of 42

Ce document a été numérisé par la Division de la gestion des documents et des archives de l Université de Montréal.

0 views
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Description
Direction des bibliothèques AVIS Ce document a été numérisé par la Division de la gestion des documents et des archives de l Université de Montréal. L auteur a autorisé l Université de Montréal à reproduire
Transcript
Direction des bibliothèques AVIS Ce document a été numérisé par la Division de la gestion des documents et des archives de l Université de Montréal. L auteur a autorisé l Université de Montréal à reproduire et diffuser, en totalité ou en partie, par quelque moyen que ce soit et sur quelque support que ce soit, et exclusivement à des fins non lucratives d enseignement et de recherche, des copies de ce mémoire ou de cette thèse. L auteur et les coauteurs le cas échéant conservent la propriété du droit d auteur et des droits moraux qui protègent ce document. Ni la thèse ou le mémoire, ni des extraits substantiels de ce document, ne doivent être imprimés ou autrement reproduits sans l autorisation de l auteur. Afin de se conformer à la Loi canadienne sur la protection des renseignements personnels, quelques formulaires secondaires, coordonnées ou signatures intégrées au texte ont pu être enlevés de ce document. Bien que cela ait pu affecter la pagination, il n y a aucun contenu manquant. NOTICE This document was digitized by the Records Management & Archives Division of Université de Montréal. The author of this thesis or dissertation has granted a nonexclusive license allowing Université de Montréal to reproduce and publish the document, in part or in whole, and in any format, solely for noncommercial educational and research purposes. The author and co-authors if applicable retain copyright ownership and moral rights in this document. Neither the whole thesis or dissertation, nor substantial extracts from it, may be printed or otherwise reproduced without the author s permission. In compliance with the Canadian Privacy Act some supporting forms, contact information or signatures may have been removed from the document. While this may affect the document page count, it does not represent any loss of content from the document. ,..( Université de Montréal Experimentation and Political Science:. Six Applications par Peter John Loewen Département de science politique Faculté des arts et des sciences Thèse présentée à la Faculté des études supérieures en vue de l'obtention du grade de Philosophiae Doctor (Ph.D.) en science politique Avril John Loewen,2008 Université de Montréal Faculté des arts et des sciences Cette thèse intitulée: Experimentation and Political Science: Six Applications par Peter John Loewen (LOEP ) a été évaluée par un jury composé des personnes suivantes: Denis Saint-Martin président-rapporteur André Blais directeur de recherche Patrick Fournier codirecteur Richard Nadeau membre du jury James Druckman examinateur externe Chantal Benoit Barre représentante du doyen de la FES Résumé Cette thèse démontre l'utilité de l'expérimentation en science politique à l'aide de six articles. Bien que disparates quant à leurs questions et sujets, ils sont tous liés au comportement et à la psychologie politique. Le premier article examine le rôle des considérations qui ne sont pas de l'ordre de l'intérêt personnel dans la formation des préférences pour les dépenses publiques. Je mesure l'altruisme des répondants par un jeu du dictateur. Je démontre qu'un niveau d'altruisme élevé prédit un plus fort appui pour des dépenses publiques, malgré le coût pour le répondant.. Le deuxième article s'intéresse à un paradoxe central de la participation politique: s'il est peu probable que le vote d'un simple citoyen décide du résultat d'une élection, pourquoi y a-t-il autant d'électeurs? Je démontre à l'aide d~un modèle formel que certàins votent pour le bénéfice des autres. À l'aide d'une série de jeux du dictateur, je montre ensuite que les préférences variées pour certains partisans prédisent la décision de voter ou non. Le troisième article présente les résultats d'une expérience de terrain sur les capacités de persuasion du courrier publicitaire. Durant la campagne de Michael Ignatieff pour la direction du Parti libéral du Canada en 2006, nous avons assigné de façon aléatoire du courrier aux délégués qui s'étaient engagés auprès d'autres candidats, puis nous les avons sondés. Ceux ayant reçu du courrier ont ajusté leur évaluation des autres candidats à la hausse et ont moins bien classé M. Ignatieff. Il' Le quatrième article montre comment le modèle de Bradley-Terry peut être utilisé pour analyser le pouvoir persuasif de l'argumentation. Au cours du référendum d'octobre 2007 sur la réforme électorale en Ontario, nous avons assigné à chacun des 520 répondants de notre sondage un argument en faveur et un argument contre la réforme, avant de leur demander leur opinion. Les arguments pour le système existant profitent d'un avantage général, tout comme les arguments qui font appel à la justice et à la représentation locale. Les arguments qui