Tag Archives: Latin America

Aníbal Pérez-Liñán and John Polga-Hecimovich – Getting Rid of the President

This is a guest post by Aníbal Pérez-Liñán of the Department of Political Science at the University of Pittsburgh and John Polga-Hecimovich of the Political Science Department at the US Naval Academy at Annapolis. It is based on their paper in Democratization.

Are presidential impeachments modern functional equivalents of old-fashioned military coups? The impeachment of Brazilian President Dilma Rousseff in August 2016 led to an acrimonious debate on whether her removal from office constituted a “soft coup” against an elected leader. Similar concerns were voiced after the impeachment of Paraguayan President Fernando Lugo in 2012. As calls to impeach President Donald Trump multiply, this question appears to gain increasing relevance for US politics as well.For students of presidentialism, the idea of “functional equivalence” between military coups and legal ousters (impeachments, legislative declarations of presidential incapacity, or anticipated resignations of the executive) translates into very specific questions: Are there any historical factors able to explain military coups as well as impeachments? If so, why are some presidents removed following legal procedures while others are removed by force?

In a forthcoming paper in Democratization we develop a unified theory of presidential instability to explain why presidents are removed from office through military coups or through legal procedures.

We identify two sets of historical causes. First, some factors create conditions for presidential instability, irrespective of the mode of premature exit from office. Because they motivate a political opposition to conspire against the government, those factors explain why presidents are likely to fail, but not how they fail. Second, an alternative set of causes accounts for the specific institutional manifestations of presidential instability. Those factors map onto the relative capabilities of groups inclined to pursue a military coup or the legal removal of the president.

The distinction between general motivations to remove the president and the capabilities of specific opposition groups helps us identify the role of different causal explanations in the literature.

Among the common causes of legal removals and coups, we find:

  • Poor economic conditions. Recessions undermine the president and facilitate conspiracies. Studies on military coups argue that negative economic shocks increase the risk of military rebellions, while the literature on impeachments shows that weak economies undermined Latin American presidents in the 1990s.
  • Popular protests. Mass mobilization against the government signals that the president is weak and destabilizes any elected administration. Students of military intervention find that mass protests help elites coordinate in a coup. Students of impeachment emphasize that protests encourage reluctant legislators to act against the president.
  • Radicalization. Radical actors have intense and extreme preferences; they are reluctant to bargain and remain intransigent in defense of their policy goals. Radicalism is therefore a potential cause of military coups, but also an explanation for the role of social movements forcing the resignation of presidents in places like Bolivia and Ecuador.

Given the prior conditions for instability, several factors separate legal removals from coups:

  • The regional context. A long line of research has invoked international diffusion as an explanation for democratic instability – though not necessarily government instability. The regional context may strengthen the position of coup perpetrators or otherwise direct elites towards legal strategies against the president.
  • Legislative support for the president. Two causal mechanisms are discussed in the literature: Linz’s argument that presidentialism itself is a source of instability and the argument that a legislative majority “shields” the executive against impeachment.
  • Elite support for democracy. A strong normative preference for democracy among elites forecloses the possibility of a military coup and leaves legal removal as the only acceptable strategy for the opposition. The government’s normative preferences also matter: a president dismissive of democratic rules may be unwilling to recognize the legitimacy of an impeachment procedure, driving opponents to consider the option of a coup.

To test those expectations, we use discrete-time event history models with selection.  Our sample covers all democratic regimes in nineteen Latin American countries between 1945 and 2010 (N = 729). The dependent variable measures yearly outcomes for each president:  survival, exit via military coup, or exit via legal removal. Our sample includes 21 coups and 15 legal removals. The selection model estimates the risk of president being removed from office (in any way) in the selection stage, and the risk of being removed via coup (as opposed to a legal procedure) in the outcome stage.

The statistical models allow us to estimate the risk of coups and impeachments, plotted in Figures 1 and 2.

Figure 1 underscores the role of common motivations behind coups (in the bottom row) and impeachments (in the top row), as economic recession, demonstrations, and radicalization consistently expand the risk of both outcomes.

Figure 1: Common Causes of Legal Removals and Coups (Predicted Risk)

Figure 2, on the other hand, illustrates the differential impact of variables. The first column illustrates how a large number of coups in neighboring countries expands the risk of military intervention but reduces the probability of legal removal in the observed country.  The second column shows that the risk of military overthrow remains independent from the composition of congress, but impeachment is less likely when the executive controls the legislature.  The third column shows that a military coup is unlikely when political actors are more committed to democracy. By contrast, the risk of legal removal expands as groups operating within the constitution become empowered by the opposition’s reluctance to engage in military conspiracies.

Figure 2: Causes Separating Legal Removals and Coups (Predicted Risk)

Our findings underscore that common causes of presidential instability are not necessarily causes of democratic breakdown, yet crises of government may easily escalate into crises of the democratic regime when legal venues for the removal of the president are blocked.

These findings are increasingly relevant today.

In a global context in which presidents and their adversaries – in Latin America, Eastern Europe, Turkey, and even the U.S. – have displayed growing levels of radicalism, our findings raise concerns. Radical leaders engender polarization, encouraging their opponents to overthrow the government by any means possible. Combined with economic stagnation or social protest, radicalization is likely to trigger presidential instability.

Yet other factors ultimately tip a crisis towards a non-democratic resolution. A regional environment hostile to democracy and a lack of democratic commitment from domestic elites decrease the probability of a legal impeachment and increase the likelihood of a coup.

International policymakers would be wise to consider these findings: long-term efforts to build regional organizations that discourage military intervention and steady support for democratic leaders will prevent future presidential crises from escalating into full crises of democracy.

André Borges and Mathieu Turgeon – Presidential coattails in coalitional presidentialism

This is a guest post by André Borges and Mathieu Turgeon, both of whom are assistant professors of political science at the University of Brasília. It is based on a recent article in Party Politics.

Research on coalitional presidentialism has focused mostly on post-electoral coalition formation, neglecting the  pre-electoral origins of cabinets  in many – if not most – presidential countries with multiparty systems (Albala 2014; Chasquetti 2008; Freudenreich 2016). Kellam (2015) analyzed pre-electoral coalition formation in presidential elections in eleven Latin American countries from the 1980s to the late 2000s, and found that 35% of all presidential candidates that obtained at least 10% of the national vote formed a coalition with one or more parties. Although pre-electoral coalitions in presidential elections are a rather frequent phenomenon, there is a paucity of research on the causes and consequences of these pre-electoral alliances. In particular, the literature on presidential coattails has failed to consider the potential impacts of multiparty alliances on party system formation, assuming that parties entering the presidential race as members of an alliance do not obtain electoral gains (Mainwaring and Shugart 1997; Shugart and Carey 1992; West and Spoon 2015). That is, the coattail effect benefits only parties that enter the race with a candidate of their own, as voters rely on the party of their preferred presidential candidate as an information shortcut to help them decide how to vote in legislative election (Golder 2006). But, if allied parties do not benefit from presidential coattails and they actually risk losing credibility and weakening their party base if the coalition is not perceived as adequate , why would they support a presidential candidate from another party in the first place? Even if parties believe that entering a pre-electoral coalition will increase their chances of entering the presidential cabinet, they cannot be sure of the supported candidate’s victory in the presidential contest (Freudenreich 2016).

