1. Whyte, Martin King, “Inequality and stratification in China,” The China Quarterly, No. 64, p. 690.Google Scholar

2. Ibid. pp. 690–691.

3. Ibid. p. 688, n. 11.

4. For the sake of brevity, I am concerned in this comment with only one of Whyte's several arguments and conclusions. His article takes on two related but distinct topics: the level and types of inequalities in China today, and the mechanisms and structures for transmitting these inequalities into the future. Here I am concerned mainly with the former problem, and more specifically with the problem of intra-village distribution.

5. Alker, Hayward R. and Russett, Bruce, “Indices for Comparing Inequality,” in Merritt, Richard L. and Rokkan, Stein (eds.), Comparing Nations (New Haven: Yale University Press, 1966), p. 369.Google Scholar

6. Kyuichi, Tanaka, Jihenka No Hokuski Nōson: Kahokushō Teikennai Ichinōson Jittai Chosa Hokoku (A North China Village Since the China Incident: An Investigation into the Real Situation in a Village in Ting County, Hopei Province) (Dairen: Minami Manshu tetsudo kabushiki kaisha chosabu (Investigation Depart-ment of the South Manchurian Railway), 1937).Google Scholar

7. The cash equivalency is clearly a statistical artifact; of course most crops were not converted into cash. It is used only as a way of expressing the different families’ harvests of different crops in the same unit of measurement. In computing the conversion, I have been forced to use 1931 crop prices for Ting county, because the 1939–40 prices in the Li-ts'un-tien survey were not reported clearly. The rate of inflation between 1931 and 1939–40 is not particularly important, as long as most crops inflated at the same rate, since we are interested only in the relative, not the absolute value of the crops. Liberation by Communist forces, which occurred in 1937 in Ting, did not drive up prices much in other places. Since Ting county was not occupied by the Japanese until 1940, in all likelihood inflation was low until then anyway. This information, and the prices I used in converting crops to cash value, can be found in Dorris, Carl E., People's War in North China (Univ. of Kansas, Ph.D. thesis, 1975), ch.4B.Google Scholar

8. My calculations leave out rent payments and income. But only 4% of the land in the village was rented in or out. They also exclude production costs, but we can assume that per-mou production costs were about the same, if not lower for larger holdings (which would mean that my calculations would understate inequality). I have averaged the harvest data from 1939 and 1940, to mitigate slightly against seasonal differentiation. The data for separate years do exist for those interested in making the calculation.

9. In the interests of clear presentation, the decile data in this table and those that follow for the other cases were extrapolated from the Lorenz curve, because the population does not naturally break down into deciles, but rather into quantiles of uneven size. The Lorenz curves and Gini coefficients were plotted and calculated on the basis of these quantiles of uneven size, in order to maximize their accuracy. The fact that deciles were extrapolated accounts for the fact that some deciles appear to have received the same share of income; but in fact the ranked quantiles all received at least slightly different shares.

10. Ramon Myers does provide data on household incomes in Sha-ching village of Hopei's Shunyi county. However, the data are misleading for two reasons. First, while they do include rent payments, they do not include income from rent collections, either from fellow villagers or from 460 mou of land which Sha-ching villagers owned in other villages. Secondly, since people in other villages owned 515 mou of Sha-ching land, the Myers data relate only to income as measured by consumption, but not to income as measured by production (since much of Sha-ching's production was being funnelled off to other villages in the form of rent). Since we can probably assume that the substantial missing income data were distributed relatively unequally, any calculations of income distribution in Sha-ching would no doubt seriously overstate equality there. I did calculate the Gini coefficient for net farm income in Sha-ching at 0·4406, and for overall net income (including non-farm income) at 0·267. These figures suggest some support for the frequently heard hypothesis that in traditional or transitional China poor peasants relied on non-farm income, including income from the sale of their own labour as farm hands, to maintain subsistence. The data suggest specifically that non-farm employment did provide a way of equalizing income shares. However, if our assumptions about the distribution of the missing income are correct, then the actual Gini coefficients for Sha-ching village would be much higher (i.e. would show more inequality) than the two figures calculated here. The Sha-ching data can be found in Myers, Ramon, The Chinese Peasant Economy (Cambridge, Mass.: Harvard University Press, 1970), ch. 4.Google Scholar

11. Hung, Fan, Hsi-shan nung-yeh sheng-ch'an ho-tso-she ti ch'eng-chang (The Maturation of Hsi-shan Agricultural Producers’ Co-operative) (Peking: San-lien shu-tien, 1957), pp. 114–15.Google Scholar

12. Ibid. p. 116.

13. Li T'ien-kao, “Mu-ch'ien Shan-tung sheng nung-min fu-tan ch'ing-k'uang ti tien-hsing tiao-ch'a” (“A model investigation of the current situation of peasant's burdens in Shantung province”), Ts'ai-cheng (Finance), 1957, No. 3.

