* Much of the research on which this article is based was supported by grants from the Subcommittee on the Chinese Economy of the Joint Committee on Contemporary China (Social Science Research Council – Joint Council of Learned Societies). The author is grateful for that support, and also for the useful comments of Thomas Rawski, Dwight Perkins, Al Usack, John Montias, Holland Hunter and others on an earlier draft, while retaining responsibility for remaining errors.

1. Summary of World Broadcasts, Part III: The Far East (SWB), FE/5178, 6 04 1976.Google Scholar

2. Peking Review, No. 18 (29 04 1977), p. 21: People's Daily editorial on the occasion of the National Conference on Taching.Google Scholar

3. Yuan-cheng, Lo, “Self-reliance and making foreign things serve China,” Peking Review, No. (8 07 1977), p. 9. State Planning Commission, “Great guiding principle for socialist construction,” Jen-min jih-pao (Jen-min), 12 09 1977, in Peking Review, No.(23 September 1977), p. 7.Google Scholar

4. It would be preferable on theoretical grounds to define “agriculture first” as a policy of giving APG a more-than-proportional share of state investment funds, but this definition is of little use empirically.Google Scholar

5. Chen, N. R., Chinese Economic Statistics (Chicago: Aldine, 1967), p. 26.Google Scholar

6. To keep strictly to this definition, it would be necessary to exclude chemical fertilizer, coal and other intermediate goods from the empirical analysis of the next section.Google Scholar

7. State-enterprise profits plus industrial and commercial taxes were 90 per cent of budget revenues in 1970–1972, versus 6 per cent for the agricultural tax. Eckstein, A., China's Economic Revolution (Cambridge: Cambridge University Press, 1977), p. 184. Peking Review, No. (25 02 1977), p. 26 states that “at present about 70 per cent of the farm machines are bought by the communes or brigades with their own funds. As regards those communes and brigades which are not so well-off, the state annually earmarks a certain amount of money as interest-free loans or free subsidies.”Google Scholar

8. So too is a substantial flow of intermediate output – agricultural raw materials – to industry.Google Scholar

9. Riskin, C., “China's rural industries: self-reliant systems or independent kingdoms?,” The China Quarterly (CQ), No. 73 (03 1978), p. 74.CrossRefGoogle Scholar

10. Peking Review, No. 4 (27 01 1978), p. 30.Google Scholar

11. Chung Chin, “China's road to industrialization,” Peking Review, No 14 (1 04 1977), p. 15.Google Scholar

12. Note that I am not passing judgment on the question of the advisability of such a policy, but simply defining what success in the policy's own terms would require. Such a policy would be desirable if, and only if, the incremental capital-output ratio (i.e., the cost in investment terms of an additional unit of output) is lower in agriculture than in other sectors. This is probably true for China for many sorts of agricultural investments, but is rather difficult to demonstrate empirically.Google Scholar

13. The producer goods sector inevitably grows faster than the consumer goods sector because of a long-term process of capital-deepening, virtually synonymous with the notion of “industrialization.”

14. By using value added in producer goods as our divisor, we in effect calculate APG_i/(APG + IPG) rather than APG_i/IPG. The two are equivalent in the sense that if the first exceeds one, the second must also, and vice-versa. Note that we use rural small-scale output of hydro-electricity, pig iron, coal and cement as a proxy for agricultural use of these products. There is a substantial body of evidence indicating that most small-scale output goes into agriculture, and that most large-scale output (with the exception of fertilizer) does not. See Central Intelligence Agency, China: Role of Small-Scale Industry in Economic Development (1974).Google Scholar

15. In this respect, it is suggestive that during 1963–70, a 57 million ton increase in grain output was associated with a 15·1 million ton increase in fertilizer, a better than three to one ratio, while for 1970–74 the figures were 19 and 9 million tons, or barely two to one. Still, China's use of chemical fertilizer per hectare is only half that of the U.S., one-fifth of North Korea, one-tenth that of Japan.

16. In particular, chemical fertilizer output increased very rapidly during 1976–77 as large foreign-built plants came on stream. And if tractor output is revised to take account of a switch from 7- to 10-h.p. garden tractors, 1·19 changes to 1·42.

17. For a definition and documentation of decentralization, and the argument concerning its impact on APG growth, see below.

