were wholesalers owned by foreign manufacturers. Due to the exclusive dealership arrangements undertaken by domestic manufacturers, foreign manufacturers were unable to increase their dealers through increasing sales tie-ups with Japanese manufacturers or their dealers. If foreign manufacturers wished to distribute products through the Japanese dealers, they needed to obtain an approval from the Japanese manufacturers without directly contacting these dealers. In addition to these arrangements, foreign manufacturers had limited choices of distribution channels in Japan, because the cost of establishing new dealer networks was prohibitively high given the high land prices and high wages. Consequently, foreign manufacturers relied on specialized automobile importers that had their own dealers as well as repair and maintenance facilities.

Historically, American automobile manufacturers introduced the exclusive dealership arrangements in order to obtain precise demand information to estimate output volumes in the United States. However, the decline in dealers’ profitability and the increasing number of bankruptcies in the 1930s increased dealer’s political power. Furthermore, the weakening financial position of American manufacturers and the legal arrangements to protect dealers in the 1950s strengthened the dealers’ right to choose their suppliers. Thus, at the time when Japanese manufacturers began to enter the American market, they could find sales channels relatively easily (JFTC, 1990). They could sell their products through existing dealers and thus, could expand their sales volumes with relatively low initial set-up costs. Also, the low costs of establishing their own exclusive dealers in the United States enabled them to increase the number of exclusive dealers. On the other hand, various government regulations protected the Japanese automobile market until the 1970s. By the time the formal regulations were eliminated, Japanese manufacturers had already improved the quality of their products, supported by well-organized dealer networks and nationwide repair and maintenance networks (New York Times. January 6, 1992).

If foreign manufacturers had partial or complete ownerships of Japanese manufacturers, they usually had access to these dealers and the dealers may have acted to further their interests. For example, Ford had a 24.01 percent ownership of Mazda. Ford acquired one distribution channel called “Autorama” among Mazda distribution channels which supplied solely Ford-badged cars after 1979. Among the dealers that belonged to the dealer networks of Japanese manufacturers, 70 percent were financially independent of their manufacturers, 20 percent were partially owned by them, and 10 percent were wholly owned by them. Although most of the dealers were financially independent, they maintained stable and long-term business relationships with their Japanese manufacturers. Since the late 1970s, the Anti-monopoly Act has made it illegal for manufacturers to include an exclusive dealing clause in their contract with dealers. Despite this legal action, exclusive dealership arrangements are still commonplace. Such business practices give an impression to outsiders that the Japanese distribution system limited foreign manufacturers’ access to the Japanese market. In 1991, the USTR urged Japan to open the automobile market by concluding that the Japanese automobile market was biased against imported products (Nikkei, June 5, 1991).

2. Performance of the imported automobile market

The sales volume of imported automobiles increased by 22.9 percent between 1989 and 1990, when it exceeded 0.2 million units for the first time. The market share of imported automobiles was about 5 percent in Japan’s total domestic automobile market. The market share of imported automobiles still lagged behind that of other developed countries. For example, the market share of Japanese automobiles was 15.14 percent in Germany, and 11.15 percent in the United Kingdom in 1989 (Pemberton (1992)). Although the import car market share was very small in Japan, the sales volume of imported automobiles, particularly from the former West Germany, increased remarkably. German automobile manufacturers began to control import dealers by establishing wholly-owned importers rather than relying heavily on the sole representative importers.

As the sole representative importers had a limited number of dealers and repair and maintenance facilities in comparison with Japanese manufacturers which controlled dealer networks, their main sales strategy was to create brand images on imported automobiles to set high retail prices. Such a sales strategy was feasible if the differentiation strategy enabled consumers’ demand for imported automobiles to be inelastic to their prices. Also, the strategy was reasonable when the importers were unable to achieve economies of scale in distribution. However, the strategy, by its nature, did not contribute to an increase in sales volume. Foreign manufacturers generally earned their profits by selling their products to the importers. Thus, their profits were likely to be small when the sales volumes of their products were small, since importers or dealers took the remaining distribution margins. 1/

