Table of Contents

New introduction by the authors 1 Preface and acknowledgments 7 Introduction 11 PART ONE FRAMEWORK 25 Chapter one: Understanding electoral instability 27 The measures 28 Concepts and indices 31 Aggregate volatility as a system property 33 Aggregate volatility and individual voting shifts 34 Theoretical variants 40 A general framework for analysis 42 PART TWO THE STABILISATION OF EUROPEAN ELECTORATES 57 Chapter two: Interpreting electoral change: a debate 59 The evidence of aggregate data 62 Parties and cleavages 65 Chapter three: The bias towards stability 70 Variance in electoral instability 70 Variance in class-cleavage volatility 76 The relationship between cleavage volatility and total volatility 83 Types of elections and cleavages 85 Chapter four: Electoral instability and class-cleavage persistence 1885 - 1985 96 Testing the freezing hypothesis 96 The salience of the class cleavage 103 A general stabilisation? 105 National variations 106 Trends over time: a summary 117 PART THREE THE DETERMINANTS OF ELECTORAL INSTABILITY 123 Introduction 125 Chapter five: The fragmentation of party systems 127 Which parties count? 128 Trends in fragmentation 129 Numbers of parties and electoral volatility 131 National variations 134 Chapter six: Institutional constraints and voter opportunities 141 The impact of enfranchisement 142 Voting systems and electoral instability 145 Change of electoral systems 146 Majoritarianism versus proportionality 149 From dichotomised to continuous variables 152 Indicators of disproportionality 154 Disproportionality and volatility 155 Cross-national variation 158 Institutional constraints and electoral instability 160 Chapter seven: Electoral participation 165 When turnout matters 166 Turnout increase 167 Turnout decline 168 Variation between countries 172 Electoral participation and volatility 175 Format, institutional change, and participation 176 Chapter eight: The space of competition 182 The extent of electoral elasticity 182 Spatial competition and electoral instability 184 An interpretation of negative findings 189 Another test of the spatial hypothesis 191 National variations 193 Space and segmentation 196 Chapter nine: Cleavage systems 198 The concept of cleavage 199 Social homogeneity versus closure of mobility 205 Empirical evidence 209 The dimension of cultural heterogeneity 210 The organisational dimension 214 Cultural heterogeneity and organisational density 220 Segmentation and electoral instability 223 PART FOUR COMPETITION AND IDENTITY 231 Chapter ten: Explaining electoral instability 233 The overall pattern 234 Electoral phases 240 Cross-national variation 244 Refining the model: spatial, organisational, and cultural factors 248 A parsimonious model 255 Chapter eleven: Socio-organisational bonds, institutional incentives, and political markets 261 The stabilisation of voting patterns 262 From competition to identity 263 The forces shaping electoral mobility 266 Some applications: polarised pluralism and consociational democracy 271 Towards destabilisation? 274 Appendices and data-base 1 Rules for calculation and notes on sources 281 2 Party volatilities: data-base 291 Index 319

About the Author

Stefano Bartolini is currently Director of the Robert Schuman Centre for Adanced Studies at the European University Institute in Florence. Peter Mair was Professor of Comparative Politics at the European University Institute in Florence and at Leiden University in the Netherlands. He wass co-editor of the journal West European Politics.

