Agglomeration Economies, New Industrial Clusters and Japanese Multinational Firms’ Location in East Asia

The purpose of this chapter was to offer an empirical investigation of spatial concentration, the spatial correlation between wages and market potential, and agglomeration externalities for production in Japan’s manufacturing industry across prefectures for 1980–2000. In the first part of this chapter, we measure the spatial concentration and agglomeration indices using Ellison and Glaeser (J Polit Econ 105(5):898–927, 1997), Maurel and Sedillot (Reg Sci Urban Econ 29:575–604, 1999), and Devereux et al. (Reg Sci Urban Econ 34:533–564, 2004) in Japan’s manufacturing industries for 1980–2000. Then, we found that the most agglomerated industries are traditional industries, industries with natural advantages, and some high technology industries, and the patterns of the most and least agglomerated industries reveal similar in the USA and UK. In the second part of this chapter, from the perspective of the demand side of the spatial economy, we examine the spatial correlation between wages and market potential to see whether regional demand linkages contribute to spatial agglomeration, that is, a market-potential effect. Then, we found the existence of a positive spatial correlation between wages and market potential, that is, a market-potential effect in some industries with high agglomeration over this period. In the last part of this chapter, from the perspective of the supply side of the spatial economy, we examine that agglomeration externalities on production exist in Japan’s manufacturing industry for 1985–2000, using a flexible translog production function and data on Ellison–Glaeser index. In other words, we scrutinize the existence of agglomeration effect in Japan’s manufacturing industry for this period. Then, we found the existence of a positive weak agglomeration effect for production and scale economies in Japan’s manufacturing industry over this period.

This chapter investigates the agglomeration and coagglomeration effects on production in Japan’s manufacturing industries (lifestyle-related, basic materials, and processing and assembly-type industries). We use the agglomeration and coagglomeration indices based on Ellison and Glaeser (J Polit Econ 105(5):898–927, 1997) instead of the localization and urbanization economies variables as a new approach and examine the existence and extent of agglomeration economies in Japanese manufacturing industries using plant-level four-digit SIC data and agglomeration and coagglomeration indices for 1985, 1990, 1995, and 2000. Using a flexible translog production function and cost-share equation, as suggested by Christensen et al. (Rev Econ Stat 55:29–45, 1973) and Kim (Rev Econ Stat 74:546–551, 1992), we find that positive and weak agglomeration economies in Japan’s manufacturing industry exist without the homotheticity restriction.

Japanese Foreign Direct Investment (FDI) increased sharply in the late 1980s and early 1990s, especially since the JPY appreciation following the Plaza Accord in 1985. During this time, the globalization of Japanese firms rapidly progressed. Before this, most Japanese FDI was located in the USA and European Union; Japanese FDI to East Asia proliferated after the Accord. This chapter assesses the determinants of global and regional plant locational choice using microlevel data on Japanese multinational firms located worldwide, focusing on East Asia, especially China. From the results, combined survey responses of Japanese firm managers with the empirical estimation of location choice in the Japanese electrical and electronics industry since 1986, we found that real wage, market potential, and agglomeration variables play important roles in location choices of Japanese manufacturers in East Asia and China.

In this chapter, we conduct an empirical analysis using the new economic geography (NEG) model to examine the relationship between market access (MA, Krugman market potential), supplier access (SA), industrial agglomeration, and overseas location of final and intermediate goods production sites for Japanese electrical and electronics multinational firms (MNFs) in the processing and assembly industry in East Asia from 1995 through 2009. We conducted the conditional logit estimation of location choice of Japanese electrical and electronics firms in East Asian countries after calculating the MA and SA data using bilateral trade data from 1995 to 2009. From the estimated results, we found that wage rate, infrastructure, market access as a demand factor, and supplier access as a supply factor, and vertical and horizontal agglomerations affected foreign direct investment (FDI) flows in the Japanese electrical and electronics MNFs’ final and intermediate goods production sites in East Asia during 1995–2009. Furthermore, a comparison of two periods (1995–2000 and 2001–2009) for final goods production sites shows a negative sign condition for domestic SA during 1995–2000. However, the sign condition for domestic SA was positive during 2001–2009 because the supplier access of intermediate goods increased in East Asia during this period. Sign conditions for neighboring countries’ SA during both periods were positive and the results statistically significant.

In this chapter, an empirical analysis using the new economic geography (NEG) model will be conducted to examine the relationship between market access (MA), supplier access (SA), industrial agglomeration, and overseas location of final and intermediate goods production sites for Japanese automotive multinational firms (MNFs) in the processing and assembly industry and Japanese food and beverage MNFs in the lifestyle-related-type industry in East Asia. In doing so, first, we conducted the conditional logit estimation of location choice of Japanese automotive MNFs in East Asia countries including China (mainland), Hong Kong, India, Indonesia, Korea, Malaysia, Philippines, Singapore, Taiwan, Thailand, and Vietnam after calculating the market access and supplier access data using the bilateral trade data from 1995 to 2009 in East Asia. From the estimated results, we found wage rate, infrastructure, and market access as demand factor and supplier access as supply factor and vertical and horizontal agglomerations, which show agglomeration economy, have affected FDI of final and intermediate goods production sites for Japanese MNFs in East Asia for 1995–2009. Furthermore, comparing the period 1995–2000 with the period 2001–2009 for final and intermediate goods production sites for Japanese automotive MNFs, we found that the sign condition for domestic SA for period 1995–2000 was negative, but the sign condition for domestic SA for period 2001–2009 was positive, because the supplier access to intermediate goods increases over time in East Asia for this period. In addition, the sign condition for neighboring countries SA for two periods was positive and these results were statistically significant. Second, we present an analysis of the determinants of overseas location choices of the Japanese agro-based food industry’s MNFs’ production sites in East Asia for 1985–2009 using the conditional logit model based on the NEG model. Especially, two periods—1985–1999 (weak agglomeration period) and 2000–2009 (strong agglomeration period)—were the focus of the degree of agglomeration of Japanese food MNFs in East Asia and the overseas locations of all-goods and final goods production sites. Finally, we analyzed the determinants of overseas location choices of the Japanese beverage MNFs’ production sites in East Asia over the 1995–2009 period using the conditional logit model based on the NEG model and obtained similar location factor results to those for Japanese food products.

