Elsevier

Economics Letters

Volume 75, Issue 1, March 2002, Pages 61-67
Economics Letters

Productivity convergence in Japan’s manufacturing industries

https://doi.org/10.1016/S0165-1765(01)00582-1Get rights and content

Abstract

This paper studies the question of convergence of labor productivity in 12 manufacturing industries in Japan during 1985 to 1997, using the transition matrix for analysis. Our results indicate that manufacturing industries do not display a pattern of convergence.

Introduction

Research on the issue of convergence of income has flourished recently. Most studies have concentrated on the question of convergence of GDP per capita (e.g. Barro and Sala-i-Martin, 1992, Caselli et al., 1996). Although GDP is the aggregate value added from a variety of industries, very few studies have focused on the question of convergence of labor productivity at the disaggregated level of industries.1

The present paper studies the question of convergence of labor productivity at the disaggregated level of industries in Japan. We use the transition matrix for our analysis instead of the so-called Barro regression that is very often used in studies of convergence. The reason for using the transition matrix is that this method is more informative concerning the evolution of the distribution of labor productivity, whereas the Barro regression has some shortcomings (see Quah, 1993, pp. 426–429).

The remainder of this paper is organized as follows. In Section 2 we provide an overview of the evolution of the distribution of labor productivity in Japan’s manufacturing sector. In Section 3 we test for convergence in Japan’s 12 industries. The last section is used for concluding remarks.

Section snippets

Evolution of the distribution of labor productivity in Japan’s manufacturing sector

Our data set is the Census of Manufactures prepared by the Ministry of International Trade and Industry (MITI) of Japan (IBJ-NIKKO Information Systems, 1998). Our sample data covers 47 prefectures in Japan from 1985 to 1997. Our measure for labor productivity is value added per employee. We divide industry value added by the number of employees in the industry, taking each prefecture’s labor productivity relative to the national average.

First we check the degree of inequality of labor

Testing for convergence at the industry level

In this section we analyze the evolution of the distribution of labor productivity in 12 industries by using the transition matrix. Table 2 shows the results. Table 2 is similar to Table 1, but it averages the observed 1-year transitions over every year from 1985–1986 through 1996–1997. Then, the numbers for the sample will be 564 (=47×12) for these matrices. We apply this method so as to average the characteristics of the evolution of the distribution over the entire period from 1985 to 1997.

Conclusion

Testing for convergence in Japan’s 12 industries, our results indicate that industries have a variety of ergodic distributions of labor productivity. This confirms previous findings that manufacturing does not display a pattern of convergence.

We also found very interesting evidence that industries with relatively high labor productivity will have a bipolar distribution of labor productivity. Of course, we need to accumulate more evidence in order to argue that this phenomenon is universal, but

Acknowledgements

This study benefited from the Grant-in-Aid for Scientific Research (C) by Japan Society for the Promotion of Science and the research fund of Musashi University.

References (6)

There are more references available in the full text version of this article.

Cited by (15)

  • Sectoral productivity convergence, input-output structure and network communities in Japan

    2021, Structural Change and Economic Dynamics
    Citation Excerpt :

    On the other, service industries show great diversity in their performance. Togo (2002) studies productivity convergence across 12 manufacturing industries over the 1985–1997 period. His results indicate that manufacturing industries do not show a pattern of convergence.

  • Income convergence in Japan: A Bayesian spatial Durbin model approach

    2012, Economic Modelling
    Citation Excerpt :

    To the best of our knowledge, no research has been conducted to analyze Japan's regional income disparities at the municipality level except for our early draft, Tsutsumi et al. (2009); however, there have been several works on Japan's regional income disparities at the prefecture level (e.g., Barro and Sala-i-Martin, 1992; Kawagoe, 1999; Hashiguchi, 2010). As for the convergence analysis based on the distributional approach, see Togo (2002) and Kakamu and Fukushige (2006). This study employs the Bayesian MCMC method for the parameter estimation (e.g., Geweke, 1993; LeSage, 1997; Ertur et al., 2007; Kakamu et al., 2008), and considers both spatial dependence and spatial heterogeneity.

View all citing articles on Scopus
View full text