Sectoral Productivity and Economic Gwowth in Japan : 1970-98 (in Japanese)

1. Overview of the JIP Database To measure total factor productivity (TFP) of the Japanese economy by industry, we created a database that we are calling the Japan Industry Productivity (JIP) database. The database contains annual information on 84 sectors, including 49 non-manufacturing sectors, from 1970 to 1998. It includes detailed information on factor inputs, annual input-output tables, relatively reliable deflators, and some additional data, such as R&D stock, and Japan's trade data by trade partner, etc., at a detailed sectoral level. 2. Estimate of TFP Growth of the Japanese Economy Labor quality is not taken into consideration in estimating TFP growth rates in many previous studies. When we estimate the TFP growth rate of the Japanese economy for the period of 1970-98 by taking it into consideration, we find that the TFP growth rate in consideration of labor quality was lower than that in no consideration of it. This implies that accumulation of human capital through school education and aging after 1970 contributed to improvements in labor productivity.When we take capacity utilization into consideration, we find that we tend to underestimate TFP for the periods of recession. Still, we can say that the TFP growth rates in the 1990s were lower than those before. 3. Estimate of TFP Growth By Sector When we compare TFP growth rates in service sectors and non-service ones, we find that those in non-service sectors were relatively high before the 1990s, while TFP growth rates in service sectors were relatively low before the 1990s. However, a new trend can be found in the 1990s. It should be noted that TFP growth rates in some of service sectors where deregulation was accelerated in the 1990s were getting higher. 4. Estimate of Software Investment In National Accounts of Japan, only order-made software investment is estimated. We estimate in-house software and general application software as well as order-made software. Our estimation suggests that Japan's official statistics underestimate software investment. The huge gap in software investment between the U.S. and Japanese official statistics vanishes when we take account of in-house software and general application software.