1. The development of poverty from 1984 to 1989 is controversial. According to the World Bank, Sharing Rising Incomes, Disparities in China (Washington, D.C.: The World Bank, 1997), p. 10Google Scholar, the rural poverty rate actually increased as there was little growth and inequality increased.

2. Wong, C. K., “Measuring Third World poverty by the international poverty line: the case of reform China,” Social Policy and Administration, No. 29 (1995), pp. 189–203Google Scholar; and “How many poor people in Shanghai today? The question of poverty and poverty measure,” Issues and Studies, No. 33 (1997), pp. 32–49.Google Scholar

3. World Bank, China Strategies for Reducing Poverty in the 1990s (Washington, D.C.: The World Bank, 1992), p. 5.Google Scholar

4. Riskin, C., “Chinese rural poverty: marginalized or dispersed?” American Economic Review, Vol. 84, No. 2 (1994), pp. 281–84Google Scholar. In an analysis of the household income survey, 1988, Gustafsson and Li reported results which were more in line with Riskin's view (Li, S. and Gustafsson, B., “An estimate of the extent and scale of poverty during the late 1980s in China,” Social Sciences in China, No. 6 (1996), pp. 29–44Google Scholar; and “The structure of Chinese poverty, 1988,” The Developing Economics, Vol. 36, No. 4 (1998), pp. 387–406)Google Scholar. However, there is no guarantee that what was valid at the end of the 1980s also applies later; this issue is addressed further below.

5. Chen, S. and Ravallion, M., “Data in transition: assessing rural living standards in southern China,” China Economic Review, Vol. 7, No. 1 (1993), pp. 23–56.CrossRefGoogle Scholar

6. Jalan, J. and Ravallion, M., “Transient poverty in postreform rural China,” Journal of Comparative Economics, Vol. 26 (1998), pp. 338–357.CrossRefGoogle Scholar

7. Gustafsson, and Li, , “An estimate,” and “The structure of Chinese poverty.”Google Scholar This poverty line is 32% higher than the absolute poverty line for rural China defined by Riskin, , “Chinese rural poverty.”Google Scholar

8. Ren, R. and Chen, K., “An expenditure-based bilateral comparison of gross domestic product between China and the United States, Review of Income and Wealth, Vol. 40, No. 4 (1994), pp. 377–394.CrossRefGoogle Scholar

9. Griffin, K. and Zhao, R. (eds.). The Distribution of Income in China (London: Macmillan, 1993)CrossRefGoogle Scholar. More details of the survey can be found in Eichen, M. and Zhang, M., “The 1988 household sample survey – data description and availability,”Google Scholar in ibid. pp. 331–36. See also Khan, A. R., Griffin, K., Riskin, C. and Zhao, R., “Household income and its distribution in China,” The China Quarterly, No. 132 (1992), pp. 1029–61.Google Scholar

10. Khan, A. R. and Riskin, C., “Income and inequality in China: composition, distribution and growth of household income, 1988 to 1995,” The China Quarterly, No. 154 (1998), pp. 221–253Google Scholar. See also Zhao, R., Li, S. and Riskin, C. (eds.), Re-study on Income Distribution of Chinese Households (Beijing: Publishing House of Chinese Finance and Economy, 1999) (in Chinese).Google Scholar

11. As households belonging to the floating population are likely to be more poverty prone than other urban households, we are probably underestimating urban poverty in China. However, it is less clear if the floating population is more or less poverty prone than the rural population. It thus follows that we do not know whether our estimates under- or overestimate poverty in China as a whole.

12. The provinces in the rural sample are Beijing, Hebei, Shanxi, Liaoning, Jilin, Jiangsu, Zhejiang, Anhui, Jiangxi, Shandong, Henan, Hubei, Hunan, Guangdong, Sichuan, Guizhou, Yunnan, Shaanxi and Gansu. The provinces in the urban sample are Beijing, Shanzi, Liaoning, Jiangsu, Anhui, Henan, Hubei, Guangdong, Sichuan, Yunnan and Gansu.

