PROMISING DEMOCRACY Parties, Citizens and Election Promises ELIN NAURIN
Distribution Elin Naurin Department of Political Science University of Gothenburg P.O. Box 711, 405 30 Gothenburg, Sweden E-mail: Elin.Naurin@pol.gu.se 2009 Elin Naurin ISBN: 978-91-89246-42-3 ISSN: 0346-5942 http://hdl.handle.net/2077/20800 Cover: Wood anemone in Annekärr, photo Michael Tobler, Spring 2006 Printed by Intellecta Infolog, Göteborg 2009
To the memory of my father
This dissertation is included as number 118 in the series Gothenburg Studies in Politics, edited by Bo Rothstein, Department of Political Science, University of Gothenburg.
CONTENTS List of Tables and Figures...6 Preface & Acknowledgements...9 PART 1: Introduction 1. The Pledge Puzzle...15 2. Election promises in representative democratic theory...27 PART 2: Scholars investigations of election promises 3. Definitions of fulfilled election promises in empirical research...45 4. Do parties fulfil their election promises?...57 PART 3: Citizens views on election promises 5. How widespread is the image of the promise-breaking politician?...97 6. Just general distrust?...111 7. Citizens definitions...127 PART 4: Conclusions 8. Conclusions...177 Appendencies & References A. Fulfilment of election promises in Sweden: Further description of the selection of respondents and reliability tests for analyses in Chapter 4...191 B. Original Swedish wordings of survey questions used in Chapters 5, 6 and 7...195 C. Interview guide for the longer research interviews presented in Chapter 7...201 References...205
Table 4.1. Table 4.2. Table 4.3. Table 4.4. Table 4.5. Table 5.1. Table 5.2. Table 5.3. Table 5.4. Table 5.5. LIST OF TABLES AND FIGURES TABLES Average percentages of government parties election promises at least partially fulfilled in descending order... 58 In what situations is it more likely that a political party fulfil its election promises?... 65 Election promises given by the Social Democrats in 1994 and 1998 divided according to action promises and output promises. Number of promises with row percentage in parentheses... 87 Fulfilment of the Social Democrats election promises during the period 1994-2002. Number of promises with row percentage in parentheses... 87 Fulfilment of the Social Democrats election promises during the period 1994-2002 divided into action promises and output promises. Number of promises with row percentage in parentheses... 88 The perception of MPs fulfilment of election promises in different countries. Percentages and net values... 99 The perception of MPs fulfilment of election promises. International comparisons of net values over time... 101 The perception of parties fulfilment of election promises over time in Sweden (1956-2002)... 102 The perception of parties fulfilment of election promises among different groups of respondents in Sweden... 104 The perception of politicians fulfilment of election promises. Questions asked in different ways in Sweden... 107 Table 6.1. Correlation matrix for items on political trust... 113 Table 6.2. Table 6.3. Table 6.4. Table 6.5 The perception of parties fulfilment of election promises among respondents with and without trust in politicians. Percentages... 116 Dimensions of distrust including the question about election promises and other questions about political distrust... 121 Dimensions of trust including the question about election promises and other questions about political trust... 123 How well does the promise item fit into a theoretically defined cumulative political dimension? Lower bound set low (H=.30)... 125 Table 7.1 Respondents in the longer research interviews... 133
Table 7.2. Table 7.3. Table 7.4. Table 7.5. Do those who think parties usually break their election promises give examples of broken election promises in the open ended survey question?... 145 Examples of broken election promises given in the open ended survey question. Divided into subject areas... 146 Answers given in the open survey question divided according to how concretely they are formulated... 148 Concrete examples given in the open survey question divided into definitions of broken election promises... 170 FIGURES Figure 7.1. Citizens ways of forming judgements about election promises... 139 Figure 7.2. Comparison 1: Assumptions about what has been promised, or unspoken promises, are compared to the personal perception of society... 154 Figure 7.3. Comparison 2: Outspoken promises are compared to the outcome of politics... 158
Preface and acknowledgements PREFACE & ACKNOWLEDGEMENTS Election promises are not just dirty tricks used by political parties to get votes. They give important information about how parties prioritize after winning elections. Citizens therefore have good reasons to read election manifestoes and listen carefully to what parties pledge before they make their choice on Election Day. This was the conclusion made in the first systematic comparison between election promises and parties actions in Sweden. The Social Democrats had acted on between 80 and 90 percent of the promises they had made in their 1994 election manifesto, thereby strengthening other pledge studies results that political parties tend to take election promises seriously. As a Master degree student, and as the author of the study, I got more attention than I ever could have dreamed of. One of the country s most read newspaper, Dagens Nyheter, published the results in a debate article (Esaiasson & Naurin 1999). Political editors in national and local newspapers all over the country commented on it. As I see it now, the attention and the comments that followed the publishing of the results were more interesting than the study itself. The claim that the government had kept most of its pledges from their election manifesto actually made people angry with me. It seemed as though everybody knew that political parties do not keep their promises. At gatherings I became the one who claims that politicians fulfil their promises. In the media debate that followed I was loudly accused of performing poor research and of running the government s errands. One of the editorials in the conservative newspaper Svenska Dagbladet suggested that the author of the study must be illiterate (Gür 1999). The paper even published an article with the headline A Meaningless Thesis (Broberg 1999). A commentator on the net-based paper Friktion wondered whether the Department of Political Science at University of Gothenburg has any raison d être what so ever
