Income Distribution

The very poor in China, then, were quite likely doing as well or better than the very poor of western Europe, but that was not very well in either case. When we try to estimate how many people were not very poor, and what their standards of living looked like, we run into massive difficulties.

we have still argues for rough comparability in both elite and ‘middle class' consumption, this must remain conjectural. At the top of the society (not counting a relatively small group of privileged Manchus), there is no question that many members of a loosely defined ‘gentry' class had plenty of money. Chang Chung-li's famous estimates of the income of the Chinese gentry have many problems, but are still the best we have. He estimates that in the late nineteenth century, a group of gentry and their immediate families, totalling perhaps 7.5 million people, had an income of about 645,000,000 taels, or 24% of national income (Chang 1962: 327, 329). At roughly 430 taels per family of five, such people would have had to spend only a tiny fraction of their income (under 4%) on cereals; even a luxurious diet could hardly have taken up much of their total spending. A century earlier, the numbers would have been different, of course; but the percentage of gentry in the population would not have been much different (around 2%), and most of their spending would certainly have been on things other than basic subsistence.

As I have noted elsewhere, Chang's figure of 24% of income going to something like the top 2% of the population²⁴ is not much different from figures we can derive for England and Wales, drawing on the work of Peter Lindert and Jeffrey Williamson.

The big problem, of course, is estimating the numbers and standards of living of those who were neither at the top nor at the bottom of the society. Some recent work on the clerks and runners who occupied the bottom of the hierarchy in local government offices—and were probably the least prestigious literate occupational group in late Imperial China—provides some interesting information. Bradly Reed has estimated that yamen runners in late nineteenth­century Baxian (Sichuan) made perhaps 35—65 taels per year and head runners 2—3 times that much; he puts the income of yamen clerks considerably higher, at 100—150 taels per year, and head clerks at 300—400 taels. He cites other scholars who have suggested higher figures.²⁶ We know little about the consumption habits of such people—despite, or perhaps partly because of, stereotypical complaints from their social superiors that they spent huge amounts on alcohol and prostitutes—but it would appear that they had substantial amounts of money to spend. And while the numbers of such people are not known, Reed has made an estimate of 100 clerks for his county (before it was opened to foreign trade and residence in 1890, leading to a huge boom in legal and administrative business and in the number of clerks), and anywhere from 250 to 450 runners.²⁷ While Baxian probably had more personnel than the average of China's roughly 1,500 counties, and all counties had presumably less personnel a century earlier, it seems conservative to guess that there would have been 250,000 clerks

and runners in mid- to late-eighteenth-century China—representing, with their families, perhaps 1.25 million people.

Beyond these groups, there is even less to go on. We have some wage figures for urban workers, but most are of very limited use: they are, for instance, often unclear about the extent of in-kind supplements to the cash wage. One of the few examples we have that includes an estimate of cash wages, the value of in-kind supplements, and an estimate of the cost of living for a working-class family comes from a salt works in late nineteenth-century Sichuan.²⁹ If the numbers are accurate, a family of four needed 3,600 wen in cash to live on, of which it spent roughly 600 wen for grain; it received grain worth 900 wen as part of the humblest worker's wages (brine carriers and hammermen received a bit more), plus a very small amount of other food, which we will discount. If this 4,500 in-cash and in-kind represented the family's total budget, then grain would be roughly one-third of total living expenses for these workers. It appears that this would have given the family an annual grain supply of 8.4 shi of husked grain: almost exactly the average figure we have used for China as a whole.³⁰ (It is not clear, however, how the poorest workers would have raised the 3,600 wen in cash, since their pay ranged from 1,000 to 2,200 wen, unless either their wives and/or children contributed to much of the family budget.

Other workers in this salt works must have had considerably more to spend on things beyond the bare necessities. The best-paid salt makers earned four times the cash wage of the poorest workers, the cashier three times, foremen five to six times, assistant accountants seven to eight times, and the head accountant twenty times. Throughout the scattered wage data we have, we find very large differentials between unskilled workers on the one hand and skilled workers, supervisors, and clerks on the other; this might suggest that the latter groups were relatively prosperous, as I have suggested, but there are also good reasons not to leap to this inference.³¹

Zhao Gang's historical survey of wage data provides details on monthly cash wages for various kinds of textile workers which, when converted into unhusked

rice, range from 0.29 to 4.19 shi of grain per month (3.5—50.4 per year). For less-skilled workers, about 5 shi (which equalled 2.75 shi of husked grain) a year seems typical. If we assume that such a worker also received roughly 5 shi of husked grain per year in kind (slightly higher than the grain allowance for the poorest salt workers in the nineteenth­century example),³² had a wife and two children and no other income, he would have come up slightly short of even providing his family with an all-cereals diet meeting the caloric average. But a single unskilled (or in some cases even skilled) worker could not feed a whole family elsewhere in the eighteenth-century world either and, as I have argued elsewhere, there may have been less of a gap in earning power between poor Chinese men and women in the eighteenth century than was the case in Europe.³³ If the wife in this example added to the budget half as much as the husband's cash and in-kind earnings, 75% of their combined earnings would provide an average caloric intake, making them roughly as poor as the English agricultural workers in the 1830s mentioned above.

The much larger and more important problem of estimating the numbers and incomes of the rural majority—those that were neither gentry-landlords, nor the landless labourers we have discussed above—is not even this far advanced. All we can say is that (1) agricultural yields were sufficient to feed everyone, and (as we shall see below) to produce more cloth per capita than Europeans wore—leaving aside the crucial issue of income distribution; (2) the relatively widespread ownership of land, dispersion of by-employments, and generally light tax rates may have made for a more equal distribution of income than in most commercialized agrarian societies; (3) tenants and small farm owners must have generally been better off than the landless labourers discussed above; and (4) such people were not leaving the land at rates sufficient to change the urban/rural distribution of population very much.³⁴ Income differentials do not, of course, automatically trigger large migrations, so we cannot infer much from their absence: but if even those with some property or secure tenancies in the countryside had been vastly poorer than petty urbanites, one would probably expect more movement than the record suggests. While there obviously were huge numbers of poor people in the countryside, general living standards, employment opportunities, and subsistence security seem to have been adequate to keep the overwhelming majority of them in place in the eighteenth century. As we shall see, at least subsistence security probably declined in the nineteenth and early twentieth centuries in most regions. The general standard of living may have as well, but more through the growing weight of poor regions in China-wide aggregates than through declining consumption in particular places. Before we return to this contention, we need to look at what the consumption levels of things besides basic foodstuffs were in the eighteenth century.

4.