In a recent article (Borges and Turgeon 2017), we challenge the conventional wisdom on presidential coattails and pre-electoral coalitions.  By focusing on coattails from the president-elect party—the coalition formateur—we argue that presidential coattails in coalitional presidentialism benefit not only the party of the president-elect but also the coalition party members, which has  important implications for coalition formation in presidential systems. This is what we label a diffused coattail effect.

In multiparty presidential systems, parties that are viable contenders in the presidential election are likely to “presidentialize”, shifting resources away from their legislative campaigns and focusing on the presidential race (Samuels 2002). To secure the necessary votes to win the presidency, large parties form electoral coalitions with smaller parties and adopt broad campaign strategies. Specifically, they avoid pure partisan campaign strategies and campaign, instead, on behalf of the coalition to mobilize as many voters as possible for the presidential election.

Coalition fomateurs understand that there are costs for parties to join their coalition and are disposed to make important concessions to convince them to join forces. These concessions include, in part, supporting coalition party members in simultaneous, lower-level elections and by making sure that candidates from the coalition formateur party do not “invade” the electoral strongholds of the other coalition party members. Moreover, presidential candidates campaign on behalf of the whole coalition and not only for their own party, especially in other simultaneous, lower-level electoral contests like legislative elections. In exchange, coalition party members aggregate valuable organizational and financial resources to help the formateur party reach segments of the electorate otherwise less accessible but necessary to win the presidential election.

We believe coalition party members benefit from presidential coattails because the parties involved in the coalition work together to coordinate their campaign strategies at all levels (presidential, gubernatorial, senatorial and lower chamber races). But coalitions are not all created equal and the effects they carry over election results depend, in part, on the ability of coalition party members to coordinate effectively with the formateur party. Specifically, we believe that coalition party members that coordinate more effectively with the formateur party should benefit more from presidential coattails than those who don’t. We classify coalition party members into core and peripheral coalition party members. Core coalition party members are defined as those that are close ideologically to the formateur party and that have adopted consistent strategies in the governing and electoral arenas in the past.

Coalition party members that have participated in the past governing coalition can benefit from the president’s popularity during the election by claiming credit for key government programs, tying their fortunes with that of the incumbent president. Moreover, coalition party members that have participated in previous electoral coalitions with the same formateur party should be associated more strongly to the said coalition by voters than those coalition party members that have not. Finally, we believe that coalition party members will coordinate more forcefully the closer they are ideologically to the coalition formateur because, in that scenario, both can tailor campaign messages courting ideologically similar voters.

We test two hypotheses. First, we argue that presidential coattails are diffused, benefiting the president’s party but also her coalition party members. Second, we claim that The diffused coattails effect in coalitional presidentialism should benefit more strongly core coalition party members, as compared to peripheral coalition party members.

To evaluate the two hypotheses we analyze data from Brazil and Chile. These two countries are widely studied cases of coalitional presidentialism where multiparty coalitions play a fundamental role in the governing and electoral arenas. Overall, Chile represents a most-likely case for diffused presidential coattails because its governing and electoral coalitions are stable and ideologically coherent. Brazil, on the other hand, represents a least-likely case for diffused presidential coattails because it shows much less congruence between its governing and electoral coalitions and its electoral coalitions are unstable and generally not ideologically coherent. We believe that such design allows for robust testing of our hypotheses of presidential coattails in coalitional presidentialism. Finding only weakly supportive evidence (or no evidence at all) of diffused coattails in Chile would seriously undermine or lead to outright rejection of our theoretical claims.  On the other hand, if we succeeded in finding evidence of diffused coattails in Brazil, this should strongly support the view that presidential coattails exhibit dynamics of their own in coalitional presidentialism.

Our statistical analysis of coattail effects using data on district-level electoral returns in Brazil and Chile shows that presidential coattails in coalitional presidentialism are diffused, benefiting the president’s party and her core coalition party members. Presidential coattails, however, do not affect coalition party members equally. Core coalition party members, that is, those that are more strongly associated with the coalition formateur, are the sole beneficiaries of presidential coattails. No presidential coattail effect is discernible for peripheral coalition parties.

Admittedly, we cannot make claims about the presence or not of similar diffused presidential coattails in other cases of coalitional presidentialism. We have very good reasons to believe, however, that this phenomenon extends beyond the Chilean and Brazilian cases. In particular, both Chile and Brazil are open-list PR systems. In closed-list PR systems, which are most commonly found in other cases of coalitional presidentialism, intra-coalition coordination is profoundly facilitated. Under such electoral rules, parties can more easily divide the expected seats among coalition partners by ordering the candidates’ names on party lists in each district in a way that benefits more fairly coalition party members (Cruz 2010; Leiras 2007).

Future research should explore further the broader implications of the diffused coattail effect for coalitional presidential systems and party systems, more generally. One such possibility deals with the relationship between electoral and governing coalitions. Our results, for example, suggest that the electoral success of peripheral coalition party members is not tied to that of the coalition formateur party. Consequently, their behavior within the governing coalition could be distinct than that of core coalition party members and could potentially affect the stability of governing coalitions. Thus we may ask: are peripheral coalition party members less loyal and possibly more demanding than core coalition party members? Similarly, are threats to leave the governing coalition more credible than those made by core coalition party members? These are other interesting questions to be explored.

Finally, diffused presidential coattails may also contribute to maintain or even increase party fragmentation in the lower chamber. That is, different from traditional arguments on presidential coattails and party systems, the theoretical argument and empirical evidence presented in this paper indicate that presidential coattails, when diffused, foster instead the survival and growth of small parties. Contrary to West and Spoon’s (2015) findings about electoral coalitions, it is not clear whether this will always and necessarily lead to lower fragmentation in legislative elections. These questions should be of great interest to comparativists given the spread of coalitional presidentialism in Latin America, Africa and the former Soviet Union.


Albala, Adrian. 2014. “The Timing Effect of Presidentialism on Coalition Governments: evidence from Latin America.” In 23rd IPSA World Congress, Montreal, CA.

Borges, André, and Mathieu Turgeon. 2017. Presidential coattails in coalitional presidentialism. Party Politics: 1-11.

Chasquetti, Daniel. 2008. Democracia, presidencialismo y partidos políticos en América Latina: evaluando la” difícil combinación”. Ediciones Cauce-CSIC.

Cruz, Facundo. 2010. Relaciones e interacciones partidarias en coaliciones de gobierno. Los casos de la Alianza, la Concertación y el Frente Amplio. Revista Debates Latinoamericanos 8: 15.

