 |
||
| BACK | NEXT | WEB BOOK |
Analysis of Land Use Change: Theoretical and Modeling Approaches
Helen Briassoulis, Ph.D.
| |
||
| BACK | NEXT | WEB BOOK |
3 THEORIES OF LAND USE CHANGE
This chapter presents a representative collection of theories of land use change. It is, thus, necessary to clarify: (a) the meaning or definition of "theory" as it will be used in the present context, (b) the theories of land use change that will be included (and, equivalently, those that will be excluded) and (c) the need to consider and take into account theories of land use change in the overall project of studying this subject.
The Greek word "theory" means, literally, "looking at something", "observing something". Consequently, it denotes "knowledge" – the result of observation. Theory is considered "a set of connected statements used in the process of explanation" (Johnston et al. 1994, 622). Chapin and Kaiser (1979) define theory as "a system of thought which, through logical constructs, supplies an explanation of a process, behavior, or other phenomenon of interest as it exists in reality" (Chapin and Kaiser 1979, 27).
Because theory denotes knowledge, the nature and status of theories differ among different epistemologies – discourses on how knowledge is acquired, transmitted, altered and integrated into conceptual systems (Johnston et al. 1994, 168). For example, "within positivism , a theory comprises a set of hypotheses and constraining conditions which, if validated empirically, assume the status of laws, so that theory structures understanding of the relevant portion of the empirical world through its system of interrelated laws…. Whereas within positivism, a theory is assumed to be universal in its application, within idealism , on the other hand, there are no universals – only the individual theories resident in each individual’s mind – which are used to guide action…… In realism , a theory is a means of conceptualizing reality, and thus provides a mental framework for its apprehension: the test of a theory is not its validation against empirical evidence but, rather, its coherence and, especially, its practical adequacy…Realists argue that because societies are open systems in which the same conditions are rarely reproduced, theories cannot, as positivists contend, predict the future; they can only illuminate the past and the present, and provide guidance to an appreciation of the future" (Johnston et al. 1994, 622-623).
The differences among the theories of land use change which will be presented in the following sections can be attributed, to a considerable degree, to the different epistemologies adhered to by those who have proposed them (and by those who use them). But what is a theory of land use change? Simply stated, it is a set of propositions used to understand the "what" of land use change and the "why" of this change. In other words, a theory of land use change describes the structure of the changes in the uses of land from one type to another – and explains why these changes occur, what causes these changes, what are the mechanisms of change. The "what" and the "why" of land use change are closely related although extant theories rarely address both; they refer either to the "what" or to the "why". As regards the latter, it is important to cite Sack (1990): "describing ‘what’…. can be accomplished up to a point without considering social and individual motivations. These motivations, however, must be included in the ‘why’ of human behavior, and knowing the ‘why’ is essential if we expect to change ‘what’ we do….. Much of the ‘why’ in human-nature relations can be understood only through the social side of the equation – that is, through understanding the nature of individuals and societies that create the ‘what’" (Sack 1990, 659).
Getting to the second issue, which theories of land use change are included in this chapter, it is noted that the majority of theories of land use change have to be sought in the more general theoretical frameworks of disciplines studying economic, environmental and spatial change (or, transformation). Andersson and Kuenne (1986) state that (static) spatial analysis is concerned with four reasonably distinct bodies of spatial phenomena: (a) locations, (b) interaction flows, (c) changes in availability of factors of productions and (d) spatial structures. The latter are defined as "including areal or curvilinear patterns of economic activities such as land use patterns, urban structure, transportation networks, and market or supply areas" (Andersson and Kuenne 1986, 201). Hence, for the present purposes, a broad distinction is drawn between those theoretical schemata which treat land, land use and, more importantly, land use change explicitly and those where reference to land use change is more or less indirect and implied by the broader discussion. In other words, an attempt is made to cover those theories where land is defined, at a minimum, as "a delineable area of the earth’s terrestrial surface" (see Chapter 1) as opposed to those theories in which land is either reduced to a point in space or it is totally absent.
Based on the above criterion, the genre of location theory analyses (Central Place theoretical studies included) are considered to a limited extent in this contribution. These theories are not considered theories of land use change per se as their emphasis is on particular, individual activities (usually treated as points) locating in space and not on an area of land used by various activities (see a related comment by Beckmann and Thisse 1986, 22; also, for a concise description of the "point" nature of spatial equilibrium analysis, see Takayama and Labys 1986, 171). The point is that an individual activity, say a manufacturing firm, may locate within an area whose land use may not be necessarily "industrial". Of course, the opposite is also true. Similarly, the analysis of "market areas" in location theoretic studies does not imply analysis of a particular, concrete land use or its change as the physical area designated as the market for a good or service may comprise of several types of land use (e.g. residential, commercial, open space), in general. The only exception to the location theoretic studies which are considered in the present contribution are residential location theories and models as the aggregate outcome of individual choices considered by the related theories are residential land use patterns (or, segments of the housing market).
A broader set of theories which deal with the dynamics of urban and regional spatial structure are given consideration in this contribution. These theories treat land and land use as points in space mostly (but not always) but their significance lies in that they analyze the broader spatial processes that ultimately result in land use change. The majority of these theories are agent-based; in other words, they deduce changes in spatial structure starting from the behavior of the individual household or firm. One of the reasons these theories may prove important in the future for the analysis of land use change is that they can support the building of spatially-explicit models which focus on the level of the individual decision making unit (farm, firm, household). Lastly, a set of theories that will be noted in passing but do not refer directly to the issue of land use change are those in which land use is not included at all within their stated objects of concern and analysis. These are termed frequently "aspatial" theories. They refer mostly to economic, social and other determinants of land use change and they are concerned with broader socio-economic changes that may, however, impinge on and cause land use change in one way or another. These theories include economic base theory, input-output analysis, economic development and growth theories, international trade theories, social theory, etc.
Finally, the role of theories of land use change in the study of the subject needs to be stressed. Common to all analytical tasks is the need to have a vehicle to structure the conception and explanation of reality – i.e. a theory. The analysis of land use change is no exception. Idealist and theories adopting similar epistemologies aside, theories of land use change guide thinking about land use change, indicate conceptual and operational expressions of change, its determinants and the relations between them, and suggest explanatory schemata for making sense of available empirical evidence; i.e. they support model building. Moreover, to reiterate Sack (1990) mentioned above: "knowing the ‘why’ is essential if we expect to change ‘what’ we do"; in other words, theory is a guide to policy on land use change – a strong and critical demand of the contemporary times. Inappropriate and inadequate theories of land use change may misguide policy and produce more ills than those policies are assumed to cure.
Although the use of theory in model building seems indispensable, of the several theories of land use change proposed, a relatively small number has been used to support and guide operational model building. Some theories and models have been conceived simultaneously; hence, the use of the terms "theory" and "model" either interchangeably or to denote a set of conceptual and operational statements about reality (e.g. the urban land market theory and model). In this case, the term "model" may denote mostly a formal theoretical model and not necessarily a symbolic (or, operational or empirical) model (Lonergan and Prudham 1994). But the majority of theories are still without modeling (not necessarily mathematical) counterparts and the reverse is also true. Several models are devoid of theoretical foundations. There are many explanations for this gap in the relationship between theories and models only two of which are mentioned here. One reason is the differing epistemological positions adopted by theory and model builders; usually models move in the positivist tradition while theories cover a much broader spectrum of epistemologies . A strong reflection of these differences is the way land is usually being conceptualized in theory and in models. A related reason is that reality is highly complex; land use change comes about under the influence of many macro and micro factors, acting and interacting within varying time frames and geographical space. Land use change problems are essentially metaproblems . Therefore, the reduction and simplification of this real world diversity to serve the purposes of model building is either extremely difficult, or results in a very crude representation of reality. The contrary may happen also; models have a very complicated structure that is impossible to handle within the bounds of reasonable time and other resources to provide answers to practical problems.
The sections that follow provide a broad overview of the variety of theories pertaining to the subject of land use change and present certain of them in some detail. The last section evaluates briefly the theories of land use change covered and attempts to address the question of whether a general theory of land use change is possible and meaningful or whether a synthesis of theoretical schemata is the most fruitful way of providing support to model building and policy making.
3.2. Theories of Land Use Change –
Classification
The theoretical literature on land use change contains a considerable variety of theories where land use is treated explicitly and is the direct object of theoretical inquiry. Six interrelated sources of variation, in a roughly decreasing order of importance, can be discerned:
Hence, there exist:
It is, thus, evident that, for the purposes of systematic exposition of extant theories, it is necessary to adopt a classification scheme as a presentation and discussion vehicle of these theories.
