class: center, middle, inverse, title-slide .title[ # Party Coalitions in Presidential Systems: How Defection Hurts Presidential Approval Ratings NCCLAS - 2024 ] .author[ ### Eliana Álvarez<span class="math inline">¹</span>, Juan Bogliaccini<span class="math inline">¹</span>, Peter Enns<span class="math inline">²</span>, Martín Opertti<span class="math inline">³</span> and Rosario Queirolo<span class="math inline">¹</span> ] .institute[ ### <span class="math inline">¹</span> Universidad Catolica del Uruguay <span class="math inline">²</span> Cornell University <span class="math inline">³</span> Duke University ] .date[ ### February 23, 2024 ] --- # Background ## Coalition Defection in Presidential Systems -- - Coalitions are a very common form of government in Latin American countries with presidential systems. -- - Coalition defections occur. Coalitions in Presidential Systems vary in the duration and are affected by president's value in coalition building and participation cost for other parties (Martínez-Gallardo, 2012) -- - The consequences of coalition arrangements on public opinion have been long studied in parliamentary systems but not quite in presidential systems. -- - Important differences in the meaning and consequences of coalition defection in presidential vs parliamentary systems. Presidential systems have **fixed terms** --- # Background ## Coalition Arrangments and Public Opinion -- - Parties play a crucial role in voters' opinion formation. -- - **Party cues** are signals that link a political party to a position on an issue or to an opinion in general, which is shown that influence how citizens form their opinions. -- - Coalition formation is a straightforward party action that voters use as a heuristic for updating their perceptions -- - When party form coalitions (in parliamentary settings) affect perceptions of parties' ideological positions (e.g. Fortunato and Stevenson, 2013), party preferences (e.g. Plescia, 2021) and performance evaluations. -- - **How does a coalition defection affects public opinion in a fixed-term setting?** --- # Motivation - Coalitions and defections are common in presidential systems -- - Coalition defection could act as a **powerful cue** for citizens -- - Defection can affect presidential approval given Presidents play a key role in the formation and duration of coalitions in these systems. -- - Presidential approval in Latin America has meaningful consequences -- - **Does coalition defection affect presidential approval?** --- # Theoretical Expectations - **Hypothesis 1:** After a coalition defection, there will be a rise in presidential disapproval among partisans of the defecting party. Defection will be a powerful cue for defecting party partisans and they will update their beliefs accordingly. -- - **Hypothesis 2:** After a coalition defection, there will be a rise in disapproval among partisans of the defecting party that is significantly higher than the change in disapproval among partisans of the governing party and opposition parties. --- # Case Study ## Partido Nacional Defection in Uruguay in 2002 - Multi-party presidential system, with 3 main parties. -- - 1999 election Jorge Batlle center-right Partido Colorado is elected president in the runoff, only 32% vote share, forms coalition with center-right Partido Nacional (22%) -- - Minimum Winning Coalitions (Martínez-Gallardo, 2012) -- - Strongly shared cabinet 6/13 ministers to Partido Nacional. -- - Strong economic crisis gradually led to defection -- - Highly debated (no consensus among PN) long-term deterioration in the relationship between PN and Batlle, with defection **after** worst economic moment. --- # Data and Methods -- - Individual-level data from 21 nationally representative face-to-face surveys conducted by survey firm Cifra during Batlle's administration (2000-2003) -- - Part of a larger project which allowed the dissemination of more than 200 Uruguayan public opinion surveys carried out among 1993 and 2020. -- - Surveys pre and post defection (July 2002 and November 2002) -- - Difference-in-differences, events study. --- # Results .center[ <img src="figures/desc_prepost_3_24.png" width="550px" /> ] --- # Results .center[ <img src="figures/desc_prepost_2_24.png" width="550px" /> ] --- # Results .center[ <img src="figures/aprob_serie_24.png" width="700px" /> ] --- # Results .center[ <img src="figures/aprob_serie_dif_24.png" width="700px" /> ] --- # Conclusion - Theoretical argument: coalition arrangements in Latin America impact public opinion. -- - Relatively small but significant effect, in line with hypotheses -- - Weak causal identification assumptions -- - Possible implications for theories of incentives of defection and governing after a coalitional defection. -- - Next step: evidence from another case --- # Coalition Formation and Coalition Leader Popularity .center[ <img src="figures/pou.png" width="700px" /> ] --- class: inverse, center, middle # Thanks! ### martin.opertti@duke.edu --- # Event Study .center[ <img src="figures/event.png" width="700px" /> ] --- # Differences-in-differences .center[ <img src="figures/diff.png" width="300px" /> ] --- # Differences-in-differences .center[ <img src="figures/event_study_2.png" width="700px" /> ]