Reviews

For those interested in larger patterns of electoral persistence and change among West European electorates, Bartolini and Mair's work is absolutely required reading. This is an extremely ambitious book, one that seeks to "chart patterns of stability and instability across this century of mass politics; to assess what these patterns imply with regard to the hold of traditional cleavages; [and] to explain variance in these patterns through a variety of different factors ... " (p. xviii). The authors accomplish this task principally by their compelling use of an impressive collection of electoral data representing the results of 303 elections from thirteen West European countries spanning a century of mass politics (1885-1985). This is an exciting book, whose findings are bound to refashion the debate about the shape of the West European electoral landscape. Bartolini and Mair do not waste any time getting to their main argument. For well over a decade now, study after study has proclaimed that ours is an "age of party breakdown, dealignment, and decay" (p. xvii). The overwhelming conclusion within the contemporary party literature has been to suggest that "we are witnessing a fundamental transformation in patterns of mass politics" (p. 55). However, as Bartolini and Mair demonstrate with inexorable logic and considerable elegance, the empirical evidence fails to support such exaggerated claims. Rather, as these authors take great pains to demonstrate, the real story of West European electorates is quite the opposite: one of rather massive stabilization of European electorates when viewed over a century of mass politics. They punch holes in contemporary scholarship with a multitude of startling conclusions from their exhaustive study. Take just one such conclusion as an example: "Mean volatility in Western Europe in the era of supposed change from 1966 to 1985 proved to be just 0.01 higher than it had been in the period of steady-state politics from 1945 to 1965" (p. xvii). This, as the authors point out, can hardly be called transformation. Instead, what one sees over the course of a century is a gradual stabilization within European electorates. The structuring of the vote and the resulting stabilization of electorates are viewed by the authors as essential for explaining more general processes in Western Europe of "conflict encapsulation" (p. 2), for building democratic institutions, and eventually the consolidation of democracy. A stabilized cleavage system can be a factor in support of eventual democratic consolidation because individuals are presented with "an already existing constellation of alternatives contributing to the structure of their votes and to their political integration into the society" (p. 5). One of the principal theoretical and empirical goals of Bartolini and Mair's work is to mount a frontal assault on broadly accepted tenets of recent party literature, especially its critique of Lipset and Rokkan's (1967) classic explanation for long-term persistence over change within Western European party systems. To rescue Lipset and Rokkan, the authors make two very significant contributions to the literature on political parties. First, the authors argue persuasively that electoral developments need to be traced back to the end of the nineteenth century in order to gain the long-term perspective necessary if Lipset and Rokkan's structural arguments for persistence are to be properly tested. Second, seeking to recapture the original meaning of Lipset and Rokkan's "freezing" metaphor, Bartolini and Mair point out that, by referring to the freezing of cleavage lines, Lipset and Rokkan meanf'the freezing of major political alternatives which are not necessarily represented by one specific party, but rather are often characterized by the opposition between blocks of parties" (p. 36). The authors complain that contemporary critics of Lipset and Rokkan have confused the issue by using the individual party organization as the basic unit of analysis for measuring persistence or change within the electorate. Very crucially, they insist that Lipset and Rokkan were not referring to individual party organizations but rather to fundamental political alternatives underlying national political life. By aggregating the parties into blocks, and by distinguishing between cleavage allies and cleavage opponents, Bartolini and Mair succeed in moving away from an emphasis on the individual party organization and gain a real sense of the extent of change across cleavage boundaries. Following this logic, the authors devise a relatively simple but critically important distinction between total volatility (volatility measured at the level of the individual party and summed for the system as a whole), and block volatility (that part of total volatility which occurs between the two groups of parties that have been aggregated into blocks on either side of a cleavage line). If Lipset and Rokkan are correct that "the party systems of the 1960s reflect, with but few exceptions, the cleavage structures of the 1920s" (p. 96), then declining levels of instability should follow the initial mobilization of working class in the early part of the century. And this is exactly what the data show. When measured in terms of total electoral volatility, the data reflect a striking bias towards stability over a century of mass politics. This long-term trend toward stability is even more evident when using cleavage (or block) volatility as the measuring stick, with cleavage volatility peaking precisely during the periods which represent the crucial phase of class-left electoral mobilization: 1901 to 1906 and 1916 to 1920. By contrast, from 1920 to 1965 the figures show a steady decline of cleavage volatility. Thus, Bartolini and Mair's very careful analysis lends decisive support to Lipset and Rokkan's hypothesis. Before the mass mobilization of the electorate in 1920 the electorate is characterized by