A general-purpose, unambiguous classification scheme of theories that can reflect meaningfully the six sources of variation mentioned above does not seem to exist for various reasons. The same subject is studied by many disciplines (that may traditionally have particular outlooks, spatial and temporal foci, interests in certain uses of land); theories filter from one discipline to the other – for example, from economics to geography (Cooke 1983, 83); disciplinary boundaries are blurred especially in modern times when there is also a tendency towards interdisciplinary research. Hence, a decision was made to adopt a classification scheme based on an aggregate criterion, the theorization tradition to which a theory belongs. This is taken to denote the particular way of thinking about and conceptualizing reality – land use and its change in the present case – that is mostly a function of particular disciplinary cultures and, consequently, of epistemological orientations, value systems, choice of space and time frameworks and of objects of analysis.
Based on the theorization tradition criterion, a three-fold typology is used to classify extant theories of land use change into three main categories:
Within each of these three main categories, theories can be further classified according to other, more focused and particular criteria as it is indicated in the following discussion. Table 3.1a is an attempt to draw together and present selected theories that belong to each of the three categories. Tables 3.1b, 3.1c, and 3.1d present theories within each category, according to criteria particular to the each category. As these Tables show and the ensuing discussion will reveal, it is difficult, if not impossible, to provide a neat categorization of theories as more than one criteria can be used to classify them. For example, a classification by theme or subject of analysis (e.g. development) will include economics-based theories as well as sociology and political economy-based theories. The same is true if theories are classified by their epistemological foundations or the basic concepts around which they are organized (e.g. core-periphery). Hence, some theoretical concepts will be discussed in more than one groups or categories within groups. The next sections are devoted to a brief discussion of the theories within each tradition. For each theory, the following main issues are examined: purpose (descriptive, predictive, explanatory, prescriptive), mode of theorizing (assumptions, type of land use and their determinants considered, especially the proposed mechanism of land use change), spatial scale of reference, and temporal dimension (duration, dynamics).
3.3. The Urban and Regional Economics
Theorization Tradition
The urban and regional economics theorization tradition adopts the way of thinking in economics in general. Reality is represented using concepts and procedures of an economic nature – among them, prices of the factors of production, of products and of services, transport (or, transfer) cost, marginal cost, economies of scale, externalities, and, above all, utility . All behavioral assumptions made refer to the model of the rational, economic, utility maximizing man despite efforts to replace it with less inflexible and more realistic constructs (such as Simon’s "satisficer" – Simon 1956, 1982). Real world phenomena are analyzed either from a micro-economic or from a macro-economic perspective. Hence, theories of land use change belonging to this tradition are broadly grouped into micro-economic theory-based and macro-economic theory-based. Micro-economic approaches start from individual consumer behavior and then aggregate over the behavior of all consumers to yield land use patterns produced when utility is being maximized for all consumers (usually, maximization of profits or minimization of cost or distance). In contrast, macro-economic approaches deal with aggregate behavior and indicate how aggregate patterns may be produced. A third group of theories is included which contains that belong generally to the field of Regional Science and utilize concepts from both economics and sociology. Their inclusion in the urban and regional economics theorization tradition is justified by their emphasis on economic factors and processes of spatial change. Table 3.1b presents in greater detail the microeconomic, the macroeconomic, and the regional science theoretical approaches discussed below.
3.3.1. Micro-economic theoretical approaches
Three main micro-economic theory-based theoretical schemata for the analysis of land use patterns and their changes are discussed below: J. F. von Thunen’s agricultural land rent theory, W. Alonso’s urban land market theory, and agent-based theories of urban and regional spatial structure. It should be noted that all three are considered theories as well as models because their developers proposed a theoretical structure which they translated then into a mathematical, form; i.e. a symbolic model (not necessarily operational, however).
3.3.1A Agricultural Land Rent Theory
The analysis of land use patterns and their changes in the micro-economic theorization tradition (but also in the macro-economic) has been influenced in fundamental ways by the agricultural land rent theory developed in 1826 by the North German estate owner, J. H. von Thunen (1966). The purpose of von Thunen’s exercise was to prescribe the optimum (most economical) distribution of rural land uses around a market town (Hoover and Giarratani 1984, 1999). The basic concept he used was that of the land rent which is defined as the "price for the use of a piece of land" (Hoover and Giarratani 1984, 132) or, equivalently, "the price of the services yielded by land during a specific time period" (Romanos 1976, 32).
At the regional level of analysis, to which von Thunen’s formulation refers mostly, the land use types considered are various types of agricultural land primarily and, secondarily, forest land. The analysis concerns land which is devoted to growing different types of crops (and forestry). Land was assumed to be a uniform, isotropic (of equal fertility) flat plain with movement possible in all directions around a market town located at the center of the region of interest. Land rent varies only with distance from the center. Each crop has an associated rent gradient (or, rent curve) that extends in all directions from the center (Figure 3.2a in Hoover and Giarratani, 1999) as well as the same delivered price and unit transport cost irrespective of location or rent. Moreover, the intensity of land use for each crop and the yield per acre are fixed regardless of the relative prices of land (rents), the other inputs, and the output. Perfectly competitive markets are assumed. The rule of determining the location of a particular activity (land use) with respect to the market center is that each activity (land use) occupies the zone in which the user can pay the highest rent than any one of the other users. And the rent the user of a particular land use can afford to pay depends on the value of the products produced on a parcel of land. Hence, in the jargon of land economics, the user of an activity (land use) associated with high value products can bid higher land rents and, thus, outbids other users that cannot pay the same rent. In von Thunen’s formulation, the activity (land use) with the largest amount of output per acre (highest value of output) has the steepest rent gradient and, hence, locates closest to the market center. The other activities (land uses) follow in decreasing order of the slope of their rent gradients. The resulting land use pattern is a set of concentric rings around the market center with each ring devoted to growing a particular crop (Figure 3.2a ). The envelope of the individual crop rent gradients (formed by their uppermost parts) is the bid rent curve (for the study region) (Figure 3.2a ). A last remark on this formulation: the optimum solution to the land use pattern produced following the above procedure is independent of "whether: (1) one individual owns and farms all the land, seeking maximum returns; (2) one individual owns all the land but rents it out to tenant farmers, charging the highest rents he or she can get; or (3) there are many independent land owners and farmers, each seeking his or her own advantage" (Hoover and Giarratani 1984, 143). A mathematical exposition of von Thunen’s theory can be found in Hoover and Giarratani (1984, 1999). An exposition of von Thunen’s theory with several extensions can be found in De la Barra (1989).
The von Thunen formulation makes no explicit reference to mechanisms of land use change because it is a static theory where the optimum land use pattern is assumed to be produced instantaneously. However, it is not difficult to see an implicit mechanism even under all the restrictive assumptions of the theory. If the relative prices of the crops change exogenously, this will change the relative ability of farmers (the users of land) to bid for particular locations making, thus, possible a change in location (the land use pattern preserving its circular form). The very restrictive and unrealistic assumptions of the original formulation of the agricultural land rent theory were relaxed by von Thunen himself and by researchers who used it in subsequent applications (see, for example, Alonso 1964, Romanos 1976, Wheeler and Muller 1981, Hoover and Giarratani 1984, 1999, Stahl 1986). These applications covered a wide range of spatial scales from the global (Peet 1969 cited in Johnston et al. 1994, 673) to the individual village and farm holding (Blaikie 1971 and Chisholm 1979 cited in Johnston et al. 1994, 673) as well as other land uses such as residential and commercial. More importantly, perhaps, this theory provided the foundations for Alonso’s (1964) urban land market theory (discussed below). In general, there is no doubt that von Thunen’s theory is the predecessor of both location theory and the analysis of urban and regional spatial structure.
3.3.1B Urban Land Market Theory
In the years that followed von Thunen’s seminal approach to a theory of land use, several attempts were made at analyzing various components of the urban and regional system (see, for example, Romanos 1976). However, it was only after almost 140 years that W. Alonso would present his celebrated urban land market theory that applied and refined von Thunen’s original ideas (Alonso 1964). This theory aims to describe and explain the residential location behavior of individual households and the resulting spatial structure of an urban area. The focus is on residential location; the behavior of firms is treated more briefly and abstractly. The central concept of this theory is the bid-rent function for each household and/or firm. The bid rent of a household is defined as the "maximum rent that can be paid for a unit of land (e.g. per acre) some distance from the city center if the household is to maintain a given level of utility " (Hoover and Giarratani 1984, 153) (Figure 3.2b – Figure 6.8 in Hoover and Giarratani 1999). The bid rent curve R of the actual land rents in the city reflects the outcome of a bidding process by which land is allocated to competing uses (residential demanded by households and commercial/ industrial demanded by firms).
