María Callejón. María Teresa Costa EXTERNAL ECONOMIES AND THE LOCATION OF INDUSTRY - PDF

Description
European Regional Science Association 36th European Congress ETH Zurich, Switzerland August 1996 María Callejón Departament de Política Económica i EEM. Universitat de Barcelona Fax:

Please download to get full document.

View again

of 19
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information
Category:

Abstract

Publish on:

Views: 18 | Pages: 19

Extension: PDF | Download: 0

Share
Transcript
European Regional Science Association 36th European Congress ETH Zurich, Switzerland August 1996 María Callejón Departament de Política Económica i EEM. Universitat de Barcelona Fax: María Teresa Costa Departament de Estadística, Econometria i Economia Espanyola Universitat de Barcelona Tel: EXTERNAL ECONOMIES AND THE LOCATION OF INDUSTRY ABSTRACT: The purpose of this paper is to contribute to the current applied research on the role of external economies in the location of manufacturing activities. The first section of the paper reviews a part of the traditional and recent literature on the nature and typologies of externalities that may influence industrial location in the absence of any kind of barriers to the free movement of productive factors, inputs or outputs. It shows the dispersion of approaches within the relevant literature, and how those new approaches that model market structure using the paradigm of monopolistic competition may help to improve the understanding of the nature and role of agglomerative forces in production. The second part of the paper consists on the estimation of an econometric location model based on the presence of technological external economies of two types: local own industry concentration, and local productive diversity. The estimation uses employment data of a set of manufacturing sectors in fifty Spanish provinces; and the first provisional results confirm the hypothesis that external economies are relevant in the geographical location of industry. EXTERNAL ECONOMIES AND THE LOCATION OF INDUSTRY Introduction This paper explores the determinants of industrial location. Its main objective is to analyze what factors determine the persistence of a manufacturing activity in a given location. It is attempted to provide new empirical evidence along the path opened by Glaeser et al. (1992) and especially by Henderson et al. (1995). This line of analysis tries to find out which is the role played by information externalities in the spatial agglomeration of production activities, and the relative sensibility of different manufacturing sectors, according to their technological sophistication, to these externalities. For many years economic geography has been interested in those factors that determine the spatial distribution of economic activities, but what we now call geographical economics is a new approach that involves the extension of the principles and models of economic theory to the analysis of location decisions in firms. The spatial configuration of production activities could be approached as the outcome of two opposing types of forces: agglomeration or centripetal forces, and dispersion or centrifugal forces (Fujita and Thisse, 1996). If that outcome was highly stable through time, it would be relatively easy to unveil which are the main factors determining the agglomeration of economic activity and their dispersion. But the fact is that the pattern of location requirements change through time in response to such events as the development of new markets and advances in technology which alter the relative significance of individual location factors. This fact makes more difficult to identify the causal factors in the series of data available, and also complicates the attempt to built explanatory models which, as we can see in the literature, show considerable variations. The increased interest of traditional economics in the problems of spatial location can be understood taking into account a series of actual circumstances that affect the welfare of citizens and the economic policies of governments in the developed countries. We should emphasize, first of all, the great interest raised by the new growth theory and convergence models (Romer, 1989; Barro and Sala-i-Martín, 1991) whose theoretical progresses have 1 been spurred by the evidence of strong differences in the growth rates of some Asian countries and the rest of economies. In second place there is a raising concern for spatial distribution consequences of such deep integration processes like the European Union between the integrating countries (Krugman and Venables, 1993). Third, we should not forget that regional policy, and the smoothing out of income inequalities within regions in a country, has been one of the main policy targets of governments until the present. The notion of agglomeration economies, in one way or another, is at the center of all the issues mentioned above and, consequently, in the theoretical approaches that tackle with them. If we exclude those idiosyncratic aspects which generate local comparative advantages - such as natural resources, accessibility, or institutional factors - the basic element that generates agglomerative forces is, in short, the presence of increasing returns - in other words, that long-run average cost declines when production grows. On the other hand, increasing returns may originate by the presence of scale economies internal to the individual firm or external to the firm but internal to the local territory. In terms of Fujita and Thisse (1996) the modern concept of agglomeration economies is related to the following effects: (i) scale economies; (ii) localization economies; and (iii) urbanization economies. The last two effects involve the presence of external economies among firms belonging to the same sector or activity (localization economies), or among firms belonging to different sectors (urbanization economies). In what follows this paper is divided in two parts. First of all we review very concisely the concept of production externality and its evolution in the economic literature, and in the second part we present an econometric model and estimate of spatial technological external economies - MAR and Jacobs externalities - with a data base obtained from several sectors of the Spanish manufacturing industry. Externalities in production The current theoretical approach to external economies in production activities has its origins in Marshall (1923) who undertook the first systematic analysis of the kind of externalities that were later termed as localization economies. It refers to the situation where firms of the same sector of activity find advantageous to agglomerate spatially. Marshall 2 distinguished three groups of factors generators of externalities that spur