1. See Walder, Andrew G., Communist Neo-Traditionalism: Work and Authority in Chinese Industry (Berkeley: University of California Press, 1986), ch. 6Google Scholar.

2. Ibid. chs. 4 and 6.

3. This is based on a report by Gong Xiaoxia, now a graduate student in sociology at Harvard University, who was at that time working as a Guangzhou textile worker, and who witnessed these events while travelling through Wuhan. She reports that the streets were littered with abandoned buses, some being driven along haphazard routes by citizens. Two friends of hers who worked at the Wuhan Steel Corporation told her that the strike was motivated by pay and working conditions, even though the protests became part of factional politics in the city. She had also heard, from people directly involved, of similar labour protests in Sichuan province during the same period.

4. Minghua, Li, “Observations on recent labor unrest in mainland China,” Issues and Studies, Vol. 11, No. 10 (10 1975), pp. 2–15Google Scholar, at p. 9.

5. See Forster, Keith, Rebellion and Factionalism in a Chinese Province: Zhejiang, 1966–1976 (Armonk, New York: M. E. Sharpe, 1990), pp. 214–228Google Scholar. Forster makes clear that the Hangzhou outbreak was closely tied to factional divisions dating back to the Cultural Revolution.

6. This is based on a report by Søren Clausen, now teaching at Århus University, who witnessed these events while travelling as a Beijing University student at the end of July. The movement had originated among workers in the city power plant over the persecution of a young couple who worked there. According to a wall poster read by Clausen, in 1974 the young couple had been discovered while engaged in sexual activity in the plant. Party cadres sought to make an example of such blatant bourgeois immorality by ordering a severe administrative punishment. When co-workers protested, the cadres took the case to court and had the man convicted of rape and sentenced to 10 years. When co-workers protested further that it had not been a rape, the woman was given two years of labour reform as an accomplice to a criminal act. Thus began a long struggle against the plant cadres that would blossom when factional divisions split the national and local leadership. While the protests were expressed as opposition to Deng Xiaoping's “revisionist line,” each group within the larger coalition had specific grievances against their factory leadership. Corruption, abuse of power, and living conditions were common complaints. I would like to thank Professor Clausen for providing me with a translation of some of his notes in Danish.

7. See Wilson, Jeanne L., ‘”The Polish Lesson’: China and Poland 1980–1990,” Studies in Comparative Communism, Vol. 23, Nos. 3/4 (Autumn/Winter 1990) pp. 259–280CrossRefGoogle Scholar.

8. Fu Shenqi was a factory worker who had been active in Shanghai in the 1978–1979 “democracy wall” movement, and who in 1980 sought election to the local People's Congress as a representative from his factory. One of the main planks of his platform was improvement of workers’ housing. See the materials collected in Xianggang zhongwen daxue xueshenghui (Student Association of Chinese University of Hong Kong) (ed.), Minzhu zhonghua: Zhongguo dalu minjian minzhu yundong beibuzhe wenji (Democratic China: Collected Writings of People Arrested in the Popular Democracy Movement on the Chinese Mainland) (Hong Kong: Xianggang zhongwen daxue xueshenghui/Yuandong zhiwu pinglunshe, 1982), pp. 350–52, 358–361, 370–77Google Scholar.

9. In 1978 average state sector wages were 644 yuan, in the collective sector, 506. In 1988, the respective figures were 1,853 and 1,426 yuan. ju, Guojia tongji, Zhongguo tongji nianjian 1989 (Statistical Yearbook of China, 1989) (Beijing: Zhongguo tongji chubanshe, 1989), p. 138Google Scholar.

10. ju, Guojia tongji, Zhongguo tongji nianjian 1988 (Beijing: Zhongguo tongji chubanshe, 1988), p. 804Google Scholar.

11. Zhongguo tongji nianjian 1989, p. 719.

12. Calculated from Zhongguo tongji nianjian 1989, pp. 687–88.

13. In constant 1978 yuan, average state sector incomes rose from 644 to 982 yuan; collective sector incomes from 506 to 756 yuan. Calculated from Zhongguo tongji nianjian 1989, p. 138.

