Introduction
There has been considerable debate about levels and trends in the standard of living in China during the Qing (1644—1911) dynasty. The consensus used to be that living standards were lower in China than in Europe.
Influenced by Malthus' portrait of China during the late eighteenth century as a land of misery and poverty where the desire to maintain high fertility triggered the incessant operation of the positive check, a long line of observers argued that the nineteenth century was a time of rising population pressure in China and stagnant or declining living standards.1 Scholars who have detected evidence of rising mortality during the nineteenth century in data from lineage genealogies from selected regions of China have attributed such trends to increasing population pressure and worsening conditions during the late Qing period (Liu 1985; Harrell 19950.Recent scholarship has called into question this received wisdom. Based on estimates of per capita production and consumption, Kenneth Pomeranz argues that living standards in China were probably comparable to those in Europe at least until the middle of the eighteenth century, though they may have declined during the nineteenth century (Pomeranz 2000; Chapter 1, this volume). Similarly, Bozhong Li has argued that at least in some parts of China, per capita production and consumption were stable or rose until the middle of the nineteenth century (Li 1998; Chapter 2, this volume). James Lee and Wang Feng, meanwhile, have argued that the Malthusian understanding of the Chinese demographic system was fundamentally incorrect, especially when it came to the relationship between living standards and demographic behaviour (Lee and Wang 1999). They suggest that the preventive check played a more important role in China than Malthus and his intellectual heirs realized because couples deliberately adjusted their fertility behaviour according to their economic circumstances.2 From their perspective, the sustained rise in China's population during the Qing period was a response to improving
living standards, not an unfortunate side-effect of couples' desire to maintain high fertility at any cost.
In this chapter, we assess changes in living standards of rural residents of Liaoning in northeast China during the eighteenth and nineteenth centuries. Rather than estimate per capita production or consumption, we use demographic rates as indices of the standard of living. We focus on trends in demographic rates and the sensitivity of rates to shortterm economic fluctuations. In international comparisons, high death rates are commonly treated as evidence of low living standards, and rising death rates as evidence of declining standards.3 For historical China, levels of male marriage should also have reflected the standard of living. One of the most important determinants of the chances that a man would marry was the supply of females, which in turn depended on the prevalence of female infanticide. This, of course, was affected by economic conditions (Lee and Campbell 1997). Marital fertility should also have been sensitive to the standard of living, since couples calibrated their numbers of surviving children according to their economic circumstances (Lee and Wang 1999).
In drawing inferences about living standards from the sensitivity of demographic rates to economic conditions, we make use of a concept of the standard of living advanced in Bengtsson, Campbell, Lee et al. (2004) and applied by participants in the Eurasia Project. Building on the literature on associations between real incomes and demographic rates (Lee 1981; Bengtsson and Ohlsson 1985; Galloway 1988; Lee 1990; Bengtsson 1993), Bengtsson, Campbell, Lee et al. (2004) argue that comparisons of demographic responses to short-term economic stress by region, period, household composition, individual context, and socio-economic status yield insight into differences between and within populations in the standard of living. The sensitivity of demographic rates to economic conditions reflect the ability, or inability, of households and their members to maintain minimum consumption levels during times of economic hardship.
A response of demographic rates to short-term economic stress, in other words, reflects a failure to smooth consumption, and indicates a low standard of living.From this framework, we specify how we will interpret possible outcomes. If living standards declined during the nineteenth century, we expect it to be reflected in rising rates of mortality, falling rates of fertility and male marriage, and increases in the sensitivity of demographic rates to short-term economic stress, as reflected in the prices of key grains. Ifliving standards improved, we expect the reverse. Mortality should have fallen, and fertility and male marriage rates should have risen, and all rates should have become less sensitive to economic conditions. If living standards were stable, rates and their sensitivity to economic conditions should not have changed over time.
