Increased Farm Labour Productivity in Jiangnan
One of the problems with the previous studies has been their tendency to account for labour productivity using modern standards. Though labour productivity can generally be defined as the amount of output or income produced per unit of labour,29 the standards for calculating labour productivity will vary across different societies.
Previously I have pointed out that the methods used in calculating labour productivity in pre-modern Chinese agriculture differ in four ways from those used in modern industrial societies.30 For the discussion in this chapter, farm labour productivity will be calculated using the peasant family as the unit of labour and a year as the standard for time. Male and female labour productivity will be discussed separately, since a division of labour between the sexes became much clearer during this period.The ‘trinity pattern' is the optimal pattern in the Jiangnan peasant economy because under this pattern higher yields per mu can be achieved with lower inputs. As I point out above, in the early seventeenth century, a farm ran, on average, 15 mu of cultivated land with a multi-cropping index of 140%, the second crop being wheat. The yield per mu was 1.7 shi for rice and 1 shi for wheat. If all the 15 mu of land were planted in rice, this farm would harvest 26 shi of rice and 6 shi of wheat, together equivalent to 30 shi of rice.31 In contrast, in the mid-nineteenth century, the average farm size was 9 mu with a multi-cropping index of 170%. Per mu yield was 2.5 shi of rice and 1 shi of wheat. Farm output was 23 shi of rice and 6 shi of wheat, totalling 27 shi of rice, 10% below the early seventeenth-century figures. However, if we calculate labour productivity according to the number of workers, output per worker would be 15 shi of rice in the early seventeenth century and 27 shi of rice in the mid-nineteenth century respectively.
That is, the figure for the late Ming period is only 55% of the mid-Qing figure.If this macroanalysis is insufficient, a concrete example may be helpful. Western Songjiang will be used for this purpose. This area is located in eastern Jiangnan and
is a typical agricultural region, with rice as its staple product. But before the discussion begins, I will remind the reader: the Songjiang case is very exceptional, since most peasant families never farmed as much land nor had as high yields in late Ming Jiangnan.
In western Songjiang, a peasant family could plant 25 mu of rice and harvest 2.5 shi/mu, for a total of 62.5 shi of rice in the late sixteenth century in which period one-year-one-crop pattern was dominant (He Liangjun 1959: 115). In contrast, an early nineteenth-century peasant family planted 10 mu with rice at 3 shi/mu and wheat at 1 shi/mu, for a combined total of 37 shi of rice (Jiang Gao 1963: 3b—10a). We can see that the mid-Qing's total is only 60% of the late Ming example. Moreover, production costs per mu of rice were equivalent to about 1 shi of rice in both cases. Costs for wheat were about one-fourth of those of rice, or 0.25 shi of rice/mu, in the mid-Qing case.32 Subtracting production costs, the late Ming family had a net income of 37.5 shi and the Qing family 24.5 shi, about 65% of the figure for the earlier period. Thus, labour productivity for the mid-Qing would seem to be one-third lower than for the late Ming. However, if we subject our example to further scrutiny, we come to quite a different conclusion.
First, in the term of output per worker, the late Ming farm is quite different from the mid-Qing farm. It has already been shown that there is a great difference between the numbers of farm workers in the two cases. In the late Ming case, farming was done by both the husband and wife, while in the mid-Qing period it was done just by the husband. If we calculate productivity according to an individual worker, a late Ming worker produced 31 shi, but a mid-Qing produced 37 shi, about 20% higher.
After subtracting costs of production, the net output per worker is 18.8 shi in the Ming case and 24.5 shi in the Qing case. That is, net output per worker is 30% higher in mid-Qing than in the late Ming.Second, outputs per workday in the two cases quite clearly differ from each other. During the period under study, labour inputs per mu did not change much. A mu of rice took ordinarily about fifteen workdays (including irrigation),33 while a mu of wheat took three workdays.34 For the late Ming farm, 375 workdays were required to work its 25 mu of paddy. On the mid-Qing farm, 10 mu of rice needed only 150 workdays, and 10 mu of wheat required about 30 workdays. The 10 mu of rice and wheat together required 180 workdays. Labour inputs were less than half of those of the late Ming period, the total har vest being about 60% of the example from the earlier period. This implies that production per workday is one quarter higher for the Qing case than for the Ming. More specifically, a Ming workday produces 1.7 dou" of rice and a Qing workday produces 2.1 dou. Subtracting costs of production, net daily production in the Ming is 1 dou of rice and 1.4 dou in the Qing, or 40% higher. This example from western Songjiang shows that, whether calculated on an annual or daily basis, farm labour productivity was increasing during the period under study.
