The author would like to acknowledge helpful suggestions received from Professors Reischauer, Perkins, Rosovsky, and Craig of Harvard. He is particularly grateful to Professor Myers, colleague at the University of Miami, for his detailed criticism and suggestions, which resulted in useful revisions of the original draft of this article.

¹ For their detailed views see Ohkawa, Kazushi and Rosovsky, Henry, “The Role of Agriculture in Modern Japanese Economic Development,” Economic Development and Cultural Change, Vol. 9, no. 1 (October 1960).CrossRefGoogle Scholar

² Rosovsky, Henry, “Rumbles in the Ricefields: Professor Nakamura vs. the Official Statistics,” JAS, Vol. 27, no. 2 (February 1968), p. 358Google Scholar; and Ohkawa, Kazushi, “Agriculture and Turning Points,” The Developing Economics, Vol. 3, no. 4 (December 1965), p. 478Google Scholar. In his book. Agricultural Development in China, 1638–1968 (Chicago, 1969)Google Scholar, Dwight H. Perkins also used this argument to advance his own thesis. Simply stated, the themes of his book are: China's traditional agriculture produced enough grain for her population to grow from 65 million in 1400 to 400 million in the early 19th century, and 700 million in the 1950's. Half of this increase came from the expansion of cultivated land, mainly brought about by the advancing frontiers. The other half resulted from rising land productivity achieved by additional capital and labor but with little change in technology. Population increase was indispensable for opening up the frontiers and for increasing land productivity because double cropping would have been impossible without a supply of labor. Thus, for Perkins, population growth served as the major “engine” for agricultural progress. Explaining the insignificant role of technology in China, Perkins claimed that the best technology spread steadily and even rapidly across China, so that by some unspecified time between 1400 and 1950 there was no backlog of technology. Such a situation was able to occur because China, unlike Tokugawa Japan, had no border or feudal barriers. Perkins implied several times that technological diffusion was better in China than in Tokugawa Japan.

³ Nakamura contends that there already existed surplus created by high Tokugawa productivity in agriculture. The Meiji Revolution merely created an institutional mechanism to exploit the surplus for industrialization. Nakamura sees an historical continuity in Meiji economic development. Nakamura believes that Meiji Japan, too, made sacrifices to achieve modest rates of economic growth whereas Ohkawa and Rosovsky indicate that the jump in Meiji land productivity came about without additional inputs but merely exploiting the technological backlog. See Nakamura, James I., Agricultural Production and Economic Development of Japan (Princeton, 1966).Google Scholar

⁴ See Rosovsky, “Rumbles in the Ricefields,” pp. 359–60; and Ohkawa, “Phase of Agricultural Development and Economic Growth,” in Ohkawa, K., Johnston, Bruce F., and Kaneda, Hiromitsu, eds. Agricultural and Economic Growth: Japan's Experience (Princeton, 1970), p. 9.Google Scholar

⁵ To be fair to Ohkawa, it should be noted that he speaks of increasing sectoral interdependence and contribution of other sectors such as increasing and diversifying the demand for agricultural goods. But the only source of growth Ohkawa and Rosovsky identified clearly and specifically was the technological backlog. Particularly they failed to mention specialization of land and labor as a possible source of growth. See Ohkawa, “Concurrent Growth of Agriculture and Industry,” in Dixey, Roger N., ed. International Explorations of Agricultural Economics (Ames, Iowa, 1964), pp. 202–205.Google Scholar

⁶ Hayami, Yujirō and Yamada, Saburō, “Technological Progress in Agriculture,” in Klein, Lawrence and Ohakawa, Kazushi, eds. Economic Growth: The Japanese Experience Since the Meiji Era (Chicago, 1968), pp. 143–144Google Scholar. The question is whether border guards succeeded in preventing the smuggling of better seeds. Furthermore, such examples can be construed as proof that some han were bent on acquiring the best technology. The Bakuhan system was not responsible for the slow spread of technology. One has to remember that slow technological diffusion was universal in pre-modern times. An innovation did not travel more than a mile a year even in England on the eve of the First Industrial Revolution.

⁷ I came to entertain this view after doing research on life in Tokugawa villages in preparation for my forthcoming book. The manner in which Shibusawa Eiichi's father managed his indigo farm, peasant response to the emergence of a silk and tea trade in the 1860's, and reasons why the young men in Tokugawa villages left for Edo and Osaka in the 19th century, led me to believe that these restrictive decrees were usually dead letters. It should be added that han control over commoners was intact up to the 1870's although the Bakufu's control over the han disintegrated during the sixties.

