Annual Report on the Japanese

Economy and Public Finance


- No Gains Without Reforms IV -

July 2004

Cabinet Office

Government of Japan

 [Toc]  [Prev]  [Next]  [Annual List

Section 2 Economic Divergence Between Regions and Causes

Regional income disparities will be reduced in the long term
    Particular attention has been given to the regional disparities in economic recovery during this economic recovery phase, but inter-regional economic disparities based on per capita income is narrowing from a long-term perspective. In terms of the degree of divergence from the national average of per capita prefectural income for each regional bloc after 1990, the incomes in Kanto, Chubu and Kinki as a whole were higher than average, while the incomes in Hokkaido, Tohoku, Chugoku, Shikoku and Kyushu were consistently lower than average (Figure 2-2-1). On the other hand, the divergence from the national average of per capita prefectural income for each regional bloc narrowed in most regions after 1990. Nonetheless, the average of per capita prefectural income has been on a downward trend since 1997, and regional disparities are decreasing as incomes are falling overall (Appended Figure 2-2). However, even though the disparities are narrowing, there is still a substantial disparity. For example, in 2001, Tokyo, which has the highest per capita income, had nearly double the per capita income of Okinawa, which has the lowest per capita income. For this reason, the following will discuss the reasons such disparities arise and the kinds of policies that are appropriate for reducing such disparities.
Figure 2-2-1 Per Capita Prefectural Income for Each Regional Bloc (Divergence from National Average)

Breakdown of factors in regional income disparities
    Regarding the factors that cause regional economic disparities, the Organisation for Economic Co-operation and Development (OECD) has suggested a method of analyzing regional disparities in per capita income by breaking them down into several components. This concept is applied here to analyze Japan's situation.(5)
    As for the framework of analysis, the total regional production per capita for each regional bloc is broken down into three components: labor productivity(6) (total regional production divided by the number of workers), employment rate(7) (number of workers divided by the total number of workers and unemployed job seekers) and labor participation rate(8) (total number of workers and unemployed job seekers divided by the population of the region). With respect to these components, (i) labor productivity is thought to represent the effectiveness of the regional production system, (ii) the employment rate is thought to indicate the status of regional labor demand, and (iii) the labor participation rate is thought to show the characteristics of the regional labor force (the total number of workers and unemployed job seekers here). These components are also influenced among other things by the situation of resource and factor endowment and policy-related factors. For example, labor productivity is significantly impacted by the industries in which a region specializes, and the conditions of such industrial specialization depend on the region's resource and factor endowment situations and naturally occurring corporate clusters, and it is also influenced by policies promoting technological innovation activities and educational policies. The employment rate and labor participation rate is influenced by the situation of factor endowment such as the regional demographic composition by age and gender. The former is affected by employment policy, while for the latter, policies influence differences in the labor participation rate of women.
    Based on this framework, the divergence of total regional production per capita by regional bloc from the national average for 2001 was analyzed in terms of labor productivity, employment rate and labor participation rate (Figure 2-2-2). The results of the calculation show that (i) labor productivity accounts for a large part of the divergence from the national average in regions other than Hokuriku and Kinki; (ii) the labor participation rate is a relatively significant factor in explaining the low level of total regional production per capita in Kyushu, Shikoku and Hokkaido; and (iii) although the contribution of the employment rate to regional disparities is not that substantial, the high employment rate in Hokuriku contributes to boosting income, and the low employment rate in Kinki contributes to curbing income.
    This analysis thus proved that regional economic disparities are largely due to differences in productivity and that the labor participation rate is a factor in curbing income in low-income regions. The following will analyze productivity and regional disparities attributed to labor-related factors in further detail.
Figure 2-2-2 Divergence of Gross Regional Domestic Product Per Capita (Logarithmic Value) from National Average and Breakdown of Causes (FY2001)