Elin Naurin (Friktion 1999). In contrast, the commentators of the government side saw the results as proof of the governing party s (unparalleled) moral behaviour. In his joy, the Social Democrats party secretary even referred to me and Bruce Springsteen (the Boss himself) in the same article: We made a promise we swore we d always remember. No retreat, no surrender (Stjernqvist 2003). This book is the result of my curiosity about why I stirred up so much feelings when I claimed that political parties usually fulfil their election promises. My conclusion seemed to follow rather basic assumptions from theories of representative democracy that parties act to fulfil mandates that they perceive have been given them by citizens. I started to ask myself whether I, and my fellow pledge scholars, should have done our studies differently, or whether everyone else simply was wrong. Or whether it maybe was impossible to study election promises without provoking anyone. Quite a few years later I still claim that parties take their election promises seriously, and that citizens should do so too. However, it is obvious that citizens do not mean the same thing as scholars when they dismiss election promises. Election promises can in fact be given quite different meanings. It is not so simple in the world of politics that a promise is a promise. The journey towards this final text has been long and winding. I want to direct my gratitude to all of you who have contributed to making it possible for me to finish the project. Some should be lifted forward specifically for their contribution to the book. Firstly: always by my side, with a never ending passion for life, science and politics, is Patrik Öhberg. Being my sharpest academic critic and my most dedicated supporter, you have meant everything for this book, and you will mean everything for me also in the future as my colleague and as the love of my life, (I guess that I am one of few who used my wedding day to get comments on my thesis. In my defence: it was on your suggestion). The book has also benefitted from the supervision of Peter Esaiasson and Mikael Gilljam. Thanks to Peter and Mikael this is written with a common sense attitude towards science and with a genuine enthusiasm for what can be done through what is really the art of Political Science. It has been a privilege to be apart of your team and I hope this is not the end 10
Preface and acknowledgements of our cooperation. Peter was the one who encouraged me to take on the subject of election promises when I was an undergraduate. He has kept his door open for me ever since. It sometimes feels as though he knows my work better than I do. Mikael has the important ability to disagree with Peter. He also has the talent of spreading good spirit around himself and, in one way or the other, always makes me feel good about what I do. My deepest and most sincere gratitude to both of you. Many colleagues and friends have contributed by commenting on the manuscript. A special recognition goes to Andreas Bågenholm, Sören Holmberg, Daniel Naurin and Lena Wängnerud who all have read the whole manuscript. Several colleagues have given valuable comments on parts of the manuscript. Thank you Joaquín Artés, Gregg Bucken-Knapp, Mikael Elinder, Marie Demker, Andreas Heinö Johansson, Nicklas Håkansson, Maria Jarl, Staffan Kumlin, Johan Martinsson, Nathan McCluskey, Catherine Moury, Sara Naurin, Birgitta Niklasson, Henrik Oscarsson, Jon Pierre, Helena Rohdén and Bo Rothstein. A special thanks to Christina Ribbhagen, who also assisted in the pledge study of the election period 1998-2002 and to the members of the Mechanisms of Democracy project within which this book is written. Terry Royed and Robert Thomson in the early Comparative Party Pledges Group (CPPG) have been an important source of support and inspiration for me. In June 2009 the CPPG gathered 15 election pledge scholars for a workshop at the Department of Political Science in Gothenburg. I very much look forward to our future collaborations. I am also grateful to Martin Kroh, Wouter van der Brug and Cees van der Eijk at Amsterdam University and Amsterdam School of Communications Research (ASCoR) for valuable comments and generosity during my stay there in 2002. Furthermore, the Department of Political Science at the University of Gothenburg is a friendly and creative working place. I am especially indebted to Andreas Bågenholm, Lina Eriksson, Johan Martinsson, Birgitta Niklasson and Helena Rohdén in this regard. Valuable funding from Riksbankens Jubileumsfond, the TMR network and Wilhelm & Martina Lundgren s Vetenskapsfond 1 has covered expenses associated with the project and the SOM Institute generously let me include two questions in their survey of 2000. 11
Elin Naurin Finally, I am blessed with a family who takes interest in what I do. Both my brother and my sister have commented on more or less the whole manuscript and have been my companions also in this struggle. During my work with the book our father, Torbjörn Naurin, fell ill and passed away. I loved him so much and I had so wished that he could have been together with me when I now put an end to this work. His thoughts on what I write and his love you madly cannot be replaced. Through these years, the support of my mother Pia Naurin has been crucial. Even after she became a single grandmother she has been there for me and Patrik in all times. I want to thank her for loving the written word, for never ever giving up, and for seeing me both as a professional academic, as a struggling mother and as her beloved daughter. Also the care of Marianne and Charlie Öhberg has been important for the realisation of this work. And to Miranda and William: my book is finished! Thank you for putting it in the proper perspective and, most importantly, for letting me be your loving mom. 12
Introduction PART 1: INTRODUCTION 13
14 Chapter 1
