1. Todaro, Michael, Internal Migration in Developing Countries: A Review of Theory, Evidence, Methodology and Research Priorities (Geneva: International Labor Office, 1976).Google ScholarMallce, Hein, “Agricultural labour and rural population mobility: some preliminary observations,” paper presented at the Conference on Rural-Urban Migration, Beijing, 25–27 06 1996.Google Scholar

2. Solinger, D., “The floating population leaves its rural origins: the mix of state and market,” chapter 3 of book manuscript draft, China's Floating Population, 1996.Google ScholarYang, Dennis and Hao, Zhou, “Rural-urban disparity and sectoral labor allocation in China,” Research Papers in Asian Pacific Studies, Duke University, Durham, North Carolina, 1996.Google ScholarMeng, Xin, “Regional wage gap, information flow, and rural-urban migration,” paper presented at the Conference on Rural-Urban Migration, Beijing, 25–27 06 1996.Google ScholarHare, Denise and Zhao, Shukai, “Labor migration as a rural development strategy: a view from the migration origin,” paper presented at the Conference on Rural-Urban Migration, Beijing, 25–27 06 1996.Google ScholarParish, William, Zhe, Xiaoye and Li, Fang, “Non-farm work and marketization of the Chinese countryside,” The China Quarterly, No. 143 (1995), pp. 697–730.Google Scholar

3. Hare, and Zhao, , “Labor migration as a rural development strategy.”Google ScholarGuo, Zhenglin, “Farmer migrants and their acquisition of modernity,”Google Scholar in Research Group on Migration of Female Farmers, Zhongguo dagongmei (China's Female Farmer Migrants) (Shanghai: Shanghai People's Publishing House, 1996).Google Scholar

4. Xiang, Biao, “How to create a visible ‘non-state space’ through migration and marketized traditional networks: an account of a migrant community in China,” paper presented at the Conference on Rural-Urban Migration, Beijing, 25–27 06 1996.Google Scholar It is also possible that migration is an opportunity attractive to a middle income group (Zhang, Xiaohui, Zhao, Changbao and Chen, Liangbiao, “1994: nongcun laodongli kua qutu liedong di shezheng miaoshu” (“1994: a real description of rural labour's cross-regional flow”), Zhanlüe yu guanli (Strategy and Management), No. 6 (1995) pp. 30–35Google Scholar) or that the migrant stream is a mixed group of the poor and relatively wealthy (Solinger, , “The floating population leaves its rural origins”).Google Scholar

5. Meng, Xin. “Regional wage gap, information flow, and rural-urban migration.”Google ScholarXiang, Biao. “How to create a visible ‘non-state space’.”Google Scholar

6. Liu, Jimin, “Obstacles and outlets for rural labor heading for the market,” Renkou yu jingji (Population and Economy), No. 6 (1995), p. 60Google Scholar, as quoted in Solinger, , “The floating population leaves its rural origins.”Google Scholar

7. Yang, and Hao, , “Rural-urban disparity and sectoral labor allocation.”Google ScholarKnight, John and Song, Lina, “China's labor markets,” Bulletin of Rural Labor Mobility Studies, 08 1996.Google ScholarHare, and Zhao, , “Labor migration as a rural development strategy.”Google Scholar

8. Solinger, , “The floating population leaves its rural origins.”Google Scholar

9. Zhang, , Zhao, and Chen, , “1994: a real description of rural labor's cross-regional flow.”Google ScholarParish, , Zhe, and Li, , “Non-farm work and marketization of the Chinese countryside.”Google ScholarLi, Yinhe, “A survey in the Pearl River Delta on migrant farmers,”Google Scholar in Research Group on Migration of Female Farmers, China's Female Farmer Migrants.

10. Meng, Xin, “Regional wage gap, information flow, and rural-urban migration.”Google ScholarMallee, , “Agricultural labour and rural population mobility.”Google ScholarHare, and Zhao, , “Labor migration as a rural development strategy.”Google Scholar

11. The sample villages were selected randomly on the basis of a stratified random sampling procedure. They all come from eight representative provinces (Zhejiang, Shandong, Hubei, Sichuan, Yunnan, Shaanxi, Liaoning and Guangdong) which were randomly selected from each of China's traditional geographic regions (East China - huadong, North China - huabei. Central China - huazhong, Sichuan, South-west China - xinan, North-west China - xibei, North-east China - dongbei and South China - huanan). Eight counties were selected from each province, two from each quartile of a list of counties arranged in descending order of gross value of industrial output (GVIO). GVIO was used on the basis of the conclusions of Rozelle that GVIO is one of the best predictors of standard of living and development potential and is often more reliable than net rural per capita income (Rozelle, Scott, “Stagnation without equity: changing patterns of income and inequality in China's post-reform rural economy,” The China Journal, No. 35 (01 1996), pp. 63–96Google Scholar). Two townships, one above the median GVIO and one below, were randomly selected from each county. Two villages in each township were selected in the same manner. Data problems in Liaoning and two counties in Yunnan precluded their inclusion in the analyses (one team in Yunnan failed to complete their study, and data in Liaoning were collected during pretesting and deemed incomparable to the rest of the data). Data collection in Guangdong was so expensive that the study was never started.