Finally, I will make a brief analysis of rural female labour productivity. As ‘another half of the men’, or ‘another half of the sky’ as a modern Chinese proverb says, women's labour has important influences on men's farm productivity. Taking
a western Songjiang peasant family as an example, I will make a comparison of family labour productivity under different patterns of peasant economy.
In the case of western Songjiang, as we have just seen, a late sixteenth-century peasant family had two workers—a man and a woman—and cultivated 25 mu of rice, with a yield per mu of 2.5 shi. Labour requirements for 25 mu of rice were about 375 days.
All farm work was shared by the man and the woman, except land preparation, which customarily was men's work and accounted for 50 workdays.36 Each of them, therefore, worked for 163 days. After the field practice was finished, rice needed to be husked, which was also commonly ‘man's work' and took the man 63 days.37 Altogether, the man worked for 275 days a year and the woman worked for 163 days. Since 300 days are the maximum number of workdays of a peasant for all kinds of productive work (farming and handicrafts), the man had only 25 days to work at other jobs.38 As for the woman, her maximum workdays were only about 200 days a year (Bozhong Li 19969, 1996r). Since she had already worked in the fields for 163 days, she had only 37 days to do other productive work. Here we assume that the peasant and his wife devoted all their non-farming workdays to cotton spinning and weaving, to a total of 62 days. In pre-eighteenth-century Jiangnan, it usually took about seven days to produce one bolt of cotton cloth.39 Therefore, the man and woman could produce about 9 bolts of cotton cloth within 62 workdays, which could only meet the annual cloth consumption of the family itself. The total output from 25 mu of rice was 62.5 shi of rice. Ifboth the rent of 31.3 shi and production costs (excluding ‘wages') of 12.5 shi are taken away, the net income (including ‘wages') was 18.7 shi. The net income (including ‘wages') from a bolt of cloth is worth, on average, 1.5 dou of rice in the early and mid-Qing. Using this ratio as the standard, 9 bolts were worth 1.4 shi. Together, their total net income was 20 shi.In contrast, during the mid-Qing, the family worked only 10 mu with a total output of 37 shi. Since labour requirements were 18 workdays a mu, the total labour inputs were 180 days. Adding in 37 days for husking and branning, the total becomes 217 workdays, all of which were worked by men only. This would leave the typical farmer 83 days to do other work (assuming cotton handicrafts).40 Since the wife was not required to work in the fields, she could devote all her 200 workdays to cotton spinning and weaving.
Altogether, the farmer and his wife could work 283 days in cotton handicrafts. With one bolt of cotton cloth taking six days to make during the eighteenth and nineteenth centuries, 283 workdays would result in 47 bolts of cloth. Accordingly, the income from this quantity of cloth was equivalent to 7.1 shi of rice. After subtracting rents of 15 shi and production costs (excluding ‘wages') of 6.3 shi, they could receive a net income (including ‘wages') of 15.7 shi from the output of the 10 mu. Together, the total net income was 23 shi.The difference in total incomes between the two families is 15%. Moreover, we should note that a high per mu yield of 2.5 shi of rice on a big family farm of 25 mu is very seldom seen in sources relating to Ming-Qing Jiangnan.41 In contrast, in the mid-Qing there is plenty of evidence referring to per mu yields of 3 shi of rice with 1 shi of wheat (which can be converted to 0.7 shi of rice) on a small family farm
of 10 mu or less.42 Therefore, for most areas in Jiangnan, the difference in labour productivity between farms in the late Ming and those in the mid-Qing will be even greater than those seen in Songjiang. If we use a lower yield figure for the late Ming Songjiang family farm, say 2 shi,4 the net farm income of the family would fall to 12.5 shi. Together with income from cotton handicrafts, the total would be 14 shi, 40% lower than that of the Songjiang family during the mid- Qing.44 All of the results in the preceding analysis indicate that farm productivity did increase in Jiangnan during the period concerned. In this sense, the region did experience real economic growth.45
5.