* I am grateful for comments on an earlier draft of this paper by members of the University of Michigan Research Seminar on the Chinese Economy and to J. P. Emerson, R. M. Field, M. Oksenberg and F. Yuan for providing drafts of The Provinces of the People's Republic of China: An Economic and Political Bibliography (forthcoming).

1. Works reflecting this view include Donnithorne, Audrey, The Budget and the Plan (Canberra: Australian National University Press, 1973)Google Scholar; Durand, François, Le Financement du budget en Chine populaire (Hong Kong: Sirey, 1965)Google Scholar; Schurmann, Franz, Ideology and Organization in Communist China (Berkeley: University of California Press, 1966).Google Scholar

2. I use the word local in its broad Chinese sense of all levels of government administration below the central government.

3. Tse-tung, Mao, “On ten major relationships,” Current Background (CB), No. 892, pp. 21–34.Google Scholar

4. “On government work at concluding session of National People's Congress,” NCNA (Peking), 30 06 1956Google Scholar, CB, No. 398, pp. 9–10.Google Scholar

5. Numerous references to this debate were made by high government and Party officials during this period. For example see Liu Shao-ch'i's Eighth National Party Congress report (September 1956) and Chou En-lai's report to the National People's Congress (06 1957).Google Scholar

6. State Council, “Kuan-yü kai-chin kung-yeh kuan-li t'i-chih ti kuei-ting” (“Directive concerning improvement of the industrial management system”)Google Scholar; “Kuan-yü kai-chin shang-yeh kuan-li t'i-chih ti kuei-ting” (“Directive concerning improvement of the commercial management system”); “Kuan-yü kai-chin ts'ai-cheng kuan-li t'i-chih ti kuei-ting” (“Directive concerning improvement of the financial management system”), all in Jen-min jih-pao (People's Daily) (Jenmin), 18 11 1957Google Scholar. The full text of the directive on financial decentralization has never been released. The above citation is only a summary.

7. State Council, “Kuan-yü kai-chin shui-shou kuan-li t'i-chih ti kuei-ting” (“Directive on the improvement of the tax management system”) (9 June 1958), Chung-hua jen-min kung-ho-kuo fa-kuei hui-pien (Chung-hua), Vol. 7, pp. 265–69.Google Scholar

8. State Council, “Kuan-yü wu-chia kuan-li ch'üan-hsien ho yu kuan shang-yeh kuan-li t'i-chih ti chi-hsiang kuei-ting” (“Several regulations concerning the authority over price control and the commercial management system”) (11 April 1958), Chung-hua, Vol. 7, pp. 281–83.Google Scholar

9. State Council, “Kuan-yü kai-chin liang-shih kuan-li t'i-chih ti chi-hsiang kuei-ting” (“Several regulations concerning the improvement of the grain management system”) (4 April 1958), Chung-hua, Vol. 7, pp. 281–83.Google Scholar

10. State Council, “Kuan-yü kai-chin chi-hua kuan-li t'i-chih ti kuei-ting” (“Directive concerning the improvement of the plan management system”) (24 September 1958), Chung-hua, Vol. 8, pp. 96–99.Google Scholar

11. State Council, “Kuan-yü kai-chin wu-tzu fen-p'ei chih-tu ti chi-hsiang kuei-ting” (“Several regulations concerning the improvement of the materials distribution system”) (24 September 1958), Chung-hua, Vol. 8, pp. 100–101.Google Scholar

12. Rather than continually listing these three roughly equivalent types of administrative units I will use the term province to refer to all of them. Thus I use the terms local and provincial interchangeably to mean non-central.

13. “Chung-yang ch'i-yeh 80% hsia-fang ti-fang kuan-li” (“Central government enterprises are 80% sent down to local management”), Hsin-hua pan-yüeh k'an (Hsin-hua), No. 13 (1958), p. 63.Google Scholar

14. “An important measure to develop production forces,” Jen-min, 25 06 1958Google Scholar, in Survey of China Mainland Press (SCMP), No. 1803, pp. 3–6.Google Scholar

15. In the first Five Year Plan period, 472 out of 694 large-scale investment projects were located in interior provinces which were generally less developed than coastal areas.

16. The budgetary category of culture, education, health and welfare includes expenditures for culture, education, cadre training, physical culture, sciences, news services and radio broadcasting, public health, pensions and social relief. Thus its coverage is much broader than the U.S. category of HEW. For convenience, I will use the label HEW. This is not meant to imply an equivalence with U.S. usage.