Freudenreich, Johannes. 2016. The Formation of Cabinet Coalitions in Presidential Systems. Latin American Politics and Society 58 (4): 80-102.

Golder, Matt. 2006. Presidential Coattails and Legislative Fragmentation. American Journal of Political Science 50 (1): 34-48.

Kellam, Marisa. 2015. Why Pre-Electoral Coalitions in Presidential Systems? British Journal of Political Science 47: 391-411.

Leiras, Marcelo. 2007. Todos los caballos del rey: la integración de los partidos políticos y el gobierno democrático de la Argentina, 1995-2003. Prometeo libros.

Mainwaring, Scott, and Matthew Soberg Shugart. 1997. Presidentialism and democracy in Latin America. . Cambridge University Press.

Samuels, David. 2002. Presidentialized Parties: The separation of powers and party organization and behavior. Comparative Political Studies 35 (4): 461-83.

Shugart, Matthew, and John M. Carey. 1992. Presidents and Assemblies: Constitutional design and electoral dynamics. Cambridge, UK: Cambridge University Press.

West, Karleen Jones, and Jae-Jae Spoon. 2015. Coordination and presidential coattails Do parties’ presidential entry strategies affect legislative vote share? Party Politics: 1-11.

Christopher A. Martínez – Why political institutions matter for presidential survival

This is a guest post by Christopher A. Martínez, Assistant Professor in the Department of Sociology and Political Science, Temuco Catholic University, Chile

There is no crisis here, nor problems” Fernando de la Rúa (resigned, December 2001)
I’ll continue to govern” Lucio Gutiérrez (dismissed by Congress, April 2005 )

Since 1979, thirteen South American chief executives have been unable to complete their constitutional terms. These failed presidencies occur when a popularly elected president is forced to leave office early, though the ouster is not followed by a democratic breakdown. Some presidents have been impeached (e.g., Collor and Rousseff in Brazil, Pérez in Venezuela, Cubas and Lugo in Paraguay); others could not withstand massive and widespread street protests (e.g., Alfonsín and De la Rúa in Argentina, Siles Zuazo and Sánchez in Bolivia, and Fujimori in Peru); while other leaders were unseated via unorthodox mechanisms (e.g., Bucaram, Mahuad, and Gutiérrez in Ecuador). Being forced to leave office early represents a dramatic deviation from a central goal of all political leaders, which is to maintain power. Thus, failing to fulfil a presidential term should be an exceptional political event in a presidential democracy.

I used survival analysis to quantitatively study 65 South American presidencies between 1979 and 2012. My results show that the most important forces driving presidential survival are institutional ones: legislative support for the president, and a country’s democratic tradition. Interestingly, inflation, economic recessions, and scandals have no significant impact on presidential survival, whereas violent social mobilisations exhibit a rather weak effect.

Some presidents are “safer” than others: Why a country’s democratic tradition matters

Previous studies have not established whether democracy had any impact, be it positive or negative, on the occurrence of presidential failures. Rather than focusing on current levels of democracy, in my research I used a new measure of democracy which represents a country’s past records with democratic and authoritarian politics: democratic tradition. Figure 1 illustrates how different a country’s current level of democracy (Polity2) and democratic tradition truly are. For instance, Argentina, Chile, Paraguay and Uruguay all had low levels of democracy in 1980. Nevertheless, only Chile and Uruguay stand out for their accumulated experience with democracy.

When considering democracy’s effects over extended periods of time, my findings show that the larger the democratic tradition of a country, the more likely the president will “survive.” That is to say, countries with a well-established democratic track provide a “safer,” less threatening environment for presidential survival. Unlike countries with poor democratic experiences, in these type of polities, political actors—presidents, legislators, parties, among others—are discouraged from pursuing questionable courses of action and are more likely to settle political disagreements through negotiation and accommodation, all of which reduces the risk of early government termination.

Figure 1: Democratic Stock and Polity2 Score of Democracy, 1900-2012

Legislative support is central to “survive” in office

In theory, chief executives in presidential systems do not require a legislative majority to stay in office; however, presidents need legislators’ support more than they may think. Passing relevant legislation is a central task for most executives, but hanging on to power is unquestionably a far more important goal for any president. If presidents are to complete their terms in office, they must ensure the backing of a disciplined contingent of members of congress. This “legislative shield” (Pérez-Liñán 2007) would especially come in handy during dire economic circumstances and intense social mobilisations, as loyal legislators may undermine the opposition’s attacks and criticism against the executive. As in previous studies, my research maintains legislative support for the president as the most consistent and strongest predictor of presidential survival in South America.

What do political scandals do?

Results have been mixed about the relationship between political scandals and failed presidencies. Unlike Hochstetler (2006) and Pérez-Liñán (2007), my findings show that corruption scandals do not reduce presidential survival in South America. Presidents’ involvement in scandals may be frowned upon and weaken their approval ratings, yet they do not directly or necessarily force them to step down. True, some presidents have been deposed because of corruption accusations (for example, Collor in Brazil and Pérez in Venezuela). Nevertheless, political scandals are not exceptional in the region, and many of them have not triggered presidential interruptions (e.g., Menem in Argentina, Samper in Colombia, among others). What scandals can do, especially in cases of fragile president-party relations, is to undermine the ruling coalition and/or reduce the president’s chances to form a new one. Such an instance is what I argue occurred with Fernando De la Rúa’s bribery scandal in Argentina (Martínez 2017) and Lucio Gutiérrez’s alleged links with a drug trafficker in Ecuador (Martínez forthcoming).

Social mobilisations

Though it may come as a surprise, my results show that street protests have only a weak—if any—effect on presidential survival. This is true of both general strikes and social mobilisations aimed at the executive. On the other hand, violent demonstrations such as riots do increase the risk of early presidential removals; nonetheless, their impact is significantly weakened when one analyses a president’s legislative support. That is to say, when it comes to “surviving” in office, the role of congress outweighs any type of social mobilisation, even the bloody ones. An alternative explanation for the weaker-than-expected effects of public demonstrations is that it is their intensity, rather than their simple occurrence, that matters.

Final remarks

Even though a president’s popularity may be negatively affected by economic recessions, street protests, and political scandals, their “survival” in office largely hinges upon legislative support and democratic tradition. The role of congress is likened to the proverbial two-edged sword: it may either shield the president or turn against him/her. Presidents, thus, ought to cultivate smooth relations with their ruling partners should they indeed want to hold onto office. Moreover, chief executives ruling over countries with a weak democratic tradition may have fewer chances to “survive” to begin with, as most political actors in those countries may be more accustomed to bend the rules of the game, which would heighten the risk of presidential failures.

Christopher A. Martínez holds a PhD in Political Science from Loyola University Chicago. He is Assistant Professor in the Department of Sociology and Political Science, Temuco Catholic University, Chile. His current research interests include the executive branch, government survival, institutional performance and democratic consolidation in Latin America. He can be reached at christopher.martinez@fulbrightmail.org and @martineznourdin.

