The purpose of this paper is to investigate the relationship between production and income distribution in Japanese manufacturing after WW II. For the purpose, we adopt a structuralist Goodwin model proposed by Barbosa-Filho and Taylor (2006). Goodwin (1967), described a mechanism of business cycles from the view point of distributive conflict between capitalists and workers. Barbosa-Filho and Taylor extended the original Goodwin model. The characteristic of their model is to focus on a fluctuation of capacity utilization. In their model, capacity utilization (u) and the labor share (ψ) are determined by, (u^・ψ^・)=(φ_u θ_u φ_ψ θ_ψ)(u ψ). On the (u.ψ) plane, they draw two curves. First is an effective demand curve (u^^・=0). If it has a negative slope, it shows that demand is "profit-led". Second is a distributive curve (ψ^^・=0). If it has a positive slope, it demonstrates a "profit squeeze". Following this model, Barbosa-Filho and Taylor proceed to VAR econometric estimation for US economy. As a result, it is shown that (φ_u θ_u φ_ψ θ_ψ)=(- + - -) in USA. Therefore, in US, the effective demand curve's slope is negative, and the distributive curve's is positive. This econometric result suggests that "profit-led" effective demand and "profit squeeze" are observed in US economy. And then, there are stable counter-clockwise cycles involving capacity utilization u and labor share ψ. In this paper, we try to apply Barbosa-Filho and Taylor's analysis to Japanese manufacturing after WW II. We estimate coefficients (φ_u θ_u φ_ψ θ_ψ) in Japanese manufacturing by VAR model. As a result, it is shown that Japanese manufacturing after WW II has following structure: (φ_u θ_u φ_ψ θ_ψ)=(- - - -). φ_u<0, and, θ_ψ<0 demonstrate that this dynamic system has stability. φ_ψ<0 implies that effective demand is profit led. About above three coefficients, estimation for America and Japan are same. But values of θ_u are different. In Japanese manufacturing, θ_u is negative in contrast to American one. We think that institutional differences between America and Japan cause this result. In Japan, quantity of the input of labor is adjusted to the output fluctuation less elastically than US. As a result, the labor share tends to fluctuate counter-cyclically to business cycles in Japan. Based on the values of coefficients mentioned above, it is shown that the effective demand curve and the distributive curve have also negative slope, and that (u, ψ) sinks to the point of inter-section of two curves, in Japanese manufacturing. Accordingly, we must think that Japanese business cycles are caused not by (φ_u θ_u φ_ψ θ_ψ), but by other factors. As we have said, the structures which combine production and income distribution are different between US and Japan. But the dynamic systems of the two countries have also stability, and suggest that various institutions can exist.
View full abstract