⁸ Yuri, Masamichi, Shishaku Yuri Kimimasa den (A biography of Viscount Yuri Kimimasa) (Tokyo, 1940), p. 83.Google Scholar

⁹ Yasuoka, Shigeaki, Nihon hōken keizai seisaku shi ron (A historical study of economic policies during the Tokugawa period) (Tokyo, 1959), pp. 35 and 38.Google Scholar

¹⁰ Horie, Yasuzō, Kinsei nihon no keizai seisaku (Economic policies of the han during the Tokugawa Period) (Tokyo, 1942), pp. 64 and 113.Google Scholar

¹¹ Rosovsky, “Rumbles in the Ricefields,” p. 358.

¹² Shinriki became the predominant variety first in western Japan and then in other regions. Most of the high yielding varieties of rice did not come from the Tokugawa technological backlog but were created after the 1870's. See Yasuda, S., Meiji iko ni okeru nōgyō gijutsu no hatten (Development of agricultural technology since Meiji (Tokyo, 1952), pp. 6–7 and 14.Google Scholar

¹³ When Japan secured an imperial foothold in Manchuria after the Russo-Japanese War, the supply of soybean mash increased. The eras of soybean mash and Shinriki use coincided. It is interesting to note that most of the chemical fertilizer plants were located in the Kinki and Tokyo area.

¹⁴ Seiichi Tōbata, et al., “Nihon nōgyō hattatsushi” (A history of Japanese agricultural development) (Tokyo, 1954), IV, 233.

¹⁵ Hayami and Yamada, p. 151. The Kinki region presently covers the prefectures of Kyoto, Osaka, Nara, Wakayama, Mie, Hyōgo, and Shiga.

¹⁶ Ōshima, Nobujirō, Nihon kōtsū shiron (A study of Tokugawa Japan's transport) (Tokyo, 1964), pp. 331–332Google Scholar. The Kinai is an historical name meaning an area where the sovereign resides. The Kinai was the birth place of Japan's ancient civilization, as well as its rice and cotton culture. It seems that the Kinai was endowed widi favorable and natural factors, as in ancient times natural factors were much more important in determining land productivity than technology. The Kinai is the economic heart of the Kinki region.

¹⁷ sha, Ryūmon, ed., Shibusawa Eiichi denki shiryō (Collection of biographical materials on the life of Shibusawa Eiichi), 58 vols. (1955–1965), X, 384Google Scholar. See also Ukita, Noriyoshi, “Kinai nōgyō no tokushitsu to sono hempō,” (The characteristics of the Kinai agriculture) in Shibata, Minoru, ed., Kinki ken (The Kinki region) (Tokyo, 1969), p. 67.Google Scholar

¹⁸ Ukita, Noriyoshi, “Kinsei no Kinai nōson,” (The Kinai agriculture during the Tokugawa Period) in Fujioka, Kenjirō, ed. Kinki rekishi jiri kenkyū (A study of the historical geography of the Kinki) (Tokyo, 1958), pp. 232–236. The official productive capacity of the two largest han in the Kinai were only 200,000 and 100,000 koku of rice.Google Scholar

¹⁹ Generally speaking, the Bakufu was favorably disposed toward the chōnin in its domains, taxing them very little and relying instead on the land taxes, minting, and currency debasements for its fiscal needs. The Bakufu also protected the chōnin from political harassment by the daimyo. When the daimyo attempted to cheat the chōnin, a law suit ensued, and the Bakufu sided with the chōnin. More importantly, the tenryō was the place where the daimyo marketed their taxes in kind, and where a big share of national spending occurred, Some historians of the Tokugawa period have even claimed that the chōnin took refuge in the tenyrō. For these views see Takekoshi, Yosaburō, Nihon keizai shi (An economic history of Japan), 8 vols. (Tokyo, 1925), Vol. 1, pp. 312–313Google Scholar; Sakudō, Yotarō, Kinsei nihon no kahei shi (A history of the Tokugawa currency) (Tokyo, 1958), pp. 144–170Google Scholar; Yamaji, Aizan, Gendai kinken shi (The power of the modern businessmen) (Tokyo, 1908), p. 30.Google Scholar

²⁰ Some examples are high-yield seeds, knowledge of when best to plant them, knowledge of how to prepare the soil, most productive tools, etc. Unlike industrial technology, the transferability of farming technology is limited: what is best for farmers of one region is not necessarily the best for farmers in another locale. The best time to sow will vary from region to region; crops will be subject to different diseases and pests.

²¹ Nawa, Tōichi, Nihon bōsekigyō no shiteki kenkyū (A historical study of the Japanese cotton spinning industry) (Tokyo, 1948), pp. 56–57.Google Scholar

²² Furushima, Toshio, Nihon nōgyō gijutsu shi (A history of farming technology) (Tokyo, 1946), II, pp. 547–571.Google Scholar

²³ As the land taxes were commutated from 1871, farmers were given freedom of choice in using their land and labor but were subjected to market discipline. In the past, agricultural products were marketed in Osaka and Edo by han, but with the commutation of taxes farmers had to sell their products. Consequently, there appeared local and regional markets.

The great inflation accelerated monetization and commercialization of the Meiji economy. One could argue that by the end of the inflation the Meiji economy had undergone a “commercial revolution.” The rising farm prices in conjunction with the new land tax system (chisō kaisei) must have stirred entrepreneurship in agriculture. All in all, these changes accelerated commercialization and specialization which had been going on since the middle of the 19th century.

As everyone knows, market size defines the degree of specialization. The building of the railroad network during the Meiji period greatly expanded the markets for farm products, including rice. In addition, improvements and changes in lifestyle prompted farmers to specialize further. Farmers in Aomori Prefecture began to raise apples for distant markets partly because the Japanese began to eat apples, and partly because the railroads could transport apples to distant markets at reasonable costs. The rise in per capita income gave rise to the confectionary, shōyu-miso, and beer industries. The rise of such industries created a demand for wheat and soybean which, in turn, induced farmers to specialize in them. Without providing direct evidence, they allege that technological diffusion was accelerated with the Meiji Revolution. I am in agreement although I do disagree with their statements on the extent and causes of acceleration. The acceleration after the Meiji Revolution was caused not so much by the disappearance of the feudal barriers as by the development of the railroad. The railroad increased factor mobility, as well as fertilizer distribution and population movement. People simply traveled more than ever before. This must have enhanced technological diffusion, although I have no direct evidence to indicate the extent. I also feel that market expansion, like free trade, introduces competition which, in turn, prompts farmers to adopt the best technology. These points of views can be supported by Tetsudō In, ed. Hompo tetsudō no shakai oyobi keizai ni okeru eikyō (Economic and social effects of the railroads) (Tokyo, 1916), Vol. 2.

²⁴ In this excellent article, differing from Ohkawa and Rosovsky, he indicates an increase in the labor input during the Meiji era. See Crawcour, E. Sydney, “Japan, 1868–1920,” in Shand, Richard T., ed. Agricultural Development in Asia (Berkeley, 1969), p. 16.Google Scholar

²⁵ The size of the official retinue and the trips' duration varied from daimyo to daimyo depending on their status and location. As late as the 1850's, a daimyo with only 100,000 koku domain made the Sankin kōtai trip with a retinue of several hundred. Once a lord of Chōshū made the trip with a retinue of two thousand, taking a month. See Kamekichi Takahaski, Nihon kindai keizai keisei shi (The formation of Japan's modern economy) (Tokyo, 1968), I, pp. 46–50.

²⁶ Shimazaki Tōson (1872–1943), in his long historical novel, Yoake mae (Before the dawn) (Tokyo, 1936) gives a very vivid picture of the Sankin kōtai system. He shows how it served as a sensitive national nerve system and stimulated the Tokugawa economy. This novel has great historical value as it was based on the records of his family, which had been for generations station masters on the Sankin kōtai highway. In accounting for the First Industrial Revolution it is often cited that England was blessed with inexpensive waterways and that no point in England was further than seventy miles from the coasts. One can say almost the same about Tokugawa Japan.

²⁷ Kenkawa, Taichi, Hompō menshi bōseki shi (A history of the Japanese cotton spinning industry), 5 vols. (Tokyo, 1937–1941), II, p. 259.Google Scholar

²⁸ Takahashi, I, pp. 46–58. One of the Mito han's compounds covered 400 acres, while a small han's compounds covered four acres. Owari han maintained 600 dwellings in Edo. See Yoake mae, II, p. 195. Honjō states that the sankin kōtai expenses often constituted 70% of the han's budgets, and that the han usually stationed about 20% of its officials in Edo. See Honjō, Eijirō, Bakumatsu no shin seisaku (New policies during the end of the Tokugawa Period) (Tokyo, 1940), p. 145Google Scholar. The size of these compounds and their grandeur depended on a daimyo's status and vanity.

