1. By contrast, the role of grassroots political leaders has been extensively analysed.

2. Some prominent examples are: Lardy, Nicholas, “State intervention and peasant opportunities,” in Parish, William L. (ed.), Chinese Rural Development: The Great Transformation (Armonk, N.Y.: M. E. Sharpe, 1985)Google Scholar; and Agriculture in China's Economic Development (Cambridge: Cambridge University Press, 1983); Lyons, Thomas P., Economic Integration and Planning in Maoist China (New York: Columbia University Press, 1987)Google Scholar; Sicular, Terry, “Agricultural planning and pricing in the post-Mao period,” The China Quarterly, No. 116 (December 1988), pp. 693–95.Google Scholar

3. Terry Sicular writes: “Economic signals sent out by the central government can differ from those received by farmers. Intermediate levels of government filter and modify signals sent by the centre. Local officials adopt policies selectively, often emphasizing those measures that benefit the local government and ignoring those that do not. Local leaders also continue directly to intervene in production and marketing, thus constraining farmers’ range of choices. The effects of changes in prices and incentives, then, depends not only on farm response but also on the response of local governments. Local government reactions must therefore be taken into account when formulating price and incentive policies.” (“Agricultural planning,” p. 703.)

4. In 1986 Shulu county was redesignated a municipality (shi), and renamed after its central town of Xinji. This occurred under a national policy of indentifying rural counties that had been relatively successful industrializers as prospective centres of further growth in their regions. The redesignation did not appear to be the effect or the cause of any special state assistance, nor did it involve any boundary changes or perceptible changes in administrative relationships or power vis-à-vis higher or lower levels of the political system. But for the sake of historical continuity and expository simplicity, we will refer to it as Shulu county throughout.

5. This is discussed in Marc Blecher and Vivienne Shue, Tethered Deer: Government and Economy in a Chinese County (Stanford: Stanford University Press, forthcoming).

6. Huang, Philip C. C., The Peasant Economy and Social Change in North China (Stanford: Stanford University Press, 1985), pp. 111–120 and 125–137.Google Scholar

7. Myers, Ramon, The Chinese Peasant Economy: Agricultural Development in Hopei and Shantung, 1890–1949 (Cambridge, Mass.: Harvard University Press, 1970), p. 179.Google Scholar

8. By this time cotton was a minor crop occupying 1 or 2% of arable land. (Hebei jingji shouce (Handbook of Hebei Economy; hereafter HBSC) (Shijiazhuang: Hebei renmin chubanshe, 1986), p. 264; Huang, The Peasant Economy, p. 128.) Huang's estimate is based on Kraus, Richard A., Cotton and Cotton Goods in China, 1918–1936 (New York: Garland, 1980).Google Scholar Yet it accords almost exactly with that of the Hebei jingji shouce. But none of these three sources offers any explanation of this huge change in the fortunes of Hebei cotton by the late Qing.

9. Zhonggui, Li and Fengtai, Zhang, Shulu xiang tuzhi (Local Gazetteer of Shulu) (1906), 12 juan, pp. 35–37.Google Scholar

10. “Hebei sheng Shulu xian difang shiji qingkuang diaocha baogao” (“Report of an investigation into the actual conditions in Shulu county, Hebei province”), in Jicha diaocha tongji congkan, Vol. II, No. 3 (1936), p. 110. Thanks to Kathleen Hartford for referring us to this source and supplying a copy.

11. Hebei Provincial Statistics Bureau (ed.) Hebei tongji nianjian 1987 (Hebei Statistical Yearbook; hereafter HBNJ) (Beijing: Chinese Statistics Press, 1987), pp. 463–485.Google Scholar

12. Hebei ribao (Hebei Daily News), 24 August 1986.

13. See n. 8.

14. HBSC, p. 264; Huang, The Peasant Economy, p. 128.

15. For instance, the Hebei provincial government announced on 21 September 1949 that in no place should the price of one jin of standard cotton be lower than that of nineyin of millet (HBSC, p. 737).

