Impacts on Urban Geography
There has been considerable support both for and against the overall effectiveness of central place theory in describing and explaining the locational relationships of central places and their market areas. On the one hand, the theory has been criticized for its emphasis on description rather than explanation. The classical models are extremely rigid and deterministic. They predict a partial equilibrium under a constant set of conditions. The models describe an essentially static set of locational relationships with no reference to how the spatial pattern might change over time. The social and cultural environment of the relationships is taken for granted making them independent of the political and social forces operating at various spatial scales.
In contrast to these limitations, central place theory has stimulated research in urban geography and related disciplines such as retailing/marketing geography by providing a framework for empirical and applied studies. The broader notions of the theory, particularly the hierarchy of functions and centers, the interdependence of settlements and market areas, threshold population, and spatial competition, have given extremely useful terms of reference and organizational concepts within which to examine systematic regularities in the size, number, and spacing of settlements, retailing centers, and public service facilities. A great deal of research has been built on these core ideas. The vast literature which has grown up around the classical central place theory may be broadly classified into two categories: (1) theoretical studies concerned primarily with evaluating and modi fying the various postulates of the theory and (2) empirical and applied studies which have been more concerned with utilizing elements of the theory for descriptive and planning purposes.
Modifications of the Classical Models
There is a considerable amount of research on theoretical refinements of the classical central place models. According to Berry et al., this research can be subdivided into two groups of approaches aiming at: (1) extensions of the classical models, and (2) development of alternative approaches. The former studies attempt to modify the traditional models without changing their underlying assumptions, while the latter intend to make the classical models more realistic and to increase their theoretical generality.
Subsequent to Christaller and Losch, considerable efforts have been made to modify and extend the classical central place models, particularly with regard to their nature and properties. These efforts have been directed toward developing more general hierarchical models, refining the spatial equilibrium (market area structures), examining the size and configuration of centers in a central place system. In addition, some attempts have been made to integrate the central place theory with fundamental concepts of urban geography such as the rank size rule. The rank size rule indicates that central places, at least in their population sizes, follow a kind of gradient rather than being arranged into a hierarchy. Beckmann resolved this inconsistency by suggesting that each level of the urban place hierarchy might be subject to random influences. If this provision for population variation is added to central place theory then the patterns of center sizes would approximate the rank size distribution.
To make the classical models more realistic a number of attempts have been undertaken to scrutinize the central place system under various conditions including: a nonuniform population (or demand) density, spatial variation in the incomes of consumers, interdependencies between levels of urban hierarchy, price and transportation cost differentials, and variations in consumer behaviors. The alternative approaches to the classic central place models are mostly of a partial equilibrium nature, attempting to describe and explain some aspects of the central place system while disregarding others. One of the main problems with these approaches is that the economic rationality behind actions of individual consumers and firms remains blurred. Researchers engaged in developing so called new economic geography have addressed this problem. These 'new' approaches propose a framework that adds dynamic factors to the classical models and incorporates the behaviors of individual consumers and firms into the central place framework. Krugman demonstrated that with some adjustments, the static central place models could be made sufficiently dynamic to explain the development of a hierarchical urban system. Fujita and his colleagues provided further insight into the nature of urban systems. They suggested that in a region with multiple industries that differ in terms of scale economies and/or transport costs, the urban system self organizes into a highly regular hierarchical system along with a gradually increasing population size of the region. This finding supports Christaller's central place hierarchy.
Empirical and Applied Studies
The theories outlined in the works of Christaller and Losch have generated a large volume of literature reporting on empirical research into the spatial regularities suggested by the central place models. These empirical studies, however, have only partially been successful in demonstrating the existence of urban hierarchy. Although the theory has been verified by empirical evidence derived from many regions of the world, the empirical studies have tended to focus on regions with specific characteristics. They have concentrated on rural regions with relatively homogeneous natural and cultural environments (such as the United States Midwest, the flat Gangetic Plain of India, Szechwan in China, and Saskatchewan in Canada) and therefore have failed to distinguish a complete range of upper order settlements such as national metropolitan centers and global level cities. Christaller's original hierarchy as established in southern Germany included seven orders of central places. There is some evidence to show that Christaller's hierarchy of urban places has changed in response to the process of globalization and development of telecommunication technologies. Hall suggests that the two bottom levels of Christaller's hierarchy (i.e., market hamlets and township centers) have gradually disappeared from the hierarchy. At the same time, a global and a subglobal level have been added at the top of the hierarchy of urban places.
Interestingly, most empirical studies have been concerned with the distribution of consumer services and there has been little or no reference to the informationdependent producer services. The economic landscape as prescribed by the central place models has changed due to the development of transport and telecommunication technologies. The technological advances have eroded the frictional effect of distance on consumer behavior, and consequently there has been a major shift in the organization of urban systems from spatial structures in which central places are graded with multilevel hierarchy (according to the models of Christaller and Losch) toward systems shaped by the exchange of specialized information. Graham and Marvin suggest that a 'hub and spoke' urban network interconnected by fast transport infrastructures and telecommunication systems has gradually replaced the multi level functional hierarchy of central places. This claim echoes Pred's work, who argues that along with the structural changes of advanced economies the urban–hinterland relationships have become more horizontal in terms of the hierarchy of central places, and they have become global in scale.
The classical central place models have had an important impact on spatial diffusion research. There have been a number of theoretical and empirical studies attempting to link spatial diffusion and central place theory. The results of these empirical studies suggest that the relations between central place hierarchy and spatial diffusions follow a two stage process: (1) a 'filtering down' of innovations through the urban hierarchy and (2) a contagious 'spreading out' from urban centers to other locations in their hinterland. Thus, the innovation potential of a central place is related to its position in the urban hierarchy and the amount of interaction it has with centers that have already adopted the innovation.
The usefulness of central place models is supported by the successful applications of central place approaches for analyzing the locational patterns of the private and public sector services within urban areas. The most common and perhaps the most successful applications have been studies of the spatial patterns of retail activities. The classical studies are those by Berry and his colleagues. These studies confirm the applicability of the central place concept for analyzing the intra urban patterns of business centers. For example, the principles of central place theory have been applied to Chicago and its suburbs, where a systematic hierarchy of centers and subcenters (albeit strongly influenced by the existence of major radial transportation lines) was identified. More recently, central place models have provided a foundation for a geographic information system (GIS) based analysis of the spatial patterns of marketing and retailing and for an empirical testing of the hierarchy of central places.
Central place models are considered to be useful descriptive tools, with some prescriptive value for regional/urban planning schemes. The main stimulus for using central place theory in developing regional plans has been related to the efficiency of central place arrangements that may have a positive effect upon the economic development of a given region. Central place theory has been successfully applied to regional plans in various regions and countries (e.g., USA, Canada, Europe, India, and Middle East). For example, the spatial pattern of new settlements on the drained polders of Holland was planned around the postulates of central place theory. Similarly, Israeli settelements on the Lakhish plains were developed according to a three level hierarchy.