18. In the case of hydro, pig iron and irrigation equipment, a negative growth rate in the face of positive PG growth is even stronger evidence in support of this hypothesis than a growth rate ratio which is positive and low.

19. And because of cement's low value-to-weight ratio, the logical location of the added capacity is rural.

20. Tsui-nung, Fang and Yi-hua, Chang, “Strive for modernization of agriculture,” Peking Review, No. 23 (9 06 1978), p. 7.Google Scholar

21. Cf. Mao's “Letter on farm mechanization” (12 03 1966): “Those localities where materials (iron and steel), machine tools and farm machines are under state control but are produced locally and where output far exceeds the state targets (say by 100% or more) should be permitted to buy 30 to 50% of that portion above the target for their own use. Unless this practice is established it will be impossible to bring the initiative of the local authorities into play.” This letter, published for the first time in the midst of the current recentralization, refers only to “state-owned” local enterprises (excluding commune- and brigade-run enterprises and county enterprises not entering higher-level plans). The allowance for local purchase is miserly at best. How much more might the “initiative of the local authorities” be brought into play if the stress were on local ownership and local distribution of output within “self-reliant, independent systems.”Google Scholar

22. For this reason and others, bonuses and other material incentives logically correspond to centralized control. Bonuses to workers appear to have risen over time: 2½ to 3½% of wages in 1956–57 (Donnithorne, p. 211), 6–10% in 1964 (Bettelheim (1965), p. 55), and “10 to 20 yüan a month” or well over 10% today (Peking Review, No. 16 (21 04 1978), p. 6).Google Scholar

23. In India, the poorest 20% of the population spend 85% of their income on food; the richest 10%, only 2%. Income differentials in China are far less extreme, but a rural-urban income gap does exist.

24. An example is the cement industry, where even small-scale plants are now moving from vertical to rotary kilns, but only gradually, and with a typical capacity of 50,000 rather than 300,000 tons.

25. Two instances: “The question of speed in socialist construction is a very important one. … It is a question of life and death for our country … Lenin said: ‘Either perish or overtake and outstrip the advanced countries economically as well’.” SPC, “Great guiding principle for socialist construction,” Peking Review, No. 39 (23 9 1977), p. 7.Google ScholarAgain, Yü Ch'iu-li to the fourth session of the Standing Committee of the Fourth National People's Congress: “We must redouble our efforts and advance at a faster speed. Neither the international situation nor the domestic situation at present permits us to move slowly.” Peking Review, No. 45 (4 11 1977), p. 6.Google Scholar

26. Strong, Anna Louise, “Interview with Po I-po on economic readjustment,” Ta-kung pao (15 01 1964), Survey of China Mainland Press, (SCMP) No. 3152, cited in Riskin, “China's rural industries,” p. 85.Google Scholar

27. To say that an enterprise is “controlled” by, say, the provincial government means that the provincial government must approve the annual (and, usually, quarterly) production plans of the enterprise in advance (and also plans of various other sorts). For enterprises of any significant size – i.e., for perhaps the largest 5,000 enterprises and producing 50% or more of industrial output – local control has always been tempered by the requirement that annual plans be reported as well to central government ministries. To say that a commodity or commodity group is “balanced” by, say, the Ministry of Metallurgy means that as part of the plan-making process, that ministry will draw up a list of sources and uses of that commodity, and if (as is likely) imbalance appears, the ministry has the power to adjust the plans of enterprises or other units on either side of the ledger. This power carries with it a broader mandate to maintain a balance between supply and demand for the commodity in question throughout the plan period as well as during the annual or quarterly planning process itself. Thus the balancing power really confers on a ministry the same power to equate supply and demand which in free enterprise economies is dispersed among participants in the marketplace for that particular commodity.