1. THE PLEDGE PUZZLE Everyone knows a promise is a promise. It is hard to imagine any serious relationship that would go unaffected if promises were broken. This book deals with a controversial kind of promise, namely those that are given by political representatives to citizens before elections, so called election promises. The focus is what seems to be a puzzling controversy between scholars and ordinary citizens concerning whether or not politicians actually keep their promises. While research presents an image of responsible parties acting on most of their election promises, citizens are presumed to hold the opposite view that parties usually break their promises. Since election promises are central to representative democratic theory and practice, this divergence raises important questions about how good democratic representation is defined and achieved. So called election pledge research, where parties pre-election statements are compared to governments actions, is found in different countries and under different time periods (see for example Royed 1996; Thomson 2001; Mansergh & Thomson 2007; Artés & Bustos 2008). The levels of congruence between promises and actions differ between the different systems that are studied, but there is a clear common denominator in the conclusions that are drawn: promises given in election manifestoes are taken seriously by the governing political parties. Even in systems where parties are found to fulfil just over half of their election promises, scholars conclusions are mainly optimistic. In one of the earliest studies on election promises, American parties were shown to fulfil around 60 percent of the promises they gave in their national platforms during the period of 1944-1978. The authors of the study drew the conclusion that, in comparison to what would have been expected if citizens were asked to make the same judgement, 60 percent fulfilled election promises was an encouraging result: contrary to the conventional wisdom, these plat- 15
Chapter 1 form pledges are redeemed/ /The vote for party is also a vote for policy (Pomper & Lederman 1980:174, see also Pomper 1968). The same positive undertones come up in all later studies. Parties in the U.K., Canada, Greece, New Zealand and during some periods in Ireland have been found to fulfil between 70 and 80 percent of the promises they make in their election manifestoes (Royed 1996; Rose 1980; Rallings 1987: Kalogeropoulou 1989: McCluskey 2008). Under the rule of coalition governments in the Netherlands and in Ireland, the average rate of fulfilment is lower, around 50-60 percent. Still, as for example Robert Thomson says, also in such systems there is a stronger programme to policy linkage than the conventional wisdom would lead us to suspect (Thomson 1999:223, see also Thomson 2001; Mansergh & Thomson 2007; Costello & Thomson 2008). The point of comparison in all these studies, or the reason for the positive conclusions, is that the conventional wisdom about election promises makes us expect much lower levels of pledge fulfilment than a majority. It is even an often repeated assumption in research on election promises that most citizens instead hold the belief that parties usually break their promises (see for example Mansergh & Thomson 2007:324; Artés & Bustos 2007:23; Pomper & Lederman 1980:174-6). However, even though it serves as the most important point of comparison, the conventional wisdom about election promises has not been empirically investigated. It is therefore difficult to say more specifically what this perceived to be puzzling situation actually consists of, or how great the difference in opinions between citizens and research really is. This book will shed light on this possible Pledge Puzzle. More specifically, it will start by investigating whether there actually is a Pledge Puzzle in the sense that citizens hold the opposite view as compared to the results of research. First, a more thorough description of the research results is given. Thereafter, the conventional wisdom is quantified and it becomes obvious that there is a difference in perception between scholars conclusions and citizens descriptions of parties as notorious pledge breakers. The rest of the book therefore engages in the discussion about how we can explain this puzzling situation. More specifically, two possible explanations of the Pledge Puzzle are intuitively reasonable: either citizens are unaware of the fact that parties tend to act on their election 16
The Pledge Puzzle promises (i.e. they do not have correct information) or citizens define election promises differently than is done in the empirical investigations. The question of what information reaches citizens is interesting, but will be left aside here. As I see it, to be able to discuss what information reaches citizens, we first have to find out what citizens actually mean when they say that parties usually break their election promises. I therefore ask: do scholars and ordinary citizens mean different things when they talk about political representatives election promises? The importance of investigating the Pledge Puzzle The arguments for analysing the Pledge Puzzle can be described in both theoretical, practical and normative terms. The theoretical importance comes from the fact that election promises are given important roles in the relationship between citizens and representatives in commonly used theoretical models of representative democracy. Election promises specify parties intentions for the future, they guide parties actions between elections, and they serve as benchmarks at the end of the election period in the evaluations of how well parties are able to implement their decisions. Therefore, at least in theory, citizens can use election promises both to form opinions about parties future politics and to evaluate how well parties have performed. The Pledge Puzzle suggests that when these theoretical demands are empirically investigated, one can come to different conclusions, which raises suspicions over the research s choices of definitions. The practical importance comes from the fact that election promises are also central in actual political representation. The empirical evaluations of the fulfilment of election promises (that will be further described ahead) are the most important indicators of the role of election promises in democratic states. Research also indicates that media give attention to election promises when they report on election campaigns (Krukones 1984; Strömbäck 2001; Costello & Thomson 2008). We also know that election promises have a role in so called negative campaigning, where parties and candidates give focus to their competitors inabilities to get things done. A study of Swedish parties even shows that the most used 17