16. Kenneth Walker also found significant levels of Hebei grain imports for this period, though he delineates a somewhat different set of data based on his survey of contemporary reports in the Hebei ribao (Hebei Daily News): 1953 – 598,000 metric tons; 1954 – 905,000; 1955 – 1,205,000; 1956 – 2,115,000; 1957 – 750,000. The average is 1,114,000 metric tons, significantly higher than the 838,000 given by the HBSC data set reported in Table 2. See Walker, Kenneth, Food Grain Procurement and Consumption in China (Cambridge: Cambridge University Press, 1984), p. 87.Google Scholar The discrepancy between the data sets is not systematic from year to year, and therefore cannot be accounted for by the difference between husked and unhusked equivalents. It remains difficult to reconcile in the absence of more extensive information about the ways in which the statistics were collected and reported in the 1950s and the 1980s. Nevertheless, the point about very large grain imports remains incontrovertible. (It also finds support from Lardy, Agriculture, pp. 36, 62.) Both data sets also show that grain imports surged in 1956, the year of peak cotton sown area (until 1980).

17. Lardy, Agriculture, p.39.

18. HBSC, p, 748.

19. Nationally, grain sown area dropped only 9.4% from 1957 to 1959, and then actually recovered 5% in 1960 before dropping again to roughly 1959 levels through 1965. National Statistics Bureau (ed.), Zhongguo tongji nianjian-1981 (Statistical Yearbook of China; hereafter ZGNJ) (Beijing: Chinese Statistics Press, 1982), p. 138.Google Scholar The reasons for the much sharper drop in Hebei cannot be determined definitively on the basis of the present research. Figure 1 indicates that the drop in grain sown area was not the result of displacement by key cash crops (although cotton and, to a lesser extent, oil-bearing crop areas were held roughly steady in these years of declining grain). Rather, the land taken out of grain in this period seems to have disappeared from agriculture. This suggests the possibility that Hebei was more subject than China as a whole to the manias for farmland infrastructure projects and rural industrialization, which would have drawn off the labour needed to plant and tend crops.

20. Lardy, “State intervention,” pp. 42–43.

21. Central Committee of the Chinese Communist Party, “Guanyu shougou zhongyao jingji zuowu shixing liangshi jiangli de zhishi” (“Directive on the implementation of grain incentives for procurement of important cash crops”), 3 April 1961Google Scholar; Fangquan, Mei, “Woguo liangshi he mianhua wenbu xietiao fazhan wenti de yanjiu” (“A research note on the question of steady and balanced development of grain and cotton [production]”), Nongye jingji wenti (Problems of Agricultural Economics), 1988, p. 22.Google Scholar There is no way of ascertaining the extent to which this policy was implemented in Hebei or anywhere else; certainly, it would have been extremely difficult to do so as long as grain remained in very short supply.

22. Statistical Bureau of the People's Republic of China, China Statistical Yearbook, 1989 (Beijing: Chinese Statistical Information and Consultancy Service Centre, 1990), p. 158.

23. ZGNJ 1983, pp. 154–159.

24. These dropped from 44.9% of total output in 1959 to 15.0% in 1962 and 12.3% in 965.

25. Shulu officials said that, owing to disruptions in statistical work during the Cultural Revolution, time series data were more reliable and readily available for a few communes than for the county as a whole. We requested data for one poor, one middling and one prosperous commune (later redubbed xiang (township)). The three selected by county officials – Muqiu, Tianjiazhuang and Fanjiazhuang – do indeed capture that range. They also show some of the geographical variability of rural Shulu: Muqiu is in the historically poor, flood-prone south, Tianjiazhuang is a peri-urban locality adjacent to the county seat of Xinji (into which it was incorporated as the “Western Administrative District” (cheng xi banshichu) in 1986), and Fanjiazhuang is in north central Shulu.

