Right after the General Conference of the COE last year, I went to China for two weeks with Prof. Luo Huan-zhen (Nihon University), and Dr. Ma De-bin (Hitotsubashi University) to collect Republican-era agricultural statistics and to meet with experts of related issues. On this trip, most of our time was devoted to visiting universities, research institutes, and libraries in Nanjing and Shanghai. In the first half of this paper, I will summarize the process by which we collected various documents (concurrently reporting on the conditions of Chinese libraries), and in the second half, assess the agricultural statistics compiled.
The first place we visited was the library of the Institute of Economic Research at the Beijing Academy of Social Sciences. Whenever I go to China in connection with the COE project, I always visit this Institute. The usual pattern is that we first request copies of the documents needed, then return to pick them up just before returning to Japan. In fact, over the last few years, I have used this library more frequently than the National Diet Library of Japan. The process of searching for materials has become more efficient with each visit. Part of the incentive for looking up sources quickly is that if one does not find the documents quickly, one might end up at lunch time (from around 11AM to 2PM), the result being a waste of time waiting for lunch break to end.
The second place we visited was Nanjing. Once in Nanjing, we went to the library of the Nanjing Agriculture University's Economics and Trade Institute. Last year, Prof. Luo had discovered what looked like a draft of a survey report related to John Lossing Buck's Land Utilization in China; thus, one of the main objectives was to assess this particular document. We concluded that the document was a penultimate draft or worksheet for the published version of Land Utilization in China, Vol. 3 (Statistics). Buck's draft was sitting out on the shelves covered in dust, and it was not in a very good state of preservation. The room itself was closer to a storage room than an archive, and was usually locked up, meaning there was little or no indication that anyone had used the room before. Since we were able to enter a room that was rarely opened at best, we looked at some of the other documents stored in the room. As there was no index, we had to examine each document to assess what they contained. While we were looking through the documents at random, we accidentally came across the Sichuan Province Crop Reports and some other documents. Since they were mimeographs, they had deteriorated considerably. An archive in Japan would not have permitted the copying of documents in such condition. However, due to Prof. Luo's skillful negotiation, we were able to obtain copies. Some parts are illegible, but nevertheless, the reports were valuable finds, helping us fill the paucity of documents for the 1930s and 1940s. During the March visit, Dr. Ma found the Sichuan Province Crop Reports and Fujian Province Crop Reports for 1938-1942 in good condition. Further searches may turn up similar sources.
The building adjoining the Economics and Trade Institute is the China Agricultural Heritage Research Center, one of the major centers of research on Chinese agricultural history. The academic association's journal editorial office is located within this building. We were able to meet with some young scholars. When we discussed the documents we found at the Economics and Trade Institute, the young scholars did not know anything about it. We encountered first hand the insufficient document organization, accessibility, and usage in China.
The next day, we visited the Number Two National Archives at Nanjing (Government Archives). This archive holds many Republican period policy documents, so we were expecting some archival finds. I will describe some of the findings in more detail later, but considering that it was a brief visit, the results were acceptable. In general, archives in China are very difficult to enter. I had to fax in my university identification before I visited China in order to obtain approval to enter. Moreover, the attitude and service of the staff there left an extremely unpleasant impression. Admittedly, a stay of just a few days was not enough to dig out the relevant sources from the massive collection housed there. A few months at least would be required to generate satisfactory results. We happened to meet another researcher from Japan at the Number Two National Archives. An exchange of name cards indicated that he was Prof. Kishi Toshihiko of Shimane International College. When we explained our research objectives, we discovered that Prof. Kishi had accessed the COE home page through the Internet, and knew about the research projects of the China Group of the COE. We ate lunch together, then went our separate ways on that day, but when we again met by coincidence at the Shanghai Library several days later, we were both surprised.
Our next destination was Zhejiang University at Hangzhou. This was Prof. Luo's alma mater, and a number of his former classmates gathered to welcome us warmly. We were also able to meet with two Republican period specialists. According to economic historian Prof. Jin Pu-sen, the kinds of documents we were looking for could also be found in Chongqing and Chengdu. We heard from Liu De-qiang that he had heard the same thing from Yue Wei, formerly of the State Statistical Bureau. However, when we conducted interviews in Shanghai and Beijing, we heard contradicting opinions that the documents which had been at Chongqing were returned to Nanjing after the Sino-Japanese War, so there were generally no documents left at Chongqing. Ultimately, we will have to conduct research ourselves to determine the situation. Another professor, Guo Shi-you, a specialist in political history, looked at my card, and mentioned that he knew Nakamura Tadashi (professor emeritus of Tokyo Gakugei University). If things had stopped there, this would not have been particularly notable, but by sheer coincidence, we ran into Prof. Nakamura at Narita Airport when we landed back in Japan!
