CIVILIAN INDUSTRIAL PRODUCTION IN THE PEOPLE'S REPUBLIC OF CHINA: 1949-74 By ROBERT MICHAEL FIELD I. INTRODUCTION The index of industrial production for the People's Republic of China presented in this paper is a substantial revision of the index I prepared for the Joint Economic Committee in 19661 and brought up to date in 1971.2 The three principal changes are as follows: First, the number of commodities included in the index for the years since 1957 has been increased from 11 to 27. Although the Chinese have not resumed the systematic release of economic data, they have published or broadcast an increasing amount of information in the last 3 or 4 years. This information has made it possible to estimate the output of a larger number of commodities. In particular, coverage of the machine-building industry has improved enough to permit the calculation of a separate machinery index. While the increased coverage strengthens the index, it is well to bear in mind that the number of series is only about one-tenth of the number included in an index of Soviet industrial production prepared for the Joint Economic Committee in 1973.4 Second, the use of a new index number technique has made it possible to include physical output series for which the data are not complete. Because the coverage of the physical output data after 1957 is much worse than it is for the years 1949-57, the index for the years since 1957 had previously been calculated by the Kaplan-Moorsteen method. A shortcoming of the Kaplan-Moorsteen method is that it does not use series for which data are available only sporadically, such as the Chinese series for electric generators (which is available for 1949-58, 1964-65, and 1972) or silk cloth (which is available for 194959, 1962-63, and 1974). With the new method, all available information is used in the index. And third, the index has been restructured to facilitate comparison with a reconstruction of the official index of industrial output. The version of the index that I published in 1972 was divided into the categories of fuels, industrial materials, machinery, and light indus 1 Robert Michael Field, "Chinese Communist Industrial Production." An Economic Profile of Mainland China, Joint Economic Committee of the U.S. Congress, Washington, D.C., 1967. pp. 269-295. 2 Robert Michael Field, "Chinese Industrial Development: 1949-1970," People's Republic of China: An Economic Assessment, Joint Economic Committee of the U.S. Congress, Washington, D.C., 1972. pp. 61–85. 3 The estimates are presented in tables B-1, B-2, and B-3. 4 Rush V. Greenslade and Wade E. Robertson, "Industrial Production in the USSR," Soviet Economic Prospects for the Seventies, Joint Economic Committee of the U.S. Congress. Washington, D.C.. 1973. pp. 270-282. For a description of the technique, see appendix C. Norman M. Kaplan and Richard H. Moorsteen, Indexes of Soviet Industrial Output, Santa Monica, 1960, pp. 61-68. ( 146 ) try. Because there were no official data with which to compare these categories, their precise coverage was not important. Since then, scattered data on the gross value of industrial output for individual provinces have been compiled and used to reconstruct the gross value of industrial output for the country as a whole during the 1960's and 1970's, and the total has been divided into producer and consumer goods for at least some recent years. Thus, I have switched to the categories of machinery, other producer goods and consumer goods in order to bring the structure of my index more closely into line with that of the official index. The official index and my index both attempt to measure changes over time in the real output of factories, mines, and public utilities, and both use the same broad categories. There are, however, important differences between them. First, there are differences in coverage. The official index includes the production of military hardware and the repair of machinery and equipment, neither of which are in my index. Second, there are differences in concept. The official index is derived from data on the gross value of industrial output. The official gross value data is generally considered to be methodologically deficient because it is collected by the "factory reporting method." Under this system, each enterprise reports the gross value of its output in constant prices, net of intraenterprise transfers. Because deductions are not made for semifinished inputs purchased from other enterprises, changes in the degree of vertical integration affect the reported level of output independently from changes in the real level of production. My index is an approximation of the value added in industry. Different types of weights had to be applied to different parts of the index because data are not available in sufficient detail to calculate true value-added weights. The relative shares of the wage bill paid to workers in 1956 are used as the best approximation of value added for industry excluding handicrafts. Because neither value-added nor wage bill data are available for handicrafts, the official gross value data are used to weight the categories within handicrafts and to combine industry and handicrafts into the overall index. What impact do these conceptual differences have on the measurement of industrial output? In the 1950's, when 1952 constant prices were used to compile the official data, the official indexes are higher than mine, branch by branch, and thus appear to have a consistent upward bias. In the 1960's and 1970's, when 1957 prices were used, the indexes are quite close. It was not possible to make the comparison