Entry and Exit of Companies and Establishments, and Productivity at the Industry Level

Author Name KIM Young Gak  (Graduate School of Economics, Hitotsubashi University) /KWON Hyeog Ug  (Nihon University College of Economics) /FUKAO Kyoji  (Faculty Fellow, RIETI)
Creation Date/NO. May 2007 07-J-022
Download / Links


In order to search for the causes of sluggish productivity in Japan during the 1990s, in this paper we analyze factors causing increases in productivity, creating microdata at both the establishment and company levels covering almost all industries in both the manufacturing and non-manufacturing sectors.

With regard to the manufacturing sector we find a strong possibility that the reason for the slowdown in productivity growth in the 1990s was not a rapid decline in the metabolic function of industries as the zombie firm hypothesis maintains (the metabolic function in the manufacturing sector had been consistently low since the 1980s), but rather because 1) factors such as a decline in learning effects as a result of a decrease in the opening of new establishments and the growing impact of "shadow-of-death" effects brought about by establishments nearing their closure dates led to declines in internal effects (productivity improvement of individual establishments), and 2) the shift of production bases to other parts of Asia, mainly in the electrical equipment industry, brought about the closure of establishments in which productivity was high. With regard to the non-manufacturing sector, although we were only able to conduct analysis covering the short period since 1997 we found that productivity dynamics differ greatly between industries.

For most non-manufacturing industries the metabolic function has been stagnant, as demonstrated by the hugely negative redistribution effects. In the construction and transportation industries in particular there were remarkable declines in employment at highly productive large companies, and owing to their large scale this contributed to the decline in productivity growth in the non-manufacturing sector as a whole. The metabolic function was also sluggish in the electricity, gas, water, and broadcasting industries. On the other hand, in the telecommunications, retail, and wholesale industries, not only do we identify substantial positive internal effects but also major positive redistribution effects, for example many companies with low productivity in the retail and wholesale industries scaling down employment, and many highly productive companies in the telecommunications industry increasing employment.