As in von Thunen’s theory, a monocentric, flat, continuous and uniform urban area is assumed. The city center is the central business district (CBD) where households work and shop. A household’s utility (or satisfaction) is assumed to depend on: housing (of a given lot size), distance from the city center (reflected in the transportation cost) and all other goods (Chapin and Kaiser 1979, Romanos 1976). The household allocates its fixed budget among these three components with the aim to maximize its utility. Its preferences determine the trade-offs it is willing to make among the above three items. The price of housing and of other goods is independent of the quantities purchased. The price of housing and of commuting depend on distance from the city center. There is a distance decay relationship between land rent and distance from the CBD. The further a household lives from the city center, the more it will have to spend on commuting and the less it will be able to spend on housing. Based on these assumptions, the bid-rent curves are downward sloping (rent decreases with distance from the city center to offset transport costs) and single-valued; i.e. for a given distance from the CBD only one rent bid is associated with a given level of utility. The steepness of the slope of the bid rent curve depends on transport costs and the household’s (or the firm’s) demand for space. Steeper curves are associated with higher transport costs and/or less demand for space (hence, higher value attached to accessibility). Flatter curves are associated with lower transfer costs and/or higher demand for space (and, hence, preference for more outlying locations). Finally, lower bid-rent curves are associated with greater utility as, assuming fixed budgets, at any given distance from the CBD, if a lower rent bid is accepted, more goods can be consumed (Hoover and Giarratani 1984, 154).
Alonso’s theory distinguishes two stages in the residential location process. In the first stage, the theory derives individual equilibria for households (and firms) on the basis of their bid-rent functions (one for each level of utility/satisfaction). Households, possessing perfect knowledge of the actual land rent structure in the city (curve R in (Figure 3.2b) and of the transport costs, choose a location that maximizes their utility subject to their budget constraint; this is the point where the lowest bid-rent curve touches the actual rent curve ((Figure 3.2b). At a second stage, the equilibrium for the entire urban market is derived through a market clearing mechanism that starts from the CBD and involves potential users bidding for land and landlords selling or renting the land to the highest bidder. The most central locations go to the highest bidders (steepest bid-rent functions). Remaining available land goes to the next bidder and the process continues until the last user is located at the edge of the city. The price of land at the edge of the city is adjusted to agree with the actual price there (basically, the value of agricultural land close to the edge of the city). However, as Romanos (1976, 71) notes, because Alonso does not assume a perfectly competitive market and bid-rents are not unique but members of families of bid-rent curves, the theory cannot provide an equilibrium market solution as it was the case with von Thunen’s theory. In order to derive equilibrium land use patterns, additional assumptions must be made about the level of utility of the bidders or the number and types of bidders (Strazheim 1986). In the simplest case, equilibrium in the land market results if all bidders have identical incomes and preferences; hence, a common set of bid rent curves and land rents within the city coincide with this set (Strazheim 1986). The discussion of Alonso’s urban land market theory (and its extensions) resumes in Chapter 4 in the presentation of the formal (mathematical) model of the urban land market.
Alonso’ urban land rent theory provides a static description and explanation of urban (mainly residential) land use. In the present context, its significance lies in its explicit treatment of the actual amount of housing consumed; hence, it is a theory of (residential) land use. The bidding process is a realistic account of the way land is allocated to various competing users and it has been used in the theoretical and modeling exercises that ensued Alonso’ original contribution (see, for example, Romanos 1976, Brueckner 1986, Strazheim 1986). However, the mechanism of land use change is implicit; it has to be elicited from the factors which the model assumes to determine the steepness and the height of the bid-rent curves. These depend on preferences for various locations within the city (measured as distances from the city center) and income. Hence, when preferences and income change, the land use system will move to another equilibrium position. Other potentially important influences on the bid-rent curves such as socio-cultural and political forces are not accounted for directly by the theory.
While Alonso’s theory has been used extensively in analyses of urban spatial structure as well as in impact analysis of urban policies (see, for example, Bockstael and Irwin 1999), it suffers from several restrictive assumptions which limit its usefulness in approximating observed land use patterns as well as in analyzing land use change. The two most important perhaps are the assumption of a single, exogenously-given center (the monocentric city assumption) and the importance assigned to accessibility to this single center in explaining urban spatial structure. The theory does not consider a number of interrelated factors which, on the one hand, capture the particular forms that characterize modern urban agglomerations and, on the other, account for the dynamics of urban land use change. The most important of them include: the existence of more than one centers in metropolitan areas, externalities (e.g. traffic congestion, air pollution), increasing returns to scale, imperfect markets, the durability and inflexibility of the housing stock, technological change (see, among others, Romanos 1976, Quigley 1985, Arnott 1986, Krugman 1995, Bockstael and Irwin 1999).
The monocentric city assumption has several implications for the land use patterns derived from this theory. First, as the size of an urban area increases, employment can by no means be concentrated in the CBD. Real world evidence reveals the continuous decentralization of employment centers in large metropolitan areas and the development of polycentric cities coupled with a declining or minimal growth of the CBD. At the same time, the role of land developers in determining the location and timing of ex-urban development is critical in the evolution of urban spatial structure. Hence, the urban land market theory assigns to the CBD greater importance than it actually deserves. Secondly, the analysis of land uses in the rest of the urban area is not adequate as other nonresidential uses are present (Romanos 1976). The heterogeneity of the natural environment in which the city is embedded is also ignored as a factor which may distort the analytical tractability of the bid-rent functions. Thirdly, the monocentric city assumption is associated with another assumption; that of constant returns to scale in the production of goods and services in the CBD. However, increasing demand for these goods and services, implies increased demand for transportation to the CBD, hence, increasing diseconomies associated with congestion, pollution, etc. that lead to decreasing returns of central city production. Therefore, the assumption of constant returns is at variance with the importance attributed to the city center.
Several attempts have been made to modify the original theory to account for the presence of more than one centers, more than one places of employment, and the existence of externalities, among others, whose discussion is beyond the intent of this contribution (see, for example, Solow 1973, Romanos 1976, Shieh 1987). These modifications, however, have aimed mostly at providing improved operational versions of the theory – i.e. of the urban land market model – while leaving its basic tenets intact. Therefore, the basic limitations of the theory, the absence of a dynamic explanatory schema of land use change, still remain. A brief discussion of more recent theoretical attempts at analyzing the evolution of urban spatial structure is undertaken in the next section.
3.3.1C Agent-Based Theories of Urban and Regional Spatial Structure
A broader set of theoretical schemata have been proposed for the description and explanation of the evolution of urban spatial structure which focus on the agents operating in urban contexts and the interactions among them which influence the resulting spatial patterns. This group of "agent-based" theories does not always treat explicitly land use as in Alonso’s theory (i.e. as amount of space consumed by an agent) and their emphasis is mostly on agents’ characteristics as well as on the processes through which and the conditions under which agents interact in space. In other words, they are indirect theories of land use change for the present purposes. The following is a schematic presentation of certain of their particular features and the gist of their approach to the analysis of the evolution of urban spatial structure which render them more realistic frameworks for the study of land use change. The reader may find more detailed accounts and reviews of these theories in, among others, Krugman (1995), Henderson and Mitra (1996), Anas et al. (1998), Fujita et al. (1999).
The agent-based theoretical approaches differ from the micro-economic approach of the urban land rent theory in that they stress particular features of these agents which relate to their linkages and interactions in space; broadly speaking, they take into account the market structure of the urban setting. Several of the ideas contained in these approaches can be found in earlier theories of urban and regional spatial structure and development (e.g. Christaller 1966, Pred 1966, Myrdal 1957, Henderson 1974) as well as in broader theories stressing the role of human agency in general in the evolution of spatial form (for a collection of references see, Pred 1985). However, it seems that the synthesis of these ideas into more rigorous theoretical schemata as well as their use in building models of change began in the early 1980s. Agents are assumed to operate in the context of markets mostly. Imperfect competition is allowed in several approaches (see Krugman 1995). To explain the clustering or the dispersion of certain uses in space which is observed in the real world, the concepts of external economies, externalities, backward and forward linkages between activities, and durability of urban development are employed in the broader schema of centripetal and centrifugal forces impinging on agents’ behavior.
Centripetal forces account for the cohesion and clustering of certain activities in space (Hoover and Giarratani 1984, 1999; Krugman 1995). These forces derive from the existence of economies of scale (increasing returns to scale) and agglomeration economies in certain locations. Activities linked by means of forward (being suppliers of goods and services to other activities) and backward (being buyers of goods and services) linkages are held together in a given location. Forward and backward linkages among activities are called also vertical linkages (Hoover and Giarratani 1984, 1999). A circular relationship develops also between the location of the market and the location of an activity; i.e. activities concentrate where the market is large and the market is large because it contains a large number of activities (see Krugman 1995).
At the same time, however, centrifugal forces work against the clustering of activities in space and cause their dispersion. These concern the horizontal relationships among activities (Hoover and Giarratani 1984, 1999) and involve the competition among activities for markets and/or inputs as well as the cost of transport to the sources of inputs or to the markets. Increased competition for certain locations (having locational advantages for certain activities or containing scarce resources) drives land rents up. Some activities are driven away to sites where land rents are lower. Some others may remain in the same location, however, because of the presence of other advantages it offers which render it profitable for its user. Other centrifugal forces are associated with various kinds of diseconomies or negative externalities caused by either the clustering of activities or owing to the particular nature of certain activities which reduce the potential benefits to be reaped from being at a specific location.