12. Leaders were instructed to classify villagers using their most important jobs and to ignore transient, minor employment. Ignoring short-term, less important farm work will lead to a systematic underreporting of off-farm participation. However, farmers sometimes have two main lines of work, so there is some double counting. Although such a reporting method accurately records the pattern of primary employment activity, it would also lead to over-reporting of off-farm labour participation rates. It is unclear to what extent the two errors offset one another.

13. The survey also attempted to estimate the number of permanent outmigrants (yongjiu buanzou de ren/hu). For the purposes of this study permanent outmigrants are those who leave the village for employment purposes only, not because of marriage, education or retirement. Permanency, in contrast to long-term status, signals that the migrant has no intention of returning to or residing in the village in the near future. One signal of permanency may be that leaders officially transfer a household's or individual's residency permit (hukou) to another jurisdiction or unit. Such a transfer is not an adequate sign of permanency in itself because the residency permit may move even though the person continues to live in the village and has no intention of leaving. Alternatively, people can attain permanent migrant status without changing residency. For example, if a household's residency officially remained in the village, but it sold or leased its house and physically moved out of the village for several years, enumerators counted the members of such a household as permanent outmigrants. Leaving the village permanently was such a rare event that the survey tabulated the total number of households leaving the village in the periods 1989–95 and 1980–88. For the remainder of the article, migrants refer only to long-term, not permanent, migrants.

14. Zhongguo tongji nianjian (China Statistical Yearbook) (Beijing: State Statistical Bureau, 1996).Google ScholarParish, , Zhe, and Li, , “Non-farm work and marketization of the Chinese countryside.”Google Scholar

15. Huang, Jikun, “A report on the state of China's agriculture,” Food and Agricultural Organization, Rome, 1996.Google Scholar

16. Rozelle, , “Stagnation without equity.”Google Scholar

17. Zhongguo xiangzhen qiye nianjian (China Township and Village Enterprise Yearbook) (Beijing: Ministry of Agriculture, 1996).Google Scholar

18. Rawski, Thomas “The process of China's reform,” paper presented at the Stanford Series of Talks on East Asian Culture and Modern Society, Stanford University, 06 1997.Google Scholar

19. Entwisle, Barbara, Hendersen, Gail, Short, Susan, Bouma, Jill and Fengying, Zhai, “Self employment in China,” American Sociological Review, 1995.Google Scholar

20. Solinger, , “The floating population leaves its rural origins.”Google ScholarZhang, , Zhao, and Chen, . “1994: a real description of rural labor's cross-regional flow.”Google ScholarMallee, , “Agricultural labour and rural population mobility.”Google Scholar

21. Ze, Zhu, “The present condition of the ‘labor tide’ its elements, and countermeasures,” Zhongguo nongcun jingji (Chinese Rural Economy), No. 12 (1993), p. 33.Google ScholarZhang, , Zhao, and Chen, , “1994: a real description of rural labor's cross-regional flow.”Google Scholar

22. Mallee, , “Agricultural labour and rural population mobility.”Google Scholar

23. Ibid.Tan, Shen, “A study in Zhongshuicun village, Jiahe county, Hunan province,”Google Scholar in Research Group on Migration of Female Farmers, China's Female Farmer Migrants. Wan, Donna, “Women in China and the off farm employment decision,” unpublished senior honors thesis, Department of Economics, Stanford University, 1993.Google Scholar

24. Chan, Anita, and Senser, Robert A., “China's troubled workers,” Foreign Affairs, Vol. 76, No. 2 (06–04 1997), pp. 104–117.CrossRefGoogle ScholarSolinger, , “The floating population leaves its rural origins.”Google Scholar

25. Research Group on Migration of Female Farmers, China's Female Farmer Migrants.

26. Mallee, , “Agricultural labour and rural population mobility.”Google ScholarZhang, , Zhao, and Chen, . “1994: a real description of rural labor's cross-regional flow.”Google ScholarXiang, Biao, “How to create a visible ‘non-state space’”Google ScholarLohmar, Bryan, Zhao, Changbao and Rozelle, Scott, “Rural to rural migration: new channels of off-farm employment in rural China,” working paper, Department of Agriculture and Resource Economics, University of California at Davis, 1998.Google Scholar

27. In China, the education level of the labour force tends to be negatively correlated with age, given the progress in providing education during the last few decades. Hence, it is difficult to distinguish on the basis of descriptive statistics if we are seeing the education, age cohort or both effects. If randomly selected villages tend to have similar age structures, this shortcoming will be partially compensated for in the multivariate analysis.

