This paper proposes a survey of sources and works on Vietnam's historical statistics available in the different specialised libraries and archives, mainly in Vietnam and France, and investigates the method to be implemented for estimating Vietnam's national accounts under French rule (1895-1954). This task constitutes a first step toward a quantitative appraisal of Vietnam's economic dynamics from the middle of the nineteenth century to the end of the twentieth century.
While most of the sources on the period are related to former French Indochina, our aim is to produce time series corresponding to present day Vietnam. Under French rule, Vietnam remained implicitly, for the colonial administration, a political and economic reality. The three geographical parts of Vietnam had each got a different political status : Cochinchina was a colony, Annam and Tonkin protectorates, and in consequence Annam's Emperor did not even come to possess even a nominal authority over the south. But a large part of the former Annam Empire Confucian administrative framework and manpower remained unchanged, including the colony of Cochinchina. In terms of the prevailing economic reality, Vietnam was seen as being different from Cambodia and Laos. This is exemplified by the fact that trade between Cambodia and Vietnam was monitored by the colonial administration although no custom tariff or restriction existed ( while trade between Annam or Tonkin and Cochinchina was not reported ). Public receipts and expenditures were registered with the exception of Cambodia and Laos. Furthermore part of the currency was issued in Phnom Phen, and was thus concerned solely with Cambodia, although the coins and notes were exactly the same and circulation of the currency was free within the limits of Indochina.
Reconstituting long - term time series for Vietnam's national accounts and for other economic, demographic and social indicators for the modern period, between 1858 and 1945, and up until 1954, after Vietnam's formal independence, would seem to be at first glance a relatively easy task. They are many available sources, both public and private, and part of the work of collecting the data has already been done. But, due to the weakness of the French colonial administration, especially in Annam and Tonkin, it appears that few data were collected before 1893. Estimating Vietnam's national accounts for the first colonial period, French military occupation and administration of Cochinchina and further invasion of Annam and Tonkin, would require an extensive use of indirect methods applied to fragmentary data. For this reason, we will not study this period during the first step of this research. After 1895, the French colonial administration officially took control of the whole territory of present-day Vietnam. Revolts and political unrest did not resume but the civil colonial authorities were able to find enough collaborators among the Vietnamese traditional elite to develop a comprehensive tax system, and to monitor rice and mining production, external trade, and even, to some extent, population.
(1) The collection of time series data was characterised by imprecision and lacunae which was caused by the institutional framework and from Vietnam's economic situation in the modern period. The French colonial administration ruled the country, with the assistance of the former Annam Empire administration, composed of civil servants trained in traditional schools and selected through the Confucian examination process. This is one of the main reasons why the census of population was not easily implemented. The colonial administration was de facto bilingual, with serious difficulties in communication between the French civil servants who were usually unable to speak or read the Vietnamese language, and the lower rank natives employees who performed their task as usual, and were generally unable to read French, at least not before 1910.
(2) Despite the French colonial administration's efforts to establish and publish statistics, we find few comprehensive collections of published data. If we compare them with data available on Taiwan or Korea under Japanese rule, or India under British rule, it seems that the colonial administration had almost no knowledge of the economic conditions of the country. It is true that the concern of the colonial administration focused on legal and strategic issues rather than economy. In addition, they were very few French civil servants in Vietnam. The lack of manpower trained for establishing and publishing statistical data was the main obstacle preventing the publishing of comprehensive yearbooks. This notwithstanding a tremendous amount of data was collected at the local level by lower or medium rank Vietnamese civil servants and made available and translated at the request of the top rank civil administrators. We can therefore find in different files, from the General Government archives, extremely accurate data regarding issues such as wages, prices and non-agricultural production or trade in specific provinces.