the spatial concentration of firms of a given sector: The information flows relative to the specific skills and knowledge of the sector that spread easily among neighboring firms and give way to a process of accumulation of knowhow specific to that sector. Marshall was referring to what we now call knowledge spillovers or technological spillovers of a intraindustrial character. Existing agglomerations of firms of a given sector stimulates the establishment of new firms that perform complementary activities and supply specialized inputs, services, or machinery to the concentrated sector. When a productive activity reaches a local minimum critical mass, it allows to push forward the division of labor and this involves costs savings for the aggregate of local firms. This set of relationships corresponds to the traditional concept of interindustrial linkages. The concentration of firms in a given location generates a pooled market of workers with industry-specific skills. Both employers and workers benefit from advantages; the former because the pool implies a more elastic supply of labor that allows them to adjust firm employment, according to the firm s business cycle, at a reasonable cost; the latter because they reduce the risk of depending for their jobs of a single employer. This externalities that Marshall described have constituted only a starting point. Thereafter the economic literature has offered diverse interpretations and varieties of external economies and diseconomies in production. Perhaps the best well-known of them is Scitovsky s (1954) distinction between technological external economies and pecuniary external economies; distinction that did not appear in Marshall. Meade (1952) had defined technological external economies as direct (non-market) interdependence among producers that affect the production function. Expressed formally, the output (x1) of a firm depends not only on the factors of production used by the firm (l1, k1,..) but also on the output (x2) and factor utilization (l2, k2,...) of other firms: x1 = F(ll, k1,... ; x2, l2, k2...) (1) 3 In symmetry with Meade s concept, Scitovsky states that pecuniary external economies follow from the interdependencies among producers through the market mechanism that affects input prices and the profit function. Formally, it means that profits of one producer (P1) are affected by the actions of other producers: P1= G(x1, l1, k1,... ; x2, l2, k2,...) (2) This definition of external economies includes direct or non-market interdependencies among producers but is much broader. Krugman has used both Marshall s notions and Scitovsky definitions, but he has chosen to concentrate mainly on the concept of pecuniary externalities 1. We will comment in the following paragraphs on the theory of pecuniary externalities even though the empirical part of this paper deals with technological externalities given that we find interesting and useful to discuss on the delimitation of both concepts and also given that with the available empirical data both effects often appear entangled. The concept of pecuniary external economies has evolved considerably since its early formulation by Scitovsky, but the fundamental bricks were already laid down by him. Scitovsky viewed that general equilibrium theory with perfect competition is not suited to deal with non technological external economies, but pointed out that pecuniary externalities was a central concept in the theory of industrialization of underdeveloped countries which was basically a disequilibrium theory. The most important insight of the literature on development theory that flourished in the 40 s and 50 s was that the expansion of some modern branches of the manufacturing industry only can take place simultaneously to the expansion of other different branches. Relevant examples of this conception are the concept of big push of Rosenstein-Rodan (1943), the pôle de croissance of Perroux (1955), the circular and cumulative causation of Myrdal (1957), the backward and forward linkages by Hirshman (1958), or the vicious circle model by Nurske (1958). Nowadays, the analytical advances in economic theory, specially in the field of industrial organization, allows to be more explicit and rigorous about market structure in spatial models. In Krugman (1995) words:... During the 1940 s and 1950 s a core of ideas emerged regarding external economies, strategic complementarity, and economic development that remains intellectually valid and may continue to have practical applications. 4 In both development and geography the crucial problem, in particular, was the inability of the field s pioneers to be explicit about market structure - that is, about the conditions of competition in hypothetical economies they were describing. Particularly promising are those recent contributions to geographical economics which apply models of monopolistic competition to explain productive complementarities and agglomerative forces in general. Most of those contributions draw from the seminal work by Dixit and Stiglitz (1977). Of special interest to our subject are: the modern version of the model of the big push by Murphy, Schleifer and Vishny (1989) that has also been used by Krugman (1991b); the multiplier process due to monopoly price distortion by Matsuyama (1995); and the agglomeration economies due to increasing returns to variety by Fujita and Thisse (1996) and also Matsuyama (1995). Increasing returns leading to externalities can be captured in simple models of variety in intermediate goods considering that the production function of any given firm in an industry is of the type: Q = L Σx i 1 α α (3) Where the competitive firm has constant returns in homogeneous input (L) and a composite of differentiated intermediate inputs (xi). Assuming that all differentiated inputs represent the same share of the composite input (X): Q L n X 1 α 1 1 = = n L X n α α α α (4) Thus, the production function exhibits increasing returns in the number of specialized inputs, and confirms the idea that the division of labour improves productive efficiency. The composite of differentiated inputs will include also non-traded specialized inputs such as legal, financial, maintenance and communications services with basically local supply. If there were no scale economies, then productivity could rise indefinitely by using more and more varieties of differentiated inputs. The limit to the increase in variety of intermediate goods used can be easily modeled by reasonably assuming that the production function of intermediate goods exhibits increasing returns - there are fixed costs in the 5 production of each differentiated input. Consequently, the size of market (demand for the final good) and the extent of the division of labor are interdependent variables that generate