In this chapter, we will measure the impacts of the industrial hollowing between 2005 through 2015 (without declining population society) and the declining population society on the Japanese four regional economies and clarify the effects of new industrial clusters with innovation to revitalize Japan’s economy in a declining population society (Porter, On competition, Harvard Business School Press, Cambridge, MA, 1998, Econ Dev Q 14(1):15–34, 2000), using a four-region computable general equilibrium (CGE) model. The following four scenarios are analyzed through the simulations of a four-region CGE model. First, if the industrial hollowing of Japanese manufacturing occurs between 2005 through 2015, Japan’s real GDP will decline by 7.6% relative to base year 2005. Particularly, the real GRP in Southern Kanto region will decline drastically by 9.8%, despite the fact that real GRPs in regions outside the Kanto region will decline by only 6.6%. Second, unless the Japanese government will conduct some policies in a declining population society, Japan’s economic growth will decline to negative growth of 3.5%. However, economic growth could reach a level of approximately 0.2–0.3% relative to base year 2005 if some policies by Japanese government are taken to increase productivity and raise production subsidies to create the production volume required to meet the additional demand generated by the current Abe administration’s “Japan Revitalization Strategy.” Third, the construction of new industrial clusters will be effective in preventing the exacerbation of interregional economic welfare disparities in a declining population society. The simulations conducted in this analysis suggest that construction of new industrial clusters with coagglomeration and innovation will revitalize regional economies and will lead to narrower interregional economic disparities. Fourth, in order to recover from the influence of industrial hollowing and regional disparities in a declining population society, three policy simulations together in formation of new industrial clusters were carried out in three scenarios for a subsidy policy, a commodity-grade products policy, and a productivity improvement policy with industrial agglomeration and R&D investment of FDI back flow. When evaluating the results, we found that a contribution degree of a productivity innovation policy was the biggest among three policies.

Demographic changes such as a decline in the labor force and increase in the elderly population in Japan since 2005 have not been uniform across regions. These population dynamics are issues of concern because such dynamics restrain regional industries and impede the regrowth of regional economies in Japan. Based on the degree of depopulation, we divide the time span from 2005 to 2030 into periods of mild (2005–2015) and rapid (2015–2030) depopulation. This demographic change has a significant impact on primary and food/beverage industries. The domestic supply of Japan’s primary and food/beverage industries declined between 2005 and 2015 during the period of mild depopulation. In this chapter, we evaluate the impacts of the decline in labor force, domestic market, and total-factor productivity (TFP) on the primary and food/beverage industries and regional and international economies during mild and rapid depopulation periods. As a countermeasure to the rapid decline in population and TFP in Japan’s declining population society in 2015–2030, the policies of demand-side consumption tax reduction and supply-side new agro-based food industry clusters with digital technology are proposed, and their economic effects are empirically analyzed using the four-region computable general equilibrium (4SCGE) model. We found that the long-term sustainable economic development of primary and food/beverage industries and regional economies under rapid depopulation requires a demand-side consumption tax reduction policy to stimulate short-term domestic demand as well as a supply-side new agro-based food industry cluster policy under the digital technology transformation (DX) to expand long-term domestic production.

In the former part of this chapter, based on the premise that in addition to the aging and declining population, fiscal consolidation of both the national and local governments will be achieved by 2030, we assess the impacts of Scenario A for the development of new industrial clusters as potential sources of innovation, Scenario B for the productivity improvement of digital technology diffusion, and Scenario C for the development of new industrial clusters plus productivity improvement of digital technology diffusion. Furthermore, as the COVID-19 outbreak in 2019 had a significant negative impact on the regional economy, in the latter part, we assess the impacts of the COVID-19 pandemic on the regional economies and the industries under declining population as the base case (Scenario D0) and then implement the above three policy scenarios for 2020–2040: Scenario D1 for the productivity improvement of digital technology diffusion, Scenario D2 for the development of new industrial clusters, and Scenario D3 for the development of new industrial clusters plus productivity improvement of digital technology diffusion and analyze the economic impact of each of them as countermeasures. In order to achieve these objectives, we conduct a simulation of the ripple effect of new industrial clusters with digital technology on the regional economies in the Tokyo area and the other areas for 2020–2040, using a dynamic four-region CGE (D4SCGE) model.

Three major objectives were accomplished in this book. The first objective was to determine whether and to what extent industrial agglomeration existed in Japan’s manufacturing industry during the 1980s and 1990s, and what the economic effects of this agglomeration were on production in Japan’s manufacturing industries (lifestyle-related, basic materials, and processing and assembly-type industries) for 1985–2000. The second objective was to empirically elucidate the factors that led Japanese multinational firms (MNFs) to locate overseas in East Asia during this period. Then, with a focus on East Asia, we conducted an empirical analysis of the determinants of Japanese MNFs’ overseas location choices in East Asia for 1985–2009, employing the conditional logit model based on the new economic geography (NEG), with a particular emphasis on agglomeration, market access, and supplier access factors. The third objective was to build new industry clusters and elucidate their economic effects to overcome the negative economic impact of declining population and the COVID-19 pandemic using a dynamic four-region computable general equilibrium (D4SCGE) model.