Chapter 1 invective in political debates among parties in televised media during the 2000 th century was that other parties are not trustworthy and do not honour their words (Esaiasson & Håkansson 2002; see also Ansolabehere & Iyengar 1995; Lau & Pomper 2004). Election promises and fulfilled election promises are therefore notions of importance in public debates, making investigations of what is actually meant by the same highly relevant. The normative arguments for why we should investigate a possible Pledge Puzzle have to do with scholars abilities to facilitate communication between citizens and representatives. To be able to make rational decisions on Election Day, citizens need a lot of information. In the ideal world they would have good knowledge both about how the competing political parties have performed earlier and about what the parties plan to do after election. Scholars in political science agree that this is rarely, if ever, the case (see for example Luskin 1987; Bartels 1996; Pierce 1999; Oscarsson 2007). One important normative argument for investigating a perceived Pledge Puzzle is therefore that, in this rather blurry representative process, political scientists should communicate the information that actually exists so that citizens votes stand the best chance of being as rational and efficient as possible. Even though we have come far in the evolution of mass communication, the importance of guidance is as important now as ever in the history of representative democracy. It is even sometimes claimed that traditionally used shortcuts to information are less useful for today s citizens. For example, in most Western democracies, party identification has declined over the last decades (Dalton & Wattenberg 2000; Thomassen 2005; Oscarsson & Holmberg 2008). Citizens no longer feel genuinely attached to a specific party to the same extent. Other similar group affiliations are also in decline. Individuals generally feel less and less attached to specific groups such as workers and farmers, which means that informational shortcuts that have traditionally been used to form opinions are now less relevant (Popkin 1991; Lupia 1994; Oscarsson & Holmberg 2008). In this situation, election promises could be seen as a tool to direct communication between citizens and representatives, making an investigation of what is meant by the notion among the different actors important. 18
The Pledge Puzzle Inspiring in normative discussions about how we can improve communication between representatives and citizens is American political scientist Richard Fenno. With long experience of following politicians around, Fenno expresses commonsensical ambitions to have people (including himself) understand how politicians work and think. One of his main points is the importance of filling up the knowledge gap among citizens. If people do not know how politics works, it is difficult to improve the relationship between citizens and representatives. He writes: The present unhealthy relationship between citizens and politicians exists side by side with a knowledge gap, or an understanding gap, on the part of citizens. As the knowledge gap is filled, it is quite likely that a new appreciation of, if not more positive attitudes toward, elective politicians will be inspired. If that does not happen, the present relationship will most likely deteriorate (Fenno 1990:2). The difference between citizens and scholars way of describing election promises indicates that there is the kind of understanding gap that Fenno describes, but on both sides. It seems plausible that citizens are not familiar with the main results of research, but this lack of knowledge is paralleled by scholars, and thereby parties, poor knowledge about how citizens actually use the notion of election promises. And as Fenno continues: while the political system can doubtless withstand a good bit of distrust, disrespect and disregard for its elective politicians, just as surely its tolerance has limits. From that perspective, any prospect for improvement deserves nourishment (Fenno 1990:2). Election promises have a potential to facilitate the communication between citizens and representatives. However, to be able to push forward election promises as tools that can help citizens and representatives to a more comprehensible representative process, it is important to find out how citizens actually interpret the notion. The negligence and bad reputation of election promises Interestingly, it has not been the general image in the literature on representative democracy that election promises are particularly useful. Election promises have not been seen as mainly positive contributions to democratic representation. Instead, declarations clearly expressed by 19
Chapter 1 politicians have not seldom been treated as problems for, rather than as contributions to, a healthy relationship between voters and parties. Most of these critical voices towards election promises are founded in the theoretical debate about whether or not political representatives should be bound by authoritative demands by voters. Always cited in these discussions is Edmund Burke, the most famous opponent of the idea of a bounded mandate given by voters to representatives. In his speech to the electors of Bristol in 1774, just after he was elected by the same, Burke stated that a representative should follow his own judgement instead of the public will: his unbiased opinion, his mature judgement, his enlightened conscience, he ought not to sacrifice to you, to any man, or to any set of men living / / Your representative owes you, not his industry only, but his judgement; and he betrays, instead of serving you, if he sacrifices it to your opinion (Burke 1901, as cited in Kelley 1983). This view of the mandate as something that hinders representatives from doing a good job has long been an important part of theoretical discussions about representative democracy (Schumpeter 1942; Kelley 1983; Dahl 1990; Manin 1997). Since election promises are almost always seen as operationalisations of such bounded mandates, they are treated with equal scepticism. There is also another kind of scepticism towards election promises in the