13. We define income of owner-occupied housing in rural China similarly to the definition in Gustafsson, and Li, , “An estimate,” and “The structure of Chinese poverty.”Google Scholar

14. The procedure for arriving at disposable income which is comparable across China in reported in ibid.

15. Compare the practice of Eurostat, the statistical authority of the European Union, which publishes estimates of poverty for the various member states based on poverty lines defined as 50% of mean income for the Union, as well as based on 50% of mean income for each member state. Eurostat, , Income Distribution and Poverty in EU 12–1993 (Luxembourg: Statistics in Focus, 1997).Google Scholar

16. Foster, J., Greer, J. and Thorbecke, E., “A class of decomposable poverty measures,” Econometrica, Vol. 52, No. 3 (1984), pp. 761–66.CrossRefGoogle Scholar

17. Our picture of the development over time is consistent with the official estimates discussed above. However, our methodology results in a poverty population for China somewhat larger than the one found in official estimates. Our poverty rate is also slightly higher than the alternative “absolute poverty line” reported by the World Bank (Sharing Rising Incomes) but it is lower than for the same source's alternative “higher poverty line.” If our poverty line is increased by 25% poverty rates of 18.42% (1988) and 16.88% (1995) are found; and putting it at 75% of the preferred line results in rates of 5.63% (1988) and 5.06% (1995).

18. The issue of how to update the poverty line has received attention in the literature on poverty in industrialized countries (see Citro, C. and Michael, R., Measuring Poverty. A New Approach (Washington, D.C.: National Academy Press, 1995)).Google Scholar

19. The eastern region includes Beijing, Tianjin, Hebei, Liaoning, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong, Hainan and Guanxi. The middle region includes Shanxi, Inner Mongolia, Jilin, Heilongjiang, Anhui, Jiangxi, Henan, Hubei and Hunan. The western region includes Sichuan, Guizhou, Yunnan, Tibet, Shaanxi, Gansu, Qinghai, Ningxia and Xinjiang.

20. An altitude of less than 200 metres above sea level is termed plains, an altitude of 200–500 metres is classified as hills and an altitude of more than 500 metres is classified as mountains.

21. If the poverty line is set to 125% there is actually an increase (from 20.83 to 23.96%) while if it is set to 75% the estimates are virtually identical (6.45 and 6.49%).

22. Similar conclusions are reached when estimating poverty rates for categories having different access to irrigated land. Those with more than average irrigated land have poverty rates of 9.23% in 1988 and 5.74% in 1995, while those with less than average irrigated land have poverty rates of 22.77% in 1988 and 21.77% in 1995.

23. For further details of how economic reform has affected China's minorities and the gap in living standards between them and the majority population see Mackerras, C., “The impact of economic reform on China's minority nationalities,” Journal of Asia Pacific Economy, No. 30 (1998), pp. 61–79.Google Scholar

24. Gustafsson, and Li, , “An estimate,”Google Scholar and “The structure of Chinese poverty.”

25. Li, T. and Zhang, J., “Returns to education under collective and household farming in China,” Journal of Developing Economics, No. 56 (1998), pp. 307–335Google Scholar; Meng, X. and Wu, H., “Household income determination and regional income differential in rural China,” Asian Economic Journal, Vol. 12, No. 1 (1998), pp. 65–88.CrossRefGoogle Scholar

26. Replicating Table 6 but using a poverty line set to 75% or 125% of the preferred alternative gives in both cases a less favourable picture of the development of poverty in large households.

27. We cannot rule out the possibility that if we had applied a different equivalence scale when computing equivalent income that the age-poverty profile would have looked somewhat different.

28. Danziger, S. and Gottschalk, P., America Unequal (New York: Russell Sage; and Cambridge, MA: Harvard University Press, 1995), ch. 5.Google Scholar

29. The interaction term in Table 8 stems from the fact that the disadvantages are not simply additive.

30. Further analyses are required to find out to what extent those differences are purely locational, and to what extent they can be explained by differences in access to land, household size, education and other household variables.