Roles of Technology and Nontechnology Shocks in the Business Cycles

The empirical study of technology shocks is intensively conducted to evaluate plausibility of the technology-driven real business cycle hypothesis. A popular method is to identify technology shocks by the long-run restriction that those solely have permanent effects on labor productivity in the system consisting of labor productivity growth and hours worked. While it has an advantage of not using Solow residuals which tend to accompany measurement errors, it potentially misidentifies nontechnology shocks, which permanently affect capital-labor ratio such as a capital tax shock, as technology shocks. We show that such shock brings nonstationarity of nominal investment-output ratio and identify it through the additional restriction that it permanently affects real investment-output ratio. Data indicate that the shock works importantly in not U.S. but Japan. In the system for Japan with the shock added, hours worked responses to technology shocks become insignificant. Furthermore the technology shock loses the dominant role in Japan's lost decade. We also study an appropriate treatment of lower-frequency movements in Japan's hours worked due to inter-sectoral labor movements and working hours reductions.