German manufacturers decided to improve sales performance by lowering product prices and by increasing dealers, as the popularity of their automobiles grew during the 1980s. The appreciation of the yen, the 1989 tax reform, and several government measures to facilitate imports helped promote this movement. Also, Japanese manufacturers began to undertake sales tie-ups with foreign manufacturers. With the tie-ups, foreign manufacturers could acquire access to some dealers of Japanese manufacturers, as shown in Table 5. While some of the recent tie-ups could be a result of political pressures from abroad, some domestic manufacturers preferred selling imported automobiles to increase product variety. For example, Honda, which had no utility vehicle of its own, began to supply jeeps made by Chrysler that were increasingly popular among young Japanese (New York Times, January 6, 1992). Also, an increase in the number of imported American-made Japanese-badged automobiles, such as the Camry made by Toyota and the Accord Coupe made by Honda, was one of the factors that contributed to an increase in imports. These imported American-made Japanese-badged automobiles added another 20,000 units to the imported automobile market. In fact, the import volume from U.S. Honda already exceeds imports from General Motors. The Accord models produced in the United States were a two-door coupe and a station-wagon, both of which were designed for the American market and were not produced in Japan. Such differences created the prestigious image for imported Japanese automobiles and increased demand (Economist, 1991).

Table 5.

Sales Tie-Ups Between Domestic and Foreign Firms

article image
Source: JAMA, “1991 The Motor Industry of Japan,” The 1991 edition, pp. 28-29.

Since 1991.

Since the end of 1992.

Note: The equity participation data is at the date of March 1991.

The yen has appreciated both against the German mark and the U.S. dollar since 1985. The appreciation rate against the latter was much larger than for the former. Consequently, the average suggested prices of American-made automobiles declined while those of German-made cars did not decline as much, although there were exceptions. However, the sales volumes of BMW, Volkswagen-Audi and Daimler-Benz increased from 52,854 units in 1986 to 88,740 units in 1988 and to 128,832 units in 1990. The sales volumes of Ford and General Motors increased from 2,235 units in 1986 to 8,650 units in 1988 and to 14,543 units in 1990. While the sales growth rate of American-made cars was remarkable, the market share remained very small in comparison with that of German-made cars.

3. Price gap between imported and domestic automobiles

The sales performance of imported automobiles improved in the 1980s. However, high prices of imported automobiles remained one of the major reasons for consumers not purchasing them. The 1990 consumer survey by the JFTC showed that 47.8 percent of respondents, as one of the reasons for not purchasing imported automobiles, pointed out their higher prices compared with those of Japanese automobiles. Also, the survey showed that 22.6 percent of respondents would purchase imported automobiles if their prices declined. The insufficient after-purchase service for imported automobiles was related to the low price sensitivity of demand. The same survey reported that 21.9 percent of respondents would not purchase imported automobiles without better after-purchase service. Some argue that automobiles made by American manufacturers were unpopular in Japan because American manufacturers did not shift the steering wheel to the right-hand side and did not adjust their models to better suit the Japanese climate and Japanese consumers’ tastes (New York Times, January 1, 1992). According to the consumer survey, however, these factors accounted for only 3.3 percent of all other reasons. Thus, these factors were important but were far from being the whole story behind the low popularity of imported automobiles. Furthermore, the same survey showed that 55.6 percent of imported automobile dealers recognized that prices of imported automobiles were higher in Japan than in the United States and Europe.

The 1990 price survey by the JFTC showed that imported automobiles were more expensive in Japan than abroad. Also, the survey showed that imported automobiles were more expensive than similar Japanese products in Japan. (New York Times, January 12, 1992) argued that high prices of American-made automobiles in Japan came from costs of shipping, requirements of standardization by the Japanese Government and high markups charged by dealers. 1/ The average total distribution margins of Japanese automobiles and of imported automobiles, in terms of suggested retail prices, were around 20 percent and 20 to 25 percent, respectively. The average actual total margins of Japanese automobiles, considering discounts, were around 10.9 percent. The margins for Japanese-made automobiles in the United States and Germany were 15 percent in terms of suggested retail prices and 7.6 percent after considering discounts. Nishimura (1991) estimated total margins (wholesale and retail) of automobiles for the United States and Japan. His study showed consistently poor performance or high margins for automobiles in Japan, The total margins of Japan were 22.2 percent in 1965, 19.4 percent in 1970, 33.5 percent in 1980 and 38.6 percent in 1985. Those of the United States were 18.3 percent in 1963, 16.6 percent in 1967, 20.9 percent in 1975 and 20.2 percent in 1977. These figures suggested that, in general, margins were higher in Japan than abroad although margins data for American-made and European-made automobiles in Japan were not available. Furthermore, the ratio of actual retail prices to suggested retail prices was 95.2 percent for Japanese automobiles and 94.9 percent for imported automobiles, implying that dealers of Japanese and imported automobiles applied similar discount rates. However, large differences in suggested retail prices existed between Japanese and imported automobiles. Thus, the actual price differences between Japanese and imported automobiles could be attributed to the high suggested retail prices of imported automobiles determined by the Importers.

The 1985 survey by the MITI showed that estimated distribution margins were higher than those suggested in the JFTC survey (EPA 1985). The retail margins were around 15 to 20 percent for both imported and Japanese automobiles. In addition to the retail margins charged by dealers, imported automobiles faced additional margins charged by the sole representative importers, which were estimated at around 10 to 20 percent. Thus, the total margins for imported automobiles were around 25 to 40 percent while those for Japanese automobiles were around 15 to 20 percent. The survey decomposed retail prices of imported automobiles into their constituent elements, placing retail prices at 100. The CIF prices accounted for around 50 to 60 percent, taxes accounted for around 10 to 15 percent, and total distribution margins accounted for the remaining portion. If the margins were assumed to be constant since 1985, these data suggest that high prices of imported automobiles were due to high margins set by importers and dealers. The JFTC report pointed out several reasons why the sole representative importers charged higher prices than abroad. As the first reason, as late-comers to the Japanese automobile market the sole representative importers paid higher initial set-up costs than Japanese manufacturers to develop sales routes. As the door-to-door sales strategy was a major marketing method in Japan, finding good sales personnel was difficult and costly at the margin. Also, establishing new sales outlets was difficult and costly as land prices in urban areas remained very high [Economist 1991, and Reilly (1992)]. As a second reason, the importers did not exploit economies of scale in distribution as the sales volume was small. As a result, distribution costs of imported products tended to be higher than those of domestic automobiles. As a third reason, specifications such as gas emission regulations determined by the Japanese government were quite different from those in effect abroad. Thus, importers needed to cover the costs of modifications before selling imported automobiles to customers.

Kida and Kiji (1990) showed that the wholesale margins in the Japanese automobile industry increased and were consistently higher than the average margins of consumer goods over the past fifteen years (Table 6). The main factor for the increase was increasing wholesale operational costs. Table 7 indicates that payrolls of the automobile industry increased from 27.5 percent in 1979 to 48.8 percent in 1986. While rents increased from 1979 to 1986, the relative impact was small. The retail margins in the automobile industry exceeded the average margins of consumer goods in 1986, but declined slightly from 25.2 percent to 24.3 percent between 1979 and 1986 because retail operational costs increased only from 18 percent to 18.9 percent. The share of payrolls in operating costs did not change between 1979 and 1986. The rents increased from 3.9 percent to 5.2 percent between 1979 and 1986 while the other costs declined from 37.2 percent to 35.2 percent. The operating costs had more than 50 percent share in both wholesale and retail distribution margins, as shown in Table 5 and Table 6. Operating profits accounted for only 20 to 30 percent of the margins in both the wholesale and retail sectors. Although the data included both domestic and imported automobiles, they suggested that the high prices of imported automobiles were partly due to high costs.

Table 6.

Distribution Margin Ratios in the Japanese Automobile Industry

article image
Source: Kida and Kiji, 1990.
Table 7.

Components of Operation Costs in the Japanese Automobile Industry

(In percent)

article image
Source: Kida and Kiji, 1990.

4. The repair and parts after-purchase market for imports

There were two types of repair facilities in the after-purchase market: dealers’ repair shops and independent licensed garages. Dealers’ repair shops procured over 90 percent of their parts and materials from dealers’ affiliated wholesalers and dealt only with “genuine” parts. 1/ “Genuine” parts were the parts made by the parts manufacturer with a contract to supply parts to a certain automobile manufacturer that provided detailed specifications. “Genuine” parts were distributed through both the automobile manufacturers’ channel (OE channel) and the independent wholesalers’ channel (non-OE channel). “Non-Genuine” parts were the parts produced at the initiative of the parts manufacturer without any contract and distributed through the non-OE channel. Independent licensed garages procured parts and materials from OE and non-OE channels. The variety of products procured by garages was the same as that of dealers’ repair shops, except that each independent garage needed frequent deliveries of small lots of parts and accessories for various car makes and models because of the variety of automobiles they service. There were 14,000 licensed dealer repair shops and 62,000 independent licensed garages in 1989.

Such heavy dependence of dealer repair shops on “genuine” parts might have lowered competition among parts manufacturers. The 1991 automobile parts price survey, jointly conducted by DOC and MITI, found that prices of identical or comparable uninstalled parts were 340 percent higher in Japan than in the United States. For installed parts, the prices were 198 percent higher in Japan. The report pointed out that virtually no U.S. made parts were found at auto parts outlets in either Tokyo or Osaka, which raised questions about the ability of U.S. parts suppliers to have fair access to the Japanese after-purchase market. The report further emphasized that “genuine” parts manufacturers controlled the Japanese after-purchase market, which might have significantly affected an automobile parts trade imbalance accounting for one quarter of the total U.S. trade deficit with Japan in 1990.