17. The Chinese measure, “basic capital construction investment,” (chi-pen chien-she t'ou-tzu) includes expenditures for building and installation, purchase of equipment, tools and fixtures, and ancillary expenses. Functionally, it includes industry, agriculture, water conservancy, forestry, transportation, posts and telecommunications, trade and banking, culture, education and health, and urban public utilities. The western concept most closely approaching this aggregate is gross domestic capital formation. In the following discussion, the terms investment and basic investment both refer to this single concept.

18. Total investment includes that financed through local budgets, the central government budget, and investment financed outside the state budget. As far as possible aggregates in Table 6 include all three types. However in a few cases it is not possible to tell from the Chinese sources whether or not extra-budgetary investment is included in reported investment aggregates.

19. The provincial share of HEW expenditures was 74·2% in 1955, 69·6% in 1956 and 70·2% in 1957 (planned). Hsin-hua, No. 14 (1956), p. 3Google Scholar; ibid. No. 14 (1957), p. 23.

20. In August 1958, 39 higher education institutions were transferred from the Ministry of Higher Education to lower levels of administration. Kung-fei ti hsüehhsiao chiao-yü (Education in Communist China) (Taipei, 1966), p. 17Google Scholar. In 1951–56 higher education expenditures never reached 2% of the total national budgetary expenditures. See Ch'i-hsi, Feng, “Ts'ung kuo-chia yü-suan k'an wo kuo kuo-min ching-chi ti kao-chang” (“The growth of our national economy as seen from the state budget”), T'ung-chi kung-tso, No. 12 (1957), p. 32.Google Scholar

21. The coefficient of variation in annual percentage changes in provincial per capita HEW expenditures, a measure of the relative dispersion of the increases, has the following values: 1956 over 1955 (1956/55) ·49; 1957/56 7·18; 1958/57 2·00; 1959/58 ·72. These values indicate the pattern of differential increases in expenditures experienced in the post-decentralization period was well within the range of experience of the pre-1958 period.

22. All regression equations are reported in the Appendix (p. 59).

23. The level of significance used for tests is 10%.

24. This statistical result is not due to the effect of a few outlying observations. In a two-way contingency table dividing provinces along the mode on the criteria of dependency and changes in per capita HEW expenditures only one-third of the provinces lie on the diagonal predicted by the “decentralization hypothesis.”

25. Since the vast majority of government revenues are generated by industrial output or closely associated activities GVIO is the best single measure of fiscal capacity. Food-grain output is included in our tests as a proxy variable for gross value of agricultural output to improve our measure of provincial development levels.

26. This statistical result is not influenced by the choice of measuring provincial development levels on an aggregate or per capita basis or by the choice with regard to the inclusion of provincial food-grain output as an additional independent measure of the level of provincial development. The statistical results are also not the result of a few outlying observations. A two-way contingency table dividing provinces along the mode on the criteria of GVIO and change in investment share can be used to test the hypothesis that there is no association between level of development and change in investment share. The resulting test statistic x ² = ·80 does not allow us to reject the hypothesis that the two variables are independent (at the 10% level of significance).

27. Again, this result does not depend on whether provincial development is measured on a per capita or aggregate basis or whether or not food-grain output is included as an additional independent variable.

28. Tsai-hsing, Liu, “Rational distribution of industry during socialist construction,” Hsin chien-she, No. 11 (1956)Google Scholar in Selections from China Mainland Magazines (SCMM), No. 64, pp. 11–14.Google Scholar

29. Chinese use the term “coastal” to refer to the seven coastal provinces and the municipalities of Shanghai, Peking and Tientsin. Although Liaoning is one of these provinces, it enjoyed highly preferential treatment in the allocation of investment resources during the first Five Year Plan period. It was not intended to get additional investment as a result of the 1956 decision.

30. The importance of these factors is discussed in Jung-chi, Chu, “The full utilization of coastal industry,” Chi-hua ching-chi, No. 6 (1956) in SCMM, No. 50, pp. 20–23Google Scholar; “Exploit the potentiality of coastal industry,” Jen-min editorial, 8 07 1956Google Scholar in SCMP, No. 1335, pp. 5–8Google Scholar; Fu-ch'un, Li, speech to Eighth National Party Congress in Eighth National Party Congress of Communist Party of China, Vol. 2, (Peking: Foreign Languages Press, 1956), pp. 285–304Google Scholar; Tsai-hsing, Liu, “Rational distribution of industry.”Google Scholar On defence considerations, see also MacFarquhar, Roderick, Origins of the Cultural Revolution, Vol. 1: Contradictions among the People, 1956–57 (London: Oxford University Press and New York: Columbia University Press, 1974), pp. 68–74.Google Scholar

31. The following discussion is based on sources cited in Table 7.

32. Government Administrative Council, “Ch'üan kuo shui-cheng shih-shih yao-tse” (“Regulations on administration of national taxation”) (31 01 1950)Google Scholar, Chung-yang ts'ai-cheng cheng-ts'e fa-ling hui-pien, Series 1, pp. 181–84Google Scholar, in Kuo-chün, Chao, Economic Planning and Organization in Mainland China, pp. 223–26.Google Scholar

33. These were the seven local taxes, profits of local enterprises and later the deed tax. These were, in this period, all relatively minor revenue sources and, combined, probably constituted less than 10% of the national revenues.

34. Government Administrative Council, “Kuan-yü t'ung-i kuo-chia ts'ai-cheng ching-chi kung-tso ti chüeh-ting” (“Decision on the unification of state financial and economic work”) (3 06 1950)Google Scholar, Hsin-hua yüeh-pao, No. 4 (1950), pp. 1393–395Google Scholar in Kuo-chün, Chao, Economic Planning and Organization in Mainland China, pp. 202–208.Google Scholar

35. See “The question of local plans,” Chapter 10 of First Five Year Plan for Development of the National Economy of the People's Republic of China in 1953–1957 (Peking, 1956).Google Scholar

36. Li Hsien-nien's report on the 1954 final account showed the sum of agricultural and industrial and commercial taxes was 46·7% of total revenues. It is not possible to derive a comparable figure for the pre-1954 period because data on revenue by source is not sufficiently detailed.

37. Ch'eng-jui, Li, “Hsin ti ts'ai-cheng t'i-chih yu na-hsieh hao-ch'u” (“The new financial system has many good features”), Ts'ai-cheng, No. 12 (1957), p. 6.Google Scholar

38. Government Administrative Council, “Directive on the compilation of the 1954 draft budget” Shih-nien, Vol. 2, p. 82.Google Scholar

39. Yün, Lin, “Lun wo kuo 1958 nien ts'ai-cheng kuan-li t'i-chih ti kai-chin” (“A discussion of the reform of our country's financial management system in 1958”), Ching-chi yen-chiu, No. 10 (1958), p. 35.Google Scholar

40. In 1954 when the four categories of revenue were established, it was implicit in the discussion of the difference between the categories of “income shared at a fixed rate” and “central government adjustment income” that the former were to be shared at a rate that would be unchanged over a multi-year period while the latter would be subject to annual adjustment. This was made explicit in the 1955 directive on budget compilation: “In general the proportions of income shared at a fixed rate cannot be changed.” See Government Administrative Council, “Kuan-yü pien-tsao i-chiu-wu-wu nien yü-suan ts'ao-an ti chih-piao” (“Directive on the compilation of the 1955 draft budget”) Shih-nien, Vol. 2, p. 109.Google Scholar

41. The only exception was revenue from government bonds (2–3% of government revenues) which were not shared until 1956. During this period central fixed revenues were the customs tax, collected directly by the Ministry of Foreign Trade; profits of centrally managed enterprises, collected through the state banking system and remitted through the relevant central government ministry; the salt tax (except in autonomous regions) collected directly by the Ministry of Light Industry. Local effort in collecting shared revenues was insured by assigning them shares sufficient to cover their expenditures only if the overall quota for each source was met. Failure to meet the overall revenue goal, causing a shortfall in local revenues, required reduction of local expenditures in order to avoid deficits.

42. Calculated on the basis of data in Table 1 (p. 30).

43. Shanghai, for example, in 1956 and 1957 did not receive any of the revenues from the shared or adjustment categories.

44. Some people who have studied the directive have apparently assumed that the division of revenue into four categories, discussed in the directive, was a new feature of the fiscal system. In fact these precise categories had existed since 1954. The directive only changed the classification of particular revenue sources into the four categories.

45. With the exception, noted below, of Peking, Shanghai, Tientsin and Liaoning which were not to receive enterprise profits as sources of local income.