14. Zhongguo tongji nianjian 1989, pp. 687–88.

15. Ibid. p. 138.

16. The source is a 1986 survey of 1,011 urban households conducted in collaboration with the Institute of Sociology, Tianjin Academy of Social Sciences, and the Bureau of District Affairs, Municipal Government of Tianjin. See Walder, Andrew G. et al. , “The 1986 survey of work and social life in Tianjin, China: Aims, methods, and documentation,” Working Paper No. 25, Center for Research on Politics and Social Organization, Harvard University Department of Sociology, 1989Google Scholar.

17. For example, in 1984, average per capita housing space was 25% higher in county towns than in municipalities. See State Bureau, Statistical, A Survey of Income and Household Conditions in China (Beijing: New World Press/China Statistical Information and Consultancy Service, 1985), p. 140Google Scholar.

18. This account of the altered position of firms and managers draws on two of my earlier papers, “Wage reform and the web of factory interests,” The China Quarterly, 109 (1987), pp. 22–41CrossRefGoogle Scholar; and “Factory and manager in an era of reform,” The China Quarterly, 118 (1989), pp. 242–264CrossRefGoogle Scholar.

19. Walder, Communist Neo-Traditionalism, ch. 7.

20. See the 1984–1986 interviews cited in Walder, “Factory and manager.” The emigré workers were interviewed in mid-1990 in Cambridge, Massachusetts, and are cited individually below. Three were production workers, one a ticket seller on a bus, and another had advanced from repair worker to manager over an 18-year career. They all left China after October 1989.

21. Interview no. 140, former fitter in a Beijing machinery plant.

22. Interview no. 137, production line worker in a Beijing electronics factory.

23. Interview no. 137.

24. Interview no. 140.

25. See Walder, Communist Neo-Traditionalism, ch. 5.

26. Interview no. 137.

27. See Shirk, Susan L., “Recent Chinese labour policies and the transformation of industrial organisation in China.” The China Quarterly, 88 (1981), pp. 575–593CrossRefGoogle Scholar, and Walder, “Wage reform.”

28. See Wilson, Jeanne L., “Labor policy in China: Reform and retrogression,” Problems of Communism, Vol. 39, No. 5 (09–10 1990), pp. 44–65Google Scholar, at p. 54; and a statement in an unofficial publication by a Changsha worker and democratic activist (arrested shortly thereafter) that attacked the wage readjustment process: Zhang Jingsheng, “Baifenzhi sishi de zhigong zengjia gongzi he gongzi zhidu gaige” (”Raises for 40% of employees and the reform of the wage system”), Zhongsheng (The Clarion) [Wuhan], No. 4 (1980), reprinted in Minzhu zhonghua, pp. 473–75.

29. The process of reaching a new accommodation is described in Walder, “Wage reform,” and Walder, “Factory and manager.”

30. Interview no. 137. There is some factual basis for such beliefs. In 1988 the average annual wage in ownership forms other than the state and collective sectors (excluding private entrepreneurs) was 2,382 yuan, and in joint state–private enterprises 3,229 yuan, compared to a national average of 1,747 yuan. The number of enterprises in these special ownership forms are, however, very small. Zhongguo tongji nianjian 1989, pp. 139, 143, 146 and 149.

31. Interview no. 137.

32. Interview no. 142, a repair worker promoted gradually to director. The workers demanded the tips from customers before picking up the job, “in order to make sure the goods got there, and in good shape.” Customers rarely refused, though they sometimes would complain afterwards to the company.

33. Many, however, make little money, making a transfer to a “labour service company” a new and pernicious form of punishment (see below). This new informal practice was described in all the 1990 interviews with workers.

34. These are the reasons mentioned in interview nos. 137, 140 and 142. In a large national survey in 1986, 83% of the respondents specified a preference for a state sector job. “Gaige de shehui huanjing: bianqian yu xuanze.” (”The social environment of reform: Changes and choices”), Jingji yanjiu, 12 (1987), pp. 52–62Google Scholar, at p. 61.

35. Interview no. 140.

36. Interview no. 137.

37. Interview no. 142. This is one example of a widespread pattern of infringement on the rights of female workers; see Wilson, , “Labor policy,” pp. 50–51Google Scholar.