As in the case of imported automobiles, import car dealers and sales subsidiaries of foreign automobile manufacturers played a major role in importing parts and accessories for imported automobiles. Also, responding to an increasing number of parallel imports, the parallel importers organized the Foreign Automobile Importers Association (FAIA) to jointly purchase parts and automobiles. Such cooperation may have provided stable procurement of parts and automobiles, and hence, stable prices. There were four types of parts and accessories importers: (1) subsidiaries of foreign car manufacturers or foreign parts and accessories manufacturers (e.g., Mercedes Benz/Bosch Japan), (2) importer and foreign car dealers (e.g., Yanase, Osawa Shokai), (3) importer and independent wholesalers (e.g., Empire), and (4) parallel importers. The first was the “genuine” parts and accessories channel in which the parts and accessaries were imported from abroad and distributed through importers, foreign car dealer repair shops, sub-dealers and repair shops. In the second channel, importer and foreign car dealers directly procured parts and accessories from the foreign automobile manufacturer or through the manufacturer’s subsidiary, and distributed them through their own channel to their contract garages. In the third channel, importer and independent wholesalers procured mainly “genuine” accessories and fast moving parts from foreign exporters, then distributed to independent licensed garages, independent auto shops and other retailers. “Nongenuine” parts were directly procured from overseas local parts manufacturers. The fourth channel used the same distribution channel as the third channel.

The large number of independent repair garages, and auto shops provided the repair and maintenance function for different makes and models. There were dealers’ repair shops in which they usually repaired the automobiles produced by the manufacturer to which their dealers were affiliated. However, the exclusiveness issue at repair and maintenance stages was minor due to the existence of many independent garages. For the imported automobiles, the after-purchase market size was quite small due to the small import market share. The diversification and expansion of distribution channels for imported automobiles will probably promote the expansion of the after-purchase market. Thus, the promotion of organizing coordination among parallel importers will be desirable to attain stable procurement of automobiles and associated parts/accessories.

5. Conclusion

The Japanese imported car market expanded in the 1980s. The leading factor for the high sales performance was the aggressive marketing strategy taken by German manufacturers, such as the establishment of exclusive dealer networks. The increase in sales tie-ups with Japanese manufacturers also helped such a strategy. According to the consumer survey, the higher prices of imported automobiles compared to those of Japanese, were one of the major reasons for not purchasing imported automobiles. The high distribution markups charged by importers and dealers partly explained the high prices. The major reasons why importers and dealers set up higher markups were partly attributed to the fact that they were unable to expand distribution channels due to high initial set-up costs as late-comers, high operating costs and strict specifications required by the Japanese Government. The increase in the number of parallel importers, on the other hand, suggested that sufficiently large import distributors’ rents existed (Table 8). We expect that the diversification and expansion of distribution channels for imported automobiles will promote intra-brand competition and thus, will prevent the sole representative importers from setting high retail prices. The increase in competition will probably lead to the more open distribution system in which nonexclusive dealerships are commonplace.

Table 8.

The Parallel Imports for Mercedes-Benz and BMW

(Number of vehicles and in percent, as indicated)

article image
Source: The Japan Fair Trade Commission Report, March 1987.

While the conventional business practices under the “sales - distribution Keiretsu” system may have increased Japanese manufacturers’ efficiency, such efficiency may have been achieved at the expense of consumers’ surplus. The distribution system is probably not the only factor to explain the small amount of intra-industry trade and substantial net exports in the Japanese automobile industry. However, historical and empirical surveys suggested that a higher degree of openness in the American automobile distribution system facilitated the import penetration of Japanese automobiles. 1/ By contrast, in the Japanese automobile distribution system, exclusive dealership arrangements between Japanese manufacturers and dealers made it difficult for foreign manufacturers to establish dealer networks. An increase in land prices and wages since the 1970s accelerated the set-up costs. As a result, foreign manufacturers heavily depended on the sole representative import system, which may have provided the sole representative importers an opportunity to set high markups.

