Středoevropské politické studie / Central European Political Studies Review Ročník XX (2018), Číslo 1, s. 81–85 / Volume XX (2018), Issue 1, pp. 81–85

(c)Mezinárodní politologický ústav / International Institute ofPolitical Science

DOI: 10.5817/CEPSR.2018.1.81

Sona N. Golder, Ingacio Lago, André Blais, Elisabeth Gidengil, and Thomas Gschwend: Multi-Level Electoral Politics: Beyond the Second-Order Election Model

Oxford: ECPR and Oxford University Press. 1st ed., 2017, 211 pages.

ISBN 9780198791539


The book Multi-level Electoral Politics: Beyond the Second-Order Election Model focuses on an explanation of how voters’ and parties’ decisions are involved in different kind of elections. The group of authors are experts in comparative politics and electoral studies who set out to show that the second-order election model is a useful framework for studying multi-level electoral politics. At the same time, they are aware that the model does not reflect all conditions affecting the election result. Based on this, the book aims to “explain how party and voter behaviour in given elections is affected by the existence of multiple electoral arenas for the selections of political office holders within the country” (Golder et al. 2017: 7). This book differs from other publications focused on similar topic in many aspects. First, the analysis reflects three levels of government – European (elections to the European parliament), national, and subnational (regional). The authors have selected for their analyses three countries – Germany, France and Spain. The reason for this selection is simple – the analysis is meant to test the authors’ theory by applying it to different types of electoral system both within and across countries. Accordingly, the specifics of each country are taken into account.

Second, in contrast to the original theory, the present analysis takes into account the party level. The standard model of second-order elections does not pay attention to political parties and their behaviour, which the authors consider a shortcoming of the model, and take political parties among the main objects of focus in their book. Thirdly, the book is based on a huge dataset. The research analyses data collected

1Ph.D. candidate, Department of Political Science, Faculty of Social Studies, Masaryk University, Czech Republic. E-mail:


Golder et al.: Multi-Level Electoral Politics

during the 2011–2015 period, collected in two-way panel surveys in three countries (6 regions). Only in Bavaria, there was a five-way panel survey.

The book consists of 8 chapters, with the most important findings to be found in chapters 4–7. The first three chapters provide an introduction to the topic and the corresponding research. Chapter 1 gives an overview of the second- order model and also explains the research methodology. In the second chapter, the theoretical background of the book is presented; besides that, it generally outlines the topic of election processes. Golder et al. notice that different election processes in different electoral arenas influence each other. They try to explain it with reference to behaviour of political actors, which is a unique approach, as this variable has not yet been examined in previous studies.

The third chapter provides some specifics of the countries under scrutiny. It outlines differences in electoral rules, regime types and party systems between the selected countries. It also focuses on differences between electoral systems on each level of governance. Golder et al.’s analysis also reflects some specifics of political behaviour in each country, as well as varying economic and political circumstances.

In the next three chapters, Golder et al. present the results of their research. In the fourth chapter, they analyse mobilization, in order to find out if party behaviour is important for studying multi-level systems. By “mobilization”, the authors mean the way political parties allocate and invest their money, energy, sources etc. in different kinds of election. The analysis differs in important respects from other studies on second-order elections. As mentioned above, most of them deal with variations in voter turnout and electoral results – that is, they are focused on voter choice. Golder et al. claim that such focus on voters’ behaviour is incomplete. In their view, “voters have too small a stake in mass elections to produce, by themselves, the coattail effect running from the more important body to the less important one that are often referenced in analyses of ‘first-order’ versus ‘second-order’ elections” (Golder et al. 2017: 49). Based on this, the authors explain variations in mobilization in different kinds of elections with respect to party strategies. Among their main findings, we learn that parties invest less in European elections as compared to national and regional elections, and that there is no great difference between national and regional elections. In addition, the electoral system affects party strategies as well. In the examined countries, the electoral system in elections to the European parliament opens up the path to success to more parties. As a result, also smaller parties are able to find success in the European elections. Moreover, there difference in the importance of mobilization of voters: Small parties tend to mobilize more intensively in European elections than large parties. The analysis also confirms that in the European elections, parties generally care more about core voters and engage in mobilization to a larger extent than in campaigns during national and regional elections.


Barbora Burešová

Chapter 5 deals with turnout in multi-level systems. The second-order election model starts with a premise that turnout is lower in the European elections than in the national elections because voters perceive them as less important. This premise is given a deeper analysis in this chapter and subjected to systematic testing in the selected cases. The results of the analysis confirm that the EU is perceived to have less impact on people's well-being, which is a reason why people think of the European elections as less important. This is a part of the explanation of the lower turnout. Nevertheless, the perception of importance of the institution is not the only explanatory factor. Another important factor is mobilization, analysed in the fourth chapter. Yet another significant factor, which was strongly confirmed via differences between national and regional elections, is the feeling of attachment to community. Among others, it plays an important role in Catalonia, where regional elections are perceived as the most important. Finally, Golder et al. also discuss the role of media, as there is some evidence that the media do not cover all types of election equally. Unfortunately, this topic was not explored in the book, so there is room for further research.

