The integration of time and of space into economic theory have been unequal processes. Economists have spontaneously and almost invariably viewed the economy with a temporal dimension: economic agents calculate within a particular time frame, economic activities are transformed by innovation, production resources are accumulated, and dynamic processes induce periods of steady growth, recession or economic cycles. Space, however, is neither a major nor a permanent feature of economic thought. Questions such as proximity, location, spatial competition, spatial interaction, urban structures and urban hierarchies, the role of regions and the formation of territories have been given sporadic treatment only. The history of spatial economics is an alternating pattern of attempts at integration and periods of oblivion or of varying degrees of marginalisation of space, which has at times been “a neglected topic in mainstream economics” as Thisse and Walliser put it below, p. 11. The present-day integration of space in economic theory is itself the result of a series of episodes of continuity and of change, as with the transition from New Urban Economics to agglomeration economics analysed hereafter by Baumont and Huriot.

This paper aims at providing a possible explanation for the seemingly peripheral role played by spatial economics in modern economic theory. It considers successively five points of view according to which space is introduced in economic models and assesses their contributions and limits.

Von Thünen's monocentric model is considered as one of the foundations of spatial economics. Most of its assumptions have been transfered from agricultural to urban space by New Urban Economics. This transposition gave new impetus both to the monocentric model and to urban economics. Yet the urban monocentric model, because of its strong economic and spatial assumptions, fails to explain the formation of cities. The economics of agglomeration, and more generally New Economic Geography, propose to solve this problem by considering endogenous formation of spatial concentrations, i.e. by changing radically the approach of space. We attempt to identify the lines of continuity and the main changes from von Thünen to contemporary urban microeconomics, and to understand in what way continuity is an obstacle to innovation, and how change has made it possible to renew urban economics.

This article traces a little-known French tradition in spatial economics that was advanced by state engineers who were trained at the École des Ponts et Chaussées and worked for the State corps of engineers. Spurred in part by changes in power technology, especially the advent of the railroad, these engineers pioneered the spatial aspects of competition in a microeconomic framework. The contributions of Jules Dupuit (1804-1866) and Emile Cheysson (1836-1910) are especially singled out for analysis, and are shown to have anticipated later important developments in spatial economic theory.

This paper explains how road pricing and toll charge theory appeared and evolved on the basis of contributions from the French engineer-economist school, from Dupuit to Allais, and of Anglo-Saxon theories inspired by Pigou and Vickrey on decongestion tolls and network tools.

There is an entire branch of industrial location analysis in economics and operational research that refers back to Weber. In this paper I show that the model he developed had already been proposed by W. Launhardt in 1882 as a generalization of a route planning method presented in 1872. It contained all the technical aspects of Weber’s model. That does not mean Weber should be forgotten though, as his contribution involves far more than the simple presentation of the model. In terms of method, he defends and uses the hypothetical-deductive method. As for the content, he attempts to understand the formation of concentrations of economic activities and population, which leads to an interesting analysis of agglomeration economies.

The paper discusses the relation between the justly famous Hotelling’s [1929] paper on stability in competition and the corresponding passages of the often ignored Launhardt’s [1885] book, devoted to mathematical economics. Launhardt is sometimes credited for the anticipation of Hotelling’s spatial model and analysis of oligopolistic competition with product differentiation. The paper argues that Launhardt’s model is in fact significantly more general, in particular in introducing vertical along with horizontal differentiation, and that his analysis of strategic competition is correspondingly finer and more complete. Thus, the earlier author does not appear to be a forerunner of the later, rather the later a successor of the earlier. However, Hotelling did (objectively) innovate, in a significant way, by endogenizing locations as strategic variables, and principally by developing the concept of an abstract characteristics space, which transcends the geographical space.

The interest shown in territorialized productive organizations does not simply account for a geographical phenomenon but above all highlights the territorial dimension of development and innovation processes. In the wake of the work done on Italian industrial districts, numerous studies have been conducted that have made it possible to identify similar forms of localized productive organizations. The aim of the present paper is to attempt a synthesis of the way in which these territorialized productive organizations work. To do so, we shall select two concepts: the localized production system and the milieu. The fundamental question is then to identify which logic influences the way a localized production system works: the milieu-actuated territorial logic or the functional logic, which corresponds to industrial organization of an hierarchical nature.