Regional Factor Inputs and Convergence in Japan: A macro-level analysis, 1955-2008

Using the Regional-Level Japan Industrial Productivity (R-JIP) Database, which provides data on aggregate industry value added and production factor inputs by prefecture for 1955-2008, we examined the reasons for the decline in prefectural economic inequality from the supply side. In addition, we focused on the role of capital accumulation and changes in total factor productivity (TFP) in economic convergence. We examined how the relatively rapid capital accumulation in low-income prefectures was financed and what brought about the decline in differences in TFP. The main findings of the analysis are as follows. 1) In 1955, the most important reason for prefectural labor productivity differences was differences in TFP, followed by differences in capital-labor ratios and then by differences in labor quality. Differences in capital-labor ratios and TFP declined substantially between 1955 and 2008, leading to a dramatic reduction in prefectural labor productivity differences. On the other hand, depending on the period, prefectural differences in labor quality either did not contribute to the contraction in labor productivity differences or in fact worked in the direction of increasing such differences. 2) During the high-speed growth era from 1955-1970, the main factor underlying the decline in prefectural labor productivity differences was the decline in TFP differences. On the other hand, from 1970 onward, Japan experienced a strong decline in regional differences in inputs, so that the contribution of variation in inputs to variation in output steadily dropped after 1970. 3) Migration from poorer to richer prefectures and the decline in prefectural TFP differences from 1955 to 2008 consistently contributed to the decline in per capita gross prefectural product (GPP) differences, although the contribution of the decline in prefectural TFP differences to β-convergence—for the period as a whole—was more than twice as large as the contribution of migration. On the other hand, capital accumulation actually worked in the direction of increasing prefectural inequality in the period 1955-1970, but from 1970 onward, it consistently operated in the direction of reducing inequality. 4) The accumulation of social capital, measured in relation to working hours, in post-war Japan, was concentrated in prefectures with lower per capita GPP. Given that the accumulation of social capital likely raises the efficiency of economic activity and hence has a positive effect on TFP, the emphasis on improving social infrastructure in poorer rural areas very likely contributed to the decline in prefectural TFP differences. Meanwhile, the expansion of firms with high labor productivity into rural areas and technology transfers to technologically lagging prefectures through intra-firm technology diffusion, as well as the growing agglomeration of industry in rural areas through the expansion of manufacturing in rural areas, also likely contributed greatly to the decrease in prefectural TFP differences.