26. In fact, its ubiquity even increased: the average ratio of cotton to grain sown land rose from 0.81 in 1962 to 0.86 in 1965.

27. As a result, grain output per capita was still higher than the provincial average.

28. Lyons, Economic Integration.

29. ZGNJ 1983, pp. 154–55.

30. Mei Fangquan, “Research note,” p. 22.

31. Zhongguo nongye nianjian 1981 (Yearbook of Chinese Agriculture, hereafter NYNJ) (Beijing: Agriculture Press, 1982), p. 591.

32. Grain yields rose as well, from 1.45 to 2.12 tons/ha.

33. Renmin ribao, 1 October 1971 and 10 January 1975.

34. The appearance of an “area-oriented” pattern of grain production in the sample communes in this period corresponds to two similar findings from rather different sources. First, it parallels trends for Shijiazhuang prefecture as a whole (Marc Blecher, “Economic development and distribution in Shijiazhuang prefecture, 1966–78” (unpublished manu-script)). Secondly, it corresponds to a detailed account offered by a knowledgeable former county-level official in Guangdong province (Marc Blecher, interview conducted at Universities Service Center, Hong Kong, 1978). He reported that before 1972 the Dazhai/Xiyang programme of agricultural development emphasized bringing larger amounts of land under cultivation each year, in emulation of Dazhai's remarkable example of terracing formerly uncultivable mountainsides. According to this same account, after 1972 the emphasis shifted to improving the yields of land already under cultivation.

35. ZGNJ 1983, p. 154.

36. Walker, Kenneth R., “Grain self-sufficiency in North China, 1953–1975,” The China Quarterly, No. 71 (September 1977), p. 560.Google Scholar

37. According to Shijiazhuang prefecture officials, the central and south Hebei plains on which Shulu is located were fully cultivated by 1966 (interview, Shijiazhuang Prefecture Planning Commission, 9 July 1979).

38. As indicated at the bottom of Table 3, what we call the “minimum multicropping index” is our calculation, based on dividing the sum of the sown areas of wheat and maize by the total cultivated area of grain. The actual multicropping index for grain may have been slightly higher than this, insofar as it does not take into account sown areas of minor grains for which we do not have data. But since in Shulu wheat and com make up by far the bulk of grain production, we believe that the differential between the minimum and actual figures is bound to be very small.

39. In addition to cotton, other cash crops must have also been sacrificed to expand grain production. In all the three sample communes, increased grain cultivated area was larger than that deducted from cotton area. Taken together, land taken from cash crops other than cotton contributed about half of the expansion of grain cultivated area in the three communes. Nevertheless, in absolute terms cotton was probably the one crop whose area had been cut most substantially. Here again, the more important cotton had been in 1965, the more in percentage terms the deduction of its area contributed to the expansion of grain cultivated area. Decreased cotton area in Tianjiazhuang accounted for 60% of that expansion, and in Fanjiazhuang 49%. Muqiu, the least cotton-intensive of the three, cut more from oil-bearing crops and other cash crops (60%) than from cotton (40%).

40. Tianjiazhuang's ratio had been 0.99 in 1965, and 16.6% of cotton land was taken to plant grain in the interval. Muqiu's ratio of 0.70 had been the lowest among the three, and its loss of cotton land was the smallest (7.5%). Fanjiazhuang's ratio had been in between the other two (0.94), as was its decline in cotton area (11.6%).

41. These are, of course, only assumptions. But we derive them not from some abstract notions about human motivation, but rather from the subsequent behaviour of Shulu farmers in the post-1978 period, analysed below.

42. The average figure for the 1965–71 period of extensive growth was 0.7%.

43. It had increased less than 6% per year in 1965–71. Other factors which were associated with increased grain production in the previous period–greater allocation of I labour, more draught animals, and extension of irrigated area – registered no significant : changes in this one. There was little room for increasing irrigated area in this period: by 1976, I all of Tianjiazhuang's and 99.1% of Fanjiazhuang's land was irrigated, as was 92% of Muqiu's.