28. Meng, Xin, “Regional wage gap, information flow, and rural-urban migration.”Google ScholarMallee, , “Agricultural labour and rural population mobility.”Google Scholar

29. Mallee, , “Agricultural labour and rural population mobility.”Google ScholarHare, and Zhao, , “Labor migration as a rural development strategy.”Google Scholar

30. Xiang, Biao, “How to create a visible ‘non-state space’”Google ScholarMeng, Xin, “Regional wage gap, information flow, and rural-urban migration.”Google Scholar

31. Solinger, , “The floating population leaves its rural origins.”Google Scholar

32. Yang, and Hao, , “Rural-urban disparity and sectoral labor allocation in China.”Google Scholar

33. Todaro, , “Internal migration in developing countries.”Google Scholar

34. Solinger, , “The floating population leaves its rural origins.”Google ScholarZhang, , Zhao, and Chen, , “1994: a real description of rural labor's cross-regional flow.”Google ScholarMallee, , “Agricultural labour and rural population mobility.”Google ScholarHare, and Zhao, , “Labor migration as a rural development strategy.”Google Scholar

35. Solinger, , “The floating population leaves its rural origins.”Google ScholarZhang, , Zhao, and Chen, , “1994: a real description of rural labor's cross-regional flow.”Google ScholarXiang, Biao, “How to create a visible ‘non-state space’.”Google Scholar

36. Solinger, , “The floating population leaves its rural origins.”Google ScholarMallee, , “Agricultural labour and rural population mobility.”Google ScholarYang, and Hao, , “Rural-urban disparity and sectoral labour allocation in China.”Google Scholar

37. Hare, and Zhao, , “Labor migration as a rural development strategy.”Google ScholarSolinger, , “The floating population leaves its rural origins.”Google ScholarMallee, , “Agricultural labour and rural population mobility.”Google ScholarLi, Yinhe, “A survey in the Pearl River Delta on migrant farmers.”Google ScholarZhang, , Zhao, and Chen, . “1994: a real description of rural labor's cross-regional flow.”Google Scholar

38. Oded, , Stark, “Rural to urban migration and some economic issues: a review utilising findings of surveys and empirical studies covering 1965–1975 period,” working paper, World Employment Programme Research Series, International Labour Office, Geneva, 1976.Google Scholar

39. Todaro, , “Internal migration in developing countries.”Google Scholar

40. Solinger, , “The floating population leaves its rural origins.”Google ScholarHare, and Zhao, , “Labor migration as a rural development strategy.”Google ScholarGuo, Zhenglin, “Farmer migrants and their acquisition of modernity.”Google Scholar

41. Solinger, , “The floating population leaves its rural origins.”Google Scholar

42. Hare, and Zhao, , “Labor migration as a rural development strategy.”Google Scholar

43. Guo, Zhenglin, “Farmer migrants and their acquisition of modernity.”Google Scholar

44. Xiang, Biao, “How to create a visible ‘non-state space’.”Google Scholar

45. Li, Fan, “Waichu dagong renyuan guimo, liudong fanwei: Zhongguo nongcun laodongli liudong yanjiu zhi er”Google Scholar (“The volume, scope of mobility, and other aspects of labour migrants: research on the problem of rural labour mobility in China, part two”, Zhongguo nongcun jingji (China's Agricultural Economy), No. 9 (1994), pp. 31–35.Google Scholar

46. Stark, , “Rural to urban migration and some economic issues.”Google ScholarTodaro, , “Internal migration in developing countries.”Google Scholar

47. Meng, Xin, “Regional wage gap, information flow, and rural-urban migration.”Google Scholar

48. Panel of the Rural Department of the State Council's Development Research Centre, “The organized features of the migration of rural labor,” policy recommendation, Beijing, 1996.

49. Xiang, Biao, “How to create a visible ‘non-state space’.”Google Scholar

50. Liu, Jimin, “Obstacles and outlets for rural labor heading for the market.”Google Scholar

51. Zhao, Anna Yao-Hui, “The occupational choice of labor and household incomes in rural China,” paper presented at the Allied Social Science Association Annual Meeting, Chicago, 3–5 01 1998Google Scholar

52. Johnson, D. Gale, “Is agriculture a threat to China's growth?” working paper No. 95:04, Office of Agricultural Economics Research, University of Chicago, 04 1995.Google Scholar

53. Mallee, , “Agricultural labour and rural population mobility.”Google Scholar

54. Jackson, J. A., Migration (London: Longman, 1986).Google Scholar

55. There may also be a number of urban institutions that restrict entrance into the urban market, such as the hukou, or household registration system, the lack of social services and educational facilities for migrants and their children, and the hiring practices of urban firms. This article ignores these constraints for two reasons: urban constraints should affect migrants almost uniformly, and our village-based survey would not be able to capture these effects, even if they were important.