28. Indeed, a frequently-quoted piece of Maoist economic wisdom makes exactly this point: “In a country like China, with so large an area and so great a population and with so complicated a situation, to concentrate everything on the central government would overburden the central government organs with too heavy administrative duties and would cause the danger of relaxing attention to the general policy of the state. Properly to transfer certain power and responsibility to local governments under unified leadership, thereby to realize local expediency, would, rather than causing any harm, further consolidate the central leadership through the manifestation of local initiative.” Jen-min jih-pao editorial, “Centralized leadership and divided responsibility is the correct policy in financial work,” 5 04 1951, in SCMP, No. 91, p. 16.Google Scholar

29. An anecdotal instance of this: A Canton enterprise produced limestone, cement and other construction materials. The enterprise director, Chang, regularly disagreed with the Party secretary, Chiang, after the latter's appointment to the factory in 1961. For example, “Chang insisted on selling products to Huang-p'u Harbour not only because they could get a better price and more secure payment plan, but because it would be prestigious to supply an international port and promote foreign trade. Chiang, on the contrary, opted for sale to the communes in Sha-tien, even though they would not be able to obtain a good price, and the payment from the communes was not always secure, for it was the policy of the state and party to assist the agricultural sector.” Lee, N. S., China's Industrial Bureaucracy, 1949–1973 (Dissertation, Chicago, 1975), pp. 260–63Google Scholar, based on 1973 Hong Kong refugee interviews.

30. In interpreting Table 4, it is important to bear in mind that from its very earliest stages, this system encompassed in principle the great bulk of all commodities. That is, the 28 commodities under unified distribution in 1952 consisted of broad groupings including steel, coal, cement, machine tools, cotton cloth and sugar. [For a list of these commodities, see Adler, S., The Chinese Economy (New York: Monthly Review Press, 1957), p. 249.]Google Scholar To a substantial extent, the unified distribution system is one which balances all commodities (in broad categories) at the highest level, then balances them all again, in progressively greater detail and disaggregation, at progressively lower levels. This being so, it is not possible to say, for example with reference to 1958, that ministries had four times as much distributive power as did the State Planning Commission because they struck 320 material balances to the latter's 80. But comparisons across time at a particular level are informative: thus we can say that the increase in SPC balances from 60 to 80 from 1973 to 1977 reflects an increase in the SPC's distributive power, whether the 20 new balances represent entirely new commodities or (as is more likely) disaggregations of some of the 60 commodities of 1973.

31. In what follows, we do not intend to imply that recentralization in the 1960s, and again in the 1970s, returns China to the precise status quo ante as of 1957. Very broadly speaking, one could say that the 1960s recentralization fell short of the First Five-Year Plan peak, and that the current recentralization has by now surpassed the 1960s and approaches the 1950s. But for our present purposes, we intend, not to make statements about the absolute level of centralization, but only about the direction in which policy is moving in each period and the dates of policy reversals.

32. Chang, Parris, Power and Policy in China (University Park: Pennsylvania State University Press, 1975); Chao, K. C., Economic Planning and Organization in Mainland China: A Documentary Study (1949–1957) (Cambridge: Harvard University Press, 1959); Audrey Donnithorne, , China's Economic System (New York: Praeger, 1967).Google Scholar

33. A 01 1977 eulogy to Chou En-lai states that “in 1958 and 1970, he twice handled the issue of vesting units at the lower levels with greater power to run and build certain enterprises and the reform of the system of economic management.” (I am indebted for this reference to Thomas Rawski.)

34. See note a to Table 2.

35. Press coverage during late 1970 places the rapid expansion of local industry in 1970 and late 1969. For example, “Wo-kuo ti-fang chung-hsiao-hsing kung-yeh p'eng-po fa-chan” (“China's local, medium and small-scale industry develops abundantly”), Jen-min, 27 12 1970, p. 1. Institutional changes to facilitate this process presumably preceded it.

36. For an excellent documentation of this policy shift, see Riskin, , “China's rural industries,” pp. 94–95.Google Scholar Riskin also provides a wealth of detail on the 1960–62 period.

