Agglomeration and relocation: Manufacturing plant relocation in Korea

This study investigates the magnitude of localization economies by analysing the relocation pattern of Korean establishments in the manufacturing sector. Relocation of establishments is identified by their move across the border of wards, counties, or cities and distinguished between beyond and within their workers' commuting distance. It seems that relocation beyond commuting distance costs more than that within commuting distance since the former includes additional costs related with searching for, hiring, and training new workers. Key findings of this paper show that external benefits from agglomeration are large enough to be recognized by manufacturers through relocation beyond the border of their workers' commuting area. When the own industry's share of employment in all manufacturing industries becomes doubled through relocation, the probability for establishments to relocate over a long distance across the boundary of their workers' commuting area rises by 17 per cent. The results for sub‐samples divided by the age of establishments show that older establishments are more likely to relocate over a long distance to an area with disproportionate presence of establishments in the same industry. These results seem to fit product life cycle theory by Duranton and Puga. As the production process of the product becomes standardized, the firms producing that product tend to relocate to the specialized area where they can reduce the production costs by increasing dependence on the existing intermediate input suppliers, who are more likely to appear in a more specialized area.     

How can western China attract FDI? A case of Japanese investment

This paper adopts a conditional logit model to empirically examine the changes of Japanese investors’ location choices in response to China’s recent reorientation of its development strategy from the eastern coast to interior areas. This strategy adjustment is intended to attract FDI into China’s western region in hopes of promoting its economic growth and reducing economic inequality. Future policy suggestions on FDI attraction to western China are proposed based on the empirical details. This paper was the winning entry of the 19th Charles M. Tiebout Prize in Regional Science, and was presented at the 44th Annual Meeting of the Western Regional Science Association in San Diego, CA, February 2005.     

Agglomeration and networks in spatial economies

We consider the parallel developments in the economics of agglomeration and the economics of networks. We explore the complementarities between the productivity benefits of agglomeration and those of network linkages, arguing that networks of actors dispersed over space may substitute for agglomerations of actors at a single point.JEL Classification: D20, L14, R30     

Production location choice and risk aversion

This paper attempts to provide a general comparative static analysis on a firm's choice of production location with respect to variations in the degree of risk aversion under demand price, input price, and technology uncertainties. Our analysis shows that whether and how the plant location varies with a change in the firm's degree of risk aversion depend upon the nature of the production technology and how the input and location choice affect risk. It also demonstrates that some of our results are new, while some are generalizations of those obtained by Martinich and Hurter (1982). Received: January 2000/Accepted: July 2000     

Cyberspace and physical space in an urban economy

In this article I construct a general equilibrium urban model with cyberspace and physical space. In cyberspace, there is a virtual production sector and it produces virtual goods. On the other hand, in physical space, a physical production sector is located in the CBD and it produces physical goods. The behaviour of home and office workers is considered in a Solow and Vickrey type linear city. I derive the market equilibrium of the city by numerical simulation using specific utility and production functions. The land use pattern and the time allocation of the workers are shown. I also show the spatial structure of the telecommunications and transportation networks. Received: 1 November 1998 / Accepted: 20 August 1999     

Urban spatial structure,agglomeration economies,and economic growth in Barcelona: An intra‐metropolitan perspective*

This paper presents an intra‐metropolitan approach to analyse the impact of urban spatial structure on local economic growth. Focusing on the Barcelona metropolitan region (BMR) between 1986 and 2001, we estimate a municipal employment growth model in which dynamic agglomeration economies related to urban spatial structure are considered using distance to employment centres, to assess metropolitan effects, and distance‐weighted variables, to measure neighbourhood effects. The results obtained show the existence of neighbourhood specialization economies and metropolitan urbanization‐localization economies fostering local growth. All of this leads to the paper's main conclusion: urban spatial structure is important for economic growth in an intra‐metropolitan context.     

A note on the meaning of neo-classical location theory and its usefulness as a basis for applied research

This note will argue that the interpretation of neo-classical location-production models, as a possible basis for informing us about real-world phenomena, is not as straightforward as might be supposed. The reason for this is that such models contain implicit assumptions concerning the characteristics of products, the preferences of consumers, and the nature of product input-output hierarchies. Therefore, when we are discussing the behaviour of real-world spatial phenomena, it is necessary to make such assumptions explicit in order to determine the conditions under which such models can be useful analytical tools. It will be shown here that such conditions are so restrictive that these models are not even approximately applicable to most real-world cases. Received: 27 October 1997 / Accepted: 3 May 1999     

Factor distribution, capital intensity and spatial agglomeration

This paper investigates the effect of capital on agglomeration behavior within a framework of the core-periphery model. Capital is interregionally mobile in the short-run, whereas labor is mobile in the long-run. In conclusion, (i) capital and industries are distributed among regions more equally than workers when transport costs are sufficiently high, and the relation is reversed when transport costs are sufficiently low; (ii) a rise in capital intensity stimulates agglomeration in the economy.Received: July 2003/Accepted: February 2004I am indebted to Professors M. Fujita and K. Yamamoto for valuable comments on an earlier draft. I also wish to thank Professors T. Mori, T. Kinugasa, T. Kuroda, M. Makabenta and two anonymous referees for helpful suggestions.     