Disparity in labor productivity between regions depends on industry specialization and human capital
    First, this section will consider the differences in labor productivity, the biggest factor leading to regional economic disparities. The level of labor productivity is greatly affected by whether a region specializes in industries with high productivity or industries with low productivity. As such, the degree of industrial specialization was determined as the ratio of the number of employees in a particular industry to employment in the region as a whole. The correlation between the share of the number of employees by industry in each region and the level of productivity in the region was calculated for each industry. According to the estimates, the trend was that the higher the proportion of the number of employees in industries such as manufacturing and services, the higher the productivity for the region and the higher the proportion of the number of employees in agriculture, forestry and fisheries, construction and other industries, the lower the productivity was for the region (Table 2-2-3). This degree of industrial specialization is largely the result of a region's factor endowment or resources and naturally occurring corporate clusters. However, regional productivity is not only determined by these initial conditions and are most likely impacted by systemic factors encouraging technological innovation, the status of human capital (education level of employees) and other factors. Given this, the correlation between the proportion of the population aged 15 years or over which has completed tertiary education (junior college, technical college, university, etc.) and regional productivity was examined in order to study the relationship between regional productivity and factors besides these initial conditions. It was determined that there was a fairly strong correlation between the two (Figure 2-2-4). In this way, it can be concluded that the higher the degree of industrial specialization with high productivity and the higher the status of human capital, the higher the regional productivity.
Table 2-2-3 Industry Specialization and Regional Labor Productivity (FY2001)

Figure 2-2-4 Human Capital and Regional Labor Productivity

Regional disparities in the unemployment rate reflect the regional industrial structure and demographic composition
    Next, disparities in the employment rate and disparities in the labor participation rate will be examined. Here, the unemployment rate, as a concept almost inverse of the employment rate, is considered.(9) Section 1 underscored that employment in industries such as manufacturing and construction had been on a downward trend in the past decade or so. There were also changes in the industrial structure brought about by the substantial expansion of employment in the service industry and as a result, it was likely that the increased necessity of labor migration between industries was driving up the unemployment rate nationwide. Meanwhile, regional disparities in the unemployment rate can be largely attributed to the regional differences in industrial structure.
    Likewise with the analysis of productivity, an examination of the relationship between each region's unemployment rate and industrial specialization (divergence from the national average of the share of the number of employees by industry in the region) reveals a trend that the higher the region's degree of specialization in agriculture, forestry and fisheries, construction and manufacturing, the lower the unemployment rate. In contrast, there is no association between the degree of specialization in tertiary industries (restaurant/pubs, lodging, medical care and welfare) and the unemployment rate (Table 2-2-5). If this is explained in terms of the above-mentioned change in the industrial structure caused by the shift from manufacturing to the service industry, the unemployment rate is low in regions in which manufacturing still exists. Nevertheless, in regions with a sharply decreasing share of manufacturing, labor migration is not occurring swiftly in response to the structural change in employment, which is likely causing an increase in the unemployment rate. Additionally, given that a characteristic of the industry is that compared to tertiary industries people working in agriculture, forestry and fisheries leave their jobs infrequently, the unemployment rate tends to be low in regions with a high ratio of agriculture, forestry and fisheries to tertiary industries.
Table 2-2-5 Industry Specialization and Unemployment
    Moreover, demographic composition also influences the unemployment rate and labor participation rate. On this point, it is evident that regions, in which the ratio of young people as a proportion of the population is high, also have high unemployment rates (Figure 2-2-6). This is a reflection of the high unemployment rate among young people according to age group. Meanwhile, a correlation exists between regions with a high ratio of people aged 60 and older in the population of the region and low unemployment rate.
Figure 2-2-6 Demographic Composition, Unemployment Rate and Labor Participation Rate of Prefectures (2002)
   As for the relationship between the unemployment rate and labor participation rate, two different correlations can be predicted: (i) if the labor participation rate is high, then the labor force will increase and the unemployment rate will rise if employment is constant, and (ii) if the unemployment rate is high, then the incentive to participate in the labor force will be discouraged and the labor participation will fall. As a consequence, the relationship between the unemployment rate and labor participation rate cannot be determined by theory alone. In fact, there is a link between regions in Japan with high unemployment rates and low labor participation rates. This is strongly attributed to the fact that the incentive to participate in the labor market is discouraged due to the high unemployment rates in these regions.
Inter-regional labor migration and unemployment rates
    These regional disparities in the unemployment rates will narrow over time if there is labor migration from regions with high unemployment rates to regions with low unemployment rates. An examination of the connection between inter-regional labor migration and unemployment rate proves that as of 1990, although the correlation was weak, the trend was for the outflow of people in regions with high unemployment rates to exceed the inflow, and the inflow of people in regions with low unemployment rates to exceed the outflow. In this way, labor migration contributed to narrowing the disparities in the unemployment rates. However, the figures for 2000 indicate that inter-regional labor migration itself contracted sharply and virtually no correlation is evident between inter-regional labor migration and unemployment rate (Figure 2-2-7). According to Higuchi (2004), this decline in inter-regional labor migration is prominent among young people. This phenomenon is believed to be rooted in two trends resulting from the declining birthrate—the tendency of parents to prefer having their children nearby and the tendency for children to receive economic support from their parents.(10) Generally, when labor demand in the region falls sharply due to some kind of shock, the adjustment in the labor market comes either in the form of an increase in inter-regional labor migration or a decrease in the labor participation rate (exit from labor market). In this respect, as has already been examined, Japan's labor market is adjusted by a falling labor participation rate when the unemployment rate in the region is high. Nevertheless, if labor migration increasingly declines, these adjustments to the labor participation rate will not be enough to absorb the shock, which will then result in a growing unemployment rate.
    The policy implications derived from the analysis of regional disparities described above are as follows: (i) even though regional productivity disparities and unemployment disparities are substantially affected by the status of industrial specialization, the nature of such regional industries ultimately depends on the situation unique to each region and local initiatives; (ii) the high unemployment rate is not necessarily attributable to the advancement of the aging population; rather, addressing the employment problem among young people is the major challenges in regional employment policy; and (iii) there is a need to pay attention in the future to the decline in inter-regional labor migration possibly intensifying the unemployment problem faced by regions.
Figure 2-2-7 Correlation Between Labor Mobility and Unemployment Rates Based on Prefectural Data