literature on representative democracy a scepticism that is less connected to theoretical discussions about the impossible mandate (Kelley 1983) and more related to empirical assumptions about parties (in)abilities to fulfil election promises. It seems that, over the years, scholars have also found it hard to believe that parties actually manage to keep their election promises. Scholars who make systematic comparisons between election promises and governmental actions have recieved surprisingly little attention. The most classical condemnations of parties platforms date from before the first pledge studies were done. One such comes from Moisei Ostrogorski in his book Democracy and the Organisation of Political Parties (1902): The platform, which is supposed to be the party s profession of faith and its programme of action is only a farce the biggest farce of all the acts of this great parliament of the party. The platform represent a long list of statements relating to politics, in which everybody can find something to suit him, but in which nothing is considered as of any con- 20
The Pledge Puzzle sequence by the authors of the document, as well as by the whole convention / / [Congress members] consider themselves in no way bound to the programs laid down in the convention (Ostrogorski 1964:138-9, as cited in Pomper & Lederman 1980:129). A similar classical refutation of platforms is Elmer Eric Schattschneider s claim in Party Government (1942) that party platforms are fatuities; they persuade no one, deceive no one, and enlighten no one (Schattschneider 1942:567, as cited in Royed 1996:49-50; similar references to typical rejections of British election manifestoes are discussed by Rose 1980:56 and Mansergh 2003:16-8). Negative expectations of election promises also tend to appear in later research. Bernard Manin, author of one of the most cited books in the field, The Principles of Representative Government (1997), argues in this very book that citizens should think retrospectively rather than prospectively when they cast their votes due to the unpredictability of candidates promises. He writes: They [citizens] know (or at least it would be reasonable for them to know) that electoral pledges are not binding and that those who are elected often fail to keep them. So it may, from their standpoint, be reasonable to take no notice of the candidates programs in the belief that their records offer a better way of predicting future conduct than do their words. Furthermore, even assuming that voters choose to pay some attention to the candidates promises, they know, or should know, that the credibility of those promises is an open question. It is not reasonable on their part to suppose that candidates will necessarily honor their commitments (Manin 1997:180). Manin later edited the also appreciated Democracy, Accountability, and Representation (1999) together with Adam Przeworski and Susan S. Stokes. Manin s co-editor Susan Stokes objects to the reasoning of Manin and claims in her own book Mandates and Democracy that political mandates have been underestimated in normative discussions about representative democracy over the years. We should not, says Stokes, draw hasty inferences about the meaninglessness of mandates based on these discussions (Stokes 2001:12 and Chapter 5). Stokes instead argues that it is reasonable to expect that politicians fulfil their election promises. According to Stokes, we should direct our interest towards the puzzling fact that representatives sometimes do not fulfil their promises. Interestingly however, she does not build her reasoning on the empirical evalua- 21
Chapter 1 tions of parties election promises that were available at the time. 1 In this way, Stokes book is also an illustration of the neglected role of election promises in normative analyses of representative democracy. If Stokes study can be seen as a defence of the usefulness of the mandate, this book becomes an investigation of and discussion about the usefulness of election promises in representative democracy. In my description of the Pledge Puzzle, I will first go through the arguments that election promises have important roles in theories of representative democracy. I will also show that research is convincing when it argues that parties take their election promises seriously. Thereafter, I investigate what citizens mean when they say that parties break their election promises. The centre of my empirical interest in this book is therefore citizens perceptions of election promises. Below follows an outline of the book where the empirical investigations that will be presented are briefly introduced. OUTLINE OF THE BOOK Part 1: Introduction After this first chapter, Chapter 2 engages in a discussion about the usefulness of election promises in theoretical discussions of representative democracy. I identify five different ways in which election promises contribute to making representative democratic theory more specific and thereby also more useful. Three of these are derived from theories about 1 Instead of giving references to for example Pomper 1968; Pomper & Lederman 1980; Rose 1980; Rallings 1987; Kalogeroupoulou 1989; Royed 1996; or Thomson 1999, Stokes refers to research from the Manifesto Research Group that has actually emphasised the unimportance of specific election promises (which I will get back to in Chapter 4, see Klingemann, Hofferbert & Budge 1994:25; Hofferbert & Budge 1992:153). Stokes also mentions Fishel s study from 1985 (where only (American) presidential promises are investigated) and Krukones study from 1984, which is a study of how well media s descriptions of campaigns predict policy. Krukones study is indeed interesting and shows that American campaigns give a reasonable image of what will happen after election. However, it is an evaluation of media s abilities to give a clear image of parties future actions, rather than of representatives abilities to fulfil election promises. 22