Agglomeration and the spatial distribution of creativity

Abstract.  This article analyses the spatial distribution of "creativity"– the production of new knowledge. We analyse commercial patents granted in Sweden between 1994 and 2001 using a panel of 100 labour market areas that encompass the entire country. We relate patent activity to measures of localisation and urbanisation, to the industrial composition and size distribution of firms, and to the regional distribution of human capital. Our analysis confirms the importance of human capital and research facilities in stimulating regional patent output. Our results document the importance of agglomeration and spatial factors in influencing creativity: patent activity is increased in larger and more dense labour markets and in regions in which a larger fraction of the labour force is employed in medium-sized firms. Our results also indicate that creativity is greater in labour markets with more diverse employment bases and in those which contain a larger share of national employment in certain industries, thus confirming the importance of urbanisation and localisation economies in stimulating creativity. Our quantitative results suggest that the urbanisation of Sweden during the 1990s had an important effect upon the aggregate level of patent activity in the country, leading to increases of up to 15 percent in aggregate patents.     

R&D competition in a spatial model with technical risk*

Abstract. In this paper, we introduce stochastic R&D into a Hotelling model with two duopolists. With an emphasis on the interaction between firms' R&D decision and location choice, we compare socially and privately optimal locations and levels of R&D. We show that, depending on R&D cost and the degree of technical risk, equilibrium location varies from minimum differentiation to maximum differentiation. From a welfare perspective, the equilibrium distance between the two firms could be either closer, farther or socially best. Moreover, depending on cost structure, firms' R&D expenditures could be either higher, lower or socially optimal, from a welfare viewpoint.     

A multiple gap approach to spatial economics

In previous work investment and regional location have been analysed in terms of gaps or thresholds that should be closed or passed in order to produce operational results (growth, local development).  The topic is taken up again using a min-algebra which, at least to the knowledge of the author, has not yet be applied to (theoretical) spatial economics, though max-plus algebra has been used to tackle scheduling problems.  Examples of multiple gap situations will be given, together with their econometric implications; min-algebra will be defined, and examples of its applications worked out; topics for further research will be presented in conclusion. Received: October 2000/Accepted: September 2001     

Spent nuclear fuel and residential property values: the influence of proximity, visual cues and public information

This article examines whether public knowledge of spent fuel storage at nuclear power plants, and any local adverse risk perceptions that may have occurred, affect the sale price of single-family residential properties. We present evidence from the Rancho Seco, California, plant on residential property values using an hedonic modeling framework. We include a large number of control variables, data with a high level of spatial detail and a number of public information variables in order to model property market effects within a fifteen mile radius of the plant. Our findings indicate that proximity and visual reminders of the plant have some influence on local property markets, and that there is a small media coverage effect on single-family home sale prices. Received: 6 May 1998 / Accepted: 26 May 1999     

Entrepreneurship and spatial externalities: Theory and measurement

The paper presents an empirical analysis on the role played by urbanisation and localisation economies on factor productivity of firms. A vast literature exists on this issue, conceptually presenting reasons supporting either industry size or city size as sources of external advantages. In general, the empirical analyses are based on the estimates of aggregate city or industry production functions; the limited hypotheses characterising these studies have suggested to test another methodology, based on the estimate of a production function at the firm level, and calculate how factor productivity changes according to different degrees of urbanisation and localisation economies. The methodology is applied to firms chosen in the high-tech sector, which demonstrates a high spatial concentration in particular areas of the Metropolitan Area of Milan. The result is that factor productivity is influenced by both urbanisation and localisation economies, but the latter show an increasing positive effect on factor productivity. Moreover, the size of firms plays an important role in defining the impact of urbanisation and localisation economies on firms' outcome.     

Networks and firm location

This paper models the decision of vertically linked firms to build either partitioned or connected networks of supply of an intermediate good. In each case, there is a correlation between the locations of upstream and downstream firms. Input specificity is related to both variable costs (transport costs of the input) and fixed costs (learning costs of the use of the input). When both are low, a connected network emerges, whereas, in the opposite case, we find a partitioned pattern. In the boundary region, there are multiple equilibria, either asymmetric (mixed network) or symmetric.     