We left Hangzhou and headed toward Shanghai, where we joined up with Dr. Ma who had arrived in Shanghai before us, and visited the Shanghai Library and Shanghai Academy of Social Sciences. In Shanghai, as might be expected, not only were the buildings of the library brand new, but the user system and service were very modern, with no difference from the standards of Japan's libraries. At the Academy of Social Sciences, we met with the co-author of A History of the Modern Manufacturing Industry in Shanghai, Research Associate Huang Han-min, and asked about industrial statistics. The only disappointing point was that the library at the Academy was being renovated when we visited, so we could not look at its holdings.
We returned to Nanjing to pick up the copies we had ordered previously at the Number Two National Archives. But here, we ran into some difficulties. The copies were too dark to make out the letters. We asked for the copies to be done again, but there seem to be no likelihood of this occurring. We were thus forced to copy by hand. We managed to finish before noon, but if the process had gone into the afternoon, we would have been forced to cancel planned appointments. In the afternoon, we visited the Nanjing Library Annex where a Republican period collection is housed. There is no sign at the entrance, so it is impossible to know what is inside by just looking at the building. Moreover, to use it, one needs a letter of introduction: we were able to get one from the Agricultural Heritage Research Center staff. Once inside, we were shocked by the copy prices. For both internal and external researchers, the cost for one A4 sheet was 8 yuan=1 dollar.
Finally, we returned to Beijing. We were able to meet with the Director of the Agriculture Museum of China, Dr. Cao Xing-sui. Dr. Cao is an outstanding scholar, a member of the National Political Agreement Council, and as it turned out, was a student of a student of J. L. Buck. We were able to obtain valuable information that only someone with personal memories of the times could provide. Recently in China, there has been growing interest in the history of the Republican period. In this connection, we were told about a plan to gather specialists of the Republican period from Japan, China, and the US to conduct a major symposium in the year 2000. We also exchanged views on the COE project. Regarding the Agriculture Group's work, we received the advice that it would be best to focus our energies on utilizing the results of the micro-level farm village surveys.
We were able to visit with a number of specialists during our visit, but the common point among the Chinese specialists seemed to be that they all stressed the importance of the micro-level in their research. Not only Dr. Cao, but Nanjing Agriculture University's young researchers and students all voiced the same views. In this aspect at least, there seems to be a rather large difference between Japan and China.
Now, let us assess the agricultural statistics, using the latest finds as our backbone for discussion. I will proceed chronologically.
First, for 1912 to 1921, the main source which covers the period is the Statistical Tables of Agriculture and Commerce put out by the Chinese Republican Government's Industry and Commerce Department (later Agriculture and Commerce Department). This set of statistics is similar to Japan's Statistical Tables, and includes mining and manufacturing in addition to agriculture. Prof. Guan Quan of the China Group of the COE has used this data as a case study, and examined its reliability in detail in his "Estimating Chinese Industrial Production Values for the 1910s" (Discussion Paper No. D97-16). As in the case for manufacturing, there are many problems in terms of coverage and reliability in tne agricultural data.
In 1921, an agricultural survey of several localities in Jiangsu Province was conducted. The results were organized and published as Jiangsu Province Agricultural Survey Record (3 volumes). This was discovered by Prof. Luo at the Nanjing Library. It is expected that this will be of great use since the 1920s are comparatively thin in terms of sources.
Professor John Lossing Buck of the University of Nanjing conducted a survey on agricultural management from 1921-1925, covering 2,866 farming households in 17 localities in 7 provinces of northern and central China (Chinese Farm Economy, 1930). We can surmise that this survey was conducted as a test-run for the more full-scale Land Utilization in China. As an aside, Pearl Buck, the Nobel Prize winning novelist known for her novel, The Good Earth, was married to J. L. Buck (married in 1917, divorced in 1934).
The surveys recognized as the official Republican Period government statistics are the Statistical Estimates of Agriculture by Province, published by the Directorate of Statistics of the National Government in 1932. This organization compiled various reports from prefectural and town administrations, post offices, and farmers. The contents of the Reports include farming household numbers, arable soil area, the acreage devoted to principal crops and production volume for principal crops during a normal year. The area covered in the survey was extremely wide, totaling 25 provinces (1,781 localities ), including 4 northeastern provinces. These Reports are included in a Japanese publication so it can be accessed easily. However, as will be discussed later, it has been pointed out that these statistics suffer from the problem of cultivated area and under-reporting of production value by farmers to avoid taxation.
During 1929-33, Buck conducted an extensive survey of agricultural villages covering 22 Provinces, 308 Localities, and 16,786 farm households (but excluding the northeast areas). The objective of this survey was to twofold: one, to train students in researching land utilization, and second, to collect useful information on Chinese agriculture. Buck received financial support from the Rockefeller Foundation, and employed graduates of his University of Nanjing who had undergone specialized training. The results were published in 1937 as Land Utilization in China (Vol. 1 Text, Vol. 2 Atlas, Vol. 3 Statistics). Interestingly, there were two Japanese translations of this work published in the same year. Buck's Land Utilization survey, along with the Crop Reports, are the best agricultural statistical sources of the Republican period; thus, it remains a reference for examining the accuracy of the agriculture statistics published in the early years of the Communist regime.
Regarding the accuracy of Buck's survey, the serious sampling error has been pointed out from quite awhile ago. The selection of the sample localities and farm households was left to the judgments of individual investigators. As a consequence, they chose samples convenient for their purposes. Furthermore, as noted above, they were Nanjing University graduates: in that period, it was only wealthy families who could afford to send their children to university, so Buck himself honestly acknowledged that there was a bias in the sample. Researchers who generated estimates of Republican period agricultural production in the 1960s, notably T. C. Liu and K. C. Yeh, and D. Perkins, did not rate Buck's survey very highly. However, in an article published in the post-war period, Buck himself asserted that in terms of both agricultural production and consumption in the 1930s, his survey remained the most reliable.
The most useful documents for micro-level time-series agricultural statistics for the 1930s are the Crop Reports published by the National Agricultural Research Bureau. This was a collection of reports from agricultural condition observers residing throughout 20 provinces, constituting around 60% of the total localities . The report was published from January 1933 to June 1939, covering the years 1931-1938. The contents included information on various issues, including the acreage allocation for principal crops, production volume, land price, farm rent, possession of domestic animals, side line business, labor wages, prices, and consumption levels. The Crop Reports were re-edited and organized in a Japanese publication. The original documents housed in the United States Library of Congress have been microfilmed and catologued by the University of Tokyo's Institute of Oriental Cultures. We made our copies from the University of Tokyo's microfilm, but as there were some issues missing, we obtained copies of the missing issues from the Nanjing Agriculture University and the Institute of Economic Research at the Beijing Academy of Social Sciences. The complete set of copies has been stored in the library of the Institute of Economic Research, Hitotsubashi University. We hope these documents will be of use to researchers interested in Chinese agricultural development.
As mentioned earlier, we found the Sichuan Province Crop Reports issued in May 1946. Its territorial coverage was limited to Sichuan Province, but statistics for the land area devoted and production volume for principal crops were all recorded in almost exactly the same format used in the National Agricultural Research Bureau's Crop Reports. It also includes the final estimates for 1938 to 1945 and the results of the second estimates for 1946. Since primary sources for the second half of the 1930s through the 1940s are scarce due to the effects of the Sino-Japanese War and the civil war in China, hopefully, we will be able to extract new and useful information from these materials.
Many documents related to the mid-1940s Reports are housed at the Number Two National Archives. For example, of the documents entitled xx Crop Reports ("xx" denotes geographic area name), we found a 1944 annual for Gwangxi Province, and a 1945 annual for Zhejiang Province and others amounting to 11 provinces in total. In addition, we confirmed the existence of materials entitled xx Agricultural Q & A Report, xx Agricultural Statistical Tables, and xx Agriculture. We were able to check the contents of parts of these materials. From the preliminary reading, these look to be original documents from each local region. The format of these reports do not follow that of the Crop Reports; therefore, it seemed to us that these document will require considerable time to decipher.
I have introduced above the agricultural statistics which are currently available for use. In the following section, I would like to outline in general terms the progress of the estimates of agricultural outputs.
First, we are using the following four items as basic sources: the Statistical Tables of Agriculture and Commerce for the 1910s; the Statistical Estimates of Agriculture by Province; Buck's Land Utilization statistics; and the National Agricultural Research Bureau's Crop Reports. Among the working group members, Prof. Luo is responsible for the Statistical Tables of Agriculture and Commerce, Dr. Ma for the Crop Reports, and I for the rest of the sources.
I would like to comment here on our approach to estimating the agricultural output totals for all of China. If we look at previous works, most take the following formula for calculating output totals:
i crop production volume=cultivated area x i crop's cultivated area x i crop's unit yield.
We will follow the precendents of earlier works and use this estimation formula. We will calculate the totals for each province, then add up the numbers to generate the agricultural output totals for all of China.
These outputs will be adjusted for exports and imports totals, then domestic supply totals will be obtained; this in turn will be compared to consumption totals (per capita consumption amounts in farm villages are available in Buck's survey and the Crop Reports), and population in order to measure conformity between agricultural production/ consumption totals and population estimates. I have made such calculations using Buck's surveys. I plan to present the results in a Discussion Paper.
(Tokyo Gakugei University, Faculty of Education)