37. “Quarterly Documentation,” CQ, No. 70 (06 1977), p. 367.Google Scholar

38. “Profits from (Shanghai) industry in 1977 rose more than 10%. … This reversed the situation of the city's failure in fulfilling its revenue plan set by the state for three years running.” See Peking Review, No. 4 (27 01 1978), p. 11. Nationally, 1977 saw a 6·1% fulfilment of the revenue plan after “several years” of non-fulfilment.Google Scholar

39. The disappearance of high-technology imports does not necessarily imply a lack of concern with technology. During decentralized periods, technology policy seems to focus on diffusion from urban IPG to rural APG, by sending skilled workers to rural plants, training rural workers in skilled plants, and assembling in urban plants complete sets of machinery to be installed in rural plants. This process is well documented in the work of Jon Sigurdson (see especially “Rural industry and the internal transfer of technology in China,” paper presented at the University of Sussex, 06 1974). See also Heyman, Hans, “Acquisition and diffusion of technology in China,” Joint Economic Committee, China: A Reassessment of the Economy (Washington, 1975), pp. 678–729.Google Scholar

40. There are two exceptions to this rule in Table 7: cement and chemical fertilizer. Cement is an anomaly: the large-plant sector added 20 new factories during 1971–75 to the 70 existing in 1970, and thereby nearly doubled the capacity of the large-scale sector. But despite this increase in capacity, output increased only 4%, leading to the large ratio 28:60. If capacity utilization in 1975 had equalled that in 1971, the ratio would be below 2. Chemical fertilizer appears to be an industry in which “small-scale” has not implied “local control”; on the contrary, attacks on small-scale ammonia plants as “air-raid shelters for the county party committee,” and on those in charge of them as “red-tapists” for “working in accordance with the documents of the Central Committee” suggest that even during the Cultural Revolution these small-scale plants were under relatively firm central control. Jen-min, 26 03 1977 in SCMP, No. 6324, p. 141.Google Scholar

41. The first Party secretary of Liaoning Province blamed provincial cadres for failing to implement the agriculture first policy in these terms: “Some functionaries of certain factories and mines showed rightist conservatism and departmentalism; they maintained that fulfilment of their production tasks came into conflict with the aid to agriculture. Some regarded the communes as “poverty-stricken relatives” and held that industrial aid to agriculture was “one-way help” and an “extra burden.” Hung-ch'i, No. 13 (1 07 1960), in Chang, Power and Policy in China, p. 275.Google Scholar

42. Chih-ta, Ko and Chao, Wang, Ta-kung pao, 17 11 1961 in Chang, Power and Policy in China, p. 315.Google Scholar

43. The above information on financial processes from Ezra Vogel, “Ministry of Finance: functions, supervision, organization” (manuscript, East Asian Research Center, Harvard University, no date), p. 18.

44. Kuo-chun, Chao, “The structural aspect of planning in China,” International Studies (Bombay), Vol. II, No. 3 (01 1961), pp. 247–48.Google Scholar

45. Central Intelligence Agency, Role of Small Plants in Economic Development (1974), p. 7.Google Scholar

46. “Completely settle the heinous crimes of China's Krushchev and company in undermining agricultural mechanization,” Nung-yeh chi-hsieh chi-shu, No. 5 (08 1967), in SCMP, No. 610, pp. 23–24.Google Scholar

47. Editorial, Nan-fang jih-pao, 15 05 1962, in SCMP, No. 2757, cited in Riskin, “China's rural industries,” p. 81.Google Scholar

48. The “gang of four” are accused of “withholding materials for farmland capital construction. … On the question of commune- and brigade-run industrial enterprises, the ‘gang’ peddled the notion that these enterprises be allowed to find their own markets and develop according to their own wishes.” Peking Review, No. 6 (4 02 1977), p. 9.Google ScholarThe “clique's followers in Kiangsi” are accused of “appropriating for other use the investments and equipment allotted by the state for expanding the [Kiangsi Tractor Plant] into one with an annual capacity of 10,000.” Peking Review, No. 41 (7 10 1977), pp. 40–42.Google ScholarFrom the other side come complaints that Hsüeh-hsi yu p'i-p'an, No. 8 (14 08 1976), in Survey of the People's Republic of China Magazines, No.–888, p. 27.Google Scholar

49. China's Foreign Trade (03 1978), pp. 3–4.Google Scholar

50. New York Times, 25 04 1978.

51. The projects include 10 iron and steel complexes, 8 coal mines, 10 oil and gas fields, 6 new trunk railway lines, 5 key harbours, and 30 power stations. Peking Review, No. 26 (30 06 1978), p. 7.Google Scholar

52. Issues and Studies, Vol. XIII, No. 2 (1977), p. 12,Google Scholarcited in Field, Robert, McGlynn, Kathleen and Abnett, William, “Political conflict and industrial growth in China, 1965–77” (forthcoming, Joint Economic Committee), p. 11.Google Scholar

53. Peking Review, No. 11 (17 03 1978), p. 47.Google ScholarPeking Review, No. 28 (14 07 1978), p. 13Google Scholar, states that profit turned over to the state by industrial enterprises was 17·8% higher in 1977 than in 1976, and that profit to the state from industrial, communications and transportation enterprises through 06 1978 was 65·3% higher than 01–06 1977.

54. Yü Ch'iu-li speech to Taching Conference, 4 05 1977, in SCMP, No. 6343, p. 225.Google Scholar

55. The first corporation formed was the National Petroleum Corporation. Next will come the industries within the First Ministry of Machine-Building, where a farm machinery has now been formed, to be followed later on by an automotive corporation, and so forth. Personal communication from B. Williams.

56. Current Scene, Vol. 6, No. 17 (1 10 1968)Google Scholar, in Dittmer, Lowell, Liu Shao-ch'i and the Chinese Cultural Revolution (Berkeley: University of California Press, 1974), p. 261.Google Scholar

57. E.g., efforts in Hopei Province to organize co-operation in the production of diesel engines among 230 provincial units, with distribution of intermediate and finished products co-ordinated by the provincial bureau of machine-building (Jen-min, 5 12 1972, p. 1); or the Kiangsu Tractor Plant, which “fell under the control of the ‘gang of four's’ cohorts” in 1974 and again in 1976. Peking Review, No. 41 (7 10 1977), pp. 40–42.Google Scholar

58. Speech by Yü Ch'iu-li to the Conference on Farm Mechanization, Jen-min (29 01 1978), pp. 1–2.Google Scholar

59. The function of trusts in the reorganization of the “joint public-private” sector after 1956 was principally to co-ordinate the supply of raw materials and marketing of finished products among functionally related enterprises, and to press for the adoption of rationalizing measures such as standardization of parts and upgrading of technical processes to average or advanced national levels. Kapelinskii, et al., Development of the Economy and Foreign Economic Contacts of PRC (Moscow: 1959), Joint Publications Research Service, No. 3234 (Scholarly Books Translation Series 406), 23 05 1960, pp. 52–54.Google Scholar This approach was revived in the period 1964–65. Cf. Lee, N. S., China's Industrial Bureaucracy, 1949–1973 (University of Chicago: Ph.D. dissertation, 1975), Chapter V.Google Scholar

60. Similarity of policy can in part be explained by similarity of personnel. Yü Ch'iu-li was minister of Petroleum in the early 1960s, and K'ang Shih-en was the head of the original Taching prospecting unit. Both were attacked during the Cultural Revolution, and the Taching model which they had promoted was eclipsed. In the 1970s, Yü has re-emerged as head of the State Planning Commission, and K'ang is minister of Petroleum and Chemicals. Buck, David D., “Taching: A model industrial community in the PRC,” The China Geographer, No. 7 (05 1977), p. 28.Google Scholar Yü now completely dominates economic policy pronouncements.

61. Federal Broadcast Information Service, No. 205 (25 10 1977), p. E12;Google ScholarPeking Review, No. 52 (26 12 1977), p. 30.Google Scholar

62. Jen-min (29 January 1978), p. 1. If all garden tractors ever produced are still in operation, then the stock is now roughly 300,000 [based on Central Intelligence Agency, Economic Indicators (10 1978,) forthcoming]. To increase this by 36% between 1977 and 1980 would require an additional 103,000, or on average 34,000/year. This is less than 1977 output of 74,000. Applying the same methodology to conventional tractors and to powered irrigation equipment yields similar, though less extreme, results.Google Scholar

63. This method of expressing the plan obscures the anticipated annual growth rate of this flow, but implies that at the most the growth rate will be 14%, and potentially much less than that if the 1977 supply of rolled steel to farm machinery was substantially above 1976 and 1975 (as seems probable). Comparing these figures to the planned increase in rolled steel output shows the low priority assigned to the farm machinery sector. The target for January–June 1978 over January–June 1977 was 54% [a figure which has been slightly over-fulfilled – Peking Review, No. 28 (14 07 1978), p. 14].Google Scholar Even if we take this as a maximum estimate of the planned growth for the period, and downgrade it, planned increase in total supply of rolled steel during the years 1978–80 will be 85%–150%, while the increase in the flow of rolled steel to farm machinery will be held to 50%.