The Pledge Puzzle mandate representation and two come from reasoning connected to accountability representation. PART 2: Scholars investigations of election promises After the theoretical discussions, attention is shifted in part one of the book to the empirical evaluations of election promises (here called pledge studies ). Chapters 3 and 4 describe the scholarly part of the Pledge Puzzle. More specifically, Chapter 3 discusses the definitions that scholars use when they compare election promises to governmental actions. A number of theoretically and practically motivated focuses in scholars definitions are lifted forward and discussed. Chapter 4 describes more specifically the results of the pledge studies that here in the introductory chapter have been only briefly mentioned. The chapter presents both previous research and adds a new case to the body of research on election promises, namely the case of Sweden. With the help of the case of Sweden, I dig deeper into a political system that has only rarely been investigated within the pledge literature, namely single party minority governments. The study of Sweden has two purposes. On the one hand, I wish to develop a hands-on understanding of problems that empirical evaluations of parties election promises face. On the other, the study gives background information about the context in which most of the book s interviews with citizens will be held. Since we know so little about citizens definitions of election promises, any case would be interesting at this stage. The choice of conducting interviews with Swedes therefore mostly relies on the author s good knowledge of the context. However, Sweden is also in a general sense a suitable case for a study of citizens views of representatives. It is a consolidated democracy, with low corruption and comparably high living standards. The relationship between the represented and the representative could therefore be argued to have fairly good chances of not being blurred by such other large complications (compare Esaiasson & Holmberg 1996; Holmberg 1999; Oscarsson & Holmberg 2008). More specifically, two election periods with the dominating Social Democratic party in government are investigated 23
Chapter 1 in the study of Swedish parties election promises (1994-1998 and 1998-2002). PART 3: Citizens views on election promises When I leave Chapter 4, it is obvious that scholars have an important point when they claim that parties take their election promises seriously. In Chapter 5, the book then enters a discussion of how citizens view election promises. The aim of this chapter is to pin down how big a puzzle the Pledge Puzzle really is by asking: how widespread is the image of the promise-breaking politician among citizens? The conventional wisdom referred to in the pledge literature is for the first time given a quantitative specification. With the help of surveys, it is concluded that the image of the promise-breaking politician is indeed widespread. The negative perception of parties as usually breaking their election promises is found in most countries, over time, within most groups of people and, it seems, more or less regardless of how scholars ask the question. Chapter 5 therefore makes it clear that it is indeed relevant to talk about a Pledge Puzzle. Citizens generally do give a completely different image of parties election pledges as compared to what research does. Chapter 6 asks whether we should regard the negativity towards election promises only as expressions of peoples general distrust in politicians, or if it is possible to make a difference between a general distrust and a specific distrust in election promises. The chapter starts with measures of correlations between citizens views of election promises and their general views of politicians, and continues with index reliability analyses and so called Mokken scaling, where the unidimensionality between general political distrust and the more specific distrust in election promises is measured. It is concluded that we should not simply assume that citizens only express their general distrust when they dismiss parties election promises. It instead seems reasonable to regard the perception of election promises as something of its own. The last empirical chapter, Chapter 7, analyses citizens own explanations and examples of broken election promises. This chapter takes the words of citizens themselves seriously. With the help of longer interviews 24
The Pledge Puzzle with 17 individuals and an open-ended survey question with a representative sample of Swedish citizens, the chapter concludes that citizens do not use the notion of election promises the way that scholars do. Comparisons similar to the ones scholars make are found among citizens, but citizens specifications of when, where and by whom the promise is given differs from scholars evaluations. Citizens clearly include a greater number of and more widely defined election promises when they evaluate politicians election promises. They sometimes even compare with what they believe should have been promised, rather than with what was actually promised. What is more, it is found that citizens can base their negative perceptions on a story, or a narrative, of the promise-breaking politician rather than on actual evaluations of what is said and done in politics. PART 4: Conclusions In the last part, consisting of Chapter 8, I give an overview of the findings and arguments of the book. I also discuss what my conclusions mean to our general understanding of representative democracy and try to give ideas about where we should go from here in future research. 25
26
2. ELECTION PROMISES IN REPRESENTATIVE DEMOCRATIC THEORY Election promises are interesting as specifications of several otherwise rather vague aspects of normative democratic theory. Important theoretical notions like mandates and accountability get specific meanings when election promises are introduced into the discussion. Election promises therefore deserve a central place in literature on representative democracy. My aim with this chapter is to discuss more specifically what roles election promises can have in theories of representative democracy. MANDATES AND ACCOUNTABILITY It is common in the literature on representative democracy to make a distinction between two main ways of accomplishing democratic representation. On the one hand, scholars emphasise representation via mandates that guide political parties in their actions (mandate representation). On the other, the focus is on representation that occurs through the process where parties are held accountable for their actions retrospectively (accountability representation). The difference between these two basic models of representative democracy lies in when and where in the representative process citizens are involved. In the mandate model, citizens enter early by communicating what policies they wish to see. In accountability representation, they are instead thought to use their influence in the latter part of the process by looking back and evaluating what representatives 27
