¹ The Chinese migrated to Taiwan in several waves. When the Dutch arrived there in 1624, they reported a Chinese population of 25,000. By 1650, the number was reported to have quadrupled. A second wave of immigrants came with Cheng Ch'eng-kung in 1622. When Cheng's administration ended in 1683, the Chinese population was between 200,000 and 350,000. The major wave of immigrants came during the Ch‘ing Dynasty, after Taiwan was made a prefecture (fu) of Fukien Province in 1683. In 1811, when the pao-chia system was introduced to Taiwan and each household was required to register its members, a Chinese population of 2,003,861 was recorded. By 1887, when it was elevated from a prefecture to a province, Taiwan had a population of 3,200,000.

² From 1700 to 1900, cultivated land increased from about 50,000 to 500,000 ha., and population increased from 350,000–400,000 to over 3,000,000. For a discussion of Taiwan's population and land data, see Samuel P. S. Ho, Economic Development of Taiwan, 1860–1970 (unpublished), Chapter II.

³ In 1858, the Treaty of Tientsin designated Anping as a treaty port. In i860, the Treaty of Peking opened Tamsui as a treaty port; three years later Keelung and Kaohsiung were added as supplementary ports to Tamsui and Anping respectively.

⁴ For a decade or more after the opening of Taiwan, its external trade appears to have increased but levelled soon after 1880. (The trade data are based on records of the Imperial Chinese aritime Customs. For a number of reasons, these data need to be used with great care. For example, they do not reflect the true total movement of goods in and out of Taiwan. Excluded are the junk trade and goods that go through non-treaty ports. Particularly for the early years, some of the increases shown by the custom data may be statistical rather than real, partly because of a probable once-for-all shift in the type of carrier used (from junks to steamers), and partly because of the more complete coverage of trade statistics over the years). Tea, the industry most directly stimulated by the opening of the treaty ports, got ofif to a flying start. From an obscure position in the 1860s, tea rose to become the leading Taiwanese export in the 1870s, accounting for 50 to 60 per cent of total export by value. But when the producers failed to recognize the importance of quality control, and the growers refused to adopt better cultivating techniques, the expansion stopped and tea exports levelled off in the 1880s to eight to nine million kg. a year, even though the world market for tea continued to expand rapidly.(Average annual production in Ceylon, for example, increased from one million pounds in the period 1880–84, t 0 4'9 million pounds from 1895–99)

⁵ There are, of course, some notable exceptions. For example, contrary to the other estimate of total product, Lee's estimate of NDP shows a sharp decline in real output between 1936 and 1941. Although agricultural production did decline be-tween 1936 and 1941, the decline in total product, if it occurred at all, was probably much less serious than Lee's estimates suggest. Most likely, Lee's estimate of total product for the late 1930s and early 1940s did not take adequate account of the many new industrial activities that were introduced during that period.

⁶ DGBAS estimated that in 1937 agriculture accounted for 39.7 per cent of Taiwan's GNP. See ROC, DGBAS, Taiwan's Gross National Product and Income, 1955, p. 129.

⁷ Simon Kuzncts has shown that the rate at which a sector's proportion in total product changes can be related to the ratio of its rate of growth to that of total output by the equation fj = ai(i -f- r,) — i, where r1 is the decadal rate of growth of the ith sector, rt is the decadal rate of growth of total output, and a1 is the ratio of the share of ith sector in total product at a given date to its share a decade earlier (c.f., Kuznets, Simon, Modern Ero-nomic Growth Rate Structure and Spread (New Haven: Yale University Press, 1966), pp. 97–8).Google Scholar For Taiwan, between 1911 and 1941, the decadal rate of growth of total product was about 42 per cent and that of the agricultural sector was about 33 per cent. This implies that for agriculture, a1 was about .94. Thus, if agriculture's share in total output was about 40 per cent in J940, it was about 50 per cent three decades earlier.

⁸ Household budget studies of Taiwanese rice farmers were conducted in 1920–21, 1931–32, 1936–37, and 1941–42. For an analysis of some of these surveys, see Myers, Ramon H., “Agrarian Policy and Agricultural Transformation: Mainland China and Taiwan, 1895–1945,” Journal of the Institute of Chinese Studies of The Chinese University of Hong Kong, III, 2 (1970), pp. 529–32.Google Scholar Estimates of real wage and other indicators of economic conditions can be found in Mizoguchi, T., “Consumer Prices and Real Wages in Taiwan and Korea Under Japanese Rule,” Hitotsubashi. Journal of Economics, 13, 2 (June 1972), pp. 47–51;Google ScholarSamuel, P. S. Ho, “Agricultural Transformation Under Colonialism: The Case of Taiwan,” Journal of Economic History, XXVIII (September 1968), pp. 333–7.Google Scholar and Ho, Samuel P. S., Economic Development of Taiwan 1860–1970, Chapter VI.Google Scholar Though far from conclusive, these findings generally suggest that real per capita personal income increased during the colonial period. The question of how the gains from development were shared is more difficult to answer. Income distribution data do not exist and much more research is needed before we can draw conclusions. What is clear is that any assessment of Taiwan's colonial experience is incomplete unless the distribution question as well as other economic and social consequences of colonialism are evaluated. I do not attempt to answer these more difficult questions in this paper.

⁹ The decadal rates of growth for the developed countries arc from Simon Kuznets, op. cit., pp. 64–5.

¹⁰ Marketable surplus is defined as the difference between total agricultural output and total consumption of agricultural goods by (he agricultural population.

¹¹ For a discussion of this aspect of Taiwan's agriculture, see Lee, T. H., Intersectoral Capital Flows in the Economic Development of Taiwan, 1895–1960 (Ithaca: Cornell University Press, 1971).Google Scholar

¹² From 1896 to 1904, before Taiwan became Japan's main source of surgar, Japan on the average imported from outside the Empire 22 million yen of sugar a year, an amount greater than one-half of Japan's-trade deficit during this period. Given the pressure on its balance of payments, the need to import substitute within the Empire was clear. However, Taiwan was never able to develop a strong comparative advantage in sugar, and it probably would have been more efficient if the Empire had continued to import sugar and not allocated as much resources to sugar production as it did.

¹³ Taiwan Provincial Government, Bureau of Accounting and Statistics, Taiwan Province; Statistical Summary of the Past 51 Years, 1946, p. 802.

¹⁴ Taiwan Government-General, Bureau of Colonial Development, Taiwan Commercial and Industrial Statistics, 1932, pp. 84–5, 1936, pp. 122–3; 1941, PP. 6–7.Google Scholar

¹⁵ Factory is any enterprise that employs five or more workers or uses a prime mover. The data are from Taiwan Provincial Government, Bureau of Accounting and Statistics, op. cit., pp. 763–6 and Taiwan Government-General, Bureau of Colonial Development, Collection of Relevant Materials Based on Recent Surveys of Resources, 1933, pp. 22–9.

¹⁶ Taiwan Provincial Government, Bureau of Accounting and Statistics, op. cit., p. 1067.

¹⁷ In 1930, for example, 80 per cent of the gainfully occupied male Japanese were in five lines of economic activities: manufacturing, communica-tion and transport, trade, government, and professional service.

¹⁸Barclay, George, Colonial Development and Population in Taiwan (Princeton: Princeton University Press, 1954), Chapter VI.Google Scholar

¹⁹ Ibid, p. 154.

²⁰ Taiwan Provincial Government, Bureau of Accounting and Statistics, op. cit., pp. 1241–2.

²¹ George Barclay, op. cit., p. 60.

²² During 1880–94, the volume of Taiwan's exports fluctuated but showed no steady upward trend. An export volume index for the period 1868–94 can be found in Ho, Samuel P. S., The Economic Development of Taiwan, 1860–1970, op. cit., Statistical Appendix Table VII-17.Google Scholar

²³ Ibid, Statistical Appendix Table VII-17.

²⁴ The relationship estimated was In X_DtR = b^o + b₁ In P_xt + b₂ In Y_tR + + b₃ In P_dt + b₄D_Dt, where X_DtR is the demand for Taiwan's real commodity export, P_Xt is Taiwan's export price index, Y_tR is Japan's real GNP, P_dt is Japan's GNP price deflector, and D_Dt is a dummy variable to account for the effects of World War I, See 'Birnberg, T. and Resnick, S., “A Model of die Trade and Government Sectors in Colonial Economies,” The American Economic Review, LXIII (September 1973), pp. 572–87. Similar results have been reported by Mizoguchi, T.. See his “Foreign Trade in Taiwan and Korea Under Japanese Rule,” Hitotsubashi Journal of Economics, 14, 1 (1974).Google Scholar

²⁵ lbid.