Chapter 6 presents the results of analysis of strategic and sincere types of voting. The author’s basic aim is to refute the claim that voters vote strategically if they choose different party in different types of election, with the aim of finding a deeper explanation. They tested their theory on the individual, regional and party levels (where the latter is unique for such type of research). The results show that most voters simply vote for their candidate or party (i.e. do not vote strategically). Still, there are some who deviate from their most preferred option. There are three reasons for such behaviour, as confirmed by Golder et al.’s research. The first is that some voters try to avoid wasting their vote and decide to vote more a successful (larger) party. This trend is more frequent if the electoral system is less permissive. At the party level, larger parties profit from this. The second reason is the higher probability that larger parties have of taking control of the post of the prime minister, or at least of a governmental position. However, this is more pronounced in national elections.2 The final reason is that voters clearly realize that they spread out their vote (i.e. vote strategically). Voters who do not have a strong preference for one party spread they vote strategically, in order to distribute the impact of ideological governmental decisions to different levels (regional, national, and European).

Regarding the second-order model, there is a discussion about what kinds of elections could be used to measure the accountability of government. Golder et al. explore this topic in the seventh chapter. They test it at all three levels with respect to two topics – economic situation and perception of corruption, trying to figure out if elections at different levels could provide the mechanism of accountability. Their results show that the most important factor for voters is the

2Except for Spain, where there is such a trend also on the regional level.


Golder et al.: Multi-Level Electoral Politics

economic situation. If economic situation deteriorates, voters “punish” governmental parties by not voting for them. This can be seen on different electoral levels, but with different dynamics. While voters are able to use subnational and European elections as a vehicle for delivering a message to the national government, this mechanism does not work in the opposite direction: Voters by and large do not punish parties on national level for worse economic situation at regional levels. The perception of corruption has a similar impact.

The conclusion to this book offers a discussion of the results. The authors see them as confirming the assumption of the second-order election model, i.e. that voters pay less attention to these election because they perceive them as less important. However, this does not mean that voters always perceive national elections as the most important. Besides that, the authors also try to come up with a remedy for one important weakness of the second-order model, namely that it does not take into account parties and their behaviour. Their research takes into consideration that parties’ strategies and expectations regarding the elections play a significant role. Another important factor are voters’ emotional attachments to their community, which is again overlooked by the second-order election model.

To sum up, this book presents new findings in the field of election studies. The topic of second-order elections was a popular one in previous decades, and we are still witnessing interesting work being done on the topic, such as the papers by Charlie Jeffery and Daniel Hough (2001) or Francesc Pallarés and Michael Keating (2003). In the Czech context, the article by Havlík and Jogheeová (2010) is worth mentioning. Compared to such contributions, which are mostly case studies, the reviewed book is unique in its breadth and attention to detail. For one, it brings into comparison three different countries, including subnational regions, and adds a temporal dimension. Perhaps the major added value of the book is in the discussion of the second-order election model, especially by presenting its shortcomings exploring ways of getting around them. I am of the view that this objective has been achieved. Also intriguing in my view is the final chapter where authors discuss potential challenges that research on this topic has will have face. The book is also well structured and written in an accessible style. The appendix then provides detailed information on the questionnaires used for data collection.

As regards the weaker spots of the book, let me mention the framing of some results or conclusions which are presented as universally confirmed and valid – whereas the data had been collected only for three European Union countries. This raises the question whether they can be easily generalised to other countries, regions etc. Further, while the compilation of the dataset itself is unproblematic, there are other limits to the interpretations suggested by the authors which they unfortunately do not entertain in any serious fashion.

All in all, the book amply demonstrates how important it is for researchers in the field of election studies to take into account different levels of elections: The


Barbora Burešová

main reason is that these levels mutually impact each other. Also, Golder et al. document that it is crucial to look not only at voter behaviour but also at behaviour of political parties. Both actors are important and, again, influence each other. This is why the book provides unique findings which will prove essential for future study of multi-level elections as well as for manifold electoral studies.


Havlík, Vlastimil a Eva Jogheeová. 2010. “Volby do Evropského parlamentu v roce 2009 optikou konceptu voleb druhého řádu.” Evropská volební studia 5(2): 96–110.

Charlie, Jeffery a Daniel Hough. 2001. “The Electoral Cycle and Multi-Level Voting in Germany.” German Politics 10(2): 73–98. DOI: 0.1080/772713264

Pallarés, Francesc a Michael Keating. 2003. “Multi-Level Electoral Competition: Regional Elections and Party Systems in Spain.” European Urban and Regional Studies 10(3): 239–255. DOI: 10.1177/09697764030103005