44. The increase in chemical fertilizer application was distributed evenly over the three communes during this period. Chemical fertilizer was an important factor in China's agricultural development at this time; it was also one which was relatively expensive and most subject to control through the state supply system. That these three rather diferent communes could register such similar increases in this expensive and tightly controlled input suggests a certain evenhandedness or uniformity in the operation of state (in this case, county) policies concerning planning, finance and allocation of chemical fertilizer in a period when it was a centrepiece of agricultural development planning.

45. Calculated from Sicular, “Agricultural planning,” table 3 (p. 687).

46. From 1965 to 1971 chemical fertilizer application to cotton rose 45.5%, faster than that to grain (39.3%). But from 1971 to 1978, the former increased only 37.5% and the latter 91%. Doubtless there was also competition between cotton and grain for other inputs such as capital investment, scientific inputs (such as pesticides, which cotton also requires in particularly high amounts) and labour time.

47. Myers, The Chinese Peasant Economy, pp. 179–180.

48. Interviews: Shulu Agricultural Bureau, 30 June 1990; Wangkou township, 4 July 1990.

49. Hebei sheng Xinji shi jianjie (A Brief Introduction to Xinji Municipality, Hebei Province), (n.p., September 1987), pp. 8–9.

50. HBSCp. 267.

51. Sicular, “Agricultural planning,” pp. 677, 691.

52. See Shue, Vivienne, “Beyond the budget: finance organization and reform in a Chinese county,” Modern China, Vol. 10, No. 2 (April 1984), pp. 150–166CrossRefGoogle Scholar; Blecher and Shue, Tethered Deer, ch. 4.

53. Dexin, Zhao, Zhonghua renmin gongheguo jingji shi, 1967–1984 (The Economic History of the People's Republic of China, 1967–1984) (Zhengzhou: Henan renmin chubanshe, 1989), p. 415.Google Scholar

54. Grain quota purchase prices were increased 20%, beginning with the summer harvest of 1979, with an additional 50% premium for above-quota sales. Purchase prices of cotton, oil-bearing crops, sugar and other farm and sideline products were also raised. The average price increase for all agricultural purchases was about 22%. Smaller price rises followed in subsequent years, and the proportion of state purchases at above-quota and negotiated prices also rose from negligible levels in 1977 to 60% in 1981 (Riskin, China's Political Economy, 55. “Communiqué of the Third Plenary Session of the Eleventh Central Committee of the Communist Party of China,” Beijing Review, No. 21 (29 December 1978), p. 52.

56. ZGNJ 1983, pp. 437–38.

57. Tsou, Tang, Blecher, Marc and Meisner, Mitch, “The responsibility system in agriculture: its implementation in Xiyang and Dazhai,” Modern China, Vol. 8, No. 1 (January 1982), pp. 41–105CrossRefGoogle Scholar; Watson, Andrew, “Agriculture looks for shoes that fit: the production responsibility system and its implications,” World Development, Vol. 11, No. 8 (August 1983), pp. 705–730.CrossRefGoogle Scholar

58. HBSC, p. 39.

59. For a general discussion, see Sicular, “Agricultural planning,” p. 685 and passim. 60. Moreover, it took a rather “leftist” form. Private plots were cultivated collectively, and in the county government leadership and an economically advanced, urban-based agricultural commune, brigade-level accounting was even being defended.

61. In Hebei, cotton production jumped from 115,600 tons in 1979 to 247,200 tons in 1980. Cotton yields climbed from 210 to 450 kgTha. (HBNJ1988, pp. 450–53). Also see NYNJ 1981, p. 473.

62. HBSC, pp. 769–770.

63. Hebei ribao (Hebei Daily News), 16 May 1981; and HBSC, p. 255. Shulu moved completely to household contracting by 1983.