mentionnent les partis politiques sont moins persuasifs. Le cinquième article pose la question des divers niveaux d'altruisme chez ceux qui s'identifient à un parti politique au Canada. Nous observons les différences entre les partisans dans leur allocation aux co-partisans, à d'autres partisans, et à des individus anonymes dans quatre jeux du dictateur lors d'un sondage en ligne. Tous donnent plus aux co-partisans et moins aux autres partisans. Les Néo-démocrates sont plus altruistes en moyenne que les Conservateurs et les Libéraux. Dans le sixième article, nous nous demandons si le vote obligatoire mène aux effets de second ordre d'augmentation des connaissances et d'engagement des citoyens. Nous avons conduit une expérience de terrain avec des étudiants en âge de voter d'un cégep de Montréal. Certains étaient payés pour compléter deux sondages, d'autres étaient aussi payés pour voter lors de l'élection provinciale. Nous avons trouvé peu d'indications d'effets de second ordre. Mots clés: expérimentation; comportement politique; psychologie politique; vote; économie comportementale; campagnes électorales Summary This dissertation demonstrates the usefulness of experimentation in political science through six articles. While eclectic in their questions and subjects, the articles ah fau under the rubric of political psychology and behaviour. The first article examines the role of non-self-interested considerations in the formation of preferences for public spending. l measure the altruism of respondents through the use of a dictator game experiment in a large online survey. l demonstrate that ,L'-,'- V'~L levels of altruism predict greater support for public spending, even when it cornes at a cost to the respondent. The second article addresses the paradox of participation : if the probability of a single vote deciding an election is so low, why do we still observe large numbers of vot ers? l demonstrate in a formai model that sorne Îndividuals will vote because of the benefits accrued to others. Empirically, preferences for partisans which differ in a dictator game significantly predict the decision to vote. The ~hird article presents the results from a field experiment into the persuasive capacities of direct mail The experiment was conducted in 2006 with the Michael Ignatieff campaign for the leadership of the Liberal Party of Canada. The experiment randomly assigned a direct mailing to delegates pledged to other candidates. We then surveyed these delegates. We find that those who received mail appeared to adjust their evaluations of other candidates upwards and to move Ignatieff down in their preference rankings. iv The fourth article uses a Bradley-Terry model to analyze the persuasive power of arguments in a survey experiment conducted during the October 2007 Ontario referendum on electoral reform. We assigned each of 520 respondents one of six arguments for and one of six arguments against electoral reform and then measured their preference for reform. We show that arguments for the existing system enjoyed an advantage, arguments whieh appeal to fairness and local representation were significantly more persuasive, and arguments which mention political parties were less persuasive. The fifth article asks if partisan identifiers in Canada differ in their levels of altruism. We examine the behaviour of partisans in four dictator games in an online survey. We compare differences between partisans in their allocations to co-partisans, other partisans, and anonymous individuals. AU partisans consistently auocate the most to co-partisans and the least to other partisans. Anonymous individuals are in the middle. We also find that New Democrats are more altruistic than Conservatives or Liberals. The sixth article asks whether compulsory voting leads to the secondorder effects of increased citizen knowledge and engagement. We conducted a field experiment among voting-aged students at a Montreal CEGEP. Our intervention involved paying sorne students to complete two surveys while paying another ùoup to also vote in a provincial election. We find little evidence of second-order effects. Keywords: Experimentation; Political Behaviour; Politieal Psychology; Voting; Behavioural Economies; Campaigns v Table de matières Résumé. Summary Liste des tableaux. Liste des figures. Dédicace.. Remercients 1 Introduction 1.1 The Underside of 'Warren Miller's Cowboy Boots. 1.2 Defining Experimentation and Experimental Types Experimental Types.... III Xll XVI. XVll. xviii An Eye Out for Inferential Monsters: The Case for Experimentation in Political Science The Quantity of Experimentation in Political Science The Nature of Experimentation in Political Science The Qualities of Experimentation in Political Science Statistical conclusion validity Vi InternaI validity Construct Validity External Validity Comparing the types The Articles Dictators and Purses: Altruism and Support for Public Spending Antipathy, Affinity, and Political Participation: How Our Concern for Other Partisans Makes Us Vote For Want of a Nail: Direct Mail and Negative Persuasion in a Leadership Race (with Daniel Rubenson) Testing the Power of Arguments with a Bradley-Terry Model (with Daniel Rubenson and Arthur Spirling) Partisanship and Altruism: Results from a Dictator Game Experiment (with Angelo Elias) Does Compulsory Voting Lead to More Informed and Engaged Citizens: An experimental test (with Henry Milner and Bruce M. Hicks) 1. 