significant cleavage volatility. After the 1920s, there exist clear signs that institutionalization predominates into the 1980s. The only exception to this trend seems to be the most recent period, where electoral volatility increases beginning in the 1970s and quickens in the 1980s. However, this final phase of electoral instability gains importance only when compared to the (exceptionally stable) steady-state period between 1945 and 1965. Having overturned recent arguments suggesting a fundamental dealignment in Western electorates, and after providing powerful arguments in support of the gradual institutionalization of party identities, Bartolini and Mair devote the second half of their work to an explanation of variations in levels of volatility, given the longer-term context of stability and persistence. They are at great pains to point out that, as cleavages become institutionalized over time, their capacity to explain instability declines, and forces for change become more narrowly and evidently concentrated within the institutional context. Here they identify six key independent variables: (1) the format of the party system; (2) changes in electoral institutions; (3) changes in levels of participation; (4) policy distance separating competing party alternatives; (5) contingent factors; and (6) the degree of cleavage closure. After a very careful discussion of each of these variables at the level of theory, and an exceptionally clear exposition of the empirical evidence in support of each claim, the authors distill these six variables further, and propose a general model for explaining variations in the level of electoral volatility over time. In the interests of parsimony, Bartolini and Mair build an explanatory model based on two composite indices drawn from the above six variables: the nature of socio-organizational bonds present within a society, and the way institutional incentives shape the structure of voter preferences. The first dimension, socio-organizational bonds, is reflected both by the way cultural (ethno-linguistic and religious) cleavages became established and were organized institutionally, as well as by the degree of organizational density manifested by social and political groups. The second dimension, the way institutional incentives structure voter preferences, is largely made up of the number of parties present in a party system and other institutional factors such as changes in electoral formulas and levels of participation (especially franchise elections). These few variables comprise two composite dimensions to explain almost one-half of all the variance in levels of instability across countries during a century of mass politics. Before closing, the authors reintroduce one final crucial independent variable: policy distance separating competing party alternatives. They find that policy distance exerts considerable impact upon levels of electoral volatility precisely when conditions already pertain that favor a propensity toward electoral mobility, that is, in situations where party identities (social bondedness) are relatively weak and where the institutional incentives toward electoral change are relatively pronounced. By reintroducing this final independent variable, the authors are able to explain a key anomaly for their argument for overall long-term stabilization: increasing levels of electoral volatility in Western electorates during the 1970s and 1980s. Most recently, diminished policy distance between competing parties in many West European countries have effectively canceled out the mild increase propensity toward stability predicted by the model. The authors are quick to point out, however, that this apparent increase is largely a "mirage" (p. 305). They stress that the increase in volatility during the 1970s and 1980s is placed in high relief only when compared to the "golden age" (1945-1965) of electoral stability. If a larger time horizon is chosen, the recent rise is relativized. In fact, the authors show that almost half of the cases (six of the thirteen) registered a net decline in total volatility, with the total aggregate difference in mean electoral volatility for all elections between the two most recent periods measuring almost no change. Again, a convincing case is made for long-term electoral stabilization. This book has succeeded where many others have failed. Bartolini and Mair have obviously labored years to compile the electoral data upon which they base their work. The most serious limitation of this book in terms of theory-building is one to which the authors themselves openly confess. The data they have collected concern almost exclusively the persistence (or otherwise) of a single dimension of cleavage: class. Though the authors admit at the outset that the generalizability of their findings is questionable, they tend to forget this admission when proposing a general theory of electoral instability later on in the book. While the general model for electoral mobility they propose draws (correctly) on multiple dimensions of cleavage, the evidence they bring to support the model across a century of mass politics focuses almost exclusively on the class cleavage. This limitation is a serious one, for much of the recent literature these authors criticize demonstrates the continued salience of, for example, religious over class cleavages in West European electorates. In their neglect, the authors have left important avenues for future research to pursue. Despite this shortcoming, Bartolini and Mair's work should be viewed as a tour de force for its many compelling arguments regarding the changing shape of the political landscape in Western Europe. In this book they have successfully combined methodological rigor with theoretical insight, and doing so have recovered important earlier contributions (especially those of Lipset and Rokkan) from the maw of many recent less compelling analyses. Timothy Scully, C.S.C. Review of Politics