64. In China the drop was only 1.8%.

65. For example, in 1978 it was 54.3% in Shulu but only 11.3% in Hebei.

66. Walker, Kenneth R., “Trends In Crop Production 1978–1986,” The China Quarterly No. 116 (December 1988), p. 616.Google Scholar

67. Mei Fangquan, “Research note,” p. 22.

68. As a result, the quality of cotton purchased by the state in 1985 was, putatively, down 1.3 grades nation-wide and 1.65 grades in north China from the previous year (“Mianhua wenti diaocha,” (“Investigation into the cotton question”), Renda baokanfuyin ziliao-nongyejingji (People's University Publication Reprints-Agricultural Economy, No. 6 (1987), p. 184).

69. In one Shulu village, they ran to 4 yuan/kg., as against a state purchase price of 5.6 yuan.

70. Interview with Zhao Yingmu, Vice-Director, Shulu Agriculture Bureau, 30 June 1990.

71. Interview with Party Secretary Ma Yue, Wangkou, Shulu, 4 July 1990. He was referring specifically to the various problems cited above. But in addition, part of the problem may have been the slow pace of increase of cotton procurement prices in Shulu compared with elsewhere in China. For this 60 yuan price increase, reported by Secretary Ma, is only 27%, which is well below the official state figures which report an increase in the procurement price of cotton in 1989 of 22.7% in 1989 and another 29.1% in 1990 (ZGNJ 1991, p. 254).

72. ZGNJ 1988, p. 248.

73. OECD, Agriculture in China, p. 29. In general the new oil-bearing crop price was a weighted average of 40% of the old quota price plus 60% of the old above-quota price (Sicular, “Agricultural planning,” p. 693).

74. By contrast it dropped only 3.1% in Hebei as a whole. The difference had to do with Shulu farmers’ above average aversion to cotton in these years. Another reason for the drop in Hebei and Shulu was a serious drought, which was especially devastating to oil-bearing crops.

75. Figures 5–7 also show that sown area of oil-bearing crops rose in two of the three sample communes. Nationally sown area of oil-bearing crops registered a 20.5% increase, and output a 32.5% increase.

76. HBNJ 1989, p. 96.

77. According to officials of the Shulu Communist Party Rural Work Department, this had occurred without significant popular opposition at the time, since fruit prices remained low and the market remained undeveloped (interview, Shulu county, 11 July 1990).

78. HBNJ 1989, p. 479.

79. Ibid. p. 96.

80. NYNJ 1985, pp. 1–3, 5–7; Sicular, “Agricultural planning,” p. 694.

81. In 1984 the Rural Development Research Group, a key government think tank, came to the conclusion that the supply of grain and cotton had greatly exceeded demand in China, and that this trend would continue unless the government adopted “effective measures to cut its procurement of grain and cotton.” Zhongguo nongcun fazhan wenti yanjiuxin (Chinese Rural Development Research Group), “Woguo liangshi gongjin de xianzhuang he biandong gushi” (“The state of our country's grain supply and its changes”).

82. Economic Research Centre of the Ministry of Agriculture, “Liangshi duanqueji jingji zhengce de tiaozheng” (“Grain shortage and the readjustment of economic policy”), Nongye jingji wenti (Problems of Agricultural Economics), No. 5 (1988), pp. 6–7.Google Scholar

83. In 1978, government price subsidies were already 6 billion yuan, or 14.3% of its budgetary revenues. As early as 1986, however, the subsidies amounted to 70 billion yuan or 38.4% of the revenues. The price subsidies thus had become one of the most important sources of the rising deficit, a situation which has not abated. Lian Tianzheng, “Guanyu jinjin caizheng chizhi wenti de shizheng fenxi” (“An analysis of budgetary deficits in recent years”), CaimaojingjUFinancial Economics), No. 5 (May 1987), p. 38; Ma Xiaohua, “Caizheng fenpei geju de xin bianhua yu shenghua gaige” (“Recent changes and enhanced reforms in financial distribution”), Caijing yanjiu (Research in Financial Economics), No. 5 (May 1988), p. 14.