Thus, diversifying and expanding distribution channels are necessary to increase the market share of imported automobiles. The recent sales tie-ups with Japanese manufacturers, as shown in Table 5, is one of the ways of achieving its objective. The tie-ups with Japanese manufacturers may put foreign manufacturers in an disadvantageous position as dealers generally further the interests of domestic manufacturers. However, we expect that setting up the target of annual sales volumes between domestic and foreign manufacturers will probably reduce such problems in the short run because such a strategy will induce Japanese manufacturers and their dealers to commit themselves to the targets. In fact, domestic and foreign manufacturers have begun to jointly decide such sales volumes in recent years. Then, we can expect that as the number of dealers which distribute imported automobiles increase, the import car market share will increase with a decline in prices. Thus, the Japanese automobile distribution system will probably become similar to that in the United States. If the dealers’ profitability declines with intense intra-brand and inter-brand competition, dealers of Japanese manufacturers will have more incentives to distribute products of other manufacturers as well. As most dealers are financially independent, they will select nonexclusive dealership arrangements as long as the benefits to do so exceed the costs.


  • Ariga, K., Y. Onichi and T. Kaneko (1993), “Ryutsu System to Kakaku Keisei,” Financial Revenue, February, pp. 69-95.

  • Baldwin, R. (1988), “Hysteresis in Import Prices: The Beachhead Effect,” American Economic Review, Vol. 78, pp. 773-85.

  • Bhagwati, J. (1991), The World Trading System at Risk, Harvester Wheatsheaf.

  • Corker, R. (1989), “External Adjustment and the Strong Yen: Recent Japanese Experience,” International Monetary Fund Staff Papers Vol. 36, No. 2, pp. 464-93.

    • Search Google Scholar
    • Export Citation
  • Economic Planning Agency, “Yunyuhin no Ryutsu oyobi Shokanko,” Keizai Kikakucho, Bukkakyoku, Bukaseisakuka, March 1986.

  • Economic Planning Agency, “Nihon no Ryutsu System: Riron to Jisyo,” Keizai Bunseki, No. 123, May, 1991.

  • Economic Planning Agency, “Nihon no Sangyo Nai Boeki,” Keizai Bimseki, No. 125, June 1991.

  • Economist, “Japan’s Next Retail Revolution,” pp. 79-80 December 21, 1991.

  • Economist, “Japan’s Car Imports Must Try Harder,” pp. 68-73 June 22, 1992.

  • Flath, D. (1986), “Why Are There So Many Retail Stores In Japan,” Working Paper No. 17, North Carolina State University.

  • Interim Report by the Japanese Delegation to the Japan/U.S. Working Group on Structural Impediments Initiatives, April 5, 1990.

  • Ito, M., “The Japanese Distribution System and Access To the Japanese Market,” NBER Conference on the US and Japan: Trade and Investment, October 19-20, 1989.

    • Search Google Scholar
    • Export Citation
  • Ito, M., and A. Matsui (1987), “Nihonteki Torihiki,” in Oyo Micro eizaigaku, Tokyo: Tokyo University Press.

  • Ito, T., (1992), “Shohisya Jyushi no Keizaigaku,” Nihhon Keizai Shinbunsya.

  • Ito, T., (1992), “The Distribution System,” in The Japanese Economy, Cambridge: The MIT Press, Chapter 13, pp. 385-504.

  • Ito, T., and M. Maruyama, “Is the Japanese Distribution System Really Inefficient?,” NBER Conference on U.S. Japan Project, October 19-20, 1989.

    • Search Google Scholar
    • Export Citation
  • Japan Automobile Manufacturers Association, “Automotive Distribution in Japan,” June, 1990.

  • Japan Automobile Manufacturers Association, (1991), “1991 Nihhon no Jidosya Kogyo”.

  • Japan Automobile Manufacturers Association, (1991), “The Motor Industry of Japan”.

  • Japan Economic Institute, “The Japanese Distribution System,” JEI Report, No. 28A, June 24, 1987.

  • Japan Economic Institute, “Japan’s Price Structure,” JEI Report, No. 4A, June 26, 1990.

  • Japan Economic Institute, “The Structure of Japan’s Imports: Causes and Consequences,” JEI Report, No. 22A, June 14, 1991.

  • Japan External Trade Organization (1991): “The Japanese Distribution System,” JETRO Monitor, Vol. 5, No. 10, pp. 1-4.

  • Japan External Trade Organization, (1992), “Japan’s Automobile Import and Inspection Practices,” JETRO Monitor, Vol. 6, No. 12, pp. 1-3.

    • Search Google Scholar
    • Export Citation
  • Japan Fair Trade Commission, “Jidosha Torihiki no Koseika ni Kansuru Shido Jyokyo,” November, 1981.