Several other factors enter the decision making calculus of individual agents which may function either as centripetal or centrifugal forces that impact on their utility function and, thus, affect their locational choices. These include speculation, in particular land speculation, the durability of the physical infrastructure associated with particular activities, the cost of land conversion to other uses (the land use inertia), other local conditions, and "historical chance" (Arthur 1989). The spatial result of the interaction of centrifugal and centripetal forces on individual but interdependent agents is the generation of monocentric, polycentric, dispersed, linear, etc. urban land use patterns. In other words, the resulting spatial structure is characterized by multiple instead of a single equilibrium pattern.
The main point put forward by agent-based theories is that the decisions and actions of agents are influenced by past locational decisions and they influence future location decisions. Thus, the resulting spatial patterns (the spatial distribution of agents and of the associated activities) are endogenously determined. Variations and changes in the factors underlying these patterns mentioned before give rise to land use change; or, in broader terms, they explain the evolution of the spatial system over time. Agent-based theories have been used in building corresponding models of locational behavior but several of the interactions these theories postulate have not received empirical testing yet. Finally, most theories focus on equilibrium patterns while real world experience shows that most of the time the process of land use change is out of equilibrium (Bockstael and Irwin 1999). Nevertheless, these theories represent a considerable improvement over the monocentric model of the 1960s and exhibit a flexibility which permits the consideration of many more factors – even idiosyncratic – which account for land use change as well as for the impacts of this change.
3.3.2. Macro-economic theoretical approaches
Compared to the micro-economic theory-based approaches to the analysis of land use change which start from the behavior of the individual consumer (or, producer) and then aggregate over the population of consumers to derive the resulting land use pattern, macro-economic approaches operate at an aggregate level employing aggregate concepts, measures and forms of behavior. For the present purposes, two macro-approaches are distinguished: the spatial and the aspatial macro-economic theories. The first group refers to the body of theories known as "spatial economic equilibrium theory" (Takayama and Labys 1986, Fischer et al. 1996, Ginsburgh and Keyser 1997) while the second comprises a variety of mostly aspatial theories.
3.3.2A Spatial Equilibrium Theory
Spatial (economic) equilibrium theory is essentially the application of utility maximization theory of welfare economics to a spatially dissaggregated economy. Building on the original contributions of Alfred Weber (1929) and August Losch (1954), the field of spatial equilibrium analysis developed fully after the 1950s with the parallel development of mathematical programming that provided the tools (techniques) to express symbolically the theoretical propositions (Takayama and Labys 1986). A finite number of demand and supply regions are represented by points in space which are interconnected by various modes of transporting goods with the transport cost structure specified in some way (Takayama and Labys 1986). In broad terms, the theory seeks to determine the equilibrium prices of goods and services as well as the wage levels which satisfy an efficient distribution of demand (consumption), supply (production) and flows of goods and factors of production (labor and capital) among those points. This distribution is derived by maximizing the welfare (or, utility , or well-being) of the population located on the points (i.e. living in the demand regions) or the profits of the firms located on the supply points. Consumer welfare is measured variously as income, consumption of goods, etc. The conditions under which this equilibrium is achieved in the spatial system are derived by elaborating the mathematical relationships which express the welfare maximization problem. Assumptions commonly made in applying spatial equilibrium theory concern the distribution of population, resources, accessibility, and preferences. The common assumption is that these are uniform. In addition, a market economy is assumed where perfect competition holds, perfect technical knowledge is available, and no barriers to market entry exist. Assumptions are made also about the relationships among regions as well as about their self-sufficiency in terms of raw materials. Among the various feasible solutions of the welfare maximization problem, the theory considers as optimum those satisfying the Pareto-efficiency criterion . Several versions of spatial equilibrium theory have appeared which attempt to relax one or more of the restrictive assumptions such as the perfect market assumption (an important prerequisite to applying the theory to non-capitalist spatial contexts, at least). The theory does not specify the appropriate spatial level of analysis although it is more commonly used at higher spatial levels (regional, interregional, national and international). Spatial economic equilibrium theory provides the theoretical underpinning of the body of spatial equilibrium as well as of regional (spatial) dynamic models (Takayama and Labys 1986, Andersson and Kuenne 1986, Isard et al. 1969, Ginsburgh and Keyzer 1997, van den Bergh et al. 1996). In addition, it is employed in the context of integrated land use modeling to be discussed in the next chapter.
Evidently, spatial (economic) equilibrium theory cannot be considered a direct theory of land use as land and land use are treated at a very high level of abstraction reduced to points and spatial configurations follow geometric shapes (lines, curves, discs) (Andersson and Kuenne 1986). The same level and degree of abstraction characterizes also the representation of several other components of the spatial system. In their static versions they do not address explicitly the issue of change as they concern equilibrium demand and supply configurations and they are used either in a descriptive or in a normative fashion usually under restrictive assumptions. In their dynamic versions, change is brought about by changes in demand, product prices, transport costs, technological change, etc. Essentially, spatial equilibrium theory is heavily mathematically oriented deriving spatial relationships deductively from a set of initial assumptions and axiomatic propositions most of which concern the mathematical prerequisites necessary to solve the welfare maximization problem. In other words, it is a functionalist theory which treats spatial relationships and patterns by means of an "aspatially constricted theoretical framework, general equilibrium theory" (Cooke 1983, 116). Its relevance to the analysis of land use change, if one accepts its epistemological orientation as well as its behavioral and other assumptions, lies in its providing the broader context of changes in the economic determinants of land use change; namely, changes in the location of production and consumption, the associated changes in demand, supply, product prices and wage levels, and changes in trade among regions (or, flows in general).
A variety of other macro-economic theoretical approaches dealing with regional and supra-regional growth and development are mentioned briefly here as their object of analysis is the development of activities in space and they touch on issues of land use and its change although in an abstract and rather indirect fashion. Regional disequilibrium theories and Keynesian type regional development theories are discussed briefly below.
3.3.2B Regional Disequilibrium Theories
Representative among the regional disequilibrium theories are Myrdal’s well known "cumulative causation theory" (Myrdal 1957) and Perroux’s "growth pole theory" (Perroux 1955, Boudeville 1966). The "cumulative causation theory" recognizes the unequal endowment of regions in terms of human and natural resources and skills and posits that development starts from regions with higher endowments. Industrialization of one region implies transfer of capital from the agricultural to the industrializing region and, hence, increasing inter-regional wealth disparities. The process becomes cumulative as the developed regions dominate the underdeveloped draining them of products and factors of production (the "backwash" effect in Myrdal’s terminology). There is the opposite tendency, however, of growth diffusing from the developed to the underdeveloped regions in the form of, e.g. increased agricultural demand, causing the industrialization of the latter (the "spread effect"). Overall, however, superior regions continue to grow and dominate all other. Despite the abstract treatment of land and land use (as well as of time), this theory implies a mechanism of land use change within and between origin and destination regions – from developed agricultural to abandoned agricultural or to industrialized agricultural, from non-industrial to industrialized regions, etc. However, because regional imbalances may take an indeterminate number of forms and the origin of the regional development process is not explained – development is assumed to start spontaneously from some region with an original advantage (Cooke 1983, 121), the cumulative causation theory neither postulates orderly land use patterns nor can offer rigorous explanatory assistance in concrete cases of land use change.
The "growth pole theory", originally conceived by Perroux (1955) and later expanded by Boudeville (1966), moves along somewhat similar lines in that growth is assumed to originate in some region where a propulsive industry is located and then spreads to the surrounding regions. However, it is possible that contiguous regions are deprived of their factors of production and markets because of the growth of the growth pole. This theory describes processes of growth which have land use implications in both the growth pole and its contiguous regions but these implications are not spelled out explicitly by the theory. Moreover, the mechanisms that account for the growth of the pole are not specified, a fact that detracts from the explanatory power of the theory itself and its relevance to analyzing the determinants of land use change in this context.
3.3.2C Keynesian Development Theory
Another group of regional development theories is based on the Keynesian macro-economic theoretical framework such as the Harrod-Domar models, the export-base model, the factor-export models, neoclassical multiregional growth analysis (Cooke 1983, Hoover and Giarratani 1984, 1999, Andersson and Kuenne 1986, Bennett and Hordijk 1986). Their most important characteristic for our purposes is that they are "purely aspatial" theories; they do not even abstract from space, they ignore it. Because of this lack of spatial specificity they cannot be used to analyze directly land use change in specific regions and locations. Another feature of these theories is that they are demand oriented and ignore the supply side of the regions of interest – land suitability for various uses being a limiting factor on regional development (especially sustainable regional development); or, negative externalities being generated when land is put to uses for which it is not suitable. The mechanism of change (implicit in their static versions or explicit in their dynamic versions) is, naturally, exogenous changes in the demand for regional goods and services which may, subsequently, lead to social, physical and other changes. Keynesian-type macro-economic growth theories have found formal expressions in corresponding mathematical models of aggregate economies – those mentioned above as well as various versions of Input-Output models . If one accepts their assumptions and epistemological position, these theories can provide directions for the changes of the macro-economic determinants of land use change – incomes, investments, consumption, imports and exports. In fact, these theories underlie contemporary global models of land use change (at the aggregate world level mostly) which will be examined in the next chapter.