²⁹ Yamagata, Bantō, “Yume no shiro,” in Takimoto, Seiichi, ed. Nihon keizai sōsho (Library of Tokugawa economic writings) (Tokyo, 1914–1917), XXV, pp. 322–323. Yamagata was one of the two leading chōnin-scholars in Osaka and wrote extensively on the economic and financial problems of his day. “Yume no shiro” is his major work.Google Scholar

³⁰ Here one is reminded of Sombart's thesis that luxury gave birth to capitalism. See Sombart, Werner, Luxury and Capitalism, tr. Dittmar, W. R. (Ann Arbor, Michigan, 1967), p. 114.Google Scholar

³¹ This view is widely held by writers of today as well as by Tokugawa writers. Today's writers include Honjō Eijirō, Tsuchiya Takao, and Yamaguchi Kazuo. Many Confucian scholars observed that the Sankin kōtai, the resultant urbanization, and the rise of a money economy plunged the daimyo into debt. Their writings are collected in Nihon keizai sōsho, 36 vols. It is essential to read the following works in order to understand the Tokugawa economy: “Seidan” by Ogyū Sorai (1666–1728), “Keizai roku” by Tazai Shudai (1680–1747), “Keizai dan” by Kaiho Seiryō (1755–1817), “Saitei no ho” by an unknown author, and “Yume no shiro.” They are to be found in Nihon keizai sōsho, Vols. 3, 6, 18, 21, and 25.

³² It is very difficult to document the fact that Tokugawa agricultural production increased extensively enough to create a surplus. Even Ohkawa, who thinks that the agricultural output did not increase significantly, states that a high surplus was created by unequal distribution of income, caused by the fact that the daimyo and their retainers, who constituted only about 7% of the population, taxed their peasants heavily. Japanese historians generally agree that peasants were exploited, since the tax rates ranged from 40 to 60%. However, it is impossible to estimate in reality the percentage of daimyo and retainer shares since it is very unlikely that what the peasants paid as taxes constituted such high percentages of what they actually produced annually. The Bakufu and the han governments also issued currencies which must have given them additional command over real goods via inflation. The fact that the daimyo taxed more than one fourth of the land outputs indicates the degree of the daimyo's grip on farmers. This high capacity to tax is one of the favorable factors that the Meiji government inherited. One wonders how Tokugawa political authorities could tax so much if the per capita income had been as low as suggested by Kuznets and odiers.

³³ Maeda, Masana, “Kōgyō iken” (How to develop the Meiji economy) in Meiji zenki zaisei keizai shiryō shūsei (Collection of early Meiji financial and economic data), ed. Ōuchi, Hyōe and Tsuchiya, Takao, (Tokyo, 1931), XVIII, pp. 112–126Google Scholar. In many cases han were responsible for maintaining high quality of craftsmanship and products. After the Meiji Restoration, the quality of rice deteriorated in some localities as han government disappeared. See Yasuda, p. 16.

³⁴ Craig, Albert, Chōshū in the Meiji Revolution (Cambridge, 1961), p. 30.Google Scholar

³⁵ The usual view among Japanese writers is that about 3 million koku of rice was marketed in Edo and Osaka, respectively. See Kazuo, Yamaguchi, Nihon keizai shi (An economic history of Japan) (Tokyo, 1968), p. 63Google Scholar; Furushima, Toshio, Kinsei nōgyō no tenkai (Agricultural development in Tokugawa Japan) (Tokyo, 1963), p. 62Google Scholar. However, Dohi Noritaka gives the figure of less than two million koku as the volume of rice marketed in Osaka about 1818–1829. In any case the volume of goods collected and distributed from Osaka constituted a large percentage of the total. I am inclined to feel that the combined market power of Osaka and Edo equalled that of London in the 18th century.

³⁶ Yamaguchi, p. 34.

³⁷ In my article, “Tokugawa Feudalism and the Emergence of the New Leaders of Early Modern Japan,” Explorations in Entrepreneurial History, IX, no. 2 (1956), I initially put forward the thesis that Tokugawa feudalism contained a built-in mechanism. This mechanism was analogous to the price mechanism under competitive yet oligopolistic conditions; Tokugawa feudalism contained particularly a ready-to-respond mechanism. I elaborate this thesis in my forthcoming book. The allegation that the Bakuhan system effectively blocked technological diffusion implies a certain growth mechanism for the Tokugawa economy. I find it extremely difficult to use this explanation to account for the most unique characteristics concerning Meiji's modern economic growth. These characteristics are not the growth rates of the Meiji economy as much as the fact that Japan entered almost “instantly” the modern economic growth process (1885–1890), even ahead of some European countries. Events in Japan moved quickly from an initial acceptance of Western trade to a national commitment to development in 1871, and subsequently to the beginning of modern economic growth. If modern economic growth in Japan had started only in the twentieth century, then the kind of growth mechanism that the Tokugawa economy might have had and the entire Tokugawa period would only be of historical interest to students of economic development.

³⁸ This point fits more properly with the concurrent growth theory than does the proposition that a single input, indigenous farming technology, was responsible for the rise in Meiji land productivity. If one cites specialization and diffusion of indigenous technology as the two sources of growth in land productivity, then it is much easier to explain the allegation that Meiji agricultural output increased with little additional capital, land or labor.