This is a guest post by Adrián Albala, University of Brasilia, Brazil. Contact: firstname.lastname@example.org. It is based on a paper published in the British Journal of Politics and International Relations
In recent years, there have been huge advances in coalition theories, particularly concerning coalitions in parliamentary systems. However, much of the literature dealing with coalitions under presidential regimes has tended limit itself to reproducing and importing models from the work on parliamentary regimes, without considering the particularities of presidentialism.
Three features of presidentialism may interfere with the coalition process. First, the main particularity of presidentialism is the “winner-takes-all” principle. This states that the election determines a clear winner.
This feature makes it almost impossible for there to be any surprise in identifying the president-elect. This is quite different from most parliamentary regimes in which the identification of a clear winner may be more difficult, and the subsequent composition of the government is, often, very hard to predict. Indeed, under parliamentary regimes elections consist of a “photograph” that depicts the strength of each party or coalition in parliament in order to determine which will have a majority for forming a government. Figure 1 sets out this difference between the two processes.
Figure 1: Election processes in parliamentary regimes/ presidential regimes
The example of Belgium in 2011-2012, where negotiations lasted almost a year and a half, during which the country had no formal government, constitutes a paradigmatic example of this feature. However, this is a recurrent feature of parliamentary politics and has happened = again in Belgium (2015), but also in the UK (2010), Ireland (2016), Spain (2016-17), Germany (2017-18), Italy (2013) and Greece (2014-2015).
The second particularity of presidentialism is the principle of the presidential mandate. This implies that both the inauguration and the conclusion of the presidential mandate are settled by the constitution. This supposes that on the day of the inauguration of the presidency the president has to have his government formed. Moreover, the termination of the mandate should not, theoretically, be dependent on a majority (re)alignment in the congress. Recent events in Brazil have shown that this principle is not deterministic: a president can be impeached for political and opportunistic motivations.
Finally, the third particularity of presidential regimes is symmetrical bicameralism. As a matter of fact, bicameral congresses under presidentialism used to be symmetric, i.e. both chambers used to have the same powers and attributes. This is quite different from parliamentary regimes, where most of the upper chambers – except in Italy – have a mere consultative role.
This is of particular importance as bicameralism supposes a two-round procedure in the policymaking process for the president, thus increasing what Lupia and Strøm (2008) call “the shadow of the unexpected”. For this reason, it is reasonable to state that for those polities with a bicameral congress, holding a bicameral majority is a relevant condition for both the policymaking process and coalition governance (Hiroi and Rennó 2014). By the same token, controlling only one of the two chambers by the president might not be sufficient to ensure that policies get approval, or even to guarantee the survival of the coalition.
When considering coalition cabinet formation and dissolution, I argue that this third particularism is strongly linked to the first two. Indeed, for too long, scholars have studied coalitions under presidential regimes as they did under parliamentarism: assuming that the executive needed to look for allies in only one chamber. However, symmetric bicameralism makes such an assumption untenable. In fact, bicameralism, particularly symmetrical bicameralism in presidential regimes, may contain significant constraints for policymaking and coalition duration. Indeed, controlling one of the two chambers may not be sufficient for the president to ensure policy approval.
This misconsideration is particularly true when reviewing the literature about coalition cabinets under presidential regimes. More particularly, an important number of works have modeled the president’s ability to govern based on his/her legislative strength, or on the distribution of portfolios following Gamson’s law (i.e based on the proportionality principal of the strength of each party in the legislative branch), also known as “coalition congruence”. However almost every study dealing with this issue has measured the legislative strength or coalition congruence, based only on their observation of the lower chamber. In other words, almost no study has ever considered the upper chamber (i. e the Senate) as a relevant actor in the coalition process. We need to consider both bicameralism and bicameral majorities as relevant variables for the understanding of coalition cabinets under presidential regimes. The only work on this topic in parliamentary regimes reaches contradictory conclusions (Eppner and Ganghof 2016; Druckman et al. 2005; Diermeier et al. 2007; Druckman and Thies 2002)
This is what this paper tries to explore, focusing on Latin American presidential regimes that have experienced coalition cabinets.
First of all, half of Latin American countries (9/18) have a bicameral legislature. Bicameralism is not a trivial issue. This number is even higher if we compute every government since the third wave of democratization in the region, which began in 1979. Indeed, I have considered 134 governments in the region, i.e., cabinets following electoral processes. Of these, 54.47% (74) were formed in bicameral polities.
Moreover, when focusing on the occurrence of coalition cabinets, the relevance of bicameralism becomes even more central. Indeed, based on a strict but common definition of coalition cabinets (see Albala 2016), I have computed 31 newly formed coalition cabinets since 1980. That is 31 cabinets that were coalitions on the day of the president assumed office (See Table 1).
Of those 31 coalitions, 29 (93,5%) were formed in bicameral polities. Only Ecuadorian coalitions were formed in polities with a unicameral congress. In other words, for every ten coalitions to be formed in Latin American presidential regimes, more than nine occurred in polities with a bicameral congress. Why has no-one ever considered this feature?
The bicameral condition
I stated above that bicameral congresses under presidential regimes used to be symmetric, that is to say that the two houses (House of Deputies and the Senate) used to share similar attributes and powers. Thus, to ensure governability, a president-elect prefers to enjoy a bicameral majority rather than a partial majority (only one chamber) or no majority.