²⁶ However, this was not achieved without cost to the Japanese agriculture. See Hayami, Yujiro and Ruttan, V. W., “Korean Rice, Taiwan Rice, and Japanese Agricultural Stagnation: An Economic Consequence of Colonialism,” The Quarterly Journal of Economics, LXXXIV (November 1970), pp. 562–89.CrossRefGoogle Scholar

²⁷ Java, Japan's most important outside supplier of sugar, definitely had an absolute advantage in the production of sugar. In the 1920s, the yield of sugar cane in Java was about five times that in Taiwan. By the 1930s, Taiwan had managed to reduce this gap, but its yield was still only one-half that of Java's. Primarily because of this, the cost of producing sugar was consistently higher in Taiwan than in Java. On the other hand, given its more sophisticated production techniques and greater investment in irrigation, Taiwan probably had an absolute advantage in rice production. Assuming a two-crop world, a realistic assumption for Taiwan and Java, these absolute cost differences would imply that Taiwan's comparative advantage was in rice rather than in sugar.

²⁸ For example, in 1935, the c.i.f. unit value of sugar imported from Java was 5.41 yen per 60 kg. and the unit production cost of sugar in modern Taiwanese sugar refineries was 6.24 yen per 60 kg.

²⁹ The tariff rates on centrifugal sugar not produced in the Japanese Empire ranged from five to 10 yen per 100 kg., the exact rate depending on the fineness of the sugar. For a more detailed discussion of Japan's sugar tariff structure see, Yanaihara, Tadao, Taiwan Under Japanese Imperialism, translated from Japanese to Chinese by Chou, Hsien-wen (Taipei: Bank of Taiwan, 1956), pp. 127–9.Google Scholar

³⁰ Taiwan Government-General, Bureau of Finance, The Annual Taiwan Trade Statistics, various issues.

³¹Kamekichi, Takahashi, Taiwan Economy in Modern Times (n.a.: Chikewa, 1937).Google Scholar

³²Huang, T., Chang, T. H., and Lee, C. C., Government Financing in Taiwan Under Japanese Regime (Taipei: JCRR, 1951), Table 20. More than half of the receipts from these bond sales were used to construct the railroad network in Taiwan. The remaining portion went to finance the construction of harbors and government buildings.Google Scholar

³³Davidson, James W., The Island of Formosa Past and Present (London: Macmillan & Co., 1903), P. 453.Google Scholar

³⁴ T. H. Lee has estimated Taiwan's net domestic product in 1911 to be about 150 million yen.

³⁵ Calculated from data taken from Taiwan Provincial Government, Bureau of Accounting and Statistics, op. cit., pp. 322–3.

³⁶Calculated from unpublished data from the 1940 population census, see Barclay, George W., A Report on Taiwan's Population (Princeton: Princeton University, Office of Population Research, 1954). Table 19.Google Scholar

³⁷For a discussion of the government's educa-tion policy, see Ho, Samuel P. S., “The Development Policy of the Japanese Colonial Government in Taiwan, 1895–1945,” in Ranis, Gustav (ed.), Government and Economic Development (New Haven: Yale University Press, 1971), PP.308–12,Google Scholar and Tsurumi, E. Patricia, Japanese Colonial Education in Taiwan: 1895–1045 (unpublished Ph.D. thesis, Harvard University, 1971).Google Scholar

³⁸ Recently Mark Elvin has described this state as the high-level equilibrium trap in agriculture (c.f., his The Pattern of the Chinese Past (London: Eyre Methuen, 1973). PP. 198–315) For a longer period, economists have described this condition as the low-level equilibrium trap.Google Scholar

³⁹ Agricultural development during the colonial period has been discussed extensively elsewhere. See Samuel P. S. Ho, “Agricultural Transformation Under Colonialism: The Case of Taiwan,” op. cit.; Myers, Ramon H. and Ching, Adrienne, “Agri-cultural Development in Taiwan Under Japanese Colonial Rule,” Journal of Asian Studies, XXIII (August 1964),CrossRefGoogle Scholar and Ho, Y. M., Agricultural Development of Taiwan 1903–1060 (Nashville, Tenn.: Vanderbilt University Press, 1966).Google Scholar

⁴⁰ See, for example, the cost survey of Ponlai and Chailai rice production conducted in 1926–27 and published as Taiwan Government-General, Bureau of Colonial Development, Economic Survey of Major Agriculture Products, No. 6 and No. 9.