56. Yang, and Hao, , “Rural-urban disparity and sectoral labor allocation in China.”Google Scholar

57. Sicular, T., “Why quibble about quota,” working paper, Department of Economics, University of Western Ontario, Hamilton, Ontario, 1995.Google Scholar

58. Mallee, , “Agricultural labour and rural population mobility.”Google Scholar

59. Solinger, , “The floating population leaves its rural origins.”Google Scholar

60. Chan, and Senser, , “China's troubled workers.”Google Scholar

61. Rozelle, Scott and Leying, Jiang, “Survival strategies during agricultural recession in China's reforming economy,” Chinese Environment and Development, Spring/Summer 1995, pp. 43–84.Google Scholar

62. Todaro, , “Internal migration in developing countries.”Google Scholar

63. Stark, , “Rural to urban migration and some economic issues.”Google Scholar

64. Weimer, Calla, “Labor mobility and institutional constraints in urban China,” working paper, Department of Economics, University of Hawaii, 1997.Google Scholar

65. Zhang, , Zhao, and Chen, , “1994: a real description of rural labor's cross-regional flow.”Google Scholar

66. Parish, , Zhe, and Li, , “Non-farm work and marketization of the Chinese countryside.”Google Scholar

67. Li, Yinhe, “A survey in the Pearl River Delta on migrant farmers.”Google Scholar

68. Mallee, , “Agricultural labour and rural population mobility.”Google Scholar

69. Hare, and Zhao, , “Labor migration as a rural development strategy.”Google Scholar

70. The analysis includes two measures from Skinner: CPZ and CS. CPZ (Core-Periphery Zone) measures the distance of a village from the “core” metropolis of the macro region and is measured from 1 to 7, 7 being the most remote. CS (City System) is an index of the urbanization of the county that the village belongs to. It ranges from 1 (highly urban, like the Shanghai suburbs) to 6 (highly rural, like the most rural areas of northern Jiangsu and eastern Anhui). Provincial dummies account for differences between provinces and control for differential policies and wage differentials (Meng, Xin, “Regional wage gap, information flow, and rural-urban migration”Google Scholar; Yang, and Hao, , “Rural-urban disparity and sectoral labor allocation in China”).Google Scholar

71. Lohmar, , Zhao, and Rozelle, , “Rural to rural migration.”Google Scholar

72. Zhao, , “The occupational choice of labor and household income in rural China.”Google Scholar

73. A Markov chain analysis is a model of transitions between two fixed states (in this case long-term migration and no long-term migration) which we used to compute the mean probability of long-term migration in 1995 given migration (or no migration) in 1988. For further explanation of the use of Markov chains in the context of labour movements see Lee, Tsoung-Chao, “Changes of employment among sectors of the fast growing economy in Taiwan: a Markov chain analysis,” in Griffiths, W., Lutkepohl, H. and Bock, M. E. (eds.), Readings in Econometric Theory and Practice (Elsevier Science Publishers B.V., 1992).Google Scholar

74. Yang, and Hao, , “Rural-urban disparity and sectoral labour allocation in China,” for example.Google Scholar

75. Personal communication from Uma Lele, senior economist World Bank to senior author, January 1998.

76. The unit of observation is the village not the individual, so our results shed little light on the debate over who within the village chooses to migrate.

77. Meng, Xin, “Regional wage gap, information flow, and rural-urban migration.”Google ScholarMallee, , “Agricultural labour and rural population mobility.”Google Scholar

78. We use two models, ordinary least squares (OLS) and tobit because the dependent variable, percentage of the village labour force in the migrant labour force, is limited. OLS results are efficient but slightly biased; tobit results arc not as efficient, but unbiased. Our conclusions rely primarily on the OLS results because only 6% of villages reported no migration (0%), so OLS bias is minimal.

79. Per capita village income in 1988 is used instead of per capita income in 1995 to avoid endogeneity. We expect that villages with high levels of migration may have higher incomes because of migration (migrants often send money home) and we wish to determine the effect of income on migration rather than the effect of migration on income.

80. Because this variable also captures all village characteristics not explicitly included in the regression, we explicitly include as many village characteristics as possible to isolate the chain effect.