46. Actually, the November directive stipulated that the sharing rates be unchanged for three years. However, in April 1958 the State Council decreed that in order to facilitate local planning for the second Five Year Plan (1958–62) the rates would be unchanged for five years rather than three. “Kuan-yü ti-fang ts'ai-cheng ti shou-chih fan-wei, shou-ju hsiang-mu ho fen-ch'eng pi-li kai-wei chi-pen-shang ku-ting wu nien pu-pien” (“Directive concerning that the scope of local financial revenue and expenditure, revenue items, and the sharing rates to be changed to be basically fixed for five years”) (11 04 1958), Chung-hua, Vol. 7. p. 234.Google Scholar

47. Autonomous regions were specifically exempted from this provision and received favourable treatment in a number of other cases.

48. Fei-ch'ing, Hsü, “T'ung-i ling-tao fen-chi kuan-li shih kuo-chia yü-suan kuan-li kung-tso ti cheng-ch'üeh fang-chen” (“Centralized leadership and multi-level administration is the correct policy for national budgetary management work”), Ts'ai-cheng, No. 19 (1959), p. 13.Google Scholar

49. This is implied in many sources and explicitly stated in “Kuan-yü Liao-ning sheng 1957 nien ts'ai-cheng chüeh-suan ho 1958 nien ts'ai-cheng yü-suan ti paokao” (“Report on Liaoning Province 1957 financial final account and 1958 financial budget”), Liao-ning jih-pao, 3 12 1958.Google Scholar

50. My analysis of the changes introduced in 1959 is tentative since a great deal of the documentary evidence on how the system was to work in 1959 is not available. Specifically, the two volumes of the series Chung-yang ts'ai-cheng fa-kuei hui-pien (Compendium of central fiscal laws and regulations) covering 1958, which probably contain directives on budget compilation for 1959, are not available in the west. My analysis is based on the provincial financial reports that are available, the short Ministry of Finance “Notice concerning the compilation of the 1959 draft state budget” cited in Table 7 (p. 42)Google Scholar and the article on “Centralized leadership” by Hsü Fei-ch'ing in Ts'ai-cheng, No. 19.

51. I-hsing, Ma, “Kuan-yü Shang-hai shih 1958 nien chüeh-suan ho 1959 nien yü-suan ts'ao-an ti pao-kao” (“Report on Shanghai Municipality's 1958 final account and 1959 draft budget”), Chieh-fang jih-pao, 14 06 1959Google Scholar. I-hsing, Ma, “Kuan-yü Shang-hai i-chiu-wu-chiu nien chüeh-suan ho i-chiu-liu-ling nien yü-suan ts'ao-an ti pao-kao” (“Report on Shanghai's 1959 final accounts and 1960 draft budget”), Chieh-fang jih-pao, 18 05 1960.Google Scholar

52. Final accounts for Liaoning are not available after 1957, but presumably this province was affected in a manner similar to Shanghai since both were specifically denied shared and adjustment revenue by the 1957 directive.

53. These sharing rates are somewhat more accurate reflections of the Centre's power to extract resources than ars the values in Table 3 (p. 32) because by 1959 a higher proportion of all government revenues were collected by provincial governments.

54. Local collections were 79% of total national revenues in 1958 (actual) and again in 1959 (planned). Ling, Ko, “Yüeh-chin tsai yüeh-chin-tu Li Hsien-nien futsung-li kuan-yü 1958 nien kuo-chia chüeh-suan ho 1959 nien kuo-chia yü-suan ts'ao-an ti pao-kao” (“Leap forward, leap forward again – study Vice-Premier Li Hsien-nien's report on the state's 1958 final accounts and the 1959 draft state budget”), Ts'ai-cheng, No. 9 (1959), p. 20Google Scholar. In 1972 the central government collected directly 20% of total national revenues. See Robinson, Joan, Economic Management China 1972, p. 26.Google Scholar

55. Ibid. p. 27.

56. Average annual growth rates for Liaoning and Shanghai in this period were 9–10% while growth rates for Ninghsia, Kwangsi, Tsinghai and Kansu were at least half again as high as this.

57. Robinson, , pp. 24 and 28.Google Scholar

58. This point is supported by Robinson and also by information obtained in an interview with the head of the Budget Bureau in the Ministry of Finance. See Eckstein, Alexander, Scarcity and Ideology in Contemporary China: Essays in Economic Development, Chap. VI–1 (Ann Arbor: University of Michigan Press, 1975) (forthcoming).Google Scholar