region]." It is not yet clear what kind of administrative structure the regions will have, or how free they will be to plan and manage their respective economies, or even whether they will accelerate or retard the flow of resources from the more developed to the less developed areas.

In any case, the gestation period for investment in relatively undeveloped regions seems to have been much longer than the Chinese had hoped. By 1985 the investment in the Southwest probably will have begun to pay off, but the regional distribution of production is not likely to have changed significantly. The data in table 3 show how slow and difficult the process has been.

TABLE 3.-DISTRIBUTION OF CHINESE INDUSTRIAL PRODUCTION, BY REGION, 1957, 1965, 1970, AND 1977

Sources: 1957: Field, Lardy, and Emerson, 1976, p. 20. 1965, 1970, 1975, and 1977: Table C-2.

III. THE INTERPLAY OF PROVINCIAL POLITICS AND INDUSTRIAL GROWTH, 1974-77

The Wuhan Iron and Steel Company was directly affected by the "gang of four."... They ruthlessly persecuted the cadres and masses and sabotaged revolution and production. In the three years of 1974–1976 they caused the Wuhan Iron and Steel Company to lose one and one-half year's output.

9a

Under the leadership of the Kiangsi tractor plant... the masses... have overcome all difficulties and, in a week, completely resumed production, ending eight months of production stoppage caused by the "gang of four." b

Due to the interference and sabotage by the "gang of four," the [Kiangsi Tractor] plant produced some 10,000 tractors less for the state in the past three years.9°

the "gang of four" struck its nose into the Hangchow silk-printing and dyeing plant... The result was a split among the cadres, workers, and staff members, forcing the plant several times to suspend its production or operate at half capacity and inflicting great losses on the party and state.9d

The close relationship between political stability and economic performance is especially striking in China's industrial sector. During periods of relative political stability such as 1970-73, the growth of Chinese industrial output has been impressive. However, the rate of growth has fallen off dramatically in times of political infighting, such as during the struggle for succession (1974-77).10 China specialists have hypothesized that the existence of factionalism-a manifestation of political instability-during periods of intense political conflict leads to the mobilization of workers who participate in disruptive activities to discredit their faction's opponents. 10 The more disruptive the

9a FBIS, Jan. 12, 1978, H5.

b FBIS, Nov. 8, 1976, G4.

c FBIS, Dec. 2, 1976, G6.

d FBIS, Nov. 18, 1976, GI.

10 The period 1974-77 gave rise to 3 major political campaigns: The campaign to criticize Lin Piao and Confucius (late 1973-74), the campaign to criticize Teng Hsiao-p'ing (1976), and the campaign to criticize the "gang of four" (October 1976 to date).

10 For example: Field, Lardy & Emerson, pp. 16-17; and Carl Riskin, "China's Economy: At the Cross Roads Yet Again?" Contemporary China vol. 1, No. 3, pp. 2-3.

factional conflict, the more serious the damage to industrial production. This section examines in detail the nature of the relationship between political conflict and industrial growth in the 19 provinces for which GVIO data were available for each of the years 1974-77."

12

First, GVIO trends for the years 1965-77 were calculated for the 19 provinces, in most cases using the years 1965-66 and 1970-73 (which were relatively stable both politically and economically) to establish the trend. For a number of provinces, different combinations of years were chosen in order to include the most stable years.' Once the trends were tailored to each province, the GVIO growth rates for the years of political turmoil (1974 and 1976) were compared with the trend growth rates and each province assigned a score ranging from −6 for a decline in GVIO in both years to +6 for growth greater than trend in both years. Next, the provinces were ranked according to their relative political stability during the period 1974-77 and assigned scores ranging from 6 for unstable, politically factionalized provinces to +6 for politically stable provinces.13

Finally, the two scores were plotted with the political stability index on the horizontal axis and the GVIO growth rate index on the vertical axis. Figure 1 presents the results, placing the provinces in their approximate order. We found a strong correlation between political stability and industrial growth in the 19 provinces examined," corroborating the hypothesis that factional infighting leads to declines in industrial production. Correlation does not prove causality; transportation problems in neighboring provinces or national decisions on the allocation of resources certainly can have an impact on provincial growth rates. As the following quadrant-by-quadrant analysis shows, however, no explanation of provincial economic performance is complete without a careful consideration of political factors.

11 The provinces are Anhwei, Chekiang, Heilungkiang, Hopeh, Hunan, Kansu, Kiangsu, Kirin, Kwangsi, Kwangtung, Kweichow, Peking, Shanghai, Shantung, Shensi, Szechwan, Tibet, Tsinghai, and Yunnan, Data on monthly industrial output for 15 of these and 7 other provinces are presented in app. B, and annual GVIO series for all 29 provinces are presented in app. C.

12 The provinces for which different years were used to calculate the trend are Chekiang (1965, and 197073), Hopch (1965, and 1970-73), Kiangsu (1965, 1969-71, and 1973), Kirin (1965-66, and 1970-72), Kansu (1965, and 1970-73), Kweichow (1965-66, and 1973), Shantung (1965, and 1970-72), Shensi (1965-66, 1970-71, and 1973), Szechwan (1965, and 1971-73), and Tibet (1965, and 1971-73).

13 The methodology is described in app. A, below.

14 The Spearman rank correlation coefficient is .695 (which is statistically significant at the .001 level), and the Kendall rank correlation coefficient is .532 (which is also statistically significant at the .001 level). For the formulas, see Sidney Siegel, Nonparametric Statistics for the Behavioral Sciences, New York, 1956, pp. 202-223.

27-427-78-18

factional conflict, the more serious the damage to industrial production. This section examines in detail the nature of the relationship between political conflict and industrial growth in the 19 provinces for which GVIO data were available for each of the years 1974-77.“

First, GVIO trends for the years 1965-77 were calculated for the 19 provinces, in most cases using the years 1965-66 and 1970-73 (which were relatively stable both politically and economically) to establish the trend. For a number of provinces, different combinations of years were chosen in order to include the most stable years.12 Once the trends were tailored to each province, the GVIO growth rates for the years of political turmoil (1974 and 1976) were compared with the trend growth rates and each province assigned a score ranging from -6 for a decline in GVIO in both years to +6 for growth greater than trend in both years. Next, the provinces were ranked according to their relative political stability during the period 1974-77 and assigned scores ranging from -6 for unstable, politically factionalized provinces to +6 for politically stable provinces.13

Finally, the two scores were plotted with the political stability Idex on the horizontal axis and the GVIO growth rate index on the Vertical axis. Figure 1 presents the results, placing the provinces in Ler approximate order. We found a strong correlation between polit

stability and industrial growth in the 19 provinces examined," rroborating the hypothesis that factional infighting leads to declines in industrial production. Correlation does not prove causality; transrtation problems in neighboring provinces or national decisions on Te allocation of resources certainly can have an impact on provincial growth rates. As the following quadrant-by-quadrant analysis shows, Sovever, no explanation of provincial economic performance is compete without a careful consideration of political factors.

The provinces are Anhwei. Cheklang, Herungklang, Hopeh, Hunan, Kansu, Kiangsu, Kirin, Kwangsi, ng, Kweichow, Peking, Shanghai, Shantung, Shers., 976 wan, Tibet, Tsinghai, and Yunnan: monthly industrial cutput for 15 of these and 7 other provinces are presented in app. B, and annual senes for all 29 provinces are presented in app. C.

==

rovinces for which di crent years were used to calculate the trend are Chekiang (1965, and 19703 1915, and 1970-73) Kiangsu (13, 1900-71, and 1973), Kiris (1945-66, and 1970-72), Kansu (1965, 3. Kwelchow (19-66, and 1975 Shantung (1945, and 1970 72), Shensi (1965-66, 1970-71, and Szechwan (1965, and 1571-13), and Tibet (145, and 1371-73).

The method flogy is described in app. A, below.

espearman rank correlation coetert is 95 (which is statistically significant at the .001 level), and Kendall rank correlation coefficient is 2 (which is also statistically sign Scant at the .001 level), For mulas, see Sidney Siegel, Nonparametric Statistics for the Behavioral Sciences, New York, 1956, pp.

23-223.

A. Politically Unstable Provinces With Declines or Slow Growth

The provinces in quadrant III can be classified into two general categories: (a) those governed by politically moderate First (or ranking) Party Secretaries (Chekiang, Kweichow, Szechwan, and Yunnan); and (b) those led by political radicals (Anhwei, Kansu, Kirin, and Shanghai). Of the provinces in the latter category, Shanghai is regarded by both Chinese and Western analysts as having been the base of radical power in China. According to the present Third Secretary, P'eng Ch'ung, the radicals (who had been in power in Shanghai since the Cultural Revolution) made a mess of industrial production during the 1970's. P'eng recently claimed that Shanghai's GVIO declined from an impressive 14.8 percent growth rate in 1970 to a mere 2 percent rate in 1976 because of the radicals' factionalism and inept management.15

15 FBIS, Dec. 8, 1977, E-7. According to a recent traveler who was briefed by officials in Peking, Shanghai's GVIO grew only 1.2 percent in 1976.