Productivity convergence of manufacturing industries in Japanese MEA

Abstract

The study presents (1) the convergence of labour productivity in manufacturing industries in Japanese MEAs during 1985 to 2000, using Markov chain transition matrix and (2) the speed of convergence to an ergodic distribution. The results indicate that (1) the labour productivities of all MEAs converge to different levels in the long-run and (2) its speed is as slow as the Solow's growth model.

Ackowledgements

We are grateful to Ken Togo, Keiichi Kubota and seminar participants at Musashi University for valuable comments on this paper. Financial support from the Ministry of Education, Science, Sports and Culture, Grant-in-Aids for Science Research (A) (2) 16203020 is grateful acknowledged.

Notes

¹ See Caselli et al. (1996), Quah (1993), Temple (1999) and so on.

² MEA is proposed by Kanemoto and Tokuoka (2002), which aggregate counties based on commuting patterns. This study follows the standard of 2000 MEA.

³ As the productivities in large MEAs tend to be high and those in small MEAs tend to be low, it is thought that the productivity is consistent with the scale of MEA.

⁴ In this study, the Markov chain transition matrices are the five-year transitions. So, the one iteration means the five-year transition.

⁵ The adjustment rate is defined as the rate of peak group.