This is a guest post by Karel Kouba and Tomáš Došek. It is based on their article in Democratization and is available here.
While full reversals of democratic order have been rare in Latin American countries since their transitions to democracy, other, less pernicious, forms of political instability have become common. Challenges to sitting presidents through the threat of impeachment or coups are the primary manifestations of governability crises (Valenzuela 2004), although others consider it as a flexibilization of the presidential regimes and thus a way of ousting unpopular presidents without a democratic regime breakdown (Marsteintredet and Berntzen 2008). We understand governability crises in a broader sense which also includes other forms of conflictive relationships between the president and the congress (Pérez-Liñán 2006).
Existing literature holds that the probability of a governability crisis or an interrupted presidency is higher in more fragmented party systems. In our recent article in Democratization, we depart from this argument in two ways. We argue that we need to focus on the level of fragmentation of presidential elections (and not only the party system itself) and that the relationship between presidential election fragmentation and governability crises is not linear but actually curvilinear (with both the least and the most fragmented elections being most conducive to political crises).
This conclusion permits the reconciliation of two apparently conflicting arguments present in the literature. The academic debate has revolved particularly around the choice of presidential electoral systems (runoff or plurality) and about how these shape the patterns of electoral competition. On the one hand, the use of runoff electoral rules, and especially the fact that the second round had taken place, is associated with higher legislative fragmentation and ideological polarization, which in turn correlates with the occurrence of presidential breakdowns making the absolute majority rule “extremely damaging to democracy” (Chasquetti 2001). On the other hand, however, the opponents of the plurality rule suggest that runoff elections promote democratic consolidation and their introduction in Latin American countries has been a positive institutional innovation (McClintock 2018). Opening up the political competition to political actors that challenge the traditional (and often undemocratic or post-authoritarian) parties as well as greater ideological moderation and wider popular acceptance of the winning candidate are among the principal mechanisms linking runoff rules to better democratic governance.
We tested the implications of our theoretical argument on a sample of 102 Latin American presidencies that have originated in competitive and direct elections between 1978 and 2013. To operationalize governability crises, we used an ordinal index developed by Pérez-Liñán (2006) creating a four-point scale between normal politics on one side and military interventions to oust the president or disband the congress at the other extreme. Running five ordered logistic regression models we show how the curvilinear relationship between presidential election fragmentation and the incidence of governability crises holds under different model specifications. In short, the quadratic term both increases the explanatory power of the model and points in the expected direction as both low and high levels of fragmentation are associated with an increased probability of crisis. The intermediate values of presidential election fragmentation, or around 3 to 4 effective presidential parties contesting the election, are most conducive to political stability. We display this relationship graphically across the range of values of the effective number of presidential candidates. This coding scheme used for the dependent variable indicating the extent of a political crisis assigns a value between 1 (i.e. stable “normal politics”) and 4 (the most extreme instability in the form military intervention).
In the article, we also posit that the causal mechanisms at both extremes are different, as suggested by the notion of equifinality (different causal paths leading to the same result, that is in this case, a governability crisis). In fact, causal mechanisms are context-specific, that is their explanations for how the same phenomenon can vary in time and space. The causal mechanism that translates high levels of party/presidential fragmentation to governability crises has been thoroughly studied and demonstrated in various cases of interrupted presidencies. Extreme fragmentation prevented presidents from having a sufficient “legislative shield” and functional government coalitions (Pérez-Liñán 2007). In combination with social mobilization (Hochstetler 2006), this weakened presidents’ positions and eventually contributed to presidential instability. This was, for example, the case of interrupted presidencies in Ecuador and Bolivia (Mejía Acosta and Polga-Hecimovich 2010; Buitrago 2010, among others).
However, the overconcentration of the presidential contest is almost as likely to destabilize politics. We identify three analytically different mechanisms that describe such processes and use short case studies to illustrate them. First, we focus on refoundationalist politics as a consequence of previous crises of representation that could trigger a governability crisis. Second, we argue that overinstitutionalized parties and party systems often maintained by plurality electoral rules prevent alternative leaders from entering the competition, and that this petrification of politics is unhealthy for democratic stability. Third, we focus on the internal conflicts within the traditional parties whose leaders are encouraged to abandon their party and form a personalist vehicle of their own to contest elections. We illustrate these scenarios with the cases of Venezuela (which combines to a certain degree the first two paths) and Honduras (which exemplifies the last two paths).
We conclude in line with McClintock’s recent work that there are risks associated with an extreme overconcentration of the party system. Thus, to the extent that concentrating the presidential contest has been advocated to avoid further legislative fragmentation and governability crises, this advice cannot be generalized across the board without caveats. Both runoff and plurality rule have their advantages supported by some formidable theoretical arguments. Consequently, the institutional advice that is consistent with our theoretical argument is the preference for a runoff rule with a reduced threshold in the first round. This middle-of-the-road rule might avoid the overconcentration of the contest between two competing blocs by facilitating access of challenger parties to the presidency, while at the same time safeguarding against the proliferation of weak candidates.
Karel Kouba is an assistant professor at the Department of Political Science, Philosophical Faculty, University of Hradec Králové, Czech Republic. He specializes in voting behaviour and electoral institutions in Latin American and post-communist countries. He can be reached at email@example.com. Website: https://www.researchgate.net/profile/Karel_Kouba
Tomáš Došek is a Ph.D. candidate in Political Science at the Instituto de Ciencia Política, Pontificia Universidad Católica de Chile, Chile. His research focuses on political parties, electoral reforms and subnational politics in Latin America. He can be reached at firstname.lastname@example.org. Website: https://sites.google.com/site/tomasdoseklatam/