Marisa Kellam – Why Pre-Electoral Coalitions in Presidential Systems?

This is a guest post by Marisa Kellam, Associate Professor, Waseda University. It is based on her recent article in the British Journal of Political Science.

Presidential politics goes hand in hand with coalitional politics in Latin America, especially in South America. As recently reported in this blog, presidents in the region often depend on the support of other parties to win election and to govern.

In this post, I will focus on pre-electoral coalitions. [1] To give some recent examples: President Bachelet in Chile, President Santos in Colombia, and former Brazilian President Dilma Rousseff all ran for re-election with a multiparty electoral alliance.  Multiple parties also supported Argentine President Macri’s candidacy. In recent elections in other countries, the incumbent party candidate defeated an opposition pre-electoral coalition, such as in Ecuador’s recent election, Bolivia’s 2014 presidential contest, or the 2013 presidential election in Venezuela.

In fact, pre-electoral coalitions in presidential elections have been a feature of Latin American democracy since the third-wave, and even before. Yet the conventional wisdom has been that these coalitions were “not binding past election day.” [2] However, increasing attention to post-electoral coalition formation in comparative presidentialism research has led to new findings that winning pre-electoral coalitions usually go on to form post-electoral governing coalitions. [3] Does the strong empirical correspondence between pre-electoral coalitions and post-electoral governments call for a revision of the conventional wisdom? My recently published article in the British Journal of Political Science speaks to this puzzle.

Why do parties form pre-electoral coalitions in presidential systems? From the perspective of a presidential candidate, it would seem to be an easy answer—the more in my camp, the merrier—that is, unless she must give something in return. When considering potential partner parties, we might assume that the presidential candidate offers them government positions—just as presidents offer coalition partners in government negotiations—except that pre-electoral agreements involve only promises not actual offers.

Although I set out to overturn the conventional wisdom on pre-electoral coalitions, I found no convincing argument to support a contrary claim that presidential candidates’ promises to distribute government positions and resources to other parties are credible commitments in presidential systems. Presidents alone control cabinet appointments—even their own parties cannot hold presidents immediately accountable. Moreover, presidents’ partners will not necessarily punish them for breaking their pre-electoral commitments. A party that wants access to resources under the president’s control is unlikely to make a loud complaint, much less to pull out of the government completely.  And if parties do not reveal the extent to which presidents fail to honor agreements to share spoils, then neither presidents nor their parties will pay a reputational cost. This isn’t to say that presidents will break their promises; it is only to make the point that candidates’ pre-electoral promises to share spoils are “cheap talk” and party leaders know this.

I find it useful to contrast these behind the door negotiations with presidential candidates’ public campaigns.  A presidential candidate and her political party pays an immediate reputational cost if she publically campaigns on a policy compromise made with another party in order to gain its support. True, a president is not bound to her campaign platforms. But even so, the pre-electoral policy agreement reveals information about her policy positions.  A pre-electoral policy agreement also gives the president and her partners a shared mandate, or common purpose, after the election. And if the president ends up reneging on that policy agreement later, the coalition partner would likely refuse to go along, consistent with their own electoral incentives and policy motivations.

These differences between patronage promises and campaign platforms provide some insight as to why parties join pre-electoral coalitions to support other parties’ candidates in presidential systems. Political parties join pre-electoral coalitions in pursuit of policy goals, but not as part of an office-seeking strategy.  To provide empirical evidence to support this argument, I compare characteristics of the parties that joined pre-electoral coalitions with those that did not.

More specifically, I compare the probability of participation in pre-electoral coalitions of programmatic parties to that of particularistic parties.  Particularistic parties are those that experts classify as having no discernible policy position on the standard, left-right macroeconomic dimension of politics; instead, particularistic parties focus on the distribution of “pork” and patronage or serve single-interests.  According to my reasoning, if these parties do not have policy goals then they should be less likely to join pre-electoral coalitions (unless one of their own members is on the president-VP ticket). Programmatic parties, in contrast, may use pre-electoral coalitions to identify and help elect presidential candidates who are closest to them in terms of policy.

I analyzed coalitions formed, and not formed, in 77 elections held in 11 Latin American countries.  I found that programmatic political parties (i.e. policy-seeking parties) were more likely than particularistic political parties (i.e. office-seeking parties) to join pre-electoral coalitions in support of another party’s presidential candidate.  As expected, I also found that the greater the ideological distance between a programmatic party and the party of a presidential candidate, the less likely they are to join that candidate’s coalition.

While the once conventional thinking that presidents have little incentive to form governing coalitions has been overturned, this does not imply that the conventional wisdom regarding electoral coalitions should also be cast aside. As I have discussed, pre-electoral coalition bargaining differs from post-electoral government negotiations, with important implications for presidential politics in multiparty systems.

In conducting this research, I realized that the reason why parties join pre-electoral coalitions in presidential systems is less obvious than it appeared at first glance. Even if pre-electoral coalitions are not binding commitments to govern together after the election, the coalition formation process itself informs political parties about their respective policy positions and creates a shared mandate.



[[1]] On a side note, the prefix “pre” seems unnecessary to me, but I use it nonetheless because the term pre-electoral coalitions is widely used in the literature.

[2] Mainwaring, Scott, and Mathew Shugart. 1997. Presidentialism and Democracy in Latin America. Cambridge UP, p. 397.

[3] As discussed previously in this blog, see Freudenreich, Johannes. 2016. “The Formation of Cabinet Coalitions in Presidential Systems.” Latin American Politics and Society 58(4): 80-102.  In my own work-in-progress with Cecilia Martinez-Gallardo, we also find a strong empirical relationship between pre- and post-electoral coalitions in Latin America.

Fernando Meireles – Latin American presidents and their oversized government coalitions

This is a guest post by Fernando Meireles, Ph.D candidate in Political Science at Federal University of Minas Gerais (Brazil). E-mail: fmeireles@ufmg.br

In many countries, presidents have a difficult time governing because their parties lack a legislative majority. In fact, because of the combination of separate elections for executive and legislative branches with multiparty systems, this situation is far from uncommon: during the last two decades in all 18 Latin American countries with presidential systems, only 26% of the time has the president’s party had a majority in the lower house. Due to this constraint, as a vast amount of research now highlights, minority presidents usually form multiparty government coalitions by including other parties in their cabinets. Again, only four Latin American presidential countries in the last twenty years were not governed by a multiparty coalition at some point since the 1980s.

However, the need to craft a legislative majority alone does not explain why presidents frequently include more parties in their governments than necessary to obtain a minimum winning coalition – forming what I call an oversized government coalition. The distribution of this type of coalition in Latin America is shown in the graph below. As can be seen, it is not a rare phenomenon.