7 A Closing Word Dictators and Purses: Altruism and Support for Public Spending Introduction Altruism and Public Spending 48 51 VII Defining Altruism Theoretical Link Previous Findings Innovation Altruism and Dictator Games 2.4 Survey Design and Participants Participants 2.5 Results Support for Public Spending. 2.6 Discussion and Conclusion Affinity, Antipathy and PoIitical Participation: How Our Concern For Other Partisans Makes Us Vote Introduction Group Politics. 3.3 A Different Calculus of Voting. 3.4 Survey and Research Design Survey Subject Profiles Antipathy, Affinity, and other variables 3.5 Antipathy, Affinity, and Party Identification Why Antipathy and Affinity are not just Party Identification Antipathy, Affinity, and Turnout . viii 3.7 Discussion and Conclusion For Want of a Naïl: Direct Mail and Negative Persuasion in a Leadership Race (with Daniel Rubenson) Introduction Direct Mail and Persuasion The Race The Experimental Study The Experiment The Survey Results and Discussion Conclusion.. ~ Testing the Power of Arguments with a Bradley-Terry Model (with Daniel Rubenson and Arthur Spirling) Introduction Data, Context and Connections Experimental Design The 'Framing' Connection Measuring the Power to Persuade: A Statistical Model The Power of Arguments: Conventional Results Testing the Power of Arguments Determining the Sources of Power The Power of Arguments: The Bradley-Terry Method Unstructured Results lx The Sources of an Argument's Power 5.6 Further Applications and Conclusion Partisanship and Altruism: Results from a Dictator Game Experiment (with Angelo Elias) Introduction Altruism and the Dictator Game The Study Subjects The Survey Dependent Variable: Dictator Game Allocations Independent Variables Results Discussion and Conclusion Does Compulsory Voting Lead to More Informed and Engaged Citizens: An experimental test (with Henry Milner and Bruce M. Hicks) Introduction Existing Knowledge and the Case for Experimentation Hypotheses and Experimental Design Hypotheses Subject Recruitment and Survey Administration Survey and Dependent Variables Sample Profile x 7.4 Results Conclusion Conclusion Introduction Summary Implications Methodological Implications Substantive Implications Future Research The Sound Down the Hall 238 Références bibliographiques 239 Annexe A: Criterion Validity of Dictator Game as a Measure of Altruism XXVI Annexe B: Dictator Game Instructions xxxii Annexe C: Question Wording and Variables for Dictators and Purses XXXV Annexe D: Question Wording and Variables for Affinity, Antipathy, and Political Participation xxxviii Annexe E: Treatment Assignment. for For Want of a Nail xli Annexe F: Sample and Subject Profile for Bradley-Terry Experiment xliv ) xi Annexe G: Additional Logit Results for Bradley-Terry Experiment xlvi Annexe H: Predicted Probabilities of FPTP dominance in Structured versus Unstructured Bradley-Terry Models xlviii Annexe 1: Question Wording and Variables for Partisanship and Altruism Annexe J: Supplementary Tables for Partisanship and Altruism liv Annexe K: Treatment Assignment for Compulsory Voting Experiment lix Annexe L: Question Wording and Variables for Compulsory Voting Experiment lxii Annexe M: Accord des coauteurs et permission de l'éditeur lxviii Curriculum Vitae lxxvi Xll Liste des tableaux 1.1 Frequency of Statistical Techniques in Observational and Experimental Articles Sample demographic and political characteristics Willingness topay for a carbon tax (Ordered Logistic Regression) Willingness to pay for free university tuition (Ordered Logistic Regression) Willingness topay more taxes to reduce waiting times for medical services (Ordered Logistic Regression) Willingness to wait longer for medical care if cancer patients can receive drug coverage (Ordered Logistic Regression) Main and conditional effects of altruism when interacted with education and income (Ordered Logistic Regression) Sample demographic and political characteristics Partisanship and Average Allocations in the Dictator Game 107 Xlll 3.3 Within-Subject Differences in Dictator Game Allocations CWilcoxon Sign-Rank Differences) Antipathy, Affinity and Turnout (Logistic Regression, Odds Ratio) Separating Antipathy and Affinity (Logistic Regression) Liberal Party leadership election results Effects of Ignatieff mail on average leadership candidate ratings (mean differences) Effects of Ignatieff mail on delegates' preference ordering (Ordered Logistic Regression) Argument Characteristics Agreement with :Yl:YlP by FPTP Arguments (mean differences) Agreement with MMP by MMP Arguments (mean differences) Logistic Regression of Argument Power a Logistic Regression of Sources of Argument Power Bradley-Terry Model of Argument Power Power of FPTP Arguments Power of MMP Arguments. 