‘Non-essential’ Consumption

If most Chinese ate fairly well by pre-industrial standards, and even the poor had some money left over, what did they spend it on? Unfortunately, we have nothing for China that is comparable to the inventories of possessions held at death for many places in Europe, but we do have other information.

I have also taken steps towards measuring some part of this consumption quantitatively Usually this had to be done by working backwards from estimates of the amount of land under various crops, multiplying by contemporary yield estimates, and subtracting exports where they are relevant. This introduces various uncertainties, but I have also taken various steps to ensure that these estimates are conservative.³⁸ Even so, they are quite surprising. It may be no shock to see that per capita tea and silk consumption were higher in China than in Europe, but consider the following data for sugar and for ordinary cloth (Tables 1.1 and Tables 1.2).

Despite the uncertainties that surround these numbers, they are quite instructive: perhaps especially those for cloth. They suggest that in Jiangnan, the richest part of China, cloth production per capita in 1750 was comparable to that in England fifty years later. We cannot easily move to an estimate of cloth consumption, since much of this cloth was shipped to other parts of China; but England, of course, also

Table 1.1 Sugar and tea consumption in Europe and China (in pounds per capita)

‘ Includes England.

^b Consumption per capita for country as a whole with heavy concentration to the Lower Yangzi, southeast coast, and Lingnan, where consumption may have been as high as 10 pounds per capita.

Sources. Production figures from Philipps (1990: 58—61), for Portuguese and Spanish colonies, Steensgaard (1990: 140), for French, Dutch, and English colonies, Braudel (1981: 251—2), Gardella (1994: 6, 38), Xu and Wu (1985: 99). European population figures from McEvedy and Jones (1978: 28), British consumption figures from Mintz (1985: 67, 73), substituting year 1700 figures for 1680 in this table. For Chinese calculations, see Pomeranz (2003: ch. 3).

Table 1.2 Selected comparisons of cloth output and consumption (in pounds per capita)

‘ Population app. 31 million in 1750, omitting salt-producing prefectures from the large definition in Wang (1989).

^b Amount consumed locally unknown. For an adjustment of Yangzi Delta textile output downward from the figures in Pomeranz (2000«) see also Pomeranz (2002).

^c Probably nearer lower end of range.

^d Ramie was not included in estimates of cloth in Pomeranz (2000a). It does appear, however, that it was still quite significant though it is usually agreed that it was steadily losing ground to cotton. In the early twentieth century, production was still about 1.5 billion pounds, or roughly 3.5 pounds per capita. Whether per capita was higher than that was 150 years earlier (either because of a smaller population or because many people had not yet fully switched to cotton) or lower (because the highland areas in which much ramie was raised were not yet as fully exploited) remains conjectural.

^e Combination of cotton, linen, and wool.

^f About 8.7 pounds per capita consumed within the United Kingdom.

Sources. See Pomeranz (2000a: appendix F) and idem for discussion of data problems.

exported a large portion of the cloth it produced. For China as a whole, it suggests per capita cloth consumption that was probably lower than that of France, but comparable to that of Germany—which would presumably be a good deal above that of Europe as a whole. This is suggestive not only because clothing is an important category of expenditure, but because, according to Engel's Law, it is one

of the first areas in which one would expect expenditures to increase once basic food needs have been met. Thus, while it is no substitute for an overall ‘market basket' of goods, it is a good preliminary indicator of comparative living standards for the general populations of China and Europe.

It is also interesting that these numbers are, once again, higher than those derived in various estimates for early twentieth-century China. I have explained at length elsewhere why I believe that China's total cotton output quite likely declined between 1750 and 1900 or, at best, did not increase much, while population roughly doubled (Pomeranz 2000a: 139—42, 334—8). If that assertion is correct, then there is no conflict between my figures and the later ones, either for cotton, or for sugar and tea.

First, though, a few more things should be said about eighteenth-century consumption levels both of necessities and non-necessities. Per capita fuel supply, as I have argued elsewhere, was probably as adequate as it was in most of Europe in the eighteenth century, despite an increasingly unfavourable ratio of people to wooded acreage in both places (Pomeranz 2000a: 222—36). That China, in particular, was able to sustain adequate supplies of fuel was in part due to various labour-intensive expedients that were much less necessary in Europe—in particular, the burning of crop residues and the gathering of small wood—as well as to a different kind of cooking, more efficient stoves, and, in the case of south China, a generally warmer climate. The extra labour probably fell mostly on women and children. With cooking fires going for less time each day, Chinese homes may have been, on average, less warm, but also less smoky. And the general pattern of fuel gathering—relying more on small groves, often within the family's own courtyard, and less on consolidated blocks of forest owned by nobles—may have made for more even distribution and greater security of fuel supply, and less conflict over it, even if average levels were no better. At present, though, there is little chance of quantifying anything beyond those average levels—and even that is speculative.

