In their history of Chinese capitalism from 1550 to 1840, Xu Dixin and Wu Chengming (1985) took the labour-intensive nature of much Chinese industry as a sign of backwardness. They argued that the prevalence of rural industry with workers whose families also farmed indicated an absence of true capitalist industrialization, which would involve proletarianization, urbanization, and higher capital-labour ratios. In so arguing, Xu and Wu spoke both for a near consensus among historians in China at the time and for a view of the present in which China’s large urban factories, not its small-scale rural industries, held the key to its industrial future. Western modernization theorists shared these views, albeit for different reasons. Today, as China’s industrial output approaches that of the United States, it still counts only 50 per cent of its population as urban, while Britain was 48 per cent urban by 1840. Moreover, most Chinese industry remains very labourintensive. This chapter argues that the labour-intensive rural industries of late imperial China were part of a characteristically East Asian pattern of growth, with legacies that matter to the industrial boom in contemporary coastal China. However, those patterns have recently shown signs of converging towards patterns of industrialization closer to many Western historical experiences: this change recalls transitions which occurred roughly 20-30 years ago in Taiwan and 50-60 years ago in Japan. Distinguishing a dynamic pattern of labour-intensive growth from one in which labour intensity is simply a sign of failure to accumulate capital and improve technology requires identifying a different kind of technological and industrial development: one in which technological change tended to be labourabsorbing, not only because labour was abundant, but because much technological change took the form of creating new products which required producers to acquire new skills. These skills, in turn, could allow workers to earn more even without improving capital-labour ratios. We are just starting to explore these different kinds of labour-intensity and their inter-relationships. The discussion below will focus primarily on Jiangnan, which has been China’s richest area for most of the last millennium. Usually called the Yangzi Delta in English, Jiangnan is actually the part of the Delta south of the river. By about 1620, it had over 20 million people; by about 1770, perhaps 32 million. It

then grew very little for several decades, lost millions of people during the Taiping Rebellion (1851-64) and just barely exceeded its 1770 population in the first census under the People’s Republic (1953). Its population has since grown considerably, though much slower than the national average (Lee and Wang 1999: 115-17; Li 1998: 20; Skinner 1977a: 213; Ho 1995). Today, scholars more often speak of a Sunan (Southern Jiangsu) region: this includes adjacent areas immediately to the West of Jiangnan that have always shared many of its characteristics, while it omits the parts of historic Jiangnan in Zhejiang province. I will refer to ‘Jiangnan’ or the ‘Yangzi Delta’ except when citing recent material that specifically refers to Sunan or some other unit. Doing otherwise would not materially alter our story. Jiangnan’s high-yielding agriculture and excellent water transportation encouraged local industries, which exported relatively high value per bulk finished goods (along the coast, up the Yangzi and along the Grand Canal) and imported bulkier products (mostly food and raw materials). But its geography was bad for industries requiring lots of energy, either mechanical or thermal. The Delta had little forest, and no coal; there was not even much coal that could be shipped to it at reasonable prices. It lacked most other minerals too (Pomeranz 2000a: 63-4, 225-6; Li 2000: 272-342). The area is mostly flat and its abundant water moves slowly; while waterwheels and related devices were widely used for irrigation, water power for industry was limited. Wind was available seasonally, and early modern Jiangnan had some of the world’s most sophisticated wind-driven pumps (Elvin 1973: 127-8; Li 2000: 275); but the alternation of very high winds (which could break pumps) with no wind at all made year-round use of wind difficult. Meanwhile, dense population, paddy-rice agriculture (which required much less animal traction than dry farming on heavy soils), and waterways that reduced the need for land transport limited demand for work animals. Commercial oil presses and flour mills in some Delta towns purchased livestock (mostly from Jiangxi) for motive power, indicating that where there were sufficient returns to scale, people would invest in increasing energy intensity. But large animals competed for grass, straw and so on with the area’s pigs and poultry (which convert starch to protein more efficiently) and with people who used straw for fuel; this made them uneconomical for many uses (Li 2000: 273-80; compare Bray 1986: 114-16, 198-202).1 Thus, Jiangnan specialized in light industry, with mostly small production units. This is unsurprising with respect to the area’s famous luxury products, but it is more noteworthy that its more commonplace goods, such as bulk textiles and processed foods (alcohol, soy sauce, pickled vegetables and so on) were also usually made by small-scale producers. Cloth production, by far the largest sector, probably became more dispersed with time. It also became more rural, and its workforce more female – though dyeing and finishing were often concentrated in towns, and urban merchants handled most cloth exports.2 Jiangnan’s mixture of agriculture, commerce and industry probably kept its living standards close to Britain’s until at least 1750, or even 1800 (e.g. Wong 1997; Li 2000; Pomeranz 2000a; 2002, 2003; Ma 2004; Allen 2005; Van Zanden 2004: 21-2).