I have stated also the principle of a fixed presidential mandate. The president’s mandate not only concerns the end of the administration but also the beginning. Thus, the process of cabinet formation under presidentialism is limited in time, running from the proclamation of the result of the election to the inauguration day, generally fixed by the constitution. This feature supposes that the president will have formed his/her cabinet by inauguration.
Then, in order to determine how the combination of those two features (bicameralism + fixed mandate) may affect the coalition formation process, I compared the parliamentary strength of the coalitions after the election day of the president, with their strength at inauguration day.
We may, thus, theoretically, expect that presidents-elect who could not get a bicameral majority on the day of their election would seek to enlarge their parliamentary support including newcomers to their electoral alliance.
Results and findings
In Table 2, I identify 29 presidents-elect, comparing their legislative strength at election day with their strength at inauguration day. I simplified the operationalization of the legislative strength into three categories: i) no presidential majority at all ; ii) a legislative majority in one House; and, iii) a bicameral majority.
The data clearly confirms the hypothesis. Indeed, only 37.93% (11/29) of the presidents-elect had a bicameral majority (2) at election day. Nevertheless, the rate raises at 65.6% (19/29) at inauguration day, indicating that 8 presidents-elect proceeded to open negotiations with other parties to form a coalition or enlarge their electoral coalition. In other words, 8 presidents-elect who could not obtain a bicameral majority via the election, decided to include new members before their inauguration in order to get a bicameral majority. Conversely, the rate of minority coalitions (full or partial) fell from 62,06% ( 0= 31,03% + 1= 31.03%) to 34,4% (10,3% +24,13%).
Additionally, among the presidents who failed to obtain bicameral majorities, the first three Chilean presidents since the return of democracy (Aylwin, Frei and Lagos) had to deal with a particular constitutional feature inherited from the Pinochet rule: the existence of 9 designated senators, mostly from the military forces, who prevented the government from reaching a majority in the Senate.
By contrast, the data shows that bicameralism has been a central feature for presidents-elect who were not able to reach a bicameral majority while running alone. Indeed, among the three cases that ran alone on election day, all of them negotiated with new partners and achieved a bicameral majority.
In the cases where the length of time between election day and and inauguration day is longer – Uruguay and Brazil – there was a high degree of coalition enlargement and only one president-elect failed to achieve a bicameral majority (Lula I). However, among the polities with the shortest of time between election day and and inauguration day we can distinguish between Bolivian presidents-elect who have always managed to obtain a bicameral majority and Argentinian presidents-elect, who led coalition cabinets, but who have never enjoyed majorities in the two houses. Therefore, the timing condition deserves further attention.
We also found that in the cases where president-elect won a majority in only one chamber, the chamber in which the president-elect was unable to reach a majority was systematically the Senate. Hence, the upper chamber seems to be harder to conquer for presidents. This finding should also open a new line of investigation.
Finally, we also found that a bicameral majority makes it easier for the coalition governance generally and, thus, constitutes a sufficient, but not necessary, condition for enduring coalition agreements.
In this work, I have highlighted the need to adapt the study of coalition cabinets under presidential regimes to the particularities of such regimes. Indeed, I have shown out that presidentialism has a critical impact on the timing of the coalition formation process. Moreover, bicameralism is a central feature for the presidents-elect. These elements, in turn, open up new fields of study. For instance, no study has ever considered the role of the vice-president. However, vice-presidents can play a key role, especially when when (e.g. Argentina, Uruguay, and Venezuela) they are also the chair of the upper house.
 There are, however, some exceptions. For example, Bolivia used to have a system that could lead to “surprises”. Indeed until 2008, when no candidate reached the absolute majority, the run off used to take place in parliament leading to parliamentary bargains. Sometimes, the president-elect was not the one who won the plurality at the popular election.
Albala, A. (2016) Coalitions Gouvernementales et Régime Présidentiel. Sarbruken: Editions Universitaires Européennes.
Diermeier D, Eraslan H, and Merlo A (2007) Bicameralism and Government Formation. Quarterly Journal of Political Science, 2(3): 227–252.
Druckman J N, Martin L, and Thies M (2005) Influence without Confidence: Upper Chambers and Government Formation. Legislative Studies Quarterly, 30(4): 529- 548.
Druckman JN., and Thies M (2002) The Importance of Concurrence: The Impact of Bicameralism on Government Formation and Duration. American Journal of Political Science, 46(4): 760-771.
Eppner S, and Ganghof S (2016) Institutional veto players and cabinet formation: The veto control hypothesis reconsidered. European Journal of Political Research. DOI: 10.1111/1475-6765.12172.
Eppner S, and Ganghof S 2015. Do (weak) upper houses matter for cabinet formation? A replication and correction. Research and Politics. 2(1): 1–5.
Hiroi T and Rennó L (2014) Dimensions of Legislative Conflict: Coalitions, Obstructionism, and Lawmaking in Multiparty Presidential Regimes. Legislative Studies Quarterly, 39(3): 357-386.
Lupia A and Strøm K (2008) Bargaining, Transaction Costs and Coalition Governance. In Strøm K, Müller W and Bergman T (eds) Cabinet and coalition bargaining. Oxford: Oxford University Press, pp. 51-84.