French economic growth, 1815-1938: Some remarks on the different estimates of the gross domestic product

Quantitative macro-economic history is only justified if in a first phase it is able to produce long-term statistical series of sufficient quality. This is the precondition for the second phase, interpretative, with or without econometric models. Comparing the methods used since the early 19th centtury to produce the existing series of the G.N.P. and its constituent elements for France, together with the possible consequences produced by congenital malformations of certain indicators with regards to historical truth, the author of the present article endeavours to show that quantitative history is meaningless and cannot develop ifs capacities unless its degree of rigour in no way falls short of the level set by ifs ambitions.