Geography of Elections

Peter J. Taylor, Ron Johnston

• Publisher: Taylor & Francis
• Year: 2014
• Print ISBN: 9781138814271
• eBook ISBN: 9781317601869
• Edition: 1
• Topic: Physical Sciences
• Subtopic: Geography
• Index: Geography

Within an international framework, this work provides a fully comprehensive approach to the geographical coverage of elections. Numerous applications of ideas and concepts from human geography are incorporated into a new political context, illustrating the manner in which electoral patterns reflect and help produce the overall geography of a region or state. Discussions of various topics are well supported by numerous maps and diagrams which help clarify arguments and serve to define elections within their basic geographical context.

Part One

Preliminaries

1 Geography and Electoral Studies

The study of elections has been a continuing source of interdisciplinary conflict, largely between political scientists and sociologists: the latter are accused of taking the politics out of voting studies (Key and Munger, 1959), whereas the former are accused of insufficient attention to sociology (Sheingold, 1973). Geographers have not been involved in this conflict, and it is no intention of ours that this book should launch a triangular contest. We make no claims for any distinctive discipline of electoral geography and have no imperial ambitions; our aim is to illustrate the richness of inter-disciplinary study of voting, from a geographical bias.

Although the roots of electoral geography can be traced to the early years of this century (Siegfried, 1913; Kareil, 1916; Sauer, 1918), research by geographers into various aspects of voting has never been more than slight in its volume. With the possible exception of French work, the heritage for current workers in electoral geography consists only of a disconnected set of isolated studies, few in number and limited in purpose. More recently, geographers have developed a greater interest in the topic of elections. An initial programmatic statement suggested a limited scope for electoral geography, based on a traditional concern with maps (Prescott, 1959), but growing awareness of the value of quantitative approaches and of the richness of work in other disciplines has widened the geographer’s horizons quite considerably (McPhail, 1971). A small annual volume of research is now published on electoral themes, which has led one author to claim – much more than we would – that ‘Geography is particularly important in that it can add an entirely new dimension to the study of elections’ (Busteed, 1975, p. 3).

As in most disciplines, new developments in geography produce excesses, and there are some studies which, on hindsight, have produced rather obvious results. Others have explored interesting corners of the field, however, and have unearthed some fascinating findings. It is their impact which justifies our writing a geography of elections, despite an obvious reliance on a wide range of material from other disciplines, notably political science and sociology.

There is no set of topics which can be designated as exclusively the domain of electoral geographers, as indicated by our catholic sources in this book. The result is thus a survey of relevant material which is clearly inter-disciplinary in what we hope is the best sense of that term. Research results and hypotheses from within and beyond the usually defined boundaries of geography are integrated here in a survey of three major topics. The only criterion for inclusion has been relevance to the theme, so that although much of the research is not geographical in origin it is clearly geographical in interest; conversely, all of the works cited, whether by geographers or not, are of interest to those who would understand the full complexity of elections.

Modern human geography is often defined as having a focus on the spatial distributions of phenomena and the processes by which these are generated. In such studies, the adjective ‘geographical’ is used as synonymous with spatial or locational effects. To summarize, human geographers are concerned with the spatial organization of society (Morrill, 1970; Abler, Adams and Gould, 1971). This organization comprises two elements: (1) the de facto organization, which is the ‘natural’ reflection of society in spatial patterns – the continuous geographical distributions; and (2) the de jure organization, which is the set of spatial patterns explicitly defined for administrative purposes (Cox, 1973). The pattern of social areas or neighbourhoods in a city is an example of de facto organization, whereas that of parliamentary constituencies or congressional districts exemplifies that of de jure organization.

Given this general definition of the geographer’s sphere of influence, we can identify the many geographical elements in a typical election. Within a city, for example, voters are registered by their home addresses. Where these are depends on which parts (sectors) of the housing market they have access to; the most affluent have most choice and select the more ‘desirable’ areas, whereas the poorest have least choice and are usually allocated to certain prescribed areas. In some societies, other variables – such as race or religion – may combine with class to influence residential choice. The result in all cases is spatial segregation by socio-economic criteria, and perhaps by others as well. It is very often the case that these same criteria are important in shaping people’s political attitudes, and hence how they vote. We might anticipate, therefore, that the spatial patterns of voting – usually for various political parties – will also indicate segregation of people with different views and will reflect the underlying socio-economic segregation. The extent of the associations between these patterns – at all spatial scales and not just within cities – is a basic feature of geographical research into elections.

Within the patterns just outlined, where a person lives may have a still subtler connection with his or her voting behaviour. The location of a home prescribes who are the occupiers’ neighbours, and in many cases whom they have as friends and acquaintances. This pattern of social contacts may be a major element of the social environment within which voters make up their minds on how to vote, for different social milieux may involve different biases in terms of political discussion and advice. Further, different areas may receive different levels of attention from parties and their candidates in the attempts to win votes, thereby extending the range of the locational or geographical influences on voting decisions.