Cournot oligopolistic competition in spatially separated markets: The Stackelberg equilibrium

Consider a two-stage non-cooperative Cournot game with location choice involving r firms. There are n spatially separated markets located at the vertices of a network. Each firm, first selects the location of a facility and then selects the quantities to supply to the markets in order to maximize its profit. Non-zero conjectural variation at the second stage in the model by Sarkar et al. (1997) is studied. When the demand in each market is sufficiently large, equilibrium in the quantities offered by each firm in the markets exists. Furthermore, each firm chooses to locate its facility at the vertices.Partially financed by Ministerio de Ciencia y Tecnología (Spain) and FEDER, grant BFM2002-04525-C02-01.Received: August 2002 / Accepted: May 2003     

Spatiotemporal analysis of the distribution of urban facilities in terms of accessibility

Abstract.  This article develops a method for analysing the distribution of urban facilities in terms of accessibility. The method treats facility accessibility in the spatiotemporal dimension, as it depends on both spatial and temporal factors, that is, the location and opening hours of facilities. The distribution of facility accessibility is evaluated by several measures indicating the extent of concentration with respect to accessibility in the spatiotemporal dimension. The measures are standardised and evaluated in a statistical framework by means of comparison with spatial, temporal and spatiotemporal random distributions of facility accessibility. The method is applied to analysis of the spatiotemporal distribution of coffee shops in Tokyo, Japan. The results revealed interesting patterns, which are helpful for analysis of marketing strategies of coffee shop chains.     

Production-location decision under duopoly with managerial incentives

The validity of the profit-maximizing assumption has long been doubted by many economists. One reason for the deviation from profit maximization that has been emphasized is the separation of ownership and management. This paper attempts to examine the spatial consequences of this separation under duopoly where managers compete in quantities, as in the Cournot model, and owners choose their managers' incentives and plant locations. A complete analysis, including the exclusion theorem, comparisons of optimal locations under the incentive equilibrium with those under profit-maximization, and comparative statics, is provided. It is demonstrated that the separation of ownership and management has significant implications for firms in relation to their location decisions. Received: October 2000/Accepted: August 2001     

Structural and spatial characteristics of personal actor networks: The case of industries for the elderly in Korea*

This paper studies the relationship between geography and dynamics of network in the evolutionary process of a newly emerging industry. Core actors are well connected with one another through policy‐related project networks, as well as temporary clusters, such as trade fairs, conventions, and seminars. Although geographical proximity is not significantly related to the occurrence of co‐operation between individuals, position of an actor in the network is significantly related to his/her geographical reach. Despite the importance of other industrial network in the Capital Region, actors in non‐Capital Regions played important positions in the networks and assumed key roles in institutional‐building processes. Resumen. Este artículo estudia la relación entre la geografía y las dinámicas de red en el proceso evolutivo de una nueva industria emergente. Los actores esenciales están bien conectados entre sí mediante redes de proyectos relacionados con políticas, así como por conglomerados temporales como ferias de exposiciones, convenciones y seminarios. Aunque no hay una relación significativa entre la proximidad geográfica y la aparición de una cooperación entre individuos, la posición de un actor en la red está relacionada significativamente con su alcance geográfico. A pesar de la importancia de otras redes industriales en la región capital, los actores del resto de regiones ocuparon posiciones importantes en las redes y asumieron papeles clave en procesos de desarrollo institucional.      

An equation by any other name is still the same: on spatial econometrics and spatial statistics

Statistics is a branch of mathematics concerned with the collection, quantification, analysis, interpretation, and presentation of real-world data, and the use of probability theory to estimate population parameters with these data. Spatial statistics is a subset of statistics that is concerned with handling the special problems associated with geographically distributed data, which include spatial point patterns, regional and lattice measurement aggregations, irregularly spaced site-specific measurements on a surface, and image analysis. Meanwhile, econometrics is concerned with the application of statistical methods to the study of economic data and problems. When coining the term spatial econometrics in 1979, Paelinck and Klaassen characterized it as a subset of econometrics that is concerned with the role of spatial dependence in regional economic model response and explanatory variables, asymmetries in spatial relationships, the specification of geographic structure governing spatial interactions, and the explicit modeling of space. We outline and discuss principal similarities (e.g., testing for the presence of spatial autocorrelation) and differences (e.g., map generalization) between spatial statistics and spatial econometrics. In doing so, our goal is to help clarify past, present, and future relationships between these two subfields.     

Potentialised partial differential equations in spatial economics: Some further results on the potentialising function

This study is a follow-up to some recent publications on potentialised partial differential equations in the context of theoretical spatial economics. In this paper potentialised wave and diffusion equations are considered with emphasis on the relation between the resulting process and the potentialising parameter. Empirical relevance in spatial (regional and urban) economics is discussed. Received: September 1998/Accepted: August 2000