(3) In addition, the researchers are confronted with problems resulting from both the political biases and the voluntary or involuntary imperfection of the statistical system. The competition among colonial powers implies that French high rank civil servants were reluctant to publish statistical data before WWI. German territorial and commercial ambitions in East Asia concerned especially Indochina and South China, an area of French influence recognised by the other powers of the time ( UK, the US and Japan ). In addition, the failures of colonial economic development policy, especially on monetary issues, could explain that several data were kept secret by the Banque de l' Indochina and the General Government of Indochina. During the late 1920s and 1930s, the French administration became aware of Japan's ambition in East Asia and its implication for Vietnam, a context which was not propitious for increasing transparency and diffusion of data on economic performances and policy. The same situation prevailed during WWII, in the context of a unique 'gentleman's agreement' on economic co-operation between the French administration and the Japanese military authorities in Vietnam.
It seems possible, nevertheless to correct these imperfections through a systematic critique of the sources and a comparison of extant quantitative and qualitative data. These limitations and problems should be acknowledged but, on the other hand, the late 1930s were actually the last normal years, as far as economic conditions and statistical system are concerned, before the substantial improvement of the late 1980s. After 1945, due to the Indochina War ( 1945-1954 ) and later the Vietnam War ( until 1975 ), it was almost impossible to develop a system of national accounts, except during the late 1950s and early 1960s in Southern Vietnam. Post 1975 statistical data following the Soviet model are thought of as extremely unreliable, even by the Material Product System standard.
Another reason for investigating Vietnam's economic statistics under French rule is that, if we compare it with the post WWII situation, the 1920s and even the 1930s were probably a relatively golden age for the Vietnamese people. According to post WWII French official reports, the living standard in Vietnam, especially Cochinchina, was one the highest in Southeast Asia. We should remain cautious because these statements were part of the propaganda used against communism during the Cold War. The size of French resident population in Vietnam, was very small, and many of them were actually wealthy Vietnamese or Chinese who had obtained full citizenship. Thus, although unequal in income and wealth distribution, Cochinchina was certainly one the most affluent Asian societies of the time, on average much more so than contemporary Thailand or Malaysia.
If this view corresponds to reality, it could be interpreted as good news assuming that the 35 years of war ( 1941-1975 ) are, from a very long - term economic perspective, a terrible exogenous shock. Given the historical trend in human capital accumulation in Vietnam before and during French rule, recent economic growth may be considered in terms of 'catch up' given Vietnam's previous economic level. This phenomenon could be compared with China's recent performances having endured more than a century of economic recession and an absolute and relative decline in living standards. In any case, Vietnam's economic conditions and performances during the 1920s and 1930s constitute indispensable benchmarks for estimating Vietnam's present growth potential.
The task of estimating Vietnam's national accounts under French rule, during the 1895-1954 period, is currently undertaken within the framework of The Asian Historical Statistics COE Project and with the financial support of Japan's Ministry of Education, and other institutions, by a team of Vietnamese, Japanese and European researchers. The co-operative work results in the possibility of sharing expertise and improving the availability of data. We can also expect that it will contribute to further discussions with specialists from other Asian countries and therefore to a better understanding of Vietnam's economic history.
The past experience of Japanese economists in the estimation of national accounts ( Japan, but also Taiwan and Korea ) is essential for the success of the project. One of the main methodological aspects of this research on Vietnam is the use of time series methodology reporting the dynamics and economic structure of other Asian Confucian or Southeast Asian countries, mainly Taiwan, Japan and Korea, as benchmarks for performing extrapolations of Vietnam's prevalent time series ( for example on the structure of the labour force ).
On the other hand, European and Vietnamese researchers, many of them historians and anthropologists, associated with several Japanese historians, economic historians and anthropologists, have a good knowledge of the French and Vietnamese archives, the social and historical framework of the French colonial administration, the statistical system in Vietnam, and the social or anthropological environment. They can therefore evaluate the reliability of the available data and identify complementary information by for example using reports on the lifestyle of ethnic minorities or Vietnamese villagers at the turn of the Century.
In this paper, we first present a survey of the different statistical sources and previous works on these data. We then discuss various methods to be implemented for estimating Vietnam's national accounts under French rule.