Housing was one area in which Chinese living standards may well have been falling behind European ones by the late eighteenth century. In part, this would have been the result of the increasing pressure on wood supply mentioned above, and the fact that China did not replace wood with stone in residential construction on anything like the scale of at least England. Some other housing amenities that were becoming more common in Europe, such as glass windows, remained quite rare in China. We should be careful not to paint these as universal patterns—recent research suggest that England's ‘Great Rebuilding' reached only a minority of the population, and that a majority of the eighteenth­century poor still lived in one to three rooms, made largely of wood and clay; brick, even for the chimney, appears to have been quite exceptional among at least the rural poor (Shammas 1990: 159—65).

neighbours (Daunton 1990: 203). For the most part, housing in continental Europe was worse. The first quantitative figures we have for Chinese housing quality comes from J. L. Buck's early twentieth-century rural surveys. While they far exceed what is cited for the seventeenth-century English rural poor in floor space per capita, and 39% of buildings even on ‘small farms' featured tile roofs,³⁹ the comparison is hardly a well-controlled one. And though I have pointed out elsewhere that Buck's figures on average furniture ownership throughout China seem to compare quite well with those for two seventeenth-century samples for the rural Netherlands (a much richer-than-average part of Europe) examined by Jan de Vries, there are many problems with comparing those figures. The best guess, I would say, is that this is an area in which the majority of Chinese may well have been keeping up with their European poor, or even continuing to outdo them, but in which a minority of better-off western Europeans may well have been improving their circumstances faster than their Chinese counterparts.

Consumption, of course, includes services as well as goods, and here comparisons become even more difficult. Mid- Qing literati and officials wrote an enormous amount (usually with a censorious tone) about the growing numbers of ordinary people frequenting tea houses, travelling (especially, but not exclusively, for pilgrimages), hiring religious specialists to perform rituals, patronizing travelling entertainers of various sorts, and so on. Few used numbers that are useful to us, but we know that some of the most popular pilgrimage sites for instance, welcomed as many as 1,000,000 people per year (Pomeranz 1997: 188). In the Yangzi Delta, Fan Shuzhi has identified a number of towns, each with only a few thousand people that had anywhere from 40 to 100 recorded tea houses in the early nineteenth century (Fan 1990: 279). For so many of these establishments to survive, a significant number of people from the surrounding villages must also have been partaking of the food, entertainment, and gambling that they offered. The boom in the number of itinerant religious specialists in the eighteenth century—which some have taken as a sign that increasing numbers of people were economically desperate⁴⁰—can just as well be read as a sign that the society in general was able to support an increased number of such people. (They received no support from any religious establishment, and survived by being paid for services and/or begging.)

Of course, people everywhere spend some of their incomes on rituals, celebrations, and entertainment. Early modern European texts are also full of complaints about the ‘wasteful' activities of the poor, and their frequency and stridency may say more about elite attitudes than about the numbers of people or levels of spending involved. Still, there may be some grounds for speculating that such activities would have comprised a larger share of plebeian consumption in eighteenth-century China than Europe. Looking at the comparison from one side, the relatively lesser importance and complexity of kinship in Europe (speaking in gross, general terms) may well have contributed to a stronger need to mark statuses by consumption of goods, and a greater penetration of fashion (which, among other things, raises the share of goods in the budget by making them socially obsolete before they are

physically so).⁴’ That, in turn, would tend to direct spending away from services. At the same time, the more complex and prominent kinship relations in which Chinese were embedded did not simply reproduce themselves—people were taught whom they had important ties with, and how to act towards them, through ritual convocations in which one incorporated as many of the people with whom one could claim such ties as one could afford.⁴² And since neither a state nor an established church registered marriages in Qing China, the burden fell on each family to mount an event capable of showing the relevant neighbours that what they were entering into was a legitimate marriage, rather than the purchase of a concubine or an illicit cohabitation. Various social and cultural historians have made the point that the combination of rapid commercialization and the erasure of almost all formal legal status distinctions by ’730—including those which had isolated entertainers and commercial sex workers from ordinary commoners-----------------------------------------------------------------------------------------------------------------

greatly increased pressure at all social levels to perform marriages in a way that clearly showed that the bride's parents were not ‘selling’ her.⁴³ While we should not put too much credence in twentieth-century complaints by self-styled ‘modernizers' (both foreign and Chinese) that the Chinese ‘wasted’ an unusual amount of their resources on such events—much less read them back into the eighteenth century—it seems to me quite likely that there was, indeed, a difference in the relative shares of goods and services in the discretionary spending of ordinary Chinese and Europeans. If so, that would make the rough comparability in those goods we can measure all the more impressive a testament to the relatively high standard of living (for a pre-industrial society) of eighteenth-century China.

5.