¹ This paragraph draws largely on T. C, Liu and K. C. Yeh, The Economy of the Chinese Mainland: National Income and Economic Development, 1933–1959 (Princeton: Princeton University Press, 1965); Dwight Perkins, “Growth and changing structure of China's twentieth century economy,” in Dwight Perkins (ed.), China's Modern Economy in Historical Perspective (Stanford: Stanford University Press, 1975); and Liang-lin Hsiao, China's Foreign Trade Statistics, 1864–1949 (Cambridge, MA: Harvard University Press, 1974).

² For current comparisons, see World Bank,World Bank Development Report1995(New York: Oxford University Press,1995).Google Scholar

³ Two books that helped precipitate the reassessment are Thomas Rawski, Economic Growth in Prewar China(Berkeley: University of California Press,1990), and Loren Brandt, Commercialization and Agricultural Development in Central and Eastern China: 1870s–1930s (New York: Cambridge University Press, 1989). These have been frequently reviewed along with Philip Huang's The Peasant Family and Rural Development in the Yangzi Delta, 1350–1988 (Stanford: Stanford University Press, 1990), and together were the focus of a series of papers that were subsequently published in Republican China, Vol. 18, No. 1 (November 1992), edited by Daniel Little. It remains the case, however, that the amount of work by economists on the Republican period is remarkably thin, and is far outpaced by the work of historians and politician scientistsGoogle Scholar

⁴ Rawski, Economic Growth in Prewar China, Table 6.11, suggests per capita growth of about 1 % over the two decades between 1914–18 and 1931–37. Between the 1890s and 1930s, a conservative estimate for the cumulative per capita growth would be between a quarter and a third.Google Scholar

⁵ Compare, for example, the estimates in J. Knight and L. Song, “The spatial contribution to income inequality in China,” Cambridge Journal of Economics, No. 17 (1993), and that

⁶ Between Asia and continental Europe, for example, the costs of shipping a ton of rice declined by two-thirds between 1870 and 1914, and by an additional half to two-thirds in the 1920s and 1930s relative to the pre-First World War rates. See E. A. V. Angier, Fifty Years Freight, 1869–1914 (London: Fairplay, 1920), and John McGee, “Ocean freight rateGoogle Scholar

⁷ Converting imports and exports into yuan from HKT, the total volume of trade in the late 1920s was 3.5 billion yuan. GNP in current prices was around 29 billion. The trade ratio declined significantly in the 1930s, a product of increased tariff protection, import substitution and falling export demand. Trade estimates are taken from Liang-lin Hsiao, China's Foreign Trade Statistics supplemented with data for Manchuria after 1931.

⁸ See Nicholas Lardy, China in the World Economy (Washington, D.C.: Institute for International Economics, 1994), pp. 2 and 18.

⁹ For an expression of such a view in the context of China, see Albert Feuerwerker, The Chinese Economy, ca. 1870–1911 (Ann Arbor: Center for Chinese Studies, 1968), p. 17Google Scholar

¹⁰ Joel Mokyr, The Lever of Riches (Oxford: Oxford University Press, 1990), p. 3.Google Scholar

¹¹ See, for example, Richard Nelson, National Innovation Systems(New York:Oxford University Press,1993).Google Scholar

¹² Liu and Yeh, The Economy of the Chinese MainlandGoogle Scholar

¹³ For Japan, see Takafusa, Nakamura, Economic Growth in Prewar Japan (New Haven: Yale University Press,1978). For India, I have not been able to find a breakdown between file traditional and modern sectors. However, between 1913 and 1937, modern manufacturing grew around 5% while all of manufacturing grew slightly more than 3%. Estimates of the growth in manufacturing output are taken from Morris, Morris D., “Growth of large-scale Industry to 1947,” and Alan Heston, “National income,” both in Vol.2 of The Cambridge Economic History of India (Cambridge:Cambridge University Press, 1983)Google Scholar

¹⁴ I emphasize quantitative because Rawski has as much to say about the qualitative nature of the growth process, especially the inter-sectoral linkages and externalities, as he does the quantitative. Reviewers have tended to focus on the latter, much to the neglect of theformer.

¹⁵ Foreigners were given the rights to establish manufacturing facilities without government consent in 1895. While foreign established firms were frequently the first to move into a new industry, by the 1920s and 1930s Chinese firms actually dominated most sectors

¹⁶ See, for example, Shannon Brown, “Cakes and oil: technology transfer and Chinese soybean processing, 1860–1895,” Comparative Studies in Society andHistory, Vol. 23, No. 3 (July 1979), pp. 550–568; and “The Ewo Filature: a study in the transfer of technology to China in the 19th century,” Technology and Culture, No. 20 (1979). More recently, business historians have begun to provide new insights into the establishment and dynamics of China's early 20th-century modern sector growth through case studies of Chinese and foreign firms. I should like to thank Sherman Cochran for allowing me to see his unpublished monograph “Adapting to the market: Western, Japanese, and Chinese businesses, 1880–1937.”