external pecuniary economies. The strictly technological externalities that are involved in the agglomerations of firms operate through the channels of communication between firms that convey crossed information flows - knowledge spillovers. Information flows can be characterized as public goods in the sense that any piece of information can be used by many agents without affecting the amount of public good available to each user. As far as firms develop specific skills they provide different types of information, and thus the total quantity of information available to each firm increases with the number of firms. Also, if the quality of the communication channels is sensible to distance, firms will tend to agglomerate in order to improve the information flows. The incentive of firms to agglomerate is, however, counterbalanced by increasing congestion costs which imply that firms pay rising land rents and rising wages. As local agglomeration of firms takes place, workers confront higher housing and transportation costs. Agglomerative forces are limited by growing congestion costs. The distinction between pecuniary external economies and technological external economies that originated in development theory can be complemented by another classification that is being used by researchers in urban and regional economics. Glaeser et al (1992) distinguish between dynamic externalities - or knowledge spillovers which are equivalent to technological externalities- and static externalities - related in general to productive linkages. Glaeser et al. stress the importance of knowledge diffusion on the growth of cities and the spatial concentration of particular industries and suggest a typology of dynamic externalities between firms which distinguishes between: (i) MAR externalities (Marshall, Arrow, Romer), defined as own-sector spillovers in imperfectly competitive markets 2 ; (ii) Porter externalities or own-sector spillovers in competitive markets 3 ; and (iii) Jacob externalities or technological spillovers between firms belonging to different industrial branches 4. Jacobs (1969) interindustry externalities refer to the positive effects on the efficiency of firms derived of exposition to information flows from different local industries. According to Jacobs this cross-fertilization of ideas is more relevant than own industry spillovers. Henderson et al (1995) adopt a slightly modified version of externalities. Dynamic externalities are defined as dealing with the role of prior information accumulations in the 6 local area on current productivity. Such accumulations lead to a build-up of knowledge - local trade secrets - available to firms just in a local space. Their interest is centered in the operational and empirical utility of the concepts and they distinguish only between MAR and Jacobs dynamic externalities: MAR (localization) economies derive from a build-up of knowledge associated with ongoing communications among local firms in the same industry ; Jacobs (urbanization) economies derive from a build-up of knowledge associated with historical diversity. On the other hand Henderson et al (1995) define the static counterparts of MAR and Jacobs externalities as the immediate information spillovers about current market conditions. However, none of the previous definitions of static and dynamic effects provide a enough clear criteria for their classification. In this paper we propose a discriminative criterion. In analogy with the concept of static scale economies and dynamic scale economies, it can be argued that dynamic externalities have irreversible effects and represent the technical spillovers and other type of knowledge that give way to an irreversible improvement in the efficiency of the firm. If the beneficiary firm moves its location to another area its efficiency does not fall, former dynamic effects are embodied in its production function. On the contrary, static externalities are reversible in the sense that its effects on firm s efficiency disappear with the cessation of the externality if the firm moves to another location. But, in any case, as we shall se below, in practice it is not ease to sort out static from dynamic effects in a variable. As we have seen in the approach of Henderson et al (1995) dynamic MAR externalities deal with information spillovers historically accumulated on technological aspects and other issues (marketing, design, management, organization) among firms of the same industry and located in a specific area. Recent literature on this subject stress that such local information spillovers disseminates relatively slowly (Jaffe, 1993). Local spillovers involve tacit knowledge which is not traded in the market but exchanged directly by means of barter 5. This market failure can be overcome by developing long-term relations between firms. The set-up of such a network of relationships is time consuming and - as it happens to all social institutions - it is not possible to transport or to imitate in a different location 6. Inversely, the exploitation of these type of externalities leads to the persistence of industrial clusters in their traditional locations. It favors agglomeration. 7 Model, estimations and results This model attempts to capture the influence of dynamic externalities in the location of manufacturing sectors and is based in the work developed by Henderson (1994) and Henderson et al. (1995). It is assumed that present employment of a sector in a location depends on market historical and present conditions. The equilibrium employment level of an industry at time t in location i (Eit) is the result of the equalization of the local wage rate to the value of marginal product of labor. If the production function takes de form Ait(.)F(Eit,...) - where the term Ait represents the state of technology - the nominal wage is Wit=Ait(.)F (Eit,...)Pit(.). The price of output Pit is given by the inverse demand function Pit(.)=P(Eit, M it). We can plausibly assume that the price function is downward sloping in local industry output (represented by Eit). The argument Mit is a vector that includes local characteristics such as accessibility to main markets and regional demand for manufactures. The term Ait is a function whose arguments represent local externalities. Dynamic externalities include, as we have seen above, the stock of local industry specific knowledge accumulated through time. This can be approximated by three historical variables: own industry employment in the base period Eio; the degree of concentration of that employment, Cio; and the industrial diversity of the environment, Dio. Static externalities derived from knowledge spillovers about current market conditions can be represented by the current local own industry siz
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks