Agricultural statistics for the Republican era can be generally divided into three main eras, (1) 1912-27, (2) 1927-31, and (3) 1931-49, according to changes in the statistical system. This paper outlines the features of each period.
In 1912, immediately following the Republican Revolution, each section of the central government established a small statistical organization. From 1913, statistical organizations were established even at the province level (in Shanxi Province, for example, there was a Statistics Office (tongjichu) for the province and statistics managers (tongji zhuren) in the prefectures, while in Hubei Province there was a statistics official (tongjigu) for the province and statistics specialists (tongji zhuren) for the prefectures). Then, in 1916, the Foreign Ministry Statistics Bureau was created to be the nation's top statistical organization, though in reality statistical operations remained scattered among the country's various government agencies and provinces.
In 1927, the Nanjing government was established, and the different central government agencies established their own offices to administer statistics. These offices took different forms. For example, the Legislature and the Railway Bureau each had a Statistics Office (tongjichu), the Business Bureau and the Military Bureau each had a Statistics Department (tongjike), while the Home Bureau had a Statistics Office (tongjishi). In 1929, the Legislature Statistics Office became the central government's primary statistical agency, but in reality it did not have the authority to supervise the statistics gathering activities of other government agencies. Moreover, while there were statistical agencies in all the sectors of the province governments, no province level office had general jurisdiction. This meant that soon after the founding of the Nanjing government the organization of statistics compilation had grown complicated and lacked coordination.
In 1931, an accounting system was introduced and a new Accounting Office was established in order to unify the administration of operating statistics, especially the financial statistics, for all government agencies. Three bureaus, Revenue, Finance, and Statistics, were established within the Accounting Office. The Statistics Bureau included five departments, Department Number One through Department Number Five (in 1946 they were reorganized into seven departments). Agricultural statistics, including the livestock industry and the fishing industry, along with the extraction (mining) and natural resource industries, were administered by Department Number Two. Statistics Offices corresponding to the new Accounting Office were also set up at the province level. The Province Statistics Offices were composed of three sections. The First Sections administered social statistics, the Second Sections administered economic statistics, and the Third Sections were responsible for statistical systems and personnel. During the same period that the new system was being established, two basic laws governing statistical operations were promulgated. These were the Republic of China Statistics Law (1932) and the Detailed Regulations on Statistics Operations (1934). In 1930, the Central Statistics Federation (chair: Chen Yu, secretary: Zhang Xinyi) was established, with the Legislature Statistics Office at its center, in order to coordinate communications among various government statistics administrators and to expedite research. The area of responsibility of the Accounting Office Statistics Bureau, based on the Republic of China Statistics Law, was broadly divided between gathering and completing operating statistics (gongwu tongji) from other governmental sections and conducting statistical surveys (population censuses and price surveys). At the beginning of the 1930s, operating statistics were classified into 40 "categories" (lei), 235 "subcategories" (gu), and 1072 "sub-subcategories" (mu) (a me corresponded to one statistical table). Items related to agriculture in this system included agriculture, foodstuffs, livestock, new cultivated land, irrigation, and forestry. In 1944, business statistics included 29 "types" (zhong), 135 subcategories, and 656 sub-subcategories, but there was no major change in the composition of agriculture-related items.
The Accounting Office Statistics Bureau published a large number of statistical materials. Some of the representative major publications (all in Chinese) included The Republic of China Statistical Yearbook, The Republic of China Statistical Summary, and Monthly Statistics. The Republic of China Statistical Yearbook (published in 1948) was the most comprehensive source, covering twenty categories of statistics (geographical environment; the political institutions; population; agriculture and forestry, livestock and water-related industries; irrigation; mining; manufacturing; commerce and prices; international trade; local government; foodstuffs; public finance; finance; transportation; education; hygiene, social affairs; and law) from the end of the 1920s through the end of 1947. The Republic of China Statistical Summary, which listed the main statistics, was published four times, in 1935, 1940, 1945, and 1947. Its contents varied, but the 1935 edition, which contained the most detailed information, included 36 areas and 330 statistical tables. Monthly Statistics was published monthly beginning in October 1931. At first, it focused on discussions of statistical methodology and on domestic and foreign statistical systems, but from around 1934 it began to include statistical data as well. For example, the May 1933 issue, a special edition on agriculture ( entitled "General Statistical Report on Agriculture in the Provinces") presented the Accounting Office Statistics Bureau's systematic macrostatistical data on agriculture. Finally, the Legislature Statistics Office put out its own Statistics Monthly.