84. See n. 4.

85. The redesignation would not change this rule. Hebei Provincial Government, “Guanyu shixing liangshi zhenggou xiaoshou diaobo baogan yiding sannian banfa de tongzhi” (“On implementing the method of a definite three-year plan to control grain procurement, purchasing, allocation and contracting”), 8 February 1982; in HBSC, pp. 201–203.Google Scholar

86. In 1986, Shulu's Agriculture Bureau, the local arm of the Ministry of Agriculture, opened an Agricultural Technical Centre.

87. Guanyu 1986 nian nongcun gongzuo bushu de jige zhongyao wenjian (Several Important Documents on the Deployment of Rural Work in 1986) (Beijing: Renmin chubanshe, 1986).

88. HBNJ 1989, pp. 95–96.

89. More specifically, from 1983 to 1986 charges in Hebei for electricity rose 100%, water 200%, and diesel fuel 300%(Renmin ribao, 3 October 1986).

90. HBNJ 1989, p. 96.

91. This may explain the puzzle, posed by Sicular, of why rising input prices were accompanied by increased cropping in input-intensive crops. Another answer may have to do with cotton's own input-intensity. Sicular, “Agricultural planning,” pp. 697–98.

92. On top of all that, the relative profitability of agricultural production in contrast with non-agricultural endeavours may have affected their cropping decisions insofar as it prompted farmers to grow crops that required less intensive tending or whose tending could fit into a regime of non-agricultural employment.

93. Cf. Japan, South Korea and Taiwan, which have in recent decades gained great experience in and institutions and information well-suited to close monitoring and regulation of the market. By contrast, from 1956 to 1978 the Chinese state tried to regulate its own private sector mainly by bludgeoning it.

94. Renmin ribao, 16 April and 26 November 1986.

95. HBNJ 1989, p. 64.

96. Interviews, Shulu county, July 1990.

97. Sicular, “Agricultural planning,” pp. 695–96.

98. Da, Huo, “Minyishi weitian” (“Hunger breeds discontent”), Zhongguo zuojia (Chinese Writers), No. 4 (1989).Google Scholar

99. In Jiucheng, which borders on Junqi, net income of cotton per ha. was 6,450 yuan, second only to that from vegetables. By contrast, in southernmost Wangkou, which was receiving new targets for cotton area, it was only 1,500 yuan.

100. Perry, Elizabeth J. and Christine Wong (eds.), The Political Economy of Reform in Post-Mao China (Cambridge, Mass.: Harvard University Council on East Asian Studies, 1985), p. 14.Google Scholar

101. Xuchu, Ouyang, “Bixu diaodong difang zhengfu fazhan liangshi shengchan de jijixing” (“It is necessary to mobilize the activism of local governments to develop grain L production”), Nongye jingji wenti (Problems of Agricultural Economics), No. 6 (1988), p. 25.Google Scholar

102. Moreover, while a local government like Shulu's maintains an interest in a basic level of grain production in order to keep its population fed, it has no such interest in cotton.

103. For example, it had linked up Wangkou's premier industrial enterprise with the provincial Industry Bureau, which provided capital and vastly expanded markets. The county had also arranged to contract out the operation of a failing state enterprise under its administration to Wangkou. These are discussed in detail in Blecher and Shue, Tethered Deer.

104. They might also seek to do so if they were pursuing certain specific local goals such as expanding production of industrial products that required local agricultural inputs. We found no examples of such products in Shulu's industrial repertoire.

105. Interview, Shulu Water Conservancy Bureau, 20 June 1990.

106. I.e., the head of the county government after it was reclassified as a municipality.

107. Interview with Party Secretary Bai Runzhang and Mayor Liu Baolu, 13 December 1986.