If government coalitions are costly to maintain, as presidents have to keep tabs on their coalition partners to ensure they are not exploiting their portfolios to their own advantage – not to mention the fact that by splitting spoils and resources between coalition partners, the president’s own party is worse off – then why are these oversized coalitions prevalent in some Latin American countries?

In a recent article in Brazilian Political Science Review, I tackled this puzzle by analyzing the emergence of oversized government coalitions in all 18 presidential countries in Latin America[1], followed by a case study focusing on Brazil, spanning from 1979 to 2012. To this end, I gathered data on cabinet composition[2] from several sources to calculate the size of each government coalition in the sample: if a coalition had at least one party that could be removed without hampering the majority status of the government in the lower house in a given year, I classified it as an oversized coalition.

Specifically, I examined three main factors that, according to previous research, should incentivize presidents to include more parties in their coalitions than necessary to ensure majority support: 1) the motivation party leaders have to maximize votes, which would make joining the government attractive to opposition parties (vote-seeking); 2) the motivation presidents have to avoid coalition defections to implement their policy agendas (policy-seeking); and 3) the institutional context, considering the effects of bicameralism, qualified majority rules, and party system format on government coalition size.

The results support some of the hypothesis suggested by the literature. First, presidents are more prone to form oversized coalitions at the beginning of their terms, which shows that the proximity to the election affects Latin American presidents’ decision to form, and opposition parties to accept being part of, large coalitions – as others studies argue, this is mainly due to parties defecting from a coalition to present themselves as opposition when elections are approaching. Second, party fragmentation also has a positive effect on the emergence of oversized coalitions, consistent with the hypothesis that presidents might include additional parties in their coalitions anticipating legislative defections. Yet on the other hand, presidential approval, party discipline, and ideological polarization do not have the same positive effects on the probability of an oversized coalition being formed.

The factor that has the most impact on the occurrence of oversized coalitions, however, is the legislative powers of the president. As the literature points out, legislative decrees and urgency bills could be used by skilled presidents to coordinate their coalitions, facilitating horizontal bargaining between coalition partners. The comparative results show that this is the case in Latin America: the difference in the predicted probability of a president with maximum legislative powers in the sample forming an oversized coalition and another with minimum powers is about 32 percent points.

By exploring the Brazilian case in more depth, I also found that bicameralism dynamics and qualified majority rules impact the emergence of oversized coalitions. With two chambers elected through different electoral rules, parties in Brazil are often unable to secure the same seat share in both houses; to make things worse for presidents, party switching is still widespread in the country. In this context, as my results uncovered, differences in the number of seats controlled by the government in the Chamber of Deputies and the Senate positively affect the emergence of oversized coalitions. Finally, as some bills require supermajorities to be approved, such as constitutional amendments, reformist presidents also tend to form and maintain larger coalitions: the maximum value in this variable predicts increases by up to 10 percentage points on the probability of an oversized coalition being formed.

Taken together, these results show a more nuanced picture of why and how presidents form multiparty government coalitions in Latin America: often, obtaining a legislative majority is not enough to implement their legislative agendas, and so they might resort to a complementary strategy: to form larger coalitions. And presidents with greater legislative power, at the beginning of their terms or facing fragmented party systems, are in the best position to pursue such a strategy. In this way, both electoral and programmatic factors, as well as the institutional context, become key to understand variations in the size and the composition of government coalitions in presidential countries.


[1] These countries are Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, El Salvador, Ecuador, Guatemala, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru, Dominican Republic, Uruguay, and Venezuela.

[2] The criteria employed to identify a government coalition is the party affiliation of the ministers of the principal ministerial portfolios in each country – taking into account that ministers are not always recruited due to their connections or their congressional influence, and that in some cases they are not recognized by their parties as legitimate representatives of the same.

Johannes Freudenreich – The Formation of Cabinet Coalitions in Presidential Systems

This is a guest post by Johannes Freudenreich, Postdoctoral research fellow at the Geschwister-Scholl-Institut für Politikwissenschaft at the University of Munich. It is based on an recent article in Latin American Politics and Society

In the beginning of the 21st century, prospects of Latin American presidential democracies were good. The dictatorships of the 1970s and 1980s had vanished, economies were constantly growing, and comprehensive social welfare programs were implemented. Many political scientists link these successes to the ability of Latin American presidents to form, maintain and manage cabinet coalitions (Cheibub 2007). The differences between presidential and parliamentary systems of government seemed to have become rather marginal. Both presidents and prime ministers achieved legislative majorities by forming broad cabinet coalitions and critics of the presidential form of democracy, such as Juan Linz (1994), seemed to be proven wrong. However, soon presidential impeachments became the new pattern of political instability in the region (Pérez Liñan 2007). Cabinet reshuffling remains constantly high and broad corruption schemes, directly linked to coalition politics, have been disclosed, such as the Mensalão Scandal in Brazil, where the ruling party of President Lula da Silva used illegal side payments to secure the legislative support of members of the ruling coalition.

My recent article in Latin American Politics and Society takes a systematic look at the formation of cabinet coalitions in presidential systems over the past 25 years. It analyzes the extent to which presidents in 13 Latin American countries have formed coalitions that increase their law-making capabilities, and whether presidents form coalitions tailored to find majorities in Congress especially when presidents have low independent influence over policy based on their institutional law-making powers.

The study complements the perspective that cabinet coalitions are largely an instrument for finding legislative majorities with the idea that presidents use cabinet posts to honor pre-electoral support. The reason is the following: presidential elections provide strong incentives for electoral coordination because they tend to favor two-candidate competition. In a multi-party setting, this means that parties have incentives to form pre-electoral coalitions to present joint presidential candidates. When negotiating pre-electoral pacts, parties are likely to agree on how to share the benefits of winning including cabinet posts. After the election, presidents find it difficult to abandon these agreements as they need the trust and support of other parties within and outside of their coalition during their presidential term. Thus, it is expected that cabinet coalitions are likely to be based on the electoral team of presidents and that other legislative parties are invited to join the cabinet only additionally to parties of the existing pre-electoral coalition.

The study further argues that parties attractive as pre-electoral coalition partners are not necessarily the ones that would achieve cabinet participation if the negotiations of cabinet posts were an unconstrained post-electoral process. For example, in a one-dimensional policy space, extreme parties, parties more extreme than the president to the median legislator, are relatively unimportant for legislative decisions and thus unlikely to be included in the cabinet for legislative reasons. In a presidential race, however, extreme parties can provide valuable votes and campaign resources and therefore have far stronger blackmailing power. Furthermore, presidential contests produce a strong antagonism between the president and the parties of the president’s electoral rivals. Since the president’s survival in office is not contingent on the support of other parties in parliament, parties that present a strong presidential candidate are likely to be excluded from the cabinet, even if their inclusion is rational from a lawmaking perspective. It is therefore expected that the party of the runner-up is generally excluded from the presidential cabinet and that the overall explanatory power of variables of legislative bargaining increases once one controls for the effects of pre-electoral coalition formation and competition.