  • Japan Fair Trade Commission, “Ryutsu Kozo Torihiki Keitai no Kokusai Hikaku to Kyoso Seisaku,” Ryutsuka Kokusai Hikaku Kenkyukai, April, 1986.

    • Search Google Scholar
    • Export Citation
  • Japan Fair Trade Commission, “Shohizai no Heiko Yunyu ni Kansuru Jittai Chosa,” Nihon Research Sogo Kenkyugyo, March, 1987.

  • Japan Fair Trade Commission, “Shijyo Access no Kaizen to Kyoso Seisaku,” Dokusen Kinshi Konwakai Shiryosyu-X, February, 1986.

  • Japan Fair Trade Commission, “Naigai Kakakusa ni Kansuru Shohizai 9 hinmoku no Ryutsu Jittai Chosa Hokokusyo,” Naigai Kakakusa ni Kansuru Chosa Kenkyukai, June, 1990.

    • Search Google Scholar
    • Export Citation
  • Japan Fair Trade Commission, “Ryutsu Torihiki Kanko to Korekarano Kyoso Seisaku: Hirakareta Kyoso to Shohisya Rieki no Tameni,” Ryutsu Torihiki Kankoto to Kyoso Seisaku ni Kansuru Kento Iinkai,” June, 1991.

    • Search Google Scholar
    • Export Citation
  • Japan Fair Trade Commission, “Keizai Kozo no Kenkyu to Sangyo Soshiki,” Dokusen Kinshi Konwakai Shiryosyu-XII, March, 1992.

  • Kearney, A. T., “Distribution System of the Japanese Auto Parts Aftermarket: Final Report,” U.S. Department of Commerce, June 1991.

  • Kida, K. and M. Kiji, “Ryutsu Kozo to Ryutsu Hiyo no Henka,” Discussion Paper No. 90-DOJ-21, MITI, Research Institute of International Trade and Industry, October, 1990.

    • Search Google Scholar
    • Export Citation
  • Lawrence, R. Z. (1987), “Imports in Japan: Closed Markets or Minds?,” Brookings Papers on Economic Activity, No. 2, pp. 517-48.

  • Lawrence, R. Z., (1991), “Efficient or Exclusionist?: The Import Behavior of Japanese Corporate Groups,” Brookings Papers on Economic Activity, No. 1.

    • Search Google Scholar
    • Export Citation
  • Lawrence, R. Z., “Keiretsu and Japan’s Rigged Distribution System,” International Economy, pp. 64-67, May-June, 1991.

  • Lincoln, E. J., (1990), Japan’s Unequal Trade, Washington D.C.: The Brookings Institution.

  • Maruyama, M., (1991), “Fact and Fiction in Japan’s Distribution System,” Economic Eye, pp. 4-8, Autumn.

  • Maruyama, M., Y. Togawa, K. Sakai, N. Sakamoto and M. Arakawa, “Distribution System and Business Practices in Japan,” EPA International Symposium on Structural Problems in the Japanese and World Economy, October 12-13, 1989.

    • Search Google Scholar
    • Export Citation
  • MITI (1988), White Paper.

  • MITI, “Naigai Kakakusa Chosa ni tsuite,” June 22, 1993.

  • Miyazawa, K. (ed.) (1991), Kokusaika Jidai no Ryutsu Kiko, Syoji Homu Kenkyukai.

  • Miwa, Y. and K. Nishimura (ed.), (1991), Nihon no Ryutsu, Tokyo: Tokyo Univ. Press.

  • Nishimura, K., “The Distribution System of Japan and the United States: A Comparative Study from the Viewpoint of Consumers,” a paper presented to the Japan Economic Seminar, Nov. 16, 1991.

    • Search Google Scholar
    • Export Citation
  • Nishimura, K., and H. Tsubouchi (1990), “Nihon no Ryutsu ha Shohisya nitotte Kouritsuteki ka? Sangyo Renkanhyo ni motozuku Shohizai Ryutsu Marginritsu no Nichibei Hikaku,” Keizaigaku Ronsyu, Vol 56, No. 2, pp. 43-59.

    • Search Google Scholar
    • Export Citation
  • Nishino, T. (1991), Hitomede Wakaru Ryutsu Keiretsu to Shijo Chizu, Nihon Jitsugyo Syuppansya.

  • Nissan Motor Corp. (1990), “Jidosya Sangyo Handbook.”

  • Noland, Marcus, “U.S.-Japan Trade Friction and its Dilemmas for U.S. Policy,” paper presented for the Japan Economic Seminar, October 8, 1993.