In the 1910s, the one and only national statistical publication was the Industry and Commerce Bureau's Statistical Tables on Agriculture and Commerce (numbers 1 through 9). The agricultural statistics included in Statistical Tables on Agriculture and Commerce were 1. number of farming households, 2. area under cultivation and cultivated area per household, 3. planting and production of rice , wheat, beans, mulberries, silkworm cocoon and raw silk thread, and tea, 4. natural disasters, and 5. conditions in desolate areas. The numbers were compiled at the prefectural level and then passed on through the provinces to the Industry and Commerce Bureau (Statistics Department). As has been frequently pointed out, there is too little information regarding the means of conducting surveys for the main materials, and the number of prefectures submitting information within a single province could vary widely in different years (the southwest regions of Guangxi, Guizhou, and Sichuan are not covered), and there are many problems, such as the frequent appearance of obviously distorted figures. For this reason, existing research is of extremely limited usefulness. For example, the raw data on regions and years in D.K. Liu's "China's Agricultural Statistics" (in China Economic Scholars (1929) China's Economic Issues (in Chinese)) is usable as it has relatively few problems, but probably the safest approach is, like Perkins (1969), to average figures across a number of years.
The most important source for 1930s agricultural macro statistics is Central Agricultural Institute of Research of the Business Bureau, Report on Conditions in Agriculture. The purpose of the Report was 1. to estimate production of major crops on a nationwide basis and forecast crop conditions, and 2. to understand economic conditions in agricultural villages in all prefectures, and to find out causes of economic development and decline. The Report was based on a survey which covered over 1200 prefectures in 22 provinces, and was compiled from agricultural reports provided by several thousand contributors (6000 persons in 1935, and 6300 in 1936). It appeared once a month. The number of prefectures covered in the reports varied, but in 1933, for example, they numbered 1068, while in 1934 the number was 1154, representing around 60 percent of all prefectures (which totaled 1744 that year). Among the criteria for selection of contributors were that they were to have resided in their areas for long periods and know the area conditions well, that they be concerned with agricultural issues, and that they demonstrate competence at providing the survey information. The actual composition of the contributors was: 25 percent were primary school teachers, 22 percent were farmers, 17 percent were teachers at agricultural or business schools or otherwise working in education-related fields, and 16 percent were prefecture or town officials. Among items covered in the survey were areas planted and amounts produced for major agricultural products, including foodstuffs (rice, wheat, sorghum, corn, millet, sweet potatoes, Irish potatoes, etc.), raw materials for clothing (raw cotton, jute and hemp , etc.), agricultural products for export (peanuts, raw silk thread, tea, wool, etc.); land (prices as well as farm rents); livestock holdings; side jobs; wages and prices; and consumption. The actual method employed for estimating production was to use the area under cultivation and the composition or units planted in an average year as the base, then to take account of any changes in planting and whether crop conditions were good or bad, as reported by the contributors (in this case, area under cultivation and average yearly production were based on figures from the Accounting Office Statistics Bureau). Further, the contributors submitted survey results every quarter for prices of household goods, and once every year for land prices, taxes and public fees, farm rents, wages, and interest.
Recent research on agricultural production using long-term statistics include Kueh (1995) and Stone and Rozelle (1995). These enlarged the agricultural production statistics for 1950 through the 1970s compiled by Tang (1984) and others, and along with covering the period from the Republic of China era (the 1930s) to the post-revolutionary era (1980 through the early 1990s) they attempt to provide an improved and more complete analysis of the causes of fluctuations in agricultural production (changes from year to year and differences among provinces). For data on the Republic of China era, these works use figures directly from the Report on Conditions in Agriculture even though they express doubts as to their quality. This is because the goal of their analysis is not to discover the total scale of agricultural production but to interpret long-term patterns of agricultural growth. From information on changes in given periods and on differences among provinces they seek to estimate the effects of technological transformation, commodification, institutional and policy changes, and climatic shifts on production. Still, Kueh and Stone and Rozelle exercised a great deal of restraint in linking the different systems of production statistics from the Republic of China and the People's Republic of China. Given the limitations of the materials available, their limited objectives were realistic, but we need to investigate what other approaches would be possible if we were to reestimate agricultural production for the Republic of China era at the macro level.
Hitotsubashi University, Department of Economics
Kueh, Y. Y. (1995), Agricultural Instability in China, 1931-1991: Weather, Technology, and Institutions, Clarendon Press.
Perkins, Dwight H. (1969), Agricultural Development in China: 1368-1968, Aldine.
Stone, Bruce and Scott Rozelle (1995), The Composition of Changes in Foodcrop Production Variability in China, 1931-1985, SOAS.
Stone, Bruce and Z. Tong (1988), "Changing Patterns of Chinese Cereal Production Variability during the People's Republic Period," in J. R. Anderson and P. B. R. Hazell, eds., Variability in Grain Yields and Implications for Agricultural Research and Policy, Johns Hopkins University Press.
Tang, Anthony M. (1984), An Analytical and Empirical Investigation of Agriculture in Mainland China, 1952-80, Chung Hua Institution for Economic Research.