This thesis is about how elected politicians stay in power and about some of its economic and social consequences, fundamental political economy questions. It takes advantage of the decentralized political structure in the Philippines to test models of voter and politician behavior. In doing so, it contributes to the literature on clientelism and retrospective voting. In Chapter One, I assess the impacts of targeted government transfers on a local incumbent's electoral performance. I use the randomized roll-out of a CCT program in the Philippines where a number of municipalities are tightly controlled by political dynasties. In a competitive political environment, incumbent vote share is 26 percentage-points higher in municipalities where the program was implemented in all villages than in municipalities where the program was implemented in half of them. The program had no impact in municipalities with low levels of political competition. In Chapter Two, I test for the presence of political business cycles in Philippine municipalities over the period 2003-2009, a context where according to the literature such cycles are likely to be observed. I find robust evidence for the presence of political business. This effect is only present when I use quarterly data and vanishes when I aggregate the data at the yearly-level. The difference is not merely driven by a decline in statistical power due to aggregation: point estimates for the overall effects are 7 times larger when I use quarterly data than when I use yearly data. This discrepancy can be explained by a drop in employment post-election that dilutes the yearly effects. In Chapter Three, we estimate the impacts of being connected to local politicians, either currently in office or in opposition, on occupational choice. We use a large administrative dataset collected between 2008 and 2010 on all individuals in 700 Philippine municipalities along with information on all candidates in the 2007 and 2010 municipal elections. We rely on local naming conventions to assess blood and marriage links between households. Using individuals connected to successful candidates in the 2010 elections that did not run in 2007 as a control group, we find that connections to current office-holders increase the likelihood of being employed in better paying occupations. Individuals connected to candidates that were close to being elected in 2007 are less likely to be employed in better paying occupations.

Why and under what conditions do governments borrow before elections? This thesis aims to shed light on this question by exploring governments' incentives that give rise to political budget cycles, i.e. fluctuations in the budget balance during election times, under different political, institutional, and economic contexts. The argument will be developed in three stages. First, the thesis will explain why politicians may choose to use debt strategically to win elections and discuss and evaluate different models that can explain political budget cycles. One model, a moral hazard type competence model is, as will be shown, particularly suited for this study. It will be extended in stages two and three. The second stage will look at the benefits and costs from public debt, with a particular emphasis on the likelihood of re-election (government popularity), party system polarisation, and sovereign risk. Sovereign risk increases the cost of borrowing and thus dampens the magnitude of political budget cycles; the effect of government popularity on strategic debt is conditional on the degree of polarisation. The third stage will take the motives to borrow as given and examine the effectiveness of debt as a strategic instrument. The less voters attribute responsibility for fiscal policy to governments, the less effective debt is as a strategic instrument. Economic volatility, regulatory density, and economic openness, this thesis argues, reduce this effectiveness and in turn the political budget cycle. Similarly, coalition government reduces responsibility associated with individual coalition partners, and thus the strategic value of public debt - yet this effect is moderated by the distribution of cabinet portfolios. The argument in this thesis is based both on formal models and on empirical, time series-cross sectional, analyses. It is arguably the most comprehensive treatment of political budget cycles and adds to an increasing literature on the contextual determinants of fiscal policy.

[pt] Esta dissertação consiste em duas partes independentes. A primeira delas é uma avaliação empírica de ciclos políticos orçamentários nos municípios brasileiros. Nessa parte, investigamos primeiramente se as receitas e despesas municipais variam com a proximidade das eleições para prefeito. Em seguida, estimamos os efeitos dos níveis de competição política e transparência governamental sobre a magnitude dos ciclos orçamentários. Nossos resultados sugerem a existência de fortes ciclos orçamentários nos municípios do Brasil, embora seja difícil identificar uma relação causal entre tais ciclos e o calendário eleitoral. Além disso, encontramos evidência de que maior competição política e menor transparência governamental geram incentivos para que os prefeitos em exercício invistam mais nos últimos anos de seus mandatos. Na segunda parte da dissertação, procuramos estimar o impacto de propaganda eleitoral sobre o desempenho dos candidatos a prefeito e governador no Brasil. Estudos anteriores estabeleceram a idéia de que gastos de campanha e propaganda política teriam impacto muito modesto sobre os resultados de pleitos eletivos. No entanto, esses estudos sofrem de problemas sérios de identificação, como omissão de variáveis relevantes e causalidade reversa. Em nosso trabalho, exploramos uma fonte de variação exógena na distribuição dos tempos de propaganda entre os candidatos produzida pela legislação eleitoral brasileira. Nossos resultados indicam que a propaganda política por meio de rádio e televisão exerce um forte efeito sobre o desempenho dos candidatos. / [en] This thesis consists on two independent parts. The first one is an empirical evaluation of political budget cycles in Brazil. The research question is twofold. We ask the factual question of whether municipal level receipts and expenditures vary with the proximity of elections for mayor. Then, we investigate the effects of political competitiveness and govern transparency on the size of the budget cycles. Our results suggest the existence of strong budgetary cycles in the Brazilian municipalities, though it is difficult to identify a causal relationship between such cycles and the electoral calendar. Also, according to our findings, higher political competitiveness and lower govern transparency generate incentives for the incumbent mayors to invest more in the last year of their terms. In the second part of the thesis, we estimate the impact of political advertising on candidate performance in Brazilian mayoral and gubernatorial elections. Previous studies have established a minimal effects conventional wisdom that campaign spending and advertising have little impact on elections outcome. However, such studies suffer from omitted variable bias and reverse causality. We explore a quasi-natural experiment produced by Brazilian electoral legislation, and show that TV and radio advertising have a strong effect on candidates´ electoral performance.

Submitted by Andre Sander Diniz (sdiniz.andre@gmail.com) on 2017-12-04T23:21:55Z No. of bitstreams: 1 TESE_AndreDiniz.pdf: 1446746 bytes, checksum: a43fc82c153a68b1435f83aa1fccac05 (MD5) / Rejected by Suzinei Teles Garcia Garcia (suzinei.garcia@fgv.br), reason: Bom dia André, Por gentileza, verifique modelos de teses na biblioteca digital, sua submissão não segue os padrões. Todas as primeiras páginas. Aguardo nova submissão. Grata. Suzi 3799-7876 on 2017-12-06T12:55:29Z (GMT) / Submitted by Andre Sander Diniz (sdiniz.andre@gmail.com) on 2017-12-07T01:44:52Z No. of bitstreams: 1 TESE_AndreDiniz (2).pdf: 1425640 bytes, checksum: e977ba1359fa366483745166952c3ebf (MD5) / Rejected by Suzinei Teles Garcia Garcia (suzinei.garcia@fgv.br), reason: Bom dia André, Por gentileza, tire o acento da palavra Getulio. Aguardo. Grata Suzi on 2017-12-07T12:46:04Z (GMT) / Submitted by Andre Sander Diniz (sdiniz.andre@gmail.com) on 2017-12-08T01:42:52Z No. of bitstreams: 1 TESE (1).pdf: 1424821 bytes, checksum: 579e86e94e00e7fef70de1763ac50458 (MD5) / Approved for entry into archive by Suzinei Teles Garcia Garcia (suzinei.garcia@fgv.br) on 2017-12-08T20:56:27Z (GMT) No. of bitstreams: 1 TESE (1).pdf: 1424821 bytes, checksum: 579e86e94e00e7fef70de1763ac50458 (MD5) / Made available in DSpace on 2017-12-11T11:55:44Z (GMT). No. of bitstreams: 1 TESE (1).pdf: 1424821 bytes, checksum: 579e86e94e00e7fef70de1763ac50458 (MD5) Previous issue date: 2017-11-28 / This thesis is composed of three essays related to fiscal policy and public debt. The first chapter analyses quantitatively the relevance of different fiscal policy responses to a debt restructuring episode, taking into account endogenous default costs associated to the financial accelerator mechanism. The second chapter presents an empirical exercise on the effects of fiscal consolidations for Latin American countries, with the study of the impact on output and some aggregate demand components, as well as an investigation of the relative importance of revenue or expenditure-based adjustments and possible non-linearities in the economy's response. The third chapter suggests a simple political economy model to rationalize the presence of political budget cycles for external debt and tests the empirical implications of the model in a panel of developed and emerging economies. / Essa tese é composta por três ensaios relacionados aos temas de política fiscal e dívida pública. O primeiro capítulo faz uma análise quantitativa da importância dos diferentes tipos de resposta da política fiscal a um episódio de reestruturação de dívida, levando em consideração custos endógenos de default associados ao mecanismo do acelerador financeiro. O segundo capítulo apresenta um exercício empírico sobre os efeitos de ajustes fiscais em países da América Latina, estudando o impacto sobre produto e alguns componentes da demanda agregrada, bem como investigando a importância de ajustes baseados em receitas ou despesas e possíveis não-linearidades na resposta da economia. O terceiro capítulo sugere um modelo simples de economia política para racionalizar a presença de ciclos eleitorais de dívida externa, e em seguida testa implicações empíricas do modelo em um painel de economias desenvolvidas e emergentes.

No description available.