View of inter-regional economic disparities
    As described above, inter-regional economic disparities still exist in Japan and income is being redistributed among regions through national and local government finances with the aim of narrowing inter-regional disparities. This has helped to reduce inter-regional economic disparities in the long term, but the extent to which inter-regional economic disparities should be narrowed bearing in mind balance with policy costs must be kept in mind. Furthermore, the kind of methods that would be effective in reducing inter-regional economic disparities must be considered.
    Simply from the perspective of economic rationale, given that people deliberately stay in low-income regions with few employment opportunities, it can be deduced that the utility derived from the attractiveness of the regional living environment and other amenities is sufficient for them. It does not necessarily mean that the utility for the residents of the community is low. Nonetheless, if any impediment to the required adjustments arises and some kind of market imperfection leads to unfavorable circumstances in the regions, it is necessary to promote policies to rectify this. These policies constitute structural reform.
    As previously discussed, the changes in the regional industrial structure and nature of employment in the past decade may have been so significant that it would be impossible to make any prediction. Given this, it would cause great economic distress to citizens, already rooted in the regions based on certain forecasts, to move to another region due to a lack of favorable employment opportunities. In addition, it is likely that other market imperfections, such as the inflexibility of the existing home market, will further drive up the cost of moving. However, having policies that help to narrow regional disparities is still meaningful. In particular, it is vital to facilitate labor migration and promote human resource development that responds to the changes in industrial structure through education and training. Moreover, with respect to inter-regional labor migration, it is necessary to facilitate migration including taking measures in the labor market.
    More specifically, however, when considering the kind of policy measures that should be used to revitalize the regional economy, it is essential to examine whether the severity of the regional economy is the result of a nationwide demand or supply shock, or caused by a wide range of factors that vary from region to region. From the discussion in this section, it has become evident that regional disparities in the long run are attributed to structural factors unique to each region, including industrial structure. In such cases, even if the same policies were implemented across the board throughout Japan, they would have almost no effect on eliminating regional disparities in the long term. In essence, the solution to problems unique to each region should be based on local initiatives and different prescriptions will be required for each region. From this perspective, the following section will examine the kinds of steps toward structural reform that have been taken in the regions and measures necessary in the future for regional revitalization.

 [Toc]  [Prev]  [Next]  [Annual List