Chapter 2 have done. Elections are used in the first model to judge future politics, and in the second to evaluate politicians earlier performances. 2 The division between a mandate model of representation and a model of representation that focuses on the sanctioning of parties retrospectively is useful to studies on representative democracy for both empirical and normative purposes. The models can on the one hand be seen as ideal types, or measurement points, towards which reality is compared. Scholars then investigate whether the real representative process does or does not function in the way that the models describe. On the other hand, the models are used as ideals, giving arguments for how representative democracy should work. 3 However, in an actual representative situation, the relationship between citizens and their political representatives is not distinctively divided between a forward-looking and backward-looking process of representation. The periods between elections are highly related to each other. A country s yearly budget is for example often tied up by running expenses. The room for action for the governing political parties is therefore limited. Research also indicates that citizens both think ahead and evaluate the past when they vote (see for ex. Stokes 2001:174-6 and Lewis-Beck & Stegmaier 2007). In a normative sense it can be reasonable to argue that 2 For literature focusing on mandates, see APSA 1950; Ranney 1954; Ranney 1975; Klingemann, Hofferbert & Budge 1994; Esaiasson and Holmberg 1996; Schmitt & Thomassen 1999; Pierce 1999; Budge, Klingemann, Volkens, Bara & Tanenbaum 2001 and all the pledge studies that will be described in Chapters 3 and 4 in this book. A focus on democratic accountability is found in Key 1966; Fiorina 1981; Ferejohn 1986; Ginsberg & Stone 1996; Manin 1997; Przeworski, Stokes & Manin 1999. An interesting summary and interpretation of the relationship between mandates and accountability is found for ex. in Stokes 2001: Chapter 6; see also Erikson, Mackuen & Stimson 2002. 3 The distinction between the empirical and the normative use of the models is hardly always clear. Already in the early debates about the political mandate in the beginning of the 1950s when the APSA report Towards a Responsible Two Party System was published, Austin Ranney lifted this as a problem. The APSA report served as a starting point for the broad discussion of the idea that citizens give a party clear mandates to act on the bases of their election programs (APSA 1950). Ranney, who was one of the most important critics of the APSA report, underlined the problem that scholars mix the discussions about how parties behave with discussions about how they should behave. He called the famous report from APSA a leading example of this kind of confusion (Ranney 1954:8). As noted by Jacques Thomassen, this has also been the case in later studies of representative democracy (Thomassen 1994:237). 28
Election promises in representative democratic theory good representation means taking both aspects into account. In Hanna Pitkin s classical definition of representation as acting in the interest of the represented in a manner responsive to them, representative democracy therefore is something complex and complicated (Pitkin 1967:209, see also Eulau & Karps 1977; Kelley 1983; Dahl 1990; Manin 1997; Przeworzki, Stokes & Manin 1999; Esaiasson 2003). Dividing representative democracy into periods before and after elections gives an illusion of representative democracy stopping and starting again on Election Day. Still, the division between a forward- and a backward-looking model for representation is useful both for classics like Pitkin s The Concept of Representation and for this chapter. In my case the two models illustrate where in the representative process election promises can have roles to play. We will see that, regardless of whether good representation is described as mandate representation or as accountability representation, election promises are tools that have a potential to make representative democratic theory less vague and more comprehensible. To begin with, three roles of election promises will be derived from the mandate approach to representation. Thereafter, two roles of election promises in accountability representation will be discussed. THREE ROLES OF ELECTION PROMISES IN MANDATE REPRESENTATION The mandate approach to representation is sometimes referred to as the most commonly used theoretical model for representative democracy (Esaiasson & Holmberg 1996:3; Pierce 1999:10). It is also the model that is most often used to put election promises in a theoretical context. The empirical literature that evaluates fulfilment of election promises chooses to more or less only refer to the mandate model (see for example Royed 1996; Thomson 1999; Mansergh & Thomson 2007; Artés & Bustos 2008). The essence of the mandate approach to representation is the communication between citizens and representatives about future politics. The underlying idea is that citizens and representatives agree on what kind of politics should be pursued after the coming election. The agreement is the 29