14. Myrdal, Jan, Report from a Chinese Village (New York: Signet, 1965). Actually, the 46 families of the village for which we have data (out of 50 in the whole village) work in two different units. 31 work in a normal production team (and constitute the entire team) and 15 work in a vegetable-raising team which also includes five families from other villages. Since the focus here is on equality in the village, and since the higher the n the better. I have combined the data from the two units. I did work out separate income distributions for the two units, and they were quite similar. Myrdal lists data on cash and grain distribution. But the value of the grain distributed to each labour-power must be deducted from the cash distribution. For purposes of analysis of income distribution, then, it is best to use cash distribution alone.Google Scholar

15. There are theoretical reasons for suspecting that calculation of per-capita income may distort reality, particularly where grain distribution is included. After all, all persons do not need the same amount of grain or cash. I did calculate income on a weighted per-capita basis. The weighting scheme I used was, like all weighting schemes, somewhat arbitrary; but it was distilled from my interviews on differential distribution of grain to different age groups and to males and females. I weighted males aged 19–65 as four units, children aged 13–18 and women 19–60 as three, children aged 6–12 and men over 65 and women over 60 as two, and children under 5 as one. There turned out to be almost no difference between the distributions figured on this weighted basis and the figures cited above, based on unweighted population calculations. Perhaps a different set of weightings would reveal more, but any weighting scheme must be both theoretically meaningful as well as empirically valid, not merely the latter. It should also be noted that the Myrdal data do not include the incomes being received in the form of remissions to 11 families from principal breadwinners with jobs at the commune or in Yenan. Most of these families have been ranked rather low in the village income hierarchy: in ranking the 46 village families by per-capita income shares, these 11 families were ranked 20, 31, 33, 38, and 40–46. So by excluding a portion of the poorer families’ incomes, the calculations based on Myrdal actually overstate inequality of income distribution. But if we estimate remissions of ¥25/family/month, the amount of missing income is only ¥275, or 1% of the total village income.

16. For the case from Tung-kuan county, I have data on 22 families who were neighbours of the informant, representing 22.5% of the households of the team. For the case from Tien-pai county, I have data on the entire team. Because in the Tung-kuan case there was no clear relationship between residence patterns and wealth, class, or lineage group, we can treat the 22 families as a random sample of the team, at least for purposes of income distribution analysis. The extremely close correspondence of the data from the two villages lends some support to this assertion about the sample. Calculations were made assuming that people who informants reported as working “occasionally” or “sometimes” worked half the number of days of a regular labourer. The Tung-kuan informant was a young single rusticated male who hailed originally from Canton. The Tien-pai informant was a local male in his late 20s.

17. Professor Kenneth Walker has argued that they did contribute to income inequality in the early 1950s, when former middle and rich peasants were able to profit from their larger holdings of land and draught animals, and when collective production was less developed than it is today. Prof. Walker's conclusions seem sound, but I am sure he would agree that they would not necessarily apply to the period of more advanced commune development where land and draught animals are all collectively owned. See Walker, Kenneth, Planning in Chinese Agriculture (London: Frank Cass and Co., 1965).Google Scholar

18. Families with many young children not yet old enough to work in the team's fields, who would be receiving low per-capita incomes as a result, can send the children out to the private plot to work, thereby raising the family's income.

19. Household animals are fed with private plot crops and crop by-products, with leftovers from household meals, and sometimes with purchased fodder. So availability of feed would tend to vary with household size and income. Kenneth Walker has also assumed that pig-raising would vary directly with the size of private plot, which we know varies with the size of the family. See Walker, Planning in Chinese Agriculture, p. 28.

20. Consider the following hypothetical example: A person with a good factory or clerical job in a town may make ¥50/month. The cost of living in town can be estimated at around ¥12/month for food (a figure reported by students as the cost of their food at school) and ¥3 for lodging. Adding ¥5 for incidentals like clothing, cigarettes, and entertainment, and ¥5 for travel expenses to visit home each month, she or he has ¥25/month left to remit home. A strong male earning 300 work-points/month would earn ¥30 in a team with a 10¢ work-point value, and ¥15 in a team with a 5¢ work-point value.

21. Whyte, “Inequality and stratification,” pp. 690–91.

22. Li, “Mu-ch'ien Shan-tung,” pp. 21–22.

23. Since 1958, Tachai brigade has steadily increased the share of its annual income allocated to collective accumulation. See Tachai Hung-ch'i (The Red Flag of Tachai) (Peking: Jen-min ch'u-pan she (People's Publishing House), 1974), p. 11.Google Scholar For another case of this same development, see Hsia-liang, Yüan, “Tsai yi-ke hsien-wei shu-chi di shen-pien” (“At the side of a county secretary”), Hsueh-hsi yü p'i-p'an (Study and Criticism), 1976, No. 6, pp. 49–57.Google Scholar

24. The number of children in a family will alter the range of the hand: mouth ratio over the life-cycle, but in no way nullify its effects.