    • Search Google Scholar
    • Export Citation
  • Ohno, K., “Export Pricing Behavior of Manufacturing: A U.S.-Japan Comparison,” IMF Staff Papers, Vol. 36, No. 3, September, 1989.

  • Petri, P. A. (1990), “Market Structure, Comparative Advantage, and Japanese Trade under the Strong Yen,” in Trade with Japan: Has the Door Opened Wide?, (ed.) Paul Krugman, University of Chicago Press for NBER.

    • Search Google Scholar
    • Export Citation
  • Pemberton, M. (1992), “Europe’s Motor Industry After 1992: A Review of Single Market Legislation and its Implications,” The Economist Intelligence Unit, Special Report No. 2090.

    • Search Google Scholar
    • Export Citation
  • Reilly E. J., “An Approach to the Japanese Market,” Japan Close-Up, No. 4, pp. 21-23, March, 1992.

  • Saxonhouse, Gary R., “Differentiated Products, Economies of Scale and Access to Japanese Market,” in Robert M. Feenstra (ed.) Trade Policies for International Competitiveness, University of Chicago Press University, January.

    • Search Google Scholar
    • Export Citation
  • Sazanami Y., S. Urata and Y. Kawai (1993a), “Trade Protection in Japan,” Institute for International Economics, draft.

  • Sazanami Y., S. Urata and Y. Kawai, (1993b), “Naigai Kakakusa no Keizaigaku,” Keizai Seminar, August, 1993b.

  • Shibayama, K. and M. Kiji (1990), “Ryutsu no Okeru Kyoso to Soshiki,” Tsusho Sangyo Kenkyusho, Kenkyu Series 3.

  • Tokyo Business Today, “Toyota Mulls Volkswagen Car Imports,” pp. 8, June, 1991.

  • Toyota Motor Corp., (1979), Toyota no Ayumi: Bumonhen.

  • Toyota Motor Corp., (1987), Sozo Kagirinaku.

  • Toyota Motor Corp., “The Automobile Industry: Toyota and Japan,” 1991 Edition.

  • U.S. Department of Commerce, “Department of Commerce/ Ministry of International Trade and Industry Automotive Parts Price Survey,” June 27, 1991.

    • Search Google Scholar
    • Export Citation
  • Yoshitomi, M., “Keiretsu: A Legitimate, Viable Capitalist Approach,” International Economy pp. 68-75, May-June 3, 1991


I would like to thank Jagdish Bhagwati, David Canning, Jay Choi and Hugh Patrick for helpful comments. I am especially grateful to Mohsin Khan, Ratna Sahay and Peter Wickham. An earlier version of this paper appears as Chapter 1 of my Ph.D. dissertation completed at Columbia University. The opinions expressed in this paper are those of the author and do not necessarily reflect those of the IMF.


Noland pointed out two factors that might have affected the trade structure of Japan: (a) direct trade and industrial policy interventions; and (b) structural characteristics of the economy. As for the former, tariff levels in Japan were 1.9 percent, being comparable to those maintained by the United States and the European Community, and quotas were largely eliminated. The data regarding nontariff barriers, on the other hand, showed that a simple average of 46.7 percent of U.S. exports to Japan encountered some form of nontariff barriers. As for the latter, structural barriers alleged to deter imports included the reliance on bureaucratic control to ensure product safety; domestic cartels and discriminatory networks of affiliated firms (Keiretsu); lack of intellectual property protection; government procurement procedures that advantaged domestic suppliers; and restrictions on the distribution channels for imported products.


This pricing strategy is not unique to the sole representative import system. For example, sales firms which are wholly-owned by foreign manufacturers are often reluctant to have discount sales to maintain suggested retail prices. This strategy is also observed in the United States and Europe.


Maruyama, et al. (1989) define the ratio of distribution margins as follows: (Mw + Mr)/R = (Mw/W)(W/R) + (Mr/R)

where Mw denotes wholesales gross margins,

Mr denotes retail gross margins,

W denotes wholesales,

R denotes retail sales.

Japan’s ratio of distribution margins is higher than each ratio of retail gross margins and wholesale, gross margins. This is because the (W/R) ratio is relatively higher in Japan.


The results of these studies may have suggested that the Japanese distribution system was not necessarily inefficient and thus, the Large Retail Store Law should not be accused of protecting inefficient small-sized retailers. Ito (1992) pointed out three problems associated with this “specific consumer’s buying behavior” view. First, Japanese consumers began to have large refrigerators that could store foods for a week. In addition, large-sized retailers improved their storage management skills, which enabled them to provide products with a high degree of freshness. Second, Japanese consumers began to prefer large-sized retail stores as more housewives began to have jobs and had less time for shopping. Third, the elimination of the law would have no impact on consumers’ buying behavior if the view were correct. Thus, there would be no reason to support the continuation of the law.