The study empirically evaluates this argumentation on the basis of so-called conditional logit models, presenting a new empirical strategy to analyze cabinet formation under this type of regime. The tests are conducted on a new dataset of 107 democratic cabinets in Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, the Dominican Republic, El Salvador, Honduras, Nicaragua, Panama, Uruguay, and Venezuela. Based on the new method and data, this study presents the most comprehensive test yet of the determinants of the partisan composition of presidential cabinets.

The most note-worthy empirical results are:

First, presidents try to form majority coalitions, but it is the upper house majority not the lower house majority which makes cabinet coalitions significantly likely to from. One potential explanation for this phenomenon is that there are generally fewer parties in the upper than in the lower chamber, due to the disproportionality of electoral systems used to elect upper chambers in Latin America. Thus, the president’s party is often overrepresented in the upper house, which makes it easier for presidents to find majorities. Furthermore, upper chambers are generally strong in Latin America (Nolte and Llanos 2004), and controlling an upper chamber is often sufficient for the president to prevent a veto override.

Second, contrary to expectations in the literature, extensive presidential decree powers decrease the probability of the occurrence of cabinets which control only a minority of seats in the lower house of congress. A potential explanation for this phenomenon is similar to the argument developed by Strøm (1990) for minority governments in parliamentary systems. Parties prefer to stay in opposition when the government has a weak independent influence on policy. The other explanation is that pre-electoral coalition formation is more prevalent when presidents’ institutional authority is high, as political actors make a relatively simple calculation about the benefits and the costs of coordination in presidential elections. The more powerful the president, the higher the incentives for pre-electoral coalition formation (Hicken and Stoll 2008; Freudenreich 2013). And if the a coalition is in power anyway, it is easier to extend this coalition to secure a majority in the lower house of congress.

Third, considerations of governability and pre-electoral bargaining describe two distinct yet compatible sets of factors that influence cabinet formation in presidential systems. Many cabinet coalitions in Latin America are congruent or extended versions of the pre-electoral coalition of the president and parties of the main presidential competitor are generally excluded from the cabinet, but these factors are distinct to the incentives of legislative bargaining. The explanatory power of variables associated with governability increases once variables of pre-electoral bargaining are included in the statistical model. For example, cabinet coalitions are more likely to form when they include the median party in the lower chamber of congress, but this effect is only statistically significant when one controls for the effects of pre-electoral bargaining.

Overall, the paper tries to show that an inclusive approach is necessary to study coalition dynamics in presidential systems. Pre-electoral commitments strongly affect cabinet formation and thereby also confound the relationship between cabinet formation, legislative bargaining and governability.


Cheibub, José A. 2007. Presidentialism, Parliamentarism, and Democracy. New York: Cambridge University Press.

Freudenreich, Johannes. 2013. Coalition Formation in Presidential Systems. Ph.D. diss., University of Potsdam.

Hicken, Allen, and Heather Stoll. 2008. Electoral Rules and the Size of the Prize: How Political Institutions Shape Presidential Party Systems. Journal of Politics 70, 4: 1109–27.

Linz, Juan J. 1994. Presidential or Parliamentary Democracy: Does it Make a Difference? In The Failure of Presidential Democracy: The Case of Latin America, ed. Linz and Arturo Valenzuela. Baltimore: Johns Hopkins University Press. 3–89.

Nolte, Detlef/Mariana Llanos. 2004. “Starker Bikameralismus? Zur Verfassungslage lateinamerikanischer Zweikammersysteme.” Zeitschrift für Parlamentsfragen 35: 113-131.

Pérez-Liñán, Aníbal 2007. Presidential Impeachment and the New Political Instability in Latin America. Cambridge University Press: New York.

Strøm, Kaare. 1990. Minority Government and Majority Rule. Cambridge: Cambridge University Press.

Catherine Reyes-Housholder – Presidentas Rise: Consequences for Women in Cabinets?

This is a guest post by Catherine Reyes-Housholder, Ph.D. candidate at Cornell University. It is based on her paper, “Presidentas Rise: Consequences for Women in Cabinets?”, published in Latin American Politics and Society, 58 (3): 3-25, 2016.

More and more scholars and citizens want to know not only how women access presidential power, but what women do with this power once they are in office. Do female presidents use their power to promote change favoring women? I tackle this question by examining gender in the executive branch in Latin America—a region that has elected female presidents more times (nines so far) than any other region of the world.

There are some theoretical reasons to believe that female presidents will use their presidential power to promote change favoring women. In a recent article in Latin American Politics and Society, I argued that female presidents are more likely than male presidents to nominate women to their cabinets.

There are two reasons for this. The first speaks to bottom-up pressures from voters and the second to top-down, elite factors. First, female presidents are more likely than their male counterparts to interpret their mandate as a call for a greater female presence in the executive branch. Voting for a female president could easily be interpreted as a desire not just for a female president, but also for more female ministers. Female presidents thus may appoint more women to their cabinets because they believe their constituencies want them to.

Turning to top-down factors, the second reason has to do with the kinds of personal qualities presidents seek when they choose their ministers. In Latin America, presidents have virtually no formal restrictions on who they can nominate (i.e. no legislative body approves the presidents’ ministerial picks). So much of cabinet decision-making is based on informal considerations.

Presidents tend to seek ministerial candidates with two specific qualities: like-mindedness and loyalty. They look for like-minded ministers because they need someone who generally agrees with their policy ideas, or is at least like-minded enough to productively disagree and produce a better solution. Presidents also need loyal ministers who will faithfully execute their legislative agenda and are unlikely to threaten their hold on power.

Why would female presidents be more likely than male presidents to perceive women as more like-minded and loyal? The homophily principle and scholarship on gendered political networks helps explain this. Gender homophily is the recurring phenomenon where, ceteris paribus, women tend to associate more with women and men tend to associate more with men. Studies on gendered political networks suggest that male-dominance tends to feed on itself, making it difficult for women to penetrate male networks. On the flip side, because elite female politicians are more likely than their male counterparts to network with women, female presidents are more likely to perceive elite female politicians as like-minded and loyal.

So there are two reasons why we should expect female presidents and female ministers to present certain affinities. First, female presidents are more likely to face bottom-up pressures to do so. Second, female presidents are more likely to view female ministerial candidates as like-minded and loyal. They therefore face elite-based incentives to name more female ministers. These bottom-up mandate and top-down elite variables may both function as mechanisms linking presidents’ sex to a use of power to enhance women’s presence in cabinets.

But there’s a catch. While male presidents often historically have named all-male cabinets, female presidents are highly unlikely to completely exclude men. This is in part because female presidents face an informal constraint in assembling their cabinets. One of the most important constraints on their ability to name female ministers is the supply of female ministerial candidates. One major determinant of the supply is “political capital resources,” which can refer to relationships with party elites and with industries or social groups related to a particular ministry (i.e. women’s groups for Women’s Ministries).

