I am indebted to Herman Freudenberger who initiated my interest in writing this paper and made helpful suggestions, and to John G. Cummins and Donald L. Huddle for their useful comments and criticism.

¹ Ho, Samuel Pao-San, “Agricultural Transformation under Colonialism: The Case of Taiwan,” The Journal of Economic History (Sept. 1968), pp. 313–40.Google Scholar

² Ho, Yhi-Min, “Agricultural Development of Taiwan 1903–1960.” (Nashville, Tennessee: Vanderbilt University Press, 1966).Google Scholar

³ It should be noted at the outset that the disagreement comes largely from the input estimates. Although Samuel Ho's output index differs from the one I compiled in coverage and in weights, the two output indices are extremely close. This is hardly surprising, inasmuch as his gross output series is based on my estimates of intermediate products. See Yhi-Min Ho, Agricultural Development …, pp. 17–18.

⁴ Here I retain Samuel Ho's notations. The formulation here is similar to the one I used; see Yhi-Min Ho, Agricultural Development …, pp. 60–63.

⁵ Ho represented land input by crop area, which is defined as physical land area adjusted for multiple cropping. Thus, an index of multiple cropping is implicitly determined by the ratio of crop area to cultivated physical land area. His labor input is based on my estimates of the gainfully occupied population in agriculture multiplied by a multiple cropping index compiled by the Chinese-American Joint Commission on Rural Reconstruction (JCRR). The fact that the JCRR's multiple cropping index is constructed with the area planted for sugarcane, adjusted to an annual basis, whereas the crop area series is unadjusted was overlooked. Ho thus employed two multiple cropping indices instead of a single consistent index in deriving his aggregate input indices.

That JCRR derived its multiple cropping index from the adjusted crop area is clearly shown in Taiwan Agricultural Statistics 1901–1955 (Taiwan: Joint Commission on Rural Reconstruction, 1956), Table 5a. The adjustment is made on the basis that sugarcane had a growing season of eighteen months prior to the introduction of new varieties with a shorter growing season. The multiple cropping index compiled by JCRR differs very little from the one I compiled, which is also Based on the adjusted crop area for sugarcane.

⁶ See, for example, Hsieh, S. C. and Lee, T. H., Agricultural Development and Its Contributions to Economic Growth in Taiwan, Joint Commission on Rural Reconstruction Economic Digest Series, No. 17 (Taiwan: Joint Commission on Rural Reconstruction, 1966)Google Scholar; Shen, T. H., Agricultural Development on Taiwan Since World War II, (Ithaca, New York: Comstock Publishing Associates, 1964)Google Scholar; Yhi-Min Ho, Agricultural Development …, Ch. 8.

⁷ See Tang, Anthony M., “Research and Education in Japanese Agricultural Development, 1880–1938,” The Economic Studies Quarterly, XIII, No. 2 and No. 3 (Feb. and May 1963), pp. 27–41, and pp. 91–99.Google Scholar

⁸ The value of all irrigation work in Taiwan is assumed to be the accumulated amount of investment in irrigation since 1901. Data on annual investment in irrigation are drawn from Rada, E. L. and Lee, T. H., Irrigation Investment in Taiwan, Joint Commission on Rural Reconstruction Economic Digest Series, No. 15 (Taiwan: Joint Commission on Rural Reconstruction, 1963), Table B-1, Table B-2, pp. 134–136.Google Scholar The series is converted into constant terms by an index of agricultural prices in Y. M. Ho, Agricultural Development …. pp. 153–54. Other fixed capital assets in agriculture are compiled from data shown in Report on the 1956 Sample Census of Agriculture, published by the Committee of Sample Census of Agriculture, Taiwan, China, 1959.

⁹ It is realized that to measure fixed capital by animal energy alone leaves much to be desired. However, for the prewar period official data on farm houses, implements and equipment are so scattered ana incomplete that they cannot be organized in any useful way. On the other hand, to measure fixed capital by area irrigated ignores private fixed capital altogether.

¹⁰ For land input, see Statistical Summary of Taiwan for the Past 51 Years, Table 196; labor input, see Yhi-Min Ho, Agricultural Development…, pp. 43–44; working capital input, see Samuel Ho, “Agricultural Transformation …,” Table 2. I wish to point out that the author's criticism of the working capital series I compiled is ill-conceived. The series I estimated is represented by commercial fertilizers consumed on farms. During the period 1901–1921 when official statistics on fertilizer consumption are unavailable I assumed that “quantity of fertilizers consumed each year is the sum of quantity imported and the amount domestically produced.” (Yhi-Min Ho, Agricultural Development …p. 57). Thus the quotation, “the sum of quantity imported (from Japan) and the amount of domestically produced” cited in the article rests on a misunderstanding. Samuel Ho's working capital input is adopted here for the reason that farm-produced fertilizers are included, whereas the series I compiled covers only commercial fertilizers.

¹¹ I wish to emphasize that the rate of technical change estimated here is based on Samuel Ho's output data.

¹² Schultz, T. W., Transforming Traditional Agriculture (New Haven: Yale University Press, 1964), Chapter 12.Google Scholar

¹³ Yamada, Saburo, “Changes in Output and in Conventional and Non-conventional Inputs in Japanese Agriculture Since 1880,” Food Research Institute Studies, VII, No. 3 (1967), pp. 371–413.Google Scholar

¹⁴ The increase in the index of quantity sold was largely due to the sharp rise in the consumption of liquor and tobacco; salt remained unimportant and relatively stable, and opium declined rather sharply throughout the period.

¹⁵ The burden of taxes and monopoly levies as a percentage of net output would have certainly increased. However, I do not believe that the change here is large enough to invalidate the argument. Again, to borrow the Japanese experience, the net income ratio in agriculture declined from 85 to 80 percent between 1878 and 1910, while Japan's agriculture was sharply increasing its use of fertilizers and other modem farm inputs. The estimates of net income ratios in agriculture for Japan are shown in Ohkawa, Kazushi and others, The Growth Rate of the Japanese Economy Since 1878 (Tokyo, Japan: Kinokuniya Bookstore Co. Ltd., 1957), Table 10.Google Scholar

¹⁶ The 60 percent figure is inclusive of tenants as well as the so-called part-owner cultivators.

According to a 1921 farm survey, 35.8 percent of the farm land was owned by 2 percent of the large farm households with cultivated land greater than 10 hectares, only 14.4 percent of the land was owned by 64 percent of the small farm households with land less than one hectare per household. See Research Department, Bank of Taiwan, The Taiwanese Economy under Japanese Rule (Taiwan: Bank of Taiwan, 1957), p. 36.Google Scholar

¹⁷ The Taiwanese Economy under Japanese Rule, p. 37.

¹⁸ See Research Department, Bank of Taiwan, Taiwan under Japanese Imperialism (Taiwan: Bank of Taiwan, 1956), pp. 15–18Google Scholar; and The Taiwanese Economy under Japanese Rule, pp. 23–30 and 36–40.

¹⁹ Taiwan under Japanese Imperialism, pp. 116–20; and The Taiwanese Economy under Japanese Rule, pp. 35–39.

²⁰ Ponlai rice was successfully developed and adopted in 1922.

²¹ The cane variety known as Java POJ2725 was introduced in 1926. The average cane yields for the new cane variety were some 83 percent higher than the old variety known as POJ161. See Research Department, Bank of Taiwan, Sugar in Taiwan (Taiwan: Bank of Taiwan, 1949), pp. 1–23.

²² For sources of information see footnotes 19 and 20.

²³ The Taiwanese Economy under Japanese Rule, p. 36; and Taiwan under Japanese Imperialism, p. 120.