Our primary substantive interest is in the response of demographic behaviour and living standards to the rise in population density in Liaoning during the eighteenth and nineteenth centuries. For some time after the founding of the Qing dynasty in 1644, northeast China was a sparsely populated frontier region. The state actively encouraged migration into the area during the seventeenth and eighteenth centuries.
A previous analysis of a smaller sample revealed that by the middle of the nineteenth century, the densely settled areas closest to what is now Shenyang, the provincial capital and at the time the prefectural capital, showed signs of rising population pressure on land (Lee and Campbell 1997). In the analysis, we will examine whether the rise in population density was accompanied by signs of decreasing living standards, including rising mortality, delayed marriage, reduced fertility, and increased sensitivity to short-term economic stress.
We also examine whether and how commercialization in southern Liaoning during the nineteenth century affected demographic behaviour and living standards there.
A substantial portion of the population for which we have data lived near the coast of the Bohai Gulf on the Liaodong peninsula in southern Liaoning. Many of the villages were in the hinterland of Yingkou, a port that was heavily involved in coastal trade in the early nineteenth century, and became a treaty port involved in international trade in 1858. Customs records suggest that in the last decades of the nineteenth century, the volume of maritime trade, both domestic and international, through Yingkou was enormous. We will investigate whether or not commercialization affected living standards in southern Liaoning by comparing its demographic rates and their sensitivity to short-term economic stress to those of other Liaoning regions. We will also assess the impact of the opening of Yingkou as a treaty port by making comparisons between the first and second halves of the nineteenth century for southern Liaoning.This study's focus on Liaoning distinguishes it from other attempts to analyse living standards in China before 1949. Most studies of levels and trends in living standards, productivity, consumption, and related issues in historical China have focused heavily on the Jiangnan region (Perkins 1969; Elvin 1973; Huang 1990; Li 1998; Chapter 2, this volume; Pomeranz 2000; Chapter 1, this volume). TheJiangnan region was one of the wealthiest and most densely populated areas in China, and accordingly very different from the rest of China. While there have also been some relevant studies of the economic history of north China (Huang 1985; Pomeranz 1993) and the southeast (Faure 1989), there have been very few analyses of the northeast. Levels and trends in regions other than Jiangnan, of course, merit far more attention than they have received because they accounted for an increasing share of China's population through the Qing period. Through empirical study of regions like the northeast, it will be possible to evaluate the suggestion by Pomeranz (2000) that living standards for China as a whole may have fallen during the Qing period not because they fell in any one region, but rather because a rising share of the population lived in regions where living standards were lower than in Jiangnan.
This study is also distinguished by its coverage on a continuous basis of the period from the late eighteenth century to the beginning of the twentieth century. Most other assessments of levels and trends in productivity, living standards, and related issues in China have focused either on the end of the nineteenth century or the beginning of the twentieth century (Brandt 1989; Faure 1989; Rawski 1989; Pomeranz 1993), or else on the eighteenth century (Li 1998; Chapter 2, this volume; Pomeranz 2000; Chapter 1, this volume). Direct measurements of trends from the
end of the eighteenth century to the beginning of the twentieth remain very rare, so that inferences about changes in conditions in China during the nineteenth century, for example, the one by Pomeranz in this volume, have been based on the interpolation of estimates from the late eighteenth and early twentieth centuries.4
We organize the remainder of this chapter into four parts. First, we provide background on the population we study and introduce the longitudinal, nominative household register data that describe it. We focus discussion on the strengths and limitations of the household registers as a source for the study of mortality, fertility, and nuptiality. We also discuss the grain price data we use to analyse the sensitivity of rates to economic conditions in the short term. Second, we describe the event-history methods we use to measure trends in demographic rates and the sensitivity of rates to economic conditions. We introduce the regression methods for limited dependent variables that we use and summarize the right-hand side variables in our model. Third, we present descriptive results on time trends in rates as well as regression results on secular trends in rates and their sensitivity to economic conditions. We conclude with some remarks on the implications of our results for our understanding of late Imperial China.
2.