Because women are less likely than men to possess “political capital resources,” the female pool ministerial candidates is generally much more shallow than the male pool. So I also argue that female presidents are more likely to “make a difference” in terms of women’s presence in cabinets when the pool of female ministerial candidates is deepest. Right after their inauguration, the pool for both male and female candidates is deeper than later on in the presidential term. As presidents later fire and hire ministers, the pool of qualified candidates will continue to shrink. I predicted that female presidents’ decision-making in naming women to cabinet is most likely to statistically differ from male presidents’ decision-making at the beginning rather than at the end of their terms.

The depth of the female ministerial pool also depends on certain characteristics of ministries. Some ministries are more associated with traditionally “feminine” roles and qualities—for example education and health. Others, namely defense and finance, are more “masculine.” There will tend to be more female ministerial candidates for “feminine” ministries because female politicians are more likely to possess political capital resources in traditionally feminine domains than traditionally masculine domains. For example, female politicians are more likely to possess political capital resources in areas of education rather than defense; they are more likely to have networked with social organizations related to schools than the military.

In short, I argue that female presidents overall are more likely than their male counterparts to name women to the cabinets. However, due to supply constraints, female presidents’ impact will likely be strongest for their inaugural cabinets and for “feminine” ministries.

I tested this theory on an original database of all inaugural and end-of-term cabinets by all democratically elected presidents from 1999-2015 in 18 Latin American countries. The dataset included 1,908 ministers. I found some evidence that presidentas in Latin America tended to name more women to their cabinets, and the most consistent evidence showed that they were more likely to name women to their inaugural cabinets and to “feminine” ministries. The dataset is located on the Harvard dataverse and on my web site www.reyes-housholder.com where you can access all the documents you would need to replicate my findings.

To conclude, there are theoretical reasons to believe and empirical evidence showing that female presidents will use at least their delegative power to improve women’s numerical representation in the executive branch.

Guillermo Rosas and Luigi Manzetti – Misery, corruption, and presidential approval

This is a guest post by Guillermo Rosas and Luigi Manzetti. It summarizes their recent paper, ‘Reassessing the trade-off hypothesis: How misery drives the corruption effect on presidential approval’ that was published in Electoral Studies, Volume 39, September 2015, pp. 26–38.

In “Reassessing the trade-off hypothesis: How misery drives the corruption effect on presidential approval”, we scrutinize a belief common among observers of Latin American politics according to which citizens in the region support corrupt governments as long as these are able to keep the good times rolling. Presidents that have arguably condoned corruption yet managed to win reelection include, among others, Carlos Menem in Argentina and Dilma Rousseff in Brazil. Observers often employ the lema “rouba, mas faz” (he steals, but he delivers) that characterized Ademar de Barros, an early 20th-century governor of the state of Sao Paolo, to celebrate the political longevity of these leaders.

Though a limited version of the trade-off hypothesis focuses on citizens that receive some direct benefit in exchange for supporting a government that otherwise condones corruption,[1] a more general claim suggests that voters might be willing to tolerate corruption as long as governments oversee high economic growth and low rates of inflation and unemployment. To test this more general flavor of the “trade-off hypothesis”, we analyzed information from over 141,000 respondents in 83 nationally representative surveys fielded by Americas Barometer[2] between 2004 and 2012. Rather than focusing on the declared vote intention of respondents, we concentrated on their levels of “presidential approval”, that is, on their assessment of their country’s president on a 5-point scale. We also avoided the use of elicited perceptions of corruption, because this indicator typically produces a large endogeneity bias as individuals tend to declare high perceptions of corruption whenever they disapprove of their government.[3] We show in our study that using an indicator of corruption victimization produces more reasonable estimates of the effect of corruption on presidential approval.[4]

Figure 1 displays partially-pooled estimates of the “corruption effect”—that is, the estimated impact of being a victim of bureaucratic corruption on an individual’s proclivity to approve of the president—in the 83 surveys that we inspect (in arriving at these estimates, we control for several individual- and survey-level confounders). For comparison, we also display as a grey horizontal bar the 95% credible interval corresponding to our estimate of a negative corruption effect pooled across all surveys. With the single exception of Bolivia in 2010, where our estimate of the corruption effect is positive but substantively small, we either uncover no effect or a negative effect across surveys. In Brazil 2008, for example, the estimated corruption effect is clearly negative, but we cannot detect a similar effect in Brazil 2010.


What explains these varying corruption effects? We focus on understanding whether a country’s economic performance deepens or assuages the effect of corruption victimization on approval.[5] Based on the logic behind Arthur Okun’s celebrated misery index (misery = inflation rate + unemployment rate), we expected inflation and unemployment to have a greater impact on the size of the corruption effect than economic growth. The rationale is that misery has a higher impact on the purchasing power of an average citizen. Inflation, in particular, is a tax that affects all citizens; more specifically, situations of high inflation are extremely disruptive, and ordinary citizens have little capacity to protect themselves from its worst effects. Similarly, high unemployment reveals that labor markets may be extremely tight, so that even those lucky enough to be employed cannot easily search for better-paid jobs. The primacy of inflation and unemployment as potential drivers of the corruption effect is also well documented in clarity of responsibility thesis according to which citizens perceive both factors to be under the control of the incumbent government.[6]

We find that inflation and unemployment have a discernible impact on the “corruption effect” in the Latin American surveys that we analyzed; in contrast, we do not find evidence that economic growth matters either way. Figure 2 shows how estimates of the corruption effect tend to become larger (i.e., more negative) as the rates of inflation and unemployment increase, while controlling for obvious survey-level covariates. We acknowledge in our study that observations from Venezuela, which suffered extraordinary high rates of inflation between 2006 and 2010, may affect our inferences inordinately, but we also believe that Latin American citizens would react similarly in the face of hyperinflation. Consequently, we suggest that information provided by the Venezuelan observations should be, at worst, only mildly discounted.
Based on presidential approval series, we conclude that Latin American voters are not easily duped. Those who experience corruption assess the country’s president negatively, and inflation and unemployment tend to lower a president’s average rates of approval. But along the lines of the more general “trade-off” hypothesis, we also find that the worse marks on presidential approval come from those that suffer corruption in economic environments characterized by inflation and unemployment.

Electoral accountability may yet foster good governance in Latin America. Naturally, there is a long causal chain going from presidential approval to vote choice to removal of governments that condone bureaucratic corruption, and even when voters disapprove of a president they might lack better alternatives among the opposition. At the very least, however, we should eliminate stereotypes about voters that prefer “corrupt but effective” leaders” over “clean but inept” politicians. Latin American voters do not forget the slights of corruption victimization as soon as they are in a macroeconomically stable environment; if there is something akin to a trade-off hypothesis, it is only because they return a harsher verdict on the executive when misery is added to the indignities of bureaucratic corruption.

[1]See Matthew S. Winters and Rebecca Weitz-Shapiro. 2013. “Lacking Information or CondoningCorruption? Voter Attitudes Toward Corruption in Brazil.” Journal of Comparative Politics 45(4):418–436.

[2]The Latin American Public Opinion Project (LAPOP) produces the AmericasBarometer, which can be found at www.LapopSurveys.org.

[3]See, among others, Mitchell A Selligson. 2006. “The Measurement and Impact of Corruption Victimisation: Survey Evidence from Latin America.” World Development 34(2):381–404.

[4]We employ an indicator that is coded 1 when the respondent reports a bribe solicitation from a public employee at any point during the twelve months that precede the interview.

[5]Elizabeth Zechmeister and Daniel Zizumbo-Colunga analyze within-country variance to address this same question. Zechmeister and Zizumbo-Colunga. 2013. “The Varying Political Toll of Concerns about Corruption in Good versus Bad Economic Times.” Comparative Political Studies 46(10):1190–1218.

[6]G. Bingham Powell Jr. and Guy D. Whitten. 1993. “A Cross-National Analysis of Economic Voting: Taking Account of the Political Context.” American Journal of Political Science 37:391–414.

Guillermo RosasGuillermo Rosas is Associate Professor of Political Science at Washington University in St. Louis. His research focuses on the economic consequences of political regimes and on the effects of political institutions on political behavior in Latin America.



Luigi ManzettiLuigi Manzetti is a research associate at the Tower Center and at the Center for Civil Society Studies of the Copenhagen Business School. He specializes in issues that include governance, corruption, and market reforms in Latin America.


Shane P. Singh and Ryan E. Carlin – Happy Medium, Happy Citizens

This is a guest post by Shane P. Singh and Ryan E. Carlin from the University of Georgia and Georgia State University respectively

As democracy’s third wave gathered steam, constitutional engineers argued fervently that “getting the institutions right” could help inoculate fledgling democracies from breakdown. Linz famously warned that reverting to presidential models of democracy in Latin America would perpetuate the cycle of democratic instability in the region. He cautioned of a “dual mandate” problem: since both executive and legislature could claim to speak for the voters, they were destined to clash. Fixed terms and an aura of sovereignty would only make the matter worse and lead to praetorian politics. Such arguments notwithstanding, newly (re-)emerged democracies throughout Latin America maintained presidentialism but reformed its institutional framework by improving its flexibility and stability in various ways. This broad institutional experimentation begs the question: are certain models of presidentialism more likely to produce consolidated democracy?

Focusing on attitudinal consolidation, in an article recently published in Political Research Quarterly, we reason that executive powers are associated with opinions about democracy. On the one hand, voters expect a degree of control over presidents, and a president that supersedes his or her mandate could spur disillusionment with the principles and performance of democratic regimes. On the other hand, an enfeebled president who lacks power could end up in perpetual conflict with the legislature, leading voters to become unhappy with policy immobilism and, ultimately, a deficit in representation. Thus, from a citizen’s perspective, the sweet spot should lie in the middle. Presidents who are either very weak or very strong will foster discouragement with democracy. But presidents who have balanced powers, and thus must engage in some give and take with the legislature, will cultivate public support for democratic regimes.

To test our expectations, we gather surveys spanning 1995-2005 from 18 Latin American countries, all with presidential models of democracy. Our dataset includes over 100,000 individuals, each of which was asked a.) whether they believe democracy is the best form of government, and b.) whether they are satisfied with the way democracy works in their country. We also gathered three measures of presidential power: an index of the legislative powers afforded to the president by the constitution, a measure of the degree to which the president is constrained by the legislature, and a measure of the president’s realized legislative success, or “box score,” calculated as the percentage of bills initiated by the president that were approved by the legislature.

As expected, we find that each measure is associated with attitudes toward democracy in a curvilinear fashion. That is, as presidential power and prerogative initially rise, individuals become more positive toward democracy, but at a certain point, further increases in presidential strength harm attitudes toward democracy. Our findings are illustrated in the figure, which makes it clear that each index of presidential power has an inverse-U shaped relationship with democratic support.

Screen Shot 2015-03-10 at 17.40.18


















As an example, the bottom-right panel depicts the relationship between legislative success and the probability that a citizen is “very”or “fairly” satisfied with democracy. Consider a president whose passage rate is just 33 percent, like Colombian president Ernesto Samper in 1996. Starting from this low success rate, a standard deviation increase is associated with about a 7-percentage point increase in the likelihood that a citizen is satisfied with democracy. Citizens value presidents who are not completely ineffectual. Compare this to a president with twice the passage rate (66 percent), such as Honduran president Carlos Roberto Reina in 1996. Starting from this middling success rate, a standard deviation increase is associated with about a 10-percentage point decrease in the likelihood of a respondent expressing satisfaction with democracy. Citizens also value presidents who are not completely dominant.

Interestingly, this curvilinear relationship is not conditional on economic performance or trust in the president. In other words, citizens do not desire very strong powers for presidents who preside over booming economies, and even citizens who fully trust the president become less positive toward democracy when he or she wields vast authority.

Our findings suggest that, when it comes to fomenting mass regime support, not all forms of presidentialism are created equal. Presidential democracies are most legitimate in the eyes of the citizens where presidents enjoy a happy medium of lawmaking power. And although presidents with vast formal powers and/or broad legislative majorities can avoid gridlock and fluidly implement their agendas, such behavior can fuel discontent with democracy and undermine its legitimacy. The same is true if the president’s every attempt to effect legislation is thwarted. Of course, the idea that a balance of power is good for democracy is nothing new. As James Madison wrote in Federalist No. 51, “Ambition must be made to counteract ambition.”

Note: a slightly different version of this post appeared at The Quantitative Peace in June of 2014.



Shane P. Singh is an Assistant Professor in the School of Public and International Affairs at the University of Georgia. He received his Ph.D. in 2009 from Michigan State University. His research focuses on the institutional and contextual foundations of political behavior and attitudes. His work has been funded by the Social Sciences and Humanities Research Council of Canada, and his research has appeared in many academic journals and edited volumes. His website is found at http://www.shanepsingh.com.


Ryan Carlin political scienceRyan E. Carlin is an Associate Professor of Political Science at Georgia State University. He received his Ph.D. in 2008 from the University of North Carolina at Chapel Hill. He is a Faculty Affiliate of GSU’s Center Latin American and Latino Studies and the Center for Human Rights and Democracy, and an Affiliated Researcher of the Latin American Public Opinion Project at Vanderbilt University. His research interests are comparative political behavior and public opinion, with a regional emphasis on Latin America. His work has been funded by the National Science Foundation, USAID, the Mellon and Ford Foundations, and the Latin American Studies Association. He is co-editor of The Latin American Voter (University of Michigan Press, forthcoming) and his work has appeared in many academic journals. His website is found at https://sites.google.com/site/ryanecarlin/.