Small- to medium-sized retailers are not subject to regulations on opening hours or closings.


They showed that the price gaps mostly came from nontariff factors. Their simulation results showed that the total quantity of imports of the products under study would double and domestic production would decline by 8.3 percent if the observed price gaps were eliminated. Furthermore, the elimination of the price gaps would increase consumers’ surplus by $152.2 billion, roughly equivalent to 5.3 percent of GNP and 9.3 percent of total private consumption. On the other hand, employment due to the reduction in domestic production would fall by 8.9 percent while overall employment would fall by 0.6 percent. The government would face a tariff revenue loss of about $3 billion and importers would lose import quota rents of about $26.1 billion. The overall efficiency gain from the elimination of price gaps would be about $30.4 billion, or 1.1 percent of GNP.


The JFTC survey found that 40.8 percent of the industries overall were regulated. With the regulated industries, 59.1 percent were considered strongly regulated. These industries included mining, construction, finance and electric and gas utilities. The products in these industries were expensive because the government provided explicit entry barriers or price controls. See JEI Report (1/26/92, p. 4).


The JFTC has been considering major reforms on the resale price maintenance practice that required retailers to sell books, music recordings and drugs at prices fixed by publishers and manufacturers. According to the reform draft, booksellers will be allowed to sell certain categories of books, magazines and other publications at any price after a certain period of time. Also, the prices of records and compact discs will be determined entirely by music stores (Nikkei (7/6/91)). The JFTC guidelines specified that the suggested retail prices by manufacturers and the practices to block parallel imports to maintain sales prices were illegal (Nikkei (7/11/91)).


Foreign manufacturers provide the exclusive sales right to importers in order to facilitate import procedures, expand sales - distribution channels, increase after-purchase service facilities and promote advertisement. The sole representative import agency system is one way for foreign manufacturers to enter targeting markets and thus, is also observed in Europe with some regulations.


An expansion of parallel imports are sometimes blocked by foreign manufacturers which attempt to maintain a “brand image” of their products by creating an excess demand. When these foreign manufacturers observe an increasing demand for their products, they often limit their supplies to importers which take a low price-large volume strategy. For example, Tiffany, Chanel and Hermes utilize this supply-control policy in many countries.


Matsushita originally introduced a new distribution system called “Renmei Sei” in 1935 in order to build a brand image of Its products by improving services. Matsushita began to buy out existing wholesalers and controlled them as exclusive trading firms in each district. At the retail level, it gradually formed affiliated retail stores called “Keiretsu stores” by ranking stores on the basis of sales shares of the manufacturer’s products and rewarding them with financial supports and management guidance. Other manufacturing firms followed this strategy in the early 1960s. The number of Keiretsu stores were 25,000 for Matsushita, 13,000 for Toshiba, 10,000 for Hitachi, 5,000 for Mitsubishi and 5,000 for Sanyo. The sales share of Keiretsu stores accounted for 41 percent of total sales in 1985 compared with 73 percent in 1975.


Large-sized manufacturers, such as Shiseido and Kanebo, controlled Keiretsu retail stores and distributed their products to these stores through their sales firms. Keiretsu stores sold only their manufacturers’ products. The sales share of Keiretsu stores accounted for 43.3 percent in 1990.


The JFTC (1990) showed that 100 percent of sampled imported car dealers completely purchased automobiles from suppliers while 80 percent of Japanese car dealers did so. In addition, 94.4 percent of the former could not return products while only 60 percent of the later could not do so. Therefore, 72.2 percent of the former replied that they would sell out products by discounting prices while 60 percent of the latter would do so. Furthermore, only 22.2 percent of the former received guidance or requests regarding sales prices from suppliers. Thus, imported car dealers usually had the retail price setting power and hence, may have increased retail margins after purchasing automobiles from suppliers.


New York Times (1/12/92) pointed out that $40,000 Cadillac Seville in Manhattan was sold at $60,000 in Tokyo. It also stressed that Japanese automobile dealers insisted on big markups to compensate for low sales volume.


See A.T. Kearney (1991), pp. 26-31 and pp. 40-41. This policy is also observed in the United States and Europe.


See Sayuri Shirai, “The History of the Distribution System in the Automobile Industry: the United States and Japan,” mimeo., Columbia University, May 1993.