3.3.3. Other Theoretical Approaches in Regional Science
The broader field of Regional Science, although dominated by economics-based theoretical approaches to spatial change, contains a rich variety of several other approaches which attempt to describe the structure and evolution of spatial systems and, consequently, underlie the particular analysis of land use change. Of the extensive and variegated literature on the subject, two theoretical streams are briefly mentioned below: (a) Social Physics and (b) Urban and Regional Mathematical Ecology.
Social Physics is a term used to denote an approach to the study of social phenomena by drawing analogies from Physics. It is mostly concerned with the notion of "interaction" among individuals and groups and aims to explain human interaction on the basis of laws governing the motion of particles in physics. Its first codification is attributed to Carey (1858) while Ravenstein (1885) applied the idea to the study of migration. This theoretical stream has provided the basis for the gravity model which is presented in detail in Chapter 4. In this context, the magnitude of the interaction between two interacting activities located at a distance d from one another in space is proportional to the "mass" of these activities and inversely proportional to the distance between them. The "masses" of the activities are approximated variously depending on the activity considered. Most common measures include population residing at the two points in the case of residential activities (or population income) and floorspace or sales in the case of shopping activities. Measures of distance vary from physical distance or time to aggregate, composite measures taking into account various aspects of the "friction of space" which usually reduce the magnitude of interaction. More details are given in chapter 4 in the context of the gravity (or, more generally, spatial interaction) models.
The theoretical framework of Social Physics has been applied to the study of urban and regional phenomena which involve interactions such as trade and migration. It has framed also the study of urban and regional spatial structure; namely, the study of the location of residential and commercial areas which are linked through the transportation network. In this framework, change occurs when either the "masses" change or when the "friction of space" separating them changes (due, e.g., to improvements in the transportation network). One of the main objections to the use of Social Physics for the analysis of urban spatial structure and change is grounded on the lack of an economic rationale for the interactions analyzed. Niedercorn and Bechdolt (1969) derived the operational expression of the theory – the gravity model – starting from micro-economic principles of utility maximization. The theoretical issue that remained, however, was the application of a macrolevel theory to the explanation of individual level phenomena (Hanes and Fotheringham 1984). Wilson (1967) introduced concepts from statistical mechanics as well as the concept of entropy to derive the gravity model starting at the macro rather than at the micro level. However, this approach also ran into the problem of using concepts from physics to analyze social phenomena.
Entropy is another concept borrowed from Physics, in particular from Thermodynamics, which has been used to analyze urban spatial structure and change (see Wilson 1970, 1974; Wilson and Bennett 1986). Entropy measures the amount of uncertainty in a system of interest. A macrostate of this system relates to a number of possible microstates which arise out of the interactions of the individuals in this system. Entropy measures express the relationship between the macrostate and the microstates that correspond to it. A value of zero indicates a completely certain system – there is only one microstate of the system which coincides with the macrostate. When all microstates are equally possible, entropy obtains its maximum value – there is complete uncertainty in the system. In the analysis of the distribution of population and uses of land in an urban system, the entropy-maximizing principle is employed. In other words, the most possible macrostate of the system is found subject to certain constraints. This approach has been used to approximate the distribution of actual land use patterns or the most likely distribution of these patterns which results from changes in the characteristics of the system (e.g. changes in the transportation network, changes in the location of people or of employment). Despite criticisms of the Social Physics approach to the study of social phenomena, mainly its lack of a rigorous grounding on economic or sociological theories, it has found several applications in the analysis of spatial structure and change as the associated gravity and entropy models have been verified in several empirical situations.
In the 1980s, other concepts from physics were employed to the analysis of urban spatial structure and growth; namely, the concept of fractal growth and fractal structures (see, for example, Tobler 1979, Batty et al. 1989, Fotheringham et al. 1989, Frankhauser 1991, White and Engelen 1993). The process of urban growth and the resulting patterns are paralleled to the growth of organisms (e.g. corals) or particles (e.g. water drops, zinc oxide particles) which leads to particular fractal patterns. The notion of diffusion-limited aggregation is applied to the case of urban settlements to simulate their growth. Diffusion-limited aggregation (DLA) "refers to a process whereby a structure grows through the accretion or aggregation of units which diffuse over space until they reach a point on the periphery of the structure where they ‘stick’." (Fotheringham et al. 1989, 56). Various behavioral assumptions are used to "guide" the DLA process which is used as a means of uncovering the order which underlies the apparent chaotic structure of modern urban settlements. Various elements such as the contiguous nature of development, the tentacular nature of urban growth, and the presence of density gradients have been explored on the basis of this approach. The broader theoretical framework of fractal analysis is the basis for the development of cellular automata models to be discussed in chapter 4. The use of the concepts of fractal growth and structure to the analysis of urban growth, like the previous concepts from physics, lacks a rationale based on economic and sociological theories. The models developed on the basis of these concepts may replicate observed patterns and growth processes but they do not get at the causes, they do not answer the "why" of these processes and patterns. They are functionalist approaches whose explanatory power is poor compared to other theories firmly grounded on economic and/or sociological principles. For other recent proposals to use concepts from Physics and Chemistry for the study of urban and regional phenomena see, for example Isard (1999).
Urban and Regional Mathematical Ecology is another theoretical stream concerned with the study of patterns and processes of urban and regional growth. It borrows ideas and concepts from Ecology as well as from the sociological theories of the Chicago School of Human Ecology (see next section 3.4.) and applies theories from Mathematics (see, for example, Wilson 1981, Dendrinos and Mullaly 1985, Nijkamp and Reggiani 1998). Cities and population residing in cities are paralleled to animal species in nature whose interactions are governed by symbiotic, predatory, competitive and other types of ecological relationships. These parallels are transferred to land uses which are seen as appearing in certain places and growing while other land uses in other locations shrink in size or disappear. Ecological relations are analyzed both within and between cities with the ultimate aim to obtain the spatial and growth patterns which result from these relations. In the intra-urban case, both the open and the closed city cases are analyzed where a dynamic version of the standard urban land rent model is derived which allows for various types of behavior besides equilibrium (Dendrinos and Mullaly 1985).
The emphasis of this theoretical stream is on the macroscopic features of urban and regional phenomena and it claims that these can be analyzed by focusing on the most important qualitative features of urban evolution observed at particular time periods. Urban and Regional Mathematical Ecology attempts to analyze the dynamic behavior of urban and regional systems such as the existence of urban cycles, the sudden growth or disappearance of settlements (discontinuities), suburbanization, slum formation, gentrification, etc. It addresses the issue of dynamic, non-linear interdependencies, stability, smooth and abrupt evolutionary change, and multiple equilibria of spatial phenomena and it aims at providing an appropriate basis for modeling these phenomena. Towards this purpose, it combines elements of mathematical population ecology and the mathematical theory of bifurcation to create a framework for the analysis of urban evolution which comes close to the general theory of evolution.
The presentation of the many applications of Urban and Regional Mathematical Ecology to the analysis of urban and regional evolution as well as to the models that have developed on their basis are beyond the intent of this contribution. The reader is referred to the references provided for further study. The basic point with respect to this theoretical stream is that it focuses on aggregate form, behavior, and processes and does not deal with land use explicitly. In addition, it can be subjected to the same evaluation as the Social Physics theoretical stream; namely, it is a functionalist and positivist type of theory which lacks economic or social theory foundations despite the fact that it can describe satisfactorily observed urban and regional phenomena.
3.4. The Sociological (and Political Economy)
Theorization Tradition
The sociological theorization tradition draws from the way of thinking in Sociology and in the broader realm of the Social Sciences (Anthropology, Psychology, Political Science and related disciplines) which, compared to Economics, is more diverse and variable. In general, theorization in this tradition emphasizes the importance of human agency, social relationships, social networks, and socio-cultural change in bringing about spatial, political, economic, and other changes. The term "social" is used here in a broad sense which encompasses all manifestations of society – modes of production , institutions, politics, culture, life styles, etc. Hence, a broad variety of factors are introduced in the analysis of spatial structure and its change whose relative importance depends on the particular discipline of the Social Sciences from which they derive. Similarly, the behavioral assumptions made and the view of reality adopted depend on the "mother" discipline as well as the theory’s epistemological position. As it was the case with the urban and regional Economics theorization tradition discussed before, some theories in the sociological tradition treat space, in general, and land use, in particular, explicitly as areas on earth’s surface with particular properties (with various degrees of abstraction, of course) which have to do with its use and change as well as with the determinants and implications of this change. In contrast, other theories are "aspatial" in the sense that although they deal with spatial relations as they change under the influence of social changes, they treat space and land use abstractly (in the background) necessitating a "translation" of their findings to concrete contexts and uses of land.
