A Note on "Long Term Changes in the National Product of Japan"

In summer 1951, ~¥'e attempted an estimate of a lon*" term series of the national net product in Japan since 1878, the result of which is to be published.2 Thls work was a preliminary one, and we conceded a large margin of error in the series estimated there, especially for the years before 1919, because our estimate for that period had to be an indirect one owing to the non~xistence of basic data which are essential to the estimation of national products. From that time onvi'ards ¥ve continued to make some improvement upon the preliminary estimate. It is the purpose of this essay to describe the re¥'ision made since that time in comparison with the first series and to explain the reason why the revision had to be made. The former estimate was approached from the viewpdint of production, nat[onal income produced being estimated by dividing all industries into three sectors, namely, primary, secondary and tertiary industries, mostly according to Colin Clark's categories. This time we confined ourselves to the figures of the primary and tertiary sectors, ¥vhile those of the secondary sector remain untouched inspite of their being obviously imperfect, because it would require too much time to arrange the data essential to the estimation of income produced in this sector. As for the wholesale price index which was used to deffate the national income series in current yen, some revision was made in the figures of early years. Since this work is in process of continuous improvement, the concluding figures ~¥'ill, of course, be left to the future. It may, to begin wit~, be convenient for readers to illustrate the various