¹⁷ Dwight Perkins, “Government as an obstacle to economic development in China,” Journal of Economic History, Vol. 27, No. 4 (December 1967).

¹⁸ Dwight Perkins, Agricultural Development in China (Chicago: Aldine, 1969). As far. as I know, these are the only serious estimates. An estimate for the 1880s was made by Zhang Zhongli on the basis of Qing land figures and a combination of 19th and early 20th-century data, and later adjusted upwards by Feuerwerker. The odd mixture of data from several periods to derive the estimate undermines its credibility and comparability with other estimates. See Zhang Zhongli, The Chinese Gentry (Seattle: University of Washington Press, 1962).

¹⁹ These estimates are largely based on data collected by the Ministry of Agriculture and Commerce, which appears to have lacked the necessary manpower to obtain reasonably accurate and consistent estimates.

²⁰ The data problems are especially severe for the Yangzi region. Perkins' data imply, for example, that rice production, the most important grain crop in the region, declined by 11.5%, and in per capita terms by more than 20%. An increase in grain imports only partially offset this

²¹ Daofu Xu, Thongguo jindai nongye shengchan ji maoyi tongji ziliao (Statistical Materials on Agricultural Production and Trade in Modern Chinese) (Shanghai: Shanghai People Publishing, 1983).

²² Since the mid-1980s, however, growth rates in agriculture have slipped back to below 3%, and grain production has only increased at an annual rate of 1.3%. Between 1952 and 1978, agricultural output increased 3.2% per annum, but only 2.1% between 1957 and 1978. Over the same period population growth averaged about 2%.

²³ Robert Allen, “Agriculture during the Industrial Revolution,” in Roderick Floud and Donald McCloskey (eds.), Economic History of Britain since 1700, Vol. 1:1700–1860 (2nd ed.) (Cambridge: Cambridge University Press, 1994), Table 5.1, p. 102.

²⁴ K. Ohkawa and M. Shinohara, Patterns of Japanese Economic Development (New Haven: Yale University Press, 1979), especially Tables A16 and A53.

²⁵ Even these figures are slightly misleading. Some portion of the higher growth before the First World War can be attributed to a “one-time” gain in productivity that was associated with the re-integration of Japan during Meiji and the exploitation of previously untapped potential. Also, shortly after the turn of the century, Japan began to protect its agricultural sector and that in the colonies. By the 1920s and 1930s, tariffs on rice from outside the Empire were over 40%. Without the tariffs, growth rates in agriculture would have been significantly lower. See Loren Brandt, “Interwar Japanese agriculture: revisionist views on the impact of the colonial rice policy and the labor surplus hypothesis,” Explorations in Economic History, No. 30 (July 1993), pp. 259–293.

²⁶ Samuel Ho, Economic Development of Taiwan (New Haven: Yale University Press, 1978).

²⁷ Peter Schran, “China's demographic evolution, 1850–1953, reconsidered,” The China Quarterly, No. 75 (September 1978), pp. 638–646, raises doubt about Perkins' estimate of population growth, which are based on late Qing totals. Schran suggests an estimate of 0.5% as opposed to 0.9%.

²⁸ Potential difficulty with some of Buck's data is illustrated by Thomas Wiens' observation that using other data compiled by Buck as a deflator produces real wage behaviour sharply at odds with series Rawski or I present. There are several problems with using Buck's deflator here however. First, they include agricultural and non-agricultural goods, and for purposes of estimating productivity in agriculture, agricultural prices are needed. Secondly, Buck's index for north China is based on data for four localities, three of which are situated in fairly isolated areas of Qinghai, Gansu and Shansi. Thirdly, for all but one of the indices We do not know the weights used in the construction, and moreover, many of the goods would not be in the market basket of a typical farm household. Wiens' observations can be found in “Trends in Late Qing and Republican rural economy: reality or illusion,” Republican China, Vol. 18, No. 1 (November 1992).

²⁹ The difficulties arise because the introduction of the railway can lead to sharp “one-time” increases as well as decreases in prices through its effect on demand and supply. For example, in a grain deficit area grain prices may fall because of access to new supplies, while prices earned for cash crops may increase.