38. Jingji ribao, 16 May 1989, as cited in Wang, Shaoguang, “Deng Xiaoping's reforms and the Chinese workers' participation in the protest movement of 1989,” unpublished paper, Department of Political Science, Yale University, 1990Google Scholar. This paper contains a long and well-documented account of the kinds of worker dissatisfactions publicized by Chinese social scientists and newspaper reporters.

39. Interview no. 137. The problem was so self-evident that leaders of the All-China Federation of Trade Unions admitted their failure with surprising frankness at their 1988 congress. Wilson, , “Labor policy,” pp. 55–56Google Scholar.

40. Interview no. 137.

41. Interview no. 137.

42. Interview no. 140. Many managers interviewed in China in 1986 gave identical accounts.

43. Interview no. 142.

44. Interview no. 141, a former ticket seller for the Beijing Public Bus Company who later became active in the Beijing Autonomous Workers Union.

45. The survey data reported in “Gaige de shehui huanjing,” pp. 60–62, suggest that inequality of opportunity was the main source of dissatisfaction with reform, something confirmed repeatedly in my recent interviews. It was clear well before 1989 that there was widespread discontent about inflation and corruption. A series of public opinion polls illustrated this dramatically. The 1986 survey just mentioned found inflation to be second only to official corruption as the matter about which citizens were most dissatisfied, and both issues far outpaced all others. An internal Tianjin city government survey in the fall of 1987 found that 98% of the respondents named inflation as their primary concern, and that almost 80% considered their income to be either unsatisfactory (26%) or barely adequate (52%). Hui, Wang, “Municipal administration in China and sociological studies,” unpublished paper, Tianjin Academy of Social Sciences. And a large 1988Google Scholar survey by the Sociology Institute of the Chinese Academy of Social Sciences and State Statistical Bureau found inflation to be by far the most serious problem in people's social lives, with 94% of respondents mentioning it. Qingfang, Zhu, “Yijiubajiu nian dalu chengshi zhigong xintai lu” (”The state of mind of mainland urban workers in 1988”), Liaowang (haiwai ban) 2,1989, pp. 7–8Google Scholar, at p. 7.

46. See Walder, Andrew G., “The political sociology of the Beijing upheaval of 1989,” Problems of Communism, Vol. 38, No. 5 (09–10 1989), pp. 30–40Google Scholar. The account here will trace events in Beijing since the movement there was by far the best documented case.

47. Interview no. 137.

48. Interview no. 140.

49. Interview no. 140. The impact was felt immediately after the large demonstration of 27 April. At a 28 April meeting called by the Municipal Trade Union headquarters of Beijing, a cadre from Guanghua Lumber Company complained that all workers wanted to do as soon as they arrived at work was to talk about “rumours.” Yan, Xuan (ed.), Jingdu xuehuo: xuechao, dongluan, baoluan, pingbao quart guocheng jishi (The Capital in Blood and Flames: A Factual Record of the Entire Process of Student Movement, Turmoil, Rebellion, and Pacification) (Beijing: Nongcun duwu chubanshe, 1989), p. 62Google Scholar.

50. The All-China Federation of Trade Unions donated 100,000 yuan to the Beijing Red Cross for the hunger-striking students. Guojia jiaowei sixiang zhengzhi gongzuo si (Political–Ideological Work Office of the State Education Commission) (ed)., Jingxin dongpo de 56 tian (A Soul-Stirring 56 Days) (Beijing: Dadi chubanshe, 1989), pp. 131–35Google Scholar. On 14 May a delegation from the union marched to the square in support of the students. Yun, Li et al. (eds.), Diankuang de shenian zhi xia (The Tumultuous Summer of the Year of the Snake) (Beijing: Guofang keji daxue chubanshe, 1989), p. 114Google Scholar.

51. Mouren, Wu et al. (eds.), Bajiu zhongguo minyun jishi (A Record of the 1989 Chinese Democracy Movement) (New York: n.p., 1989), p. 251Google Scholar, which reprints the appeal.

52. Interviews no. 138 and 141, both of whom were active in the autonomous union.

53. The same thing occurred in the early stages of the Cultural Revolution, allowing rebel factions very quickly to disrupt factories. See Walder, Andrew G., Chang Ch'unch'iao and Shanghai's January Revolution, Michigan Papers in Chinese Studies No. 32 (Ann Arbor: University of Michigan Center for Chinese Studies), ch. 2Google Scholar.

54. Mouren, Wu, Bajiu zhongguo minyun, p. 267Google Scholar; ribao, Jingji (ed.), Jieyan lingfabu zhiqian: 4.15–5.20 dongluan dashi ji (Before Martial Law was Declared: Major Events during the Turmoil of 15 April to 20 May) (Beijing: Jingji ribao chubanshe, 1989), pp. 131–33Google Scholar.

55. Interview no. 137.

56. Interview no. 140.

57. Interview no. 142.

58. jiaowei, Guojia, Jingxin dongpo de 56 tian, p. 137Google Scholar.

59. Mouren, Wu, Bajiu zhongguo minyun, p. 251Google Scholar.

60. Ibid. p. 251.

61. Ibid. p. 264.

62. These reports were all from official sources, and published after 4 June. jiaowei, Guojia, Jingxin dongpo de 56 tian, pp. 135–37Google Scholar.

63. These names were culled primarily from the following sources: zhongxin, Xianggang gonghui jiaoyu (Hong Kong Trade Union Education Centre) (ed.), Gongren qilaile: Gongren zizhi lianhehui yundong 1989 (The Workers Have Arisen: The Autonomous Union Movement of 1989) (Hong Kong: Xianggang gonghui jiaoyu zhongxin, 1990)Google Scholar; bu, Zong zhengzhi bu xuanchuan, bianjibu, Jiefang junbao (Propaganda Office of the General Political Department, Editorial Office of People's Liberation Army Daily) (eds.), Hanwei shehui zhuyi gongheguo (Protecting the Socialist Republic) (Beijing: Changzheng chubanshe, 1989)Google Scholar; Xuan Yan, Jingdu xuehuo; and Guojia jiaowei, Jingxin dongpo de 56 tian.

64. zhongxin, Xianggang gonghui jiaoyu, Gongren qilaile, pp. 152–54Google Scholar; and a hand bill issued by the Institute of the Chinese Workers' Movement Committee to Support the Students, and Temporary Headquarters of the Capital Workers' Picket Corps, “Zhi shoudu gongren shu” (”Letter to the workers of the capital”). Undated wall poster, but contents establish date as 13 May 1989. Reprinted in Zhongguo minyun yuan ziliao jingxuan (1) (Selected Source Materials from the Chinese Democracy Movement) (Hong Kong: Shiyue pinglunshe, 1989), p. 31.

65. Interview no. 141, and Yun, Li, Diankuang de shenian zhi xia, pp. 189–193Google Scholar.

66. Interview no. 141.

67. Ibid.

68. Ibid., and Zhongguo chengjian gongren zifa lianhe hui zonghuizhang, “Chengjian gongren zifa lianhe hui de zongzhi” (”Aims of the urban construction workers’ autonomous union”). Wall poster dated 21 May 1989. Reprinted in Zhongguo minyun yuan ziliao jingxuan (1), p. 32.

69. The organization used several variations of its name, causing some observers to believe at the time that there was more than one such union. In its various written pronouncements it took the names Beijing gongren zizhi lianhehui, Beijing gongren zizhi hui, and Shoudu gongren zizhi lianhe hui. This account oi gongzilian draws from a paper that is based on several dozen hand bills and interviews with two of the organization's activists: Walder, Andrew G. and Xiaoxia, Gong, “Workers in the Beijing democracy movement: Reflections on the brief history of the Beijing workers' autonomous union,” paper presented to the Berkeley China Seminar, 15 11 1990Google Scholar.

70. This demand referred to involuntary sales of state treasury bonds to workers. A large percentage of one month's wage was issued not in cash but in the form of a state treasury bond redeemable with interest at some future date. Interviews with workers in 1990 suggest that this had been common every October for a number of years.