Industrial Diversity and Metropolitan Unemployment Rate

Although it has for years had a lower unemployment rate than other industrialized countries, Japan has begun to see an increase in unemployment since its economy was hit by the recession of the late 90?s. The level of a nation?s unemployment is commonly seen as a barometer of its economy?s health, so that Japan?s increased unemployment has worried the government and prompted it to consider several policy options. Unemployment rate in Japan varies by region. In general, while large metropolitan areas such as Tokyo have lower unemployment rates, smaller metropolitan areas have higher unemployment. Strangely, however, Osaka, the second largest metropolitan area after Tokyo, is suffering from a high unemployment rate. In October of 2002, the Kansai region including the Osaka metropolitan area recorded an unemployment rate of 7.2%, much higher than the average rate (5.5%). Theoretically, regional differentials of the unemployment rate are attributed to the friction resulting from adjusting for the mismatch between demand and supply of labor markets among regions. These frictional factors consist of the costs of information, moving, transactions related to housing, and psychological costs. Frictional components are important factors but are not all. Industrial structure differences also affect regional differentials of the unemployment rate. This paper investigated the relationship between unemployment rate and industrial structure in metropolitan areas, with the aim of testing the hypothesis that more industrially diversified metropolitan areas have lower unemployment rates. Previous studies have been done on the relationship between industrial diversity and unemployment rate but these do not provide concrete agreement because of the failure to control other factors affecting unemployment rate. This paper follows the theoretical justification of Simon (1988), who argues that industrial diversity attains a lower unemployment rate by assuming that the frictional component of employment fluctuations is a random variable and independent across industries. Because fluctuations are uncorrelated across industries, frictional hiring in some industries may coincide with frictional layoffs at others. Unemployed individuals can fill concurrently occurring vacancies. Therefore, the unemployment rate will be lower in the more industrially diverse metropolis. Simon?s empirical analysis of 91 large U.S. SMSAs strongly supports the hypothesis. In this study, we analyze 117 metropolitan areas in Japan for the year 1995. Because there is no authoritative definition of a metropolitan area in Japan, we began by defining metropolitan areas and collected data for each. As for a variable expressing industrial diversity, the Herfindahl index is used, which is made of both numbers of employments and numbers of firms for ten industrial classifications. Other factors used in this analysis are size of metropolitan areas, transportation conditions, unemployment insurance, average schooling length, and so on. The basic equation for the empirical analysis of the relationship between industrial diversity and metropolitan unemployment rate is as follows: Metropolitan unemployment rate = f (Herfindahl index (industrial diversity), metropolitan size, transportation conditions, unemployment insurance, average schooling length in metropolis) In addition to this analysis, we also analyze whether or not a higher location quotient shows a lower unemployment rate. In the preliminary analysis, we found there are negative relationships in almost all industries.