This scientific survey of French presidential elections from 1988-2007 provides a new benchmark for the analysis of voting behaviour and election outcomes in presidential systems. Françoise Boucek finds that the book thus provides general evidence of change and continuity in contemporary French politics while avoiding French ‘exceptionalism’ which too often characterises volumes written by country specialists.
French Presidential Elections. Michael S. Lewis-Beck, Richard Nadeau and Eric Bélanger. Palgrave Macmillan. January 2012.
This book (also published in French) explaining what drives voters’ choices in French presidential elections is written by three electoral behaviour scholars. Michael S. Lewis-Beck has been analysing French voting behaviour for over 30 years and has authored and co-authored more than 180 articles and books, including The American Voter Revisited, Forecasting Election, Economics and Elections and The French Voter: Before and After the 2002 Elections. His French Canadian co-authors, Richard Nadeau and Eric Bélanger are based in Québec universities but consider France “a motherland in the deepest sense of the word”. They are co-authors of Le comportement électoral des Québecois (2009).
The timeliness of this volume is both a strength and a weakness. Published in 2012 - a French presidential election year – it offers a unique framework for understanding that contest. However, in the aftermath of François Hollande’s May victory and the Socialists’ majority win at the June legislative elections, the absence of the 2012 poll from the dataset leaves this reader hungry for more. It is only the second time under the 1958 Fifth Republic that the Left has won the presidency, the first being the earth-shattering victory of Socialist François Mitterrand in 1981 ,which is also missing from this book for the simple (albeit astonishing) reason that no adequate and available presidential election survey existed then. Hence, this book makes a huge contribution to the field of electoral behaviour in France.
Moreover, unlike previous electoral surveys which tend to look at single elections as discrete events, this unique comparative longitudinal survey of four French presidential elections (1988, 1995, 2002 and 2007) examines the long and short term determinants of French voting patterns over two decades. The book thus provides general evidence of change and continuity in contemporary French politics while avoiding French ‘exceptionalism’ which too often characterises volumes written by country specialists.
In terms of analytical approach and methodology the authors adapt the trailblazing ‘Michigan model’ of voting behaviour developed by Campbell, Converse, Miller and Stokes (1960) to French voting behaviour. That seminal volume sought to explain the “puzzle of 1952”: the presidential election of Dwight Eisenhower which the sociological model of voting behaviour had failed to predict. Instead, drawing on social psychology, they developed the concept of party identification to explain continuity in American voting patterns and included short-term influences (candidates and issues) to the model depicted as a ‘funnel of causality’ to explain short and medium-term political change in America.
Lewis-Beck, Nadeau and Bélanger use that “funnel of causality” (unfortunately not visually reproduced in the book) to examine, through a systematic series of logistic regression analyses (and a 33-page Appendix of regression tables) the long and short-term forces influencing French presidential voting since François Mitterrand’s 1988 re-election. Two types of long term forces are examined. The sociological variables such as religion and class (occupation) are examined first in chapters 2 and 3 followed in chapter 4 by the psychological factors – partisan identification and left-right ideological identification – crucial to the Michigan model.
The short-term variables dealing with issues and candidates are then added to the model and estimated progressively in accordance with the funnel of causality. Chapter 5 looks at issue voting on immigration and economics examined through multivariate analysis and Europe is then added to the equation. The variables corresponding to the tip of the funnel: the impact of candidate image (chapter 6) and campaign influences (chapter 7) are based on post-electoral surveys. They reveal that vote decisions are made later and later in the campaign, particularly among young voters, although better sources of survey data should in the future provide more finely-grained study of short-term effects.
Of course what makes French elections ‘exceptional’ and interesting to political scientists and ‘institutionalists’ (including this reviewer) is the double-ballot voting system. This shapes presidential choices, particularly for supporters of extremist parties on the left and right who vote strategically in the second round, to prevent their worst-preferred candidates from winning.
This volume considers the institutional context indirectly since it isn’t a book about party competition but electoral behaviour. Like in other mature democracies where large-scale electoral studies are available, the effect of social structure in France is found to be quite modest due to the weakening of social cleavages although religion is still relevant for moderate right supporters. However, social structure remains more salient in France due to the second round of voting when diverse groups of voters coalesce on the right and left. The gentrification of the socialist camp is partly due to centrist support notably for François Bayrou in 2007. However, a more ideologically polarised French electorate has emerged since 1988. Lewis-Beck reiterates perhaps less definitely than in the past that the French electoral system is still ‘stalled’ on the basis that long-term determinants account for more of the variance in candidate support than short-term effects such as leader image and the issues of immigration and Europe which are inconsistent with the exception of 2002 and the Le Pen effect.
This scientific survey of French presidential elections from 1988-2007 provides a new benchmark for the analysis of voting behaviour and election outcomes in presidential systems. Its rigorous multivariate statistical analysis tested against competing approaches and its empirical coverage of French electoral politics for the last two decades make it an ideal volume for those interested in electoral behaviour and contemporary French politics.
Françoise Boucek is a Teaching Fellow in the School of Politics and International Relations at Queen Mary, University of London, and an Associate of the LSE Public Policy Group. A political scientist, she gained her PhD from LSE and researches chiefly on party politics in advanced industrial countries. She is the co-editor (with Matthijs Bogaards) of Dominant Political Parties and Democracy: Concepts, Measures, Cases and Comparisons (London: Routledge, 2010).