64. Yü Ch'iu-li speech to Third National Conference on Farm Mechanization, Jen-min (29 01 1978), p. 1.Google Scholar

65. The publication in 12 1977 of Mao's 1966 “Letter on farm mechanization,” just prior to the Farm Mechanization Conference, set this tone as well: “The task of mechanization should be performed by the provincial, municipal and autonomous region authorities mainly through their own efforts, and the central authorities can only provide some help in materials, etc., for those areas which are deficient, but these things have to be bought with local funds when the central authorities really have reserves for sale. It won't do to start the work on the spur of the moment, with everyone stretching out his hand for help. In the absence of the necessary conditions, it would be better to postpone the matter for a few years.” Peking Review, No. 52 (26 12 1977).Google Scholar

66. This approach is sketched in Mao's Ten Major Relationships speech (25 04 1956, Part I), and appears in more detail in his Talk on the Third Five-Year Plan (6 06 1964): “In the past, the method of planning was essentially learned from the Soviet Union. … First you determine how much steel is needed, then on this basis estimate how much coal, electricity, transport force, and so on are needed. … In these last few years we have been groping our way and have found some other methods. Our policy is to take agriculture as the foundation and industry as the leading factor. Pursuant to this policy, when we map out a plan we first see what quantity of foodgrains can be produced, then estimate how much fertilizer, pesticides, machinery, iron and steel, and so on are needed.” Miscellany of Mao Tse-tung Thought (NTIS, 1974), pp. 353–55.

67. An 04 1977 author quotes Mao's 1957 On Contradictions speech (“It must be affirmed that heavy industry is the core of China's economic construction. At the same time, full attention must be paid to the development of agriculture and light industry”) and then, in his elaboration, places exclusive emphasis on heavy industry. Chung Chin, “China's road to industrialization,” Peking Review, No. 14 (1 04 1977), pp. 12–15.Google Scholar Yü Ch'iu-li's 23 10 1977 speech, in setting goals inindustry and agriculture, excludes mention of any agricultural producer goods (except farmland capital construction), and pays only lip-service to small-scale rural industry. A February 1978 article notes that “there is not a single (historical) case of a country being quite advanced in agriculture but very backward in its basic industries. … Farm mechanization is out of the question in the absence of well-developed iron and steel, petrochemicals, coal, power, chemicals and machine-building industries.” Hsing, Su, “China's industrialization: how to achieve it,” Peking Review, No. 5 (3 02 1978), pp. 11–15.Google Scholar

68. Hua, Ching, “How to speed up China's agricultural development,” Peking Review, No. 42 (20 10 1978), p. 9.Google Scholar

69. Central Intelligence Agency, Current Economic Problems and the Prospects for 1985, 9 03 1978, pp. 8–9.Google Scholar

70. Ching Hua, “How to speed up China's agricultural development,” p. 8.

71. Nathan, Andrew, “Policy oscillations in the People's Republic of China: a critique,” CO, No. 68 (12 1976).Google Scholar See also Edwin Winckler's reply in the same issue.

72. We catch a glimpse of this system at work in remarks in 1977 by Li Hsien-nien to a delegation from the United States: “Sometimes planning is a headache. The different sectors all compete with one another. For example, we have a State Planning Commission. When they have a debate there, it is extremely heated and animated, even more so perhaps than your congressional debates. Collusion and contention also exist. But it doesn't matter, because these are contradictions only among the people and once the Central Committee makes a decision all debate must cease and all must listen to and obey the decision taken.” This picture of SPC meetings is mirrored in a briefing by Ch'eng Kuo-p'ing, leading member of the Shanghai City Planning and Statistics Committee: “The SPC meets once a year to discuss overall planning for that year at the national level. … At the national level planning conference there are sometimes quarrels and debates, as many local areas want more for themselves. … So, at the planning conference matters are discussed besides the budget: labour force, organization and distribution of raw materials, etc. At the outset there is conflict and debate, but ultimately a consensus is reached.” Paul Cohen, trip notes from the Young Political Leaders Delegation, May 1977.Google Scholar