5.9 Model of Sources of Argument Power Dictator Game Allocations to Anonymous Individuals (OLS) Dictator Game Allocations to Other Partisans (OLS) Dictator Game Allocations to Co-Partisans (OLS) 199 xiv 6.4 Pooled Dictator Came Allocations (OLS) Sample Profile of compulsory voting treatment on political knowleclge, political discussion, and media usage (mean differences) Effect of treatment on knowledge, news consumption and discussion of politics for voters and non-voters (OLS) A.1 Altruism and Self-Reported Charitable Civing (Ordered Logistie Regression)... xxix A.2 Altruism and Response to those in Distress (Ordered Logistic Regression)... xxx A.3 Altruism and Willingness to Help Those Who Don't Help Themselves First (Ordered Logistic Regression)... XXXI F.1 Sample Demographies and Political Characteristics.. xlv C.1 Logistic Regression of Argument Power with Argument Matchings... xlvii H.1 Absolute difference in predicted probabilities of FPTP dominance by structured and unstructured models a..... xlix J.1 Dictator Came Allocations to Anonymous lndividuals - Low Education (OLS) liv xv J.2 Dictator Game Allocations to Other Partisans - Low Education (OLS) Iv J.3 Dictator Game Allocations to Co-Partisans - Low Education (OLS) Iv J.4 Dictator Game Allocations to Anonymous lndividuals - High Education (OLS)... ' lvi J.5 Dictator Game Allocations to Other Partisans - High Education (OLS)... Ivi J.6 Dictator Game Allocations to Co-Partisans - High Education (OLS) lvii J.7 Pooled Dictator Game Allocations - Low Education (OLS) Ivii J.8 Pooled Dictator Game Allocations - High Education (OLS). lviii .. xvi Liste des figures 1.1 Share of Experimental Articles in A JPS, APSR, and JOP Dictator Came Allocations Demand Curves for Public Spending Affinity, Antipathy, and Turnout ~.. ~ Power of arguments: unstructured results Power of arguments: structured results.. 184 For A.K. 1 am a lazy cat She is as pure as the cold driven snow and For my mother and my father To whom 1 owe more th an 1 can pay Remercients l push on open doors. That is to say, l am at a point which is supposed to be the result of great effort and sacrifice, but for me it has all seemed, if not easy, then at least certain. What keeps me from pride, then, is that l am quite certain that others have done most of the work for me. Thus my debts are large. l shall try to be as sufficient in addressing them as l have been adept in amassing them. l am indebted to Brian Thomas and James Baxter. From the time l was a child, l never lacked in intellectual confidence (whether l lacked in ability is another question!). But Thomas and Baxter were instrumental in affirming a life of intellectual pursuit and in convincing a young man in a small city that with sorne work his confidence could be justified, and that it was worth a shot. Somewhere a ledger exists and their names are written there in red ink. This red ink continued a steady fiow at Mount Allison University. To decamp to Sackville, NB was the best decision lever made and l attribute' it still to luck, whimsy and my parents' confidence and insistence that l leave my home town for my education. It had nothing to do with rational calculation. l owe a great debt to my first professors, especially Wayne Hunt, for teaching me that politics was ultimately about ideas, Janine Rogers, for teaching me about poetry and for that harrowing drive through Algonquin Park, and Frank Strain, for teaching me about market failures and for always XIX offering a place to stay. 1 thank, too, Rev. John Perkin and Charlie Hunter, who continue to share their wisdom with me. 1 should also like to thank the presidents of that institution, Ian Newbould, A. Wayne MacKay, and Robert Campbell. That 1 receive regular offers of assistance from ail of them is a testament to how lucky I was to end up at such a personal and exceptional place. And 1 th~nk finally and most sincerely Bill Cross. It was Bill who supervised my final paper, who taught me something about being a research assistant, who introduced me to so many of Canada's political scientists, and who arranged for me to work with André Blais. He opened many do ors and always made me think it was my doing. My time in graduate school has been no less indebting. l've had the great privilege to be taught by Stuart Soroka, Elizabeth Gidengil, Louis Massicotte and André J. Bélanger, whose imprint I should think is indelible. During a sojourn in America, 1 was taught and greatly influenced by Jim Johnson, Kevin Clarke, Dick Niemi, and Bing and Linda Powell. They too are in my debts, but they'll have to cross the border to collect. Outside of my formai studies, l've also been lucky to form connections and friendships with Jamie Druckman, Skip Lupia, John Aldrich, James Fowler, and Charles Blattberg. Their influence is thorough-going, and for this 1 am richer. Of,special note are the constant and unmatched intellectual critiques I have received From Patrick Fournier, my co-director. Let him move tow
Related Search
Advertisements
Related Docs
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks