A migração partidária, embora seja um evento relativamente raro na maioria dos países, trata-se de um fenômeno disseminado no Brasil. O objetivo deste artigo é apresentar empiricamente os efeitos da migração partidária sobre variáveis políticas entre os anos de 2001 e 2012, fazendo uso de um banco de dados com 4974 municípios. Além disso, parte-se de uma abordagem alternativa, uma vez que, diferente da literatura convencional que trata da migração partidária de legisladores, neste artigo a atenção se voltará para o executivo, mais especificamente, para os prefeitos, algo ainda não explorado pela literatura. A mudança partidária do prefeito é interpretada como um efeito de tratamento, em que o grupo tratado é o de municípios cujo prefeito mudou de partido durante a vigência do seu mandato. Para encontramos um município no grupo de controle que possua as mesmas chances de ser tratado aplicamos um Propensity Score Matching (PSM), e então será comparado o impacto da migração sobre os políticos migrantes. Os resultados obtidos via logit apontam que os prefeitos pertencentes a partidos coligados a esferas superiores de poder tem menor probabilidade de mudarem de partido, uma vez que estes teriam melhores condições de auxiliarem os prefeitos em sua tentativa de reeleição. Além disso, foi concluído que os prefeitos migrantes possuem maior chance de tentarem se reeleger na eleição seguinte, mas sem muito impacto sobre a chance de se reelegerem, sendo que, dentre os migrantes, os melhores desempenhos para ambas as variáveis dependentes se encontram entre aqueles que migram de um partido de fora da base do governo estadual para um partido que esteja dentro dessa base. Por fim, não foi constatato nenhum efeito estatisticamente significante da migração, tanto em si como desagregada pelos tipos de migrantes, sobre as transferências recebidas pelos municípios / Although party switching is a relatively rare event in most countries, it is a widespread phenomenon in Brazil. This paper aims to present empirically the effects of party switching on political variables between the years 2001 and 2012, using a database with 4974 municipalities. Moreover, it is based on an alternative approach, since the conventional literature deal with party switching of legislators, this paper will focus on the executive, more specifically, on mayors, something not yet explored in the literature. The mayor\'s party switching is interpreted as a treatment effect, in which the treated group is that of municipalities whose mayor changed party during the term of his mandate. To find a municipality in the control group that has the same chances of being treated, we apply Propensity Score Matching (PSM), and then the impact of the switching on the switcher politicians will be compared. The results obtained by logit show that mayors belonging to parties affiliated with the state governor\'s party or the president\'s party are less likely to change parties, since they would be better able to assist the mayors in their re-election attempt. Moreover, it was concluded that switchers mayors are more likely to attempt reelection in the next election, but without much impact on the chance of re-election, and among the switchers, the best performances for both dependent variables are among those who switch from a party outside the state government coalition to a party within that coalition. Finally, there was no statistically significant effect of switching, both in itself and disaggregated by the types of switchers, on the transfers received by the municipalities

<p>A majority of formal theoretic research in political science treats political parties as unitary actors, and endows them with decision-making powers not unlike those of strategic individuals. This is true both of most research in the spatial-theoretic tradition, as well as most game theoretic research in the field of comparative political-economy. In contrast, my dissertation examines strategic equilibria which arise when competition takes place simultaneously within parties over organizational control and between parties over political office. I first distinguish between three intra-organizational elements: a party's parliamentary group, its activist cadre, and its executive leaders. Chapters 2-4 develop a set of foundational game theoretic models which identify the equilibrium balance of power among these 3 organizational elements as a function of a country's electoral institutions and voters' relative responsiveness to marginal policy changes. In turn, this more complete understanding of intra-party competition sheds light on a number of important questions in comparative politics and comparative political-economy. For example, it helps to identify conditions under which Downsian vote-maximization is in fact a viable assumption in spatial theoretic models; conditions under which Duverger's argument that proportional representation (PR) should tend to generate multi-party competition may not apply; and, in contrast to Lijphart's famous argument, conditions under which PR may instigate rather than mediate social conflict. Ten months of intensive field research conducted in Turkey provide both the quantitative and the qualitative data which constitute the dissertation's most basic empirical material. This data includes primary and secondary source material on the history of intra-organizational competition in Turkey; observational and informant-based information on contemporary Turkish politics and the events of 2006-2008; and a data set of over 4,000 observations on party-switching in the Turkish Parliament (1987-2007).</p> / Dissertation

A migração partidária, embora seja um evento relativamente raro na maioria dos países, trata-se de um fenômeno disseminado no Brasil. O objetivo deste artigo é apresentar empiricamente os efeitos da migração partidária sobre variáveis políticas entre os anos de 2001 e 2012, fazendo uso de um banco de dados com 4974 municípios. Além disso, parte-se de uma abordagem alternativa, uma vez que, diferente da literatura convencional que trata da migração partidária de legisladores, neste artigo a atenção se voltará para o executivo, mais especificamente, para os prefeitos, algo ainda não explorado pela literatura. A mudança partidária do prefeito é interpretada como um efeito de tratamento, em que o grupo tratado é o de municípios cujo prefeito mudou de partido durante a vigência do seu mandato. Para encontramos um município no grupo de controle que possua as mesmas chances de ser tratado aplicamos um Propensity Score Matching (PSM), e então será comparado o impacto da migração sobre os políticos migrantes. Os resultados obtidos via logit apontam que os prefeitos pertencentes a partidos coligados a esferas superiores de poder tem menor probabilidade de mudarem de partido, uma vez que estes teriam melhores condições de auxiliarem os prefeitos em sua tentativa de reeleição. Além disso, foi concluído que os prefeitos migrantes possuem maior chance de tentarem se reeleger na eleição seguinte, mas sem muito impacto sobre a chance de se reelegerem, sendo que, dentre os migrantes, os melhores desempenhos para ambas as variáveis dependentes se encontram entre aqueles que migram de um partido de fora da base do governo estadual para um partido que esteja dentro dessa base. Por fim, não foi constatato nenhum efeito estatisticamente significante da migração, tanto em si como desagregada pelos tipos de migrantes, sobre as transferências recebidas pelos municípios / Although party switching is a relatively rare event in most countries, it is a widespread phenomenon in Brazil. This paper aims to present empirically the effects of party switching on political variables between the years 2001 and 2012, using a database with 4974 municipalities. Moreover, it is based on an alternative approach, since the conventional literature deal with party switching of legislators, this paper will focus on the executive, more specifically, on mayors, something not yet explored in the literature. The mayor\'s party switching is interpreted as a treatment effect, in which the treated group is that of municipalities whose mayor changed party during the term of his mandate. To find a municipality in the control group that has the same chances of being treated, we apply Propensity Score Matching (PSM), and then the impact of the switching on the switcher politicians will be compared. The results obtained by logit show that mayors belonging to parties affiliated with the state governor\'s party or the president\'s party are less likely to change parties, since they would be better able to assist the mayors in their re-election attempt. Moreover, it was concluded that switchers mayors are more likely to attempt reelection in the next election, but without much impact on the chance of re-election, and among the switchers, the best performances for both dependent variables are among those who switch from a party outside the state government coalition to a party within that coalition. Finally, there was no statistically significant effect of switching, both in itself and disaggregated by the types of switchers, on the transfers received by the municipalities

This thesis seeks to explain the phenomenon of party switching (or floor-crossing) in the understudied case of the Canadian House of Commons. It uses Müller and Strøm’s “Policy, Office or Votes?” framework at the individual level of analysis and a mixed methods approach that combines document analysis and econometrics to assess the effects of individual and institutional variables on the decision to switch parties. The results inform a wider discussion regarding individual political behavior as well as the role, influence and evolution of political parties in the Canadian state. The research demonstrates that the electorate is adept at recognizing opportunism and tends to respect MPs who switch parties on principle while punishing those that switch for more self-centered reasons.

No description available.