Given the breadth and diversity of these theories, their categorization is neither easy nor straightforward. An attempt is made here to group together theories that share a common set of concepts about spatial structure and its change. Five groups are presented in the following: functionalist/behaviorist theories, structuralist /institutionalist theories, core-periphery theories, unequal exchange theories and uneven development-capital logic theories. Needless to say that, because this classification is not unambiguous, theories in one group can be classified under another group (as it is the case with the first two classes that may contain/include theories from the other three). Table 3.1c presents in greater detail particular theories included within each one of these five groups.
3.4.1. Functionalist – Behaviorist Theories
Two classes of theories are discussed as most representative of the functionalist/behaviorist theoretical approaches to the study of land use change: the well known "family" of human ecological theories and another class of theories of (urban) spatial structure originating in the field of planning.
3.4.1A Human Ecological Theories
Human ecology is a term coined to denote a sociological approach which borrows concepts and ideas from the field of Ecology and applies them to the analysis of the relationships of humans with their physical and social environment mostly in urban areas. It was developed in the 1920s by sociologists of the Chicago School – Robert Park, Ernest Burgess, Roderick McKenzie and others. Later, it was systematized by other scholars – Amos Hawley, James Quinn, Brian Berry, J. Kasarda, among others (Johnston et al. 1994, Romanos 1976). Human ecology advances the "idea that cities are the outward manifestation of processes of spatial competition and adaptation by social groups which correspond to the ecological struggle for environmental adaptation found in nature" (Cooke 1983, 133). Basic ecological concepts used to describe social groupings and processes include "community", " invasion ", " succession ", " adaptation ", " dominance ", "disturbance, "competition", "climax equilibrium" (Johnston et al. 1994, Romanos 1976, Chapin and Kaiser 1979).
In its earliest versions, human ecology viewed the urban development process as producing and maintaining an equilibrium system and assisting the urban system to return to a stable order following any disturbance. The state of equilibrium and order resulted from a struggle for survival of different communities where the most powerful occupied the best locations in the city while the rest occupied the remaining space. The structural form of the city expressed the dominance of the fulfillment of the needs of industry and business which was then followed by the fulfillment of the residential needs of the population. The urban system was viewed to develop through processes of invasion and succession ; new interests invade certain parts of the city succeeding the former occupants who, in their turn, move to (invade) other parts, and so on. This process gives rise to particular land use patterns – the concentric ring (or, zone), the radial sector and the multiple nuclei patterns.
The concentric zone theory was proposed by Burgess (1925) to describe city patterns resulting from the ecological processes presented above. A monocentric city consists of five concentric rings containing particular urban functions; the center (the "loop") is occupied by commercial, administrative, financial, and recreational facilities. It is surrounded by a "zone of transition" which is occupied by poor and old residential property and run-down areas that have been invaded by business and light manufacturing as the CBD expands. The third zone contains the homes of the working class while the fourth is a high class residential area (white collar and middle-class families). The fifth zone is devoted to suburban and satellite development (Figure 3.2c). As the city grows, each zone extends to the next, outer zone in the process of "invasion-succession"; this is the proposed mechanism of change of the concentric zone theory which, however, does not explain the "why" of city growth. It is evident that this conception of the urban land use structure and its change bears close similarities to those suggested by von Thunen and Alonso on the basis of other (although not dissimilar) arguments.
The radial sector theory was proposed by Hoyt (1939) who argued that similar types of (residential) land uses occupy wedge-shaped sectors extending from the city center along transportation routes (Figure 3.2d). High-rent residential areas occupy certain sectors and rents decrease in all directions away from those areas. Adjoining residential areas are occupied by intermediate income classes while low-rent areas occupy other sectors extending similarly from the city center to the periphery (Chapin and Kaiser 1979, Romanos 1976). In this theory, the needs of high income groups dictate the patterns of urban expansion and residential relocation. The mechanism of change, thus, rests with the changing wealth and (spatial preferences) of those groups and, thus, cripples the theory of accommodating other forces of change – e.g. the influx of a large number of workers drawn by new manufacturing who create a demand for low and medium income housing (Romanos 1976). Hoyt’s radial sector concept bears similarities to Burgess’s concentric zone concept (Romanos 1976) and it has been criticized for its many defects (Lloyd Rodwin 1950 cited in Romanos 1976, 153-154).
Finally, the multiple nuclei theory of urban land use structure was suggested originally by McKenzie (1933) and it was expanded later by Harris and Ullman (1945) in an effort to overcome some of the restrictive assumptions of the previous two theoretical schemes (especially, the monocentric city assumption). Drawing on the observation that urban land uses are organized frequently around particular nuclei (pre-existing agglomerations or new centers of activity) rather than around a single center, they proposed a city structure that is schematically represented in Figure 3.2e. The number and functions of the nuclei differ from city to city. The mechanism of change of this structure – the emergence of new nuclei – is attributed to: (a) the need for specialized facilities by certain activities, (b) agglomeration economies, (c) agglomeration diseconomies, and (d) the effect of the city rent rate structure on attracting or repelling certain activities (Romanos 1976, Chapin and Kaiser 1979). High income groups occupy the most desirable locations while low income residents are clustered in noxious environments. The multiple nuclei theory asserts that, as industrial societies become more complex and wide-ranging in their organizational scale, the social composition of city districts changes as a function of this increasing (social) differentiation. The changes residential areas undergo – differentiation and segregation – are caused by the changing economic status, extent of acculturation to urban ways of living, and ethnic status of individuals and households.
All three theories presented above are static descriptive devices of urban land use structure with an apparent focus on residential land uses. Although not stated directly, the underlying land allocation mechanism is similar to the urban land theory’s bidding process where the determining factor is the ability of a user to pay for the price of a particular site in the city. None of these theories, however, explain the "why" of the processes of change in the land use patterns – the factors that account for the growth and decline of economic activities, the dominance of certain activities, the changes in preferences and other constraints (e.g. institutional) on land development and use. Human ecological studies that developed after the 1950s "have tended to play down the spatial focus of the Chicago School …. in favor of an emphasis on the demographic and institutional dimensions of society (Saunders 1981) although, at the same time, they have shown a strengthened interest in human interaction with the physical environment. Sociological human ecology has also moved away from those aspects of the Chicago School – embodied more in its ethnographic research monographs than in its theoretical pronouncements" (Johnston et al. 1994, 258).
Closely corresponding to the human ecological notion of succession, but not restricted to urban areas, the concept of "sequent occupance" was advanced by Whittlesey (1929) to describe the changes in the geography (landscape) of an area over long periods of time as "a succession of stages of human occupance which establishes the genetics of each stage in terms of its predecessors" (Whittlesey 1929 cited in Johnston et al. 1994, p. 549). It is worth noting Whittlesey’s qualifications of the concept. "While ‘human occupance of area, like other biotic phenomena, carries within itself the seed of its own transformation’ … such uninterrupted or ‘normal’ progression were ‘rare, perhaps only ideal, because extraneous forces are likely to interfere with the normal course, altering either its direction or rate, or both’ and ‘breaking or knotting the thread of sequent occupance’" (Johnston et al. 1994, 549). A well known application of this notion is Broek’s study of the Santa Clara Valley, California (Broek 1932). Korcelli (1982) cites applications of the notion by Hoover and Vernon (1959), Duncan, Sabagh and van Arsdol (1962) and Birch (1971) (Korcelli 1982, 96-97).
Before moving to the next class of theories in the functionalist/behaviorist group, a clarification and word of caution is in order. The term "human ecology" is being used in another sense in addition to the sociological covered in the preceding presentation. As Johnston et al. (1994) note "it is nonetheless frequently used in contemporary nature-society geographies … that continue Barrows’ focus on human adjustment to the natural environment, emphasizing the interactive and adaptive character of the human-nature interaction and its mediation by social institutions" (Johnston et al. 1994, 258). This different theoretical version of human ecology will be discussed in the context of the nature-society theories discussed below.
3.4.1B Planning Theories
Another class of theories of urban and regional spatial structure originated in planning circles. Cooke (1983) calls them "The Berkeley School" (not to be confused with the Berkeley School of nature-society theories associated with Carl Sauer to be discussed in the next section) and comprises contributions to theorizing about urban structure by Melvin Webber, Donald Foley and Stuart Chapin (see, Webber 1964). Common to theories in this class is the belief that spatial structure is a reflection of social structure and that new social norms give rise to changing spatial forms. Foley’s urban structure theorizing builds on Talcott Parsons’ hierarchical framework on which social life is structured and which consists of four systems: the social, the cultural, the personality and the physical system (Foley 1964, Cooke 1983). The four systems are related through particular relationships that lead to ordered social relations with norms playing a central role in this process. What is striking, especially for our purposes, is that, in Foley’s schema, "the physical system is analytically unimportant except in so far as it imposes functional prerequisites upon the other three systems. The functional prerequisite it imposes on the social system is one of adaptation of social organization such that system maintenance is ensured …. Foley conceptualizes a relationship between cultural values and the spatial form of the city which is more reflexive than that of Parsons. …Cultural norms may be understood as receiving partial expression in the built environment, but they are themselves modified by the cognitive feedback offered to the cultural system by the resulting physical system. In this way, abstract but fundamental social values receive spatial expression and the link between the sociological and the geographical spheres is forged….. Physical lag occurs when the built environment fails to respond to social system needs" (Cooke 1983, 88-90).