⁴¹Ho, Samuel P. S., Economic Development of Taiwan, 1860–1970, Chapter IV.Google Scholar

⁴²Ma, Frengchow C., Takasaka, T., and Yang, Ching-wen, A Preliminary Study of Farm Implements Used in Taiwan Province (Taipei: JCRR, 1955). P. 30.Google Scholar

⁴³ Ibid, p. 31.

⁴⁴ For multiple cropping to be successful, harvesting of the first crop and planting of the second must be done within a prescribed number of days. Even in a labor-abundant society, there tends to be a shortage of labor during this peak demand period. Any Innovation that permits the substitution of capital for labor during this period would tend to relax the labor constraint and allow an expansion of multiple cropping.

⁴⁵Ho, Samuel P. S., “Agricultural Transformation Under Colonialism: Reply and Further Observations.” Journal of Economic History, XXXI (September 1971), pp. 685–7.Google Scholar

⁴⁶Wright, Mary C., The Last Stand of Chinese Conservatism (New York: Atheneum, 1966), pp. 148–149.Google Scholar

⁴⁷ From his only major policy speech as Taiwan's governor-general, delivered in 1900 at a conference of ranking officials of the colonial government. For this and other excerpts from his speech, see Chang, Han-yu and Ramon, Myers, “Japanese Colonial Development Policy, 1895–1906: A Case of Bureaucratic Entrepreneurship,” Journal of Asian Studies, XXII (August 1963), p. 436.Google Scholar

⁴⁸An interesting model setting out some of these relationships formally is presented in Hymer, S. and Resnick, S., “Interactions Between the Government and the Private Sector: An Analysis of Government Expenditure Policy and the Reflection Ratio,” in Stewart, I. G. (ed.), Economic Development and Structural Change (Edinburgh: Edinburgh University Press, 1969), pp. 150–80.Google Scholar

⁴⁹Tsurumi, Patricia, “Taiwan Under Kodama Gentarō and Gotō Shimpei,” Harvard East Asian Research Center Paper on Japan, Vol. 4, September 1967, p. 101.Google Scholar

⁵⁰Chang, and Myers, , op. cit., pp. 446–48 and Tsurumi, , “Taiwan Under Kodama Gentarō and Goto Shimpei,” op. cit., pp. 122–4Google Scholar and 128–30.

⁵¹ The major government enterprises were in ihe field of transport and communications.

⁵²Ho, Samuel P. S., “The Developmental Policy of the Japanese Colonial Government in Taiwan, 1895–1945,” op. cit.Google Scholar

⁵³ The pao-chia system, a mutual responsibility scheme, is discussed extensively in Chen, Ching-Chih, “The Police and Hokō Systems in Taiwan Under Japanese Administration,” Harvard East Asian Research Center—Papers on Japan, September 1967.Google Scholar Also sec his Japanese Socio-political Control in Taiwan 1805–1045 (unpublished Ph.D. thesis, Harvard University, 1973).Google Scholar

⁵⁴For some recent data on government expenditure shares, see Gupta, S. P., “Public Expenditure and Economic Development—A Cross-Section Analysis,” Finanzarchiv (October 1968), pp. 40–1.Google Scholar Also see his “Public Expenditure and Economic Growth: A Time-Series Analysis,” Public Finance XXII, 4 (1967), pp. 423–61.Google Scholar

⁵⁵ The relationship estimated was In X_StR = a_o + a_l In P_Xt + as In P_Mt + a₃ In ARGE + a₄ InX_{St – 1} R + a₅ D_St, where X_sR is the supply of Taiwan's total real commodity exports, P_Xt is Taiwan's export price index, P_Mt, is Taiwan's import price index, ARGE is lagged accumulated real government expenditures, and D_S is a dummy variable to account for the effects of World War I. See T. Birnberg and S. Resnick, op. cit.

⁵⁶ This tax structure may, however, be undesirable on equity grounds. Furthermore, not all of the increased savings remained in Taiwan; much of it was transferred to Japan as profits and in the form of capital flows.