in as great detail after 1957 because of the smaller number of commodities for which output could be estimated and of the lack of official indexes for individual branches of industry. I have tentatively concluded, however, that the current gross value indexes are a reasonable measure of output. The two indexes should be used together, the official index with complete coverage but methodological deficiencies and my index with restricted coverage but a methodology closer to the practice of Western countries. In addition to keeping the index up to date, adding to the number of series, and improving the quality of the physical output estimates, 7 Robert Michael Field, Nicholas R. Lardy, and John Philip Emerson, A Reconstruction of the Gross Value of Industrial Output by Province in the People's Republic of China: 1949-1973. U.S. Department of Commerce, forthcoming. 8 See table B-7. 9 three areas require further research. The first is military machine building. The evidence available to me suggests that military machine building is included both in the official gross value index and in the wage bill weights that I have used in my index. I characterized the previous versions of the index as total industrial production with military hardware implicitly assumed to have grown at the same rate as civilian output. The index of military procurement presented in this volume shows that this assumption was reasonable through 1971. The decline in military procurement since 1971 makes it clear that the assumption is no longer valid. As a result, I have characterized this index as an index of civilian industrial output in which machine. building is somewhat overweighted because the value of military hardware is included in the base year. Characterizing the index as civilian calls attention to but does not solve the problem. Either the weight assigned to machine building should be reduced to make it a true index of civilian industrial output or output series for military hardware should be included. The second area requiring research is prices. Except for machine. building, 1952 prices were used to calculate the branch of industry indexes, whereas the relative shares of the wage bill in 1956 were used to aggregate them into the index for total industrial production. Thus the weight base for the index is mixed. For the machine-building industry, prices have been estimated for 1957, 1965, and 1972 as well as for 1952.10 When similar studies have been completed for other sectors of industry, the index should be calculated with 1957 prices throughout. And later, if enough information is made available on employment, the wage bill, or the gross value of output by branch of industry, the base of the index should be shifted to some more recent year. The third area is the nature of the adjustment made in the official gross value data when the Chinese shifted from 1952 to 1957 constant prices. Beginning in 1958, the preliminary processing of agricultural products and the production of handicrafts by peasants for their own consumption were no longer treated as agricultural activities but were incorporated in the gross value of industrial output instead. The year 1957 is the only year for which industrial output was reported both in 1952 and in 1957 constant prices. The figures in 1957 prices have been universally assumed to be consistent with the figures in 1952 prices, and price indexes derived from them have been used to deflate data for later years. It is possible, however, that the figures for 1957 in 1957 Sydney H. Jammes, The Chinese Defense Burden, 1965-74, a paper in this volume. 10 CIA, Prices of Machinery and Equipment in the People's Republic of China, forthcoming. 11 Industrial output in 1957 was 78.390 million yuan in 1957 prices and 70.400 million yuan in 1952 prices (see State Statistical Bureau. Ten Great Years, Peking, 1960, p. 87). The price index for 1957 derived from these figures is as follows: prices have been adjusted for coverage as well as for price and that they are consistent with the data in 1957 prices for 1958 and later years rather than with the data in 1952 prices for 1957 and earlier. Independently calculated prices indexes are needed to clarify the nature of the adjustment and thus of the relationship between the reconstructed official index and my index. II. THE GROWTH OF INDUSTRIAL PRODUCTION Industrial production in China, as measured by the indexes presented in table 1, grew at an average annual rate of 13 percent during the years 1949-74. The experience in the 1950's was quite different from that of the 1960's and 1970's. The average annual rate of growth during the period 1949-60 was 22 percent, whereas the rate for the period 1960-74 was only 6 percent. The rates of growth for these and other selected periods between 1949 and 1974 are presented in table 2. TABLE 2.-AVERAGE ANNUAL RATES OF GROWTH FOR CHINESE INDUSTRIAL PRODUCTION, A. Economic Rehabilitation, 1949-52 During the period of economic rehabilitation, the index shows that industrial production more than doubled, growing at an average annual rate of 34 percent. This rapid rate of growth was characterized by large increases in employment, with little or no growth in the net value of fixed capital assets. The capacity damaged by the war or lost through the Soviet removal of equipment from Manchuria in 1945 was repaired or replaced and put back into operation, and supplies of raw materials were improved. B. The First Five-Year Plan, 1953-57 During the First Five-Year Plan period, industrial production doubled again, reaching a level five times that of 1949, but the rate of |