Webber’s theorizing moves along similar lines emphasizing human interaction as the basis of urban communities. In his famous contribution "The urban place and the non-place urban realm" (Webber 1964b), he distinguishes between human interaction within the confines of a metropolitan area (a "place community") and human interaction that extends over scattered places on the face of the earth (a "non-place community"; their set constituting "urban realms"). Improved transportation and communication systems extend the interactions among individuals, firms, organizations and institutions to a global level making thus inadequate their analysis within an urban region (Chapin and Kaiser 1979). These dynamic interactions are traced through linkages centered upon interests rather than upon contiguity; hence, his conceptualization of "community without propinquity". The spread of communications technology "frees spatial structuring from the locational constraints exerted by localized linkage patterns and heralds the emergence of a formless spatial development in ‘nonplace urban realms’" (Cooke 1983, 91).
Similarly, Chapin’s theorizing on residential activity systems (Chapin 1965, 1968) is premised on the idea that the needs of human interaction is the key to the spatial organization of cities and adopts the sequence: basic values – activities spatial patterning. In his view, the personality system (see, Foley’s schema above) is instrumental in explaining residential mobility and competition in the housing market. Rational households "struggling to measure up to internalized norms of appropriate behavior locate in the optimal spatial location. Hence, social structure reveals itself in spatial structure as the outcome of a game in which those who adhere most closely to the rules win the best prizes" (Cooke 1983, 91).
All three theorists of the "Berkeley School" presented above follow a functionalist-structuralist approach to the explanation of urban (or, more narrowly, residential) spatial structure and its change. The theories address the dynamics of the urban system in terms of changing social needs, norms, technology. No particular spatial patterns are proposed as it was the case with the human ecology theorists as they place emphasis more on the determinants of urban land use rather than on particular spatial patterns; space, land and land use are treated in rather abstract terms (as in the concept of "nonplace urban realms") compared to the earlier formulations reviewed thus far.
3.4.2. Structuralist - Institutionalist Theories
Structuralist -institutionalist approaches emerged from a general disappointment with and as a reaction to the idealism of the functionalist-behaviorist approaches to the description and explanation of urban and regional spatial structure, a selection of which was presented before (others will be discussed in the next section). The main point of criticism they advance against the functionalist approaches is that they ignore the social and institutional constraints on individual behavior. They propose alternative conceptualizations of urban and regional spatial structure and of the determinants and processes of its change most of which are premised on the belief that the main determinant of locational behavior is power. Hence, analysis of spatial patterns has to be based in the relevant political economy (Johnston 1982, 83). Conflict between unequals, which has a class basis mostly, is an instrumental concept in these approaches many of which have taken anti-capitalist, Marxist ideological stances. Diverse theories exist which can be divided broadly into those concerned with the urban, metropolitan level and those referring to larger scales. This section discusses briefly the former group; the latter is covered in the next sections.
The common theme of most structuralist-institutionalist theories is urban development in late capitalist societies. Hence, their relevance is confined to those societies and cannot be transferred easily to the analysis of spatial development either in past periods or in contemporary societies that do not conform to this type of political economic system. In this particular context in which they operate, the analysis focuses on the (capitalist) mode of production and the ways in which it structures space and spatial relations. The latter reflect the tensions and conflicts between capital (associated with the production space) and labor (associated with the consumption space). The state plays a critical role in mediating these conflicts with the purpose of supporting those relations which contribute to the capital accumulation process. Various state functions (planning, provision of services, etc.) contribute to shaping space towards this end.
Structuralist -institutionalist theories differ between them with respect to the conceptualization of space, spatial relations, the locus and nature of conflict, and the choice of mechanisms by which the system of power relations works and shapes space. Castells’s Structuralist -Marxist theoretical contribution (Castells 1977, 1978) focuses on collective consumption and the power of the state, as the main supplier of collective consumption services, to control urban structure in ways beneficial to the interests of the owners of capital. Castells relies on Althusser’s structuralist theory of social formations (Althusser and Balibar 1970) which distinguishes three main levels in the organization of society: the economy, the state, and ideology. The economy is dominated by the capitalist mode of production whose main elements are the relations of production – owners of capital vs. laborers and the forces of production – technology, the structure of the division of labor, machines, structures, etc. Castells assumes that there is a relative autonomy between the three levels and identifies the main processes through which their interrelationships influence the development of the urban spatial structure; how the dominant mode of production shapes space. The spatial expression of the economy includes: (a) the production space – industry and offices, (b) the consumption space – with elements of the reproduction of the labor power: housing and welfare services, and (c) the exchange space – transportation and communication networks. The administration space – local government and urban planning – relates to the economic space but also to the exchange space. The organization of production is conducted mostly, although not exclusively, at the regional scale while the reproduction of the labor power is an urban level activity. Hence, a link is postulated among all three – the sphere of reproduction, the urban level and consumption. Collective consumption structures urban space. The state supplies collective consumption services through its planning apparatuses and, in this way, controls the process by urban spatial structure is formed.
The concept of urban social movements is central in Castells’ explanatory schema, the theory of reproduction. Urban struggles develop as dominated groups (the labor class) in urban areas come into conflict with state policies which aim to maintain those social relations that further the interests of the capitalist, ruling class. This is very broadly the mechanism by which Castells attempts (though not successfully as his critics contend) to explain the development process and spatial change in urban areas (for a collection and synthesis of the critiques, see Cooke 1983; also Mingione 1981).
In another perspective, Scott (1980) focuses his analysis on the urban land nexus which denotes the differential locational advantages offered by the intersection of variable land rent with the spatial requirements of households and firms. The urban development process is conceived again as resulting from the conflicts between capital (over the distribution of profits) and labor (distribution of wages). The state legitimizes capitalist social relations and assists capital accumulation through welfare and other subventions. Urban development is seen as "a function of changing capital to labor ratios among firms as they engage in technical switching to maximize profits. Increased capital intensity associated with investment in technology is accompanied by increased decentralization of location from the urban core….these changes in the production space stimulate responses in the reproduction space as households seek suburban locations closer to employment centers … The state is heavily involved in unraveling the spatial knots to which this process gives rise, especially in the spheres of reproduction and circulation, to overcome market failures in the provision of housing and transportation facilities… Urban planning performs its main functions by solving land use dilemmas ….. and smoothing the dynamics of land development" (Cooke 1983, 145). Scott attempts to integrate planning and urban theory, to analyze the peculiarities of the land market and its role in focusing urban development unevenly, and to introduce the notion of civil society to explain the development process.
David Harvey’s voluminous and influential contribution to the analysis of the urban development process in capitalist societies is impossible to summarize in a few lines. The reader is referred to his writings as well as to the analysis of his works by other scholars (see, for example, Harvey 1973, 1975a, 1975b, 1982a, 1982b, Cooke 1983). It is also difficult to categorize his theorizing in just one category – such as the institutionalist as it is done here – as it is broad and relates to several other theoretical currents within the sociological/political economy tradition; in particular, the uneven development-capital logic group of theories discussed below. Harvey adopts a Marxist theoretical and analytical framework to analyze spatial forms and processes as the result of socio-political inequalities inherent in capitalist societies. He theorizes on the recurrent crises in capitalist societies which result from the inherent contradiction between class struggle and capital’s drive towards accumulation. He explains urban development in late capitalism in terms of the generation of massive economic surpluses, underconsumption, and the state’s direct and indirect involvement in the process of modifying the built environment, among others, to support the interests of the capital. Conflicts and struggles between labor and capital are manifested in the ways the built environment is produced, manipulated, and used. Harvey has proposed a three-circuit schema to represent the connection between finance capital and the built environment. Over-accumulation of capital in the primary circuit of production causes capital to switch to a secondary circuit – investments in the built environment. Furthermore, capital may switch to a third circuit – investments in various welfare services (Harvey 1975b, 1982b). To overcome economic crises, capital seeks a spatial fix (Harvey 1982b) – spatial arrangements that help solve the crises. This is how investment in one place and disinvestment in another over time is explained as well as the ensuing spatial transformations. Harvey’s analysis, although focusing on late capitalist societies and the urban environment is broad enough to frame the qualified analysis of the land development (and, consequently, the issue of land use) in other contexts.
Structuralist -institutionalist theories in the vein of those presented above provide potentially valuable insights and analyses of the political and institutional determinants of land use and its change in urban areas of developed, capitalist countries. Functionalist undertones are apparent in several of them. Land and land use are treated in rather abstract manner as the main emphasis of these theories is not on land itself as on the forces that impinge on the uses of land. Hence, these theories do not explain always the critical connections between the political and institutional factors and the resulting land use patterns except for broad accounts of such phenomena as urbanization, suburbanization, decentralization of production and other forms of development. This caveat may be attributed, in part at least, to the relative neglect of the role of human agency and their excessive focus on aggregate forms of social organization. A theoretically interesting and operationally useful task is their "translation" into more spatially-explicit and place-specific contexts. This will necessitate a broadening of the explanatory schemata advanced by including additional factors as well as by developing more explicit linkages between the development process at various spatial levels.
3.4.3. Core-periphery theories
Core-periphery theories represent another group of theoretical approaches which refer mostly to higher than the urban (up to the global) levels. Their purpose is to describe and explain the spatial organization of human activities premised on the idea of unequal distribution of power in socio-economic and political affairs. In this way, they can be considered to provide general level theorization schemata about the uses of land and changes of land use resulting from the relations of dependence which develop between the core (a developed – but not necessarily spatially defined – region) and the periphery (underdeveloped regions). Its distinctive difference from similar theoretical approaches (dependency theory, unequal exchange theory) is perhaps that it subscribes to the goal of spatial equilibrium and ignores the uneven development of the division of labor in capitalist societies (Johnston et al. 1994, 95).
There are several variants of the basic core-periphery theoretical formulation. Its origins can be placed in earlier modernization theories (Lewis 1955 cited in Cooke 1983, 151) as well as in Rostow’s (1960) stages theory of economic growth. More generally, it can be embedded within a broad diffusion theoretic framework as the central idea is that development spreads out (diffuses) from a core region which contains the most modern economic sectors towards the periphery – regions which are at the first or pre-industrial stage of development (Cooke 1983, 151). Actually, Friedmann (1966) – with whom the core-periphery model is more closely associated – defined the core-periphery relations as the second of a four-stage sequence of the development of the space economy. These stages are: (a) pre-industrial society with localized economies; (b) core-periphery; (c) dispersion of economic activity and (d) spatial integration (Johnston et al. 1994, 95). A more radical variant of the core-periphery idea, the internal colonialism concept, has been formulated by Hechter (1975 cited in Cooke 1983, 151) who rejects the assumption of spatial equilibrium of the original idea. Hechter posits, instead, that industrialization spreads in a spatially uneven way. Groups in the core region reinforce their advantageous position and the periphery is forced into specialist, complementary functions. Continued monopolization of peripheral trade, credit and employment patterns by core institutions prevents equalization of wealth between core and periphery (Hechter 1975 cited in Cooke 1983, 152).
A more general, global version of the core-periphery dualism is the world- system theory (Wallerstein 1974, 1979), a theorization of how core, industrialized regions relate to the underdeveloped periphery on a world scale. Colonial expansion undertaken by a capitalist, industrialized core region results in a structuring of the relations of the world regions with one another according to the international division of labor. The international division of labor reflects the functionality of different areas for particular types of production. Between core and periphery, Wallerstein introduced the concept of the semi-periphery, "countries that have regressed from core status through undergoing a process of deindustrialization and those heading for core status as they experience rapid industrial development" (Cooke 1983, 153).
Evidently, this group of theories deals, in general, with broader development issues of which land development (and the implied land use change) is but one. Land and land use is treated in some of them in some way (e.g. when agricultural production and systems are analyzed) although the level of detail is necessarily coarse, a consequence of the coarse level of analysis. Most of the theories are dynamic as they indicate stages of economic development although in a prescriptive/normative sense mostly. However, they can be used to frame questions and answers about the determinants of land use change at higher spatial scales as well as to be combined with analyses of land use change at these scales.
3.4.4. Unequal exchange and dependency theories
Sharing more or less the basic core-periphery dichotomy with the previous group, a number of related development theories are discussed in this section. Unequal exchange theories build their explanatory schemata around the mechanism of exchange to account for the nature of socio-economic and other relationships developing among regions. Dependency theorists emphasize the power of external forces in imposing a situation of dependence on underdeveloped countries.
Unequal exchange theories (Emmanuel 1972, Amin 1976, 1978 cited in Cooke 1983, 154) draw from propositions of the labor theory of value (and the broader trade theory) which state that, because of differential costs of reproduction of the labor power between countries, the commodities exchanged are not equivalent in terms of "socially-necessary" labor time (technology, labor productivity and transport costs assumed constant). Usually, in developed, industrial countries real wages are higher than in the less developed countries with inferior productive facilities and large, unorganized labor reserves. The differential exchange of purchased (through commodity trade) labor power produces an exploitative relationship which is to the advantage of the high wage regions and to the detriment of low wage areas (Cooke 1983, 155). "Unequal exchange may help to maintain a permanent inability to gain from trade by the systematic extraction of value from underdeveloped economies and by the development of a permanent development gap. This may not only result in the increased penetration of imports into developed economies, but may also undermine traditional modes of production and intensify technological dependence in the underdeveloped economies" (Johnston et al. 1994, 637).
The broader idea of unequal exchange is common to the uneven development theories which are discussed below. In addition, it is observed that the notion of differences between countries in terms of "socially-necessary" labor time embodied in commodities resembles the notion of differences between countries in the "resources import content" of economic sectors and embodied in related products which is accounted for by the "environmental space" and the "ecological footprint" concepts (Hille 1997, Wackernagel 1993). Land is one of the principal resources which is "imported" by industrialized countries in the form of products produced beyond their national boundaries. More specifically, if a country or region consumes more than its own resources can produce – beyond its carrying capacity – it appropriates resources from other regions; this is the notion of the " appropriated carrying capacity " (Rees 1996). Drawing a parallel to the unequal exchange theories, commodities exchanged are not equivalent in terms of "resources import content" or of "appropriated carrying capacity". The differential exchange of purchased (through commodity trade) land and carrying capacity produces an exploitative relationship where land use and its change in underdeveloped countries is controlled by the demands of the developed countries.
Dependency theories (Frank 1967, 1972, 1979, Dos Santos 1970), although different in several respects from unequal exchange and core-periphery theories, stress the conditioning situation of dependence of underdeveloped on developed economies which is produced from the imposition of particular forms of development and technology by international enterprises in advanced industrial societies on peripheral countries. Hence, the latter are conceived as being locked in a permanently asymmetric economic relationship with the former which prevents them from breaking out from their underdeveloped status. An alternative, less deterministic, version of the dependency thesis – the associated-dependent development (Cardoso 1973) – holds that there are changing forms of dependency as factors internal to the economies of the dependent areas (e.g. the prevailing class structure and the role of the state) are as important as those external to them.
Both the unequal exchange and the dependency theories do not refer directly to land use issues but to critical determinants of land use and its change focusing specifically on the international level. As it was the case with the core-periphery theories, they do not make explicit the links between the relationships on which they focus and the resulting land use patterns and changes although these can be derived from the context of the broader analysis of the issues of inequalities and relations of dependence and exploitation. Nevertheless, the relatively high level of spatial abstraction of these theories makes them suitable, if one accepts their ideological and epistemological positions, as broad guiding frameworks for elaborating the particular influences of changing social, economic and technological relations between developed and underdeveloped countries on the land use patterns of both.
3.4.5. Uneven development – Capital logic theories
A last group of theories which overlaps considerably with all those discussed previously (with the exception of the functionalist-behaviorist theories) revolve around the general theme of uneven development. Uneven development is defined as "a systematic process of economic and social development that is uneven in space and time, and endemic to capitalism. … (It) is a basic geographical hallmark of the capitalist mode of production…. combining the opposed but connected processes of development and underdevelopment" (Johnston et al. 1994, 648-649). Uneven development is closely related to the logic of capital accumulation (hence, the label "capital logic" used by Cooke 1983) and, thus, takes place at all geographical scales. Hence, the employment of the concept in various theories which focus on particular spatial scales. The unequal exchange and dependency theories presented above may be considered as global scale applications of the concept of uneven development. Two theories from the uneven development-capital logic group are presented briefly below followed by an outline of the main elements of the broader theory of uneven development.
At the sub-global level, an influential theory is Lipietz’s theory of unequal regional exchange (Lipietz 1977, 1980) which examines the ways in which different modes of production connect across inter-regional space. Underdeveloped regions are characterized by pre-capitalist (e.g. peasant and petty-commodity) modes of production while developed regions are dominated by the capitalist mode of production. Development follows a phased course. First, control of petty commodity production passes to financial institutions as credit is extended to enable producers purchase the means of more intensive production. Then, the processing of agricultural output passes into the control of centralizing capitalist industries. The emergent spatial division of labor, based on unequal exchange between regions with different levels of technical development provides the setting for unequal exchange in terms of differences in wage levels. Underpayment and the subsequent squeeze upon living standards stimulates rural depopulation and migration to urban-industrial agglomerations. The theory describes the spatial implications (concentration and decentralization) of the course determined by the logic of capital accumulation in both the dominant and the dominated regions (Cooke 1983).
A less deterministic and restrictive, flexible theorization of uneven development which focuses on the sub-regional, local level (in industrial, capitalist societies) is offered by Massey’s (1984) theory of the spatial divisions of labor (see, also, Massey 1978, 1979, 1980 cited in Cooke 1983, 162). This theory is distinctive in that it seeks to relate the operation of the processes of capital accumulation to the areal differentiation of the space economy recognizing that