Chapter 2 result of information from the political parties side about what they want to do, and abilities on the citizens side to comprehend and use this information. More specifically, political parties clearly declare what they want to do, preferably in specific election manifestos. Citizens listen to what is said and compare the political parties views to their own opinions and vote for the party that comes closest to their preferences. The parties in power then pursue the policies promised in their election manifestoes (see for ex Royed 1996:46; Schedler 1998:195; Schmitt & Thomassen 1999:15-16; Pierce 1999:9; Budge et al 2001). The most obvious role of election promises in this mandate model of representation is to specify parties policy wishes, which will also be the place for me to start my discussion about election promises in mandate representation. 1. Election promises give parties policy wishes concrete meaning Central to a discussion of the role of election promises to representative democracy is the demand for enlightened communication between citizens and parties. Theories of representative democracy generally formulate high demands on what information the represented and the representative should have about each other. For citizens to be able to choose the party that comes closest to their opinions (as the mandate model expects them to do), good knowledge about parties policy wishes is needed. Citizens have two main ways to get information about a party s policy wishes. They can either communicate directly with the party itself by direct conversations or by reading what the party writes or they can communicate with the party through someone else. Media is an obvious such channel for indirect information for citizens in a democratic society. There are however also other so called shortcuts to information about parties. By listening to people who are, from a specific individual s perspective, trustworthy, a voter can get information about what she should think. The most important form of such a shortcut to political opinions and knowledge is the attachment to a specific party, so called party identification (Dalton & Wattenberg 2000; Thomassen 2005; Oscarsson & Holmberg 2008). Other similar shortcuts are individuals group affili- 30
Election promises in representative democratic theory ations regarding what they work with, which class they belong to, where they live and where they come from (Popkin 1991). The obvious advantage of using these shortcuts to information is that an individual does not have to make up her mind on all issues by herself; she can trust others to come to the right conclusions. A representative system of the same scoop as a national state can at least theoretically be argued to have use of these kinds of shortcuts. Still, there are also (again at least theoretical) disadvantages of forming an opinion on the basis of shortcuts. The citizen places herself in a position where the information she gets is interpreted, and maybe also changed, by the mediator, which could make it harder for her to understand what she actually wants (Dalton & Wattenberg 2000; Lupia 1994; compare Bartels 1996; Luskin 2002). Scholars argue that both party identification and other sorts of group affiliation are in decline in most Western democracies. People do not feel attached to a specific group in the same way as they have before (see for example Dalton and Wattenberg 2000). It is therefore possible to argue that actual political representation (as opposed to theoretical models) is in need of direct channels to provide information about parties policy wishes. This kind of direct communication is hard to accomplish, however. In the empirical literature, notions like informational shortage, informational asymmetry or informational scarcity illustrate that the literature identifies the lack of information as an important characteristic of representative democracies (see for example Congleton 2007; Lupia & McCubbins 1998). Manin, Przeworski, and Stokes put it in the following way: The main difficulty both in instructing governments what to do and in judging what they have done is that we, citizens, just do not know enough (Manin, Przeworski & Stokes 1999:23). In this situation, election promises can be argued to have the potential to be tools that contribute to a relationship between the party and the voter that is based on accurate information. Election promises give relatively understandable information about parties policy wishes. At least theoretically therefore, they facilitate direct communication between citizens and representatives by helping citizens understand the ideas of the different political parties. 31
Chapter 2 2. Election promises give an idea about how citizens make priorities between different policies Theories on representative democracy also place high demands on parties abilities to understand citizens. When Hanna Piktin describes representation as acting in the interest of the represented, in a manner responsive to them, she gives emphasis to representatives responsibility to grasp citizens priorities (Pitkin 1967:209). Put in a brief way, the party should both know the true interest of the people and listen to what citizens express when their interest and wishes do not go hand in hand. As for the first demand to understand the interest of the represented parties are guided by their political ideologies. The second demand responsiveness does not have an equivalently natural solution for the parties (see for example Eulau & Karps 1977). It is not as obvious how you go about being responsive to the electorate, not even the day after an election when you immediately have to make an interpretation of the wishes of the people. In contrast to scholars, who have years to analyse what citizens actually meant with their ballots, the representatives need to know how to interpret their so called mandate already the day after Election Day. What did the people mean with their ballots? What do they want us to do now? In the same way as citizens need tools to understand how the representatives think and work, representatives therefore need tools for rapidly making up their minds about what the electorate wishes to see. Election programs with clearly stated promises can serve as guides for the representatives in this respect. It is to the election manifesto that the parties turn the day after an election, simply because they do not have much of a choice if they want to be responsive to the electorate so early in the election period. Later on, the election manifesto can be combined with opinion polls, mini-demos, deliberative polls, focus groups, membership voting within the party, or other contacts with citizens. But, as a starting point, the manifesto must be seen as an important tool for the parties in trying to understand the priorities of citizens. However, interesting in this regard is the discussion about the impossible mandate. Scholars argue that it is difficult to regard elections as communicating a cohesive message from citizens to representatives. By suggesting that citizens articulate consistent wishes, the notion of 32