L'objet de la présente thèse concerne l'étude de l'impact économique de trois modalités démocratiques, à savoir (i) l'agenda électoral, (ii) l'affiliation partisane des dirigeants en place dans un cadre de multipartisme, et (iii) les changements de gouvernance constitutionnellement définis. Afin d'introduire notre étude, nous décrivons la relation qui existe entre les modalités démocratiques et la démocratie directe d'une part et la politique budgétaire d'autre part, aux Etats-Unis sur la période 1790-2014 dans l'Introduction Générale. Dans le Chapitre 2 nous étudions l'effet authentique et potentiellement néfaste des cycles politico-budgétaires. Notre méta-analyse suggère que les dirigeants nationaux manipulent effectivement le budget dans un but de réélection mais l'ampleur du phénomène est largement exagérée par la littérature. Toutefois, le biais de publication mis en lumière s'est significativement réduit lors des 25 dernières années de recherches. Dans le Chapitre 3 nous étudions comment l'affiliation partisane des gouverneurs américains affecte le statut de pauvreté des immigrants, aux Etats-Unis, sur la période 1994-2014. Pour ce faire, nous comparons le niveau de pauvreté des immigrants dans les Etats gouvernés par les Démocrates au niveau de pauvreté des immigrants constaté dans les Etats gouvernés par les Républicains. En accord avec la littérature sur l'affiliation partisane, nous trouvons que les immigrants ont plus d'opportunités de sortir de la pauvreté sous les Démocrates que sous les Républicains. Une analyse formelle de médiation révèle que nos résultats empiriques sont médiatisés par un meilleur accès au marché du travail et possiblement de meilleures rétributions du travail, pour les immigrants. Dans le Chapitre 4 nous cherchons à vérifier l'évolution des notations souveraines lors des périodes d'inauguration des nouveaux dirigeants sur un échantillon de 18 pays d'Amérique Latine et des Caraïbes ayant des systèmes présidentiels. A partir de données de panel journalières s'étendant du 1er janvier 1994 au 31 décembre 2014, nous trouvons que les notations souveraines sont de meilleure qualité durant les périodes d'inauguration anticipées comparées aux autres périodes d'inauguration. En outre, nos résultats montrent que durant ces périodes d'inauguration anticipées, les notations sont encore meilleures lorsque le dirigeant entrant (i) est économiquement de droite, (ii) dispose d'un diplôme universitaire d'un pays de l'OCDE, (iii) a un parcours professionnel traditionnel, (iv) a une approche non populiste, et (v) dispose de marges électorales de victoire importantes. Il apparaît également un biais de genre potentiel à la faveur des nouveaux dirigeants de sexe masculin. En prenant avantage d'un modèle de durée et de régressions de Cox -- modèle à risque proportionnel, nous mettons en lumière que les périodes d'inauguration des dirigeants affectent également la probabilité instantanée de dégradation des notations souveraines. Enfin, dans la Conclusion Générale nous explorons les racines des pensées antisystèmes en Occident, puis nous discutons des avantages et inconvénients de quelques formes alternatives de démocratie, à savoir (i) la démocratie directe, (ii) le tirage au sort, et (iii) le développement participatif comme bien public. Finalement, nous explorons les controverses concernant les formes alternatives de démocratie sur l'agora virtuelle que constitue Twitter. / The aim of the present dissertation is to empirically investigate the economic impact of three democratic modalities, namely (i) the electoral agenda; (ii) the partisan affiliation of incumbents in a multipartism framework; and (iii) constitutionally defined leadership changes. To introduce the scope of the dissertation we describe the democratic modalities and direct democracy behind the U.S. fiscal policy from 1790 to 2014, in the General Introduction.In Chapter 2 we study the genuine detrimental effect of political budget cycles. Our meta-analysis suggests that national leaders do manipulate fiscal tools in order to be re-elected, but to an extent that is significantly exaggerated in the literature. The publication selection bias highlighted has nonetheless been reduced during the past 25 years of research. In Chapter 3 we investigate how governors' partisan affiliation affects the poverty status of immigrants to the U.S for the period 1994-2014. To this end, we compare the poverty outcomes of immigrants in states ruled by Democratic governors relative to the outcomes for those in states ruled by Republican governors. Consistent with the literature on partisan affiliation, we find that immigrants are more likely to get out of poverty in states with Democratic governors than states with Republican governors. A formal mediation analysis reveals that the empirical results are mediated through better access to the labor market and possibly through higher wages and labor earnings for immigrants. In Chapter 4 we assess whether sovereign credit ratings change during the inauguration periods of incoming leaders, on a sample of 18 Latin American and Caribbean countries with presidential systems. Building on a daily panel dataset covering the period from January 1, 1994 to December 31, 2014, we find that credit ratings are better during anticipated inauguration periods compared to other inauguration periods. Moreover, our results reveal that, during anticipated inauguration periods, incoming leaders with (i) an economically right-wing orientation; (ii) an OECD college degree; (iii) a traditional professional background; (iv) a non-populist approach; and (v) large electoral margins of victory are associated to even better ratings. There also appears to be a potential gender bias effect in favor of male incoming leaders. Last but not least, the quality of the credit rating also matters. In the General Conclusion we explore the roots of anti-systemism in the West, then we discuss some alternative forms of democracy, and we explore the controversies regarding the alternative forms of democracy on the virtual \emph{agora} of Twitter.