Theoretical and Empirical Studies on Endogenous Business Cycle in Japan

过去两年，我们对经合组织国家的两部门资本密集度进行了估计。这两个部门由纯消费部门和资本品部门组成。对于我们的估计，我们使用了经合组织发布的投入产出表。我们还试图估算韩国的两部门资本密集度。然而，由于缺乏详细的就业数据，我们的估计并不完整。通过研究，我们发现资本密集度在1975年出现了逆转。换句话说，在日本的高速增长时期（1955年至1973年），资本货物部门比纯消费部门使用了更多的资本密集型技术。然而，在1975年之后，资本密集度发生了逆转，纯消费好的部门开始使用更多的资本密集型技术。此外，比较经合组织国家之间的资本密集度，我们发现日本1975年至1995年两部门资本密集度（纯消费部门的资本密集度与资本商品部门的资本密集度）的平均比率为1.1，与其他经合组织国家（平均比率为1.3）相比，这是非常低的。造成这一结果的主要原因是，日本生产了几乎所有生产所需的资本品，直到最近才进口了很少的资本品。我们还可以观察到，美国和西德各部门的资本密集度分别趋同于稳态水平，这在汇总数据中是无法观察到的。

Last two years, we have conducted estimation of two-sector capital intensities in the OECD countries. These two sectors consist of pure-consumption and capital goods sectors. For our estimation, we used the Input-Output tables published by the OECD. Also we tried to estimate the two-sector capital intensity in Korea. However, due to the lack of detailed employment data, our estimation is not completed. Through our research, we find that the capital-intensity reversal is observed in 1975. In other words, during the high-speed growth period in Japan ( from 1955 to 1973), the capital goods sector used more capital-intensive technology than the pure-consumption sector. However, after 1975, the capital intensity has been reversed and the pure-consumption good sector turns out to use more capital intensive technology. Furthermore, comparing the capital intensity among the OECD countries, we find that the average ratio of two-sector capital intensity (the pure-consumption sector's capital intensity over the capital good sector's capital intensity) from 1975 to 1995 in Japan is 1.1, which is very low when it comes to compare it with that of the other OECD countries (ithe average ratio is 1.3). The main reason for the result is that Japan has produced almost all capital goods needed for production and has imported very few capital goods until very recently. Also we could observed that each sector's capital intensity converges to a steady state level respectively in US and West Germany, which cannot be observed in the aggregated data.