Finally, where a voter lives determines by whom he is represented, which may or may not be the candidate(s) for whom he voted. The determining factor is the set of de jure spatial districts – often termed constituencies – which form the organizational framework for elections and send the representatives to the parliament or similar assembly. The location of constituency boundaries can influence the political complexion of the district, and hence the nature of its representation, and so geographers are interested in the boundary-drawing process. Further, the location of these boundaries can have major influences on the overall election result, through the translation of votes into patterns of representation, which can produce a peculiarly geographical influence on representation.

We have identified three main foci of geographical interest in electoral studies. First, there is the geography of voting itself; secondly, there is the set of geographical influences on voting; and finally, there is the geographical influence on representation. These three form the framework for the remainder of the book; to provide initial illustration of each so as to set the scene for the greater detail to come, we conclude the present chapter with an example of a classical study from each substantive area.

The geography of voting in Ardèche, France

Many studies of the geography of voting involve the comparison of two or more maps, thereby associating the distribution of votes – usually for a particular political party – with various other distributions. From such associations causal connections are often inferred. This cartographic approach – now somewhat superseded by statistical methods of comparing distributions – has deep roots in French geography and is often associated with the work of André Siegfried, the ‘father’ of electoral geography.

André Siegfried and his géographie électorale have not always been without their critics in political science. He is sometimes viewed as an ‘environmental determinist’, that is, a researcher who explains social phenomena by relating them causally to the physical environment. Geography has had its fair share of such simplistic thinkers but Siegfried is not one of them. Although he is sometimes credited with the statement that in northern France ‘the granite votes for the right and the chalky soil for the left’ (Dogan, 1967, p. 183), Siegfried does not imply any such direct causal relationships in his researches. There is however, ‘a correspondence between nature of the soil, agrarian landscape, type of dwelling, distribution of land ownership, degree of stratification in society, the stronger persistence of tradition, and political orientation’ (Dogan, 1967, p. 183).

Dogan presents another example in which physical geography is superficially related to voting. In Italy the higher the land the greater the conservatism. This is not because land nearer to Heaven preserves more Christian ideals, but rather it is related to land ownership and the resulting stratification – ‘Thus geography leads back to economy, and this to history’ (Dogan, 1967, p. 183). Such explanations are based upon social factors which are themselves based in part upon the physical environment.

These relationships between the physical environment, the social environment and voting are excellently illustrated by Siegfried’s study of voting in the département of Ardèche, on the west bank of the Rhône, during the period of the Third Republic (1871–1940). In this (Siegfried, 1949), he interpreted the pattern of party votes as reflecting the spatial organization of society there, which in its turn reflects variations in the physical environment. The associations which he deduced were extremely constant in election after election. Those cantons favouring parties of the left in 1871 still did so in 1936, for example. In all, Siegfried classified the 31 cantons into 17 of the left and 11 of the right, with only three which varied greatly in their support during the long period.

What factors led to the spatial polarization of cantons within Ardèche? Siegfried’s conclusions on this are summarized in Figure 1.1. The main element in the physical base of the department is its geology, which is reflected by altitudinal variations in the landscape. In turn, these variations have influenced the types of productive activity – both agricultural and the incipient manufacturing industries – and in terms of human patterns are shown as variations in farm sizes and population agglomerations. Political attitudes are coloured by these different social and economic environments, and are reflected in the voting map.

Figure 1.1 and the short paragraph above are gross simplifications of a 136-page research monograph, which includes over fifty maps of voting patterns and another twenty used to help provide an account for those patterns. The essence of Siegfried’s approach is clear, however. Through careful consideration of the appeals of the parties of the left and the right, he could identify those aspects of the social and economic milieux most likely to underlie the voting decisions, and was then able to trace those aspects back to the physical environment. The argument is supported through map comparisons. Modern studies are more likely to use statistical procedures, and to stop short of tracing the associations back to the physical environment, but neither the method nor the conclusion was irrelevant to this early example from rural France, which provides a classic prototype of the geography of voting.

Figure 1.1 The geography of voting in Ardèche, France

Geographical influences in voting in the American south

The notion of geographical influences is commonly taken to mean effects due to the physical environment, and in this way the Ardèche study may be seen as illustrating geographical influences. Most modern interpretations of the adjective ‘geographical’ relate to locational factors underlying behaviour, however, and this is how we will interpret the term here. We have already indicated that Siegfried, in effect, produced socio-economic explanations of voting in Ardèche in any case. Here we concentrate on geographical or locational influences on voting patterns.

No study better illustrates the geographical or locational influences on voting than the seminal work on politics in the American south by V. O. Key Jr (1949). Being the work of a political scientist, his analysis is steeped in the states’ political history, and he interprets twentieth-century, southern politics...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Original Copyright Page
• Dedication
• Table of Contents
• List of Figures
• List of Tables
• Preface
• Part One Preliminaries
• Part Two Geography of Voting
• Part Three Geographical Influences in Voting
• Part Four Geography of Representation
• Part Five Towards a Geography of Power
• References
• Acknowledgements
• Index