³⁰ See, for example, Li Wenzhi, Zhongguo zibenzhuyi mengya wenti taolunqi (Beijing: Sanlian, 1957), pp. 652–53; Kang Chao, Man and Land in Chinese History (Stanford: Stanford University Press, 1986); and “Tenure of land in China and the condition of the rural population,” Journal of the China Branch of the Royal Asiatic Society, No. 23 (1889), pp. 58–143

³¹ Interestingly, in the recent literature on the 19th century U.S. economy there is renewed debate over the urban-rural income gap, and the extent to which the gap is simply a product of measurement error. A comparison of urban and rural wages requires: wage series for comparably skilled workers in the two sectors; correctly valuing all in-kind payments; adjusting wages for differences in the cost of living; and constructing urban and rural deflators. The potential for systematic error is substantial.

³² The data are drawn from contracts covering households in 22 villages in Liaoning, Jihn and Heilongjiang. About half the villages are in Liaoning, and exhibit marked similarity with their north China counterparts. Nearly two-thirds of all contracts are between non-villagers. These contracts, which average over 6 months, cover nearly 85% of all labour hired, which is slightly higher than comparable estimates made for north China villages. See Loren Brandt and Arthur Hosios, “Credit, incentives and reputation,” Journal of Political Economy, Vol. 104, No. 6 (December 1996).

³³ Wages and worker age were positively correlated and increased up to the age of 40 before beginning to decline. The coefficients imply that the wage of a 40-year-old farmhand were almost double those of his counterpart 20 years his junior or senior. Wage premiums were also paid both to more highly skilled farm workers, and certain kinds of non-agricultural workers

³⁴ Differences in the interpretation of cloth consumption data highlight this point. It is generally agreed that cloth consumption increased significantly between the 1890s and 1930s, with estimates ranging between a third and a half. Is the increase related to increases in incomes? Xu Xinwu does not address this question explicitly, but attributes it to an increase in manufactured cloth output and to the fact that manufactured cloth was less durable than handicraft versions. Ignoring problems with Xu's estimate of the percentage of cloth that was manufactured, if durability was the only difference between the two type - and I am not suggesting that it is - it should be reflected in lower prices for manufactured cloth. An economic explanation would then attribute the increase in consumption to a falling relative price of manufactured cloth and substitution between the two. No increase in incomes is necessary. An examination of price data for a variety of cloth do not support the underlying claim. An alternative explanation is that the price effects were very small, and that the increase was the product of growing incomes. See Xu Xinwu, “Zhongguo tubu chanxiao guji, 1840–1936,” cited in Huang, The Peasant Family and Rural Development in the Yangzi Delta

³⁵ Much of traditional neoclassical economics is narrowly concerned with the efficiency properties of markets, and the static loss of output and welfare arising from market distortions. Static efficiency does not imply higher growth however. In fact, in traditional growth theory, the long-term steady state rate of growth is a function of exogenous technological change andpopulation growth. This explains the interest in new growth theory, and alternatives such as the new institutional economics, which endogenize the growth rate.

³⁶ Again, the data have their limitations, but there are three major factors pointing to an increase in the commercialization of agriculture: a small but important increase in the percentage of the population that was urban and non-agricultural; an increase in the percentage of total acreage in cash crops, which have higher marketing ratios; and as suggested by the experience of crops like cotton, an increase overtime in the percentage of cash crops that was marketed. A comparison of estimates for the 1930s, with an estimate that can be made using Wu Chengming's data for the 1850s also suggest the same thing. See Wu Chengming, “Woguo ban zhimindi ban fengjian guonei shichang,” Lishi yanjiu, No. 2 (1984).

³⁷ In major upper Changjiang rice markets, for example, provincial officials would frequently try to prevent the flow of rice downriver in times of rising prices. For the early 20th century, Kenneth Pomeranz argues that in parts of Shandong officials tried to block the flow of copper, and the arbitrage of local copper-silver price differentials. The difference between the rice market and the case Pomeranz analyses is that it was nearly impossible to implement a successful blockade of the Hunan rice markets. See Kenneth Pomeranz, The Making of a Hinterland: State, Society, and Economy in Interior North China, 1853–1937 (Berkeley: University of California Press, 1993).

³⁸ Empirical work suggests that markets are much more successful in their first capacity than the second, i.e. all households are not able to protect themselves perfectly from idiosyncratic exogenous shocks through the use of markets and non-market institutions. The source of these differences and their possible implications has yet to be explored.

³⁹ I use the term contract loosely here and allow it to include both implicit and explicit contracts, verbal as well as written.

⁴⁰ Philip Huang, The Peasant Economy and Social Change in North China (Stanford: Stanford University Press, 1985) and The Peasant Family and Rural Development in the Yangzi Delta.

⁴¹ Much earlier, Kang Chao, The Development of Cotton Textile Production in China (Cambridge, MA: Harvard University Press, 1977), made a related argument about development in the textile industry.

⁴² Despite access to export markets, domestic demand, much of it originating in the rural sector, remains extremely important. And in this regard, it does not matter if the growth of rural incomes is coming from an increase in working days or productivity per day. It all has the same effect on demand

⁴³ Thomas Smith, “Farm family by-employments in pre-industrial Japan,” Journal of Economic History, Vol. 29, No. 4 (December 1969).

⁴⁴ The most likely explanation concerns the monitoring costs of labour. In north and north-east China, for example, most larger farms were, in fact, extended families that typically had two or three adult males involved in farming operations. This helped to reduce some of the monitoring costs of hired hands, but there were limits. One of the advantages of the family farm is the “self-monitoring” of family labour. Traditional technology did not confer any obvious advantages on size, nor were there organizational advantages.

⁴⁵ See Dwayne Benjamin and Loren Brandt, “Markets, discrimination, and the economic contribution of women in China: historical evidence,” Economic Development and Cultural Change, Vol. 44, No. 1 (October 1995).

⁴⁶ From regression analysis, the following are the effects of an additional male or female on the household's net demand for land (measured by the amount of land the household cultivates). The effect of an additional male (or female) aged 13–17 was to increase the amount of land farmed by 4.92 (0.09) mu; for age 18–55 the corresponding figures were 10.42 (7.66) mu; and for age over 55, 8.98 (5.46) mu. These results are consistent with estimates of labour productivity.

⁴⁷ Dwayne Benjamin and Loren Brandt, “Land, factor markets, and income inequality in rural China: historical evidence,” Explorations in Economic History, forthcoming.

⁴⁸ These results also provide some empirical support for a heuristic argument offered in an earlier paper with Barbara Sands. We argued that with the impact of commercialization on inequality via its effect on asset distribution ambiguous, commercialization may have actually helped to reduce inequality because of its positive effect on wages and the levelling effect of wages on distribution, and increased factor market activity. While the links between late 19th and early 20th-century commercialization and wages remain open to debate, empirically it appears to be true that higher wages and expanded opportunities in land and labour markets are associated with lower inequality, ceterus paribus. See Loren Brandt and Barbara Sands, “Beyond Malthus and Ricardo: economic growth, land concentration, and income distribution in early twentieth century China,” Journal of Economic History, Vol. 50, No. 4 (December 1990).

⁴⁹ A major consequence of poorly functioning factor markets is that they may distort the direction and path of technological change. In an economy with poorly functioning labour markets and wages well in excess of market clearing levels, employers are motivated to move along a path of labour-saving, capital-using technological change, and to demand innovation in the same direction

⁵⁰ The logic here is very simple. Public goods have two key features: they are non-rivalrous, meaning that the marginal costs of providing them to an additional consumer are zero; and they are non-exclusive, meaning that people cannot be excluded from consuming them. Under these conditions, private agents will supply a level of the goods below that socially optimal.

⁵¹ Anthony Tang, “Research and education in Japanese economic development, 1880–1938,” Economic Studies Quarterly, Vol. 13, Nos. 2 and 3 (1963)

⁵² Contrast, for example, the view offered by Randall Stress in The Stubborn Earth (Berkeley: University of California Press, 1986), with T. H. Shen's more positive assessment in Paul Sih (ed.), The Strenuous Decade: China's Nation-Building Efforts, 1927–1937 (New York: St John's University Press, 1970).

⁵³ See, for example, Bruce Stone and Scott Rozelle, Foodcrop Production Variability in China (London: The School of Oriental and African Studies, Research Notes and Monograph Series, Vol. 9, 1995), and Jikun Huang and Scott Rozelle, “Technological change: rediscovering the engine of productivity growth in China,” Journal of Development Economics, No. 49 (1996).Google Scholar

⁵⁴ Lardy, Nicholas,Agriculture in China's Economic Development(Cambridge:Cambridge University Press,1983)CrossRefGoogle Scholar

⁵⁵ In Chinese history, there are numerous occasions in which such systems were financed using a combination of public and private funds. In Japan, a combination of public and private finance was also used, although public funding became more important later after the First World War. See, for example, Peter Perdue, Exhausting the Earth: State and Peasant in Hunan, 1500–1850 (Cambridge, MA: Harvard University Press, 1987). For data on public funding, see Yujiro Hayami and Saburo Yamada, The Agricultural Development of Japan: A Century's Perspective (Tokyo: University of Tokyo Press, 1991).