Abstract
In recent years, four approaches about executive-legislative relations in Brazil have emerged: i) the perspective that points out limitations and constraints of multiparty presidential systems; ii) the building of government coalitions; iii) coalition management; and iv) the role played by institutions including the prerogatives of party leadership inside the House. In this paper, we review the literature on these approaches, offering a guide for studies about the Brazilian multiparty presidential system.
Introduction
The generic problem facing a president in a fragmented multiparty setting is how to induce legislators, especially those who take part in the president's coalition in the congress, to act in a way that supports the president's interest? Of particular significance is if and to what extent the president has the capacity to offer the appropriate incentives to legislators in order to promote the government's interests.
Scholars of the largely disperse literature on presidential–congresssional relations have recognized and concurred that this is a problem faced by most presidents, especially those whose political party does not enjoy a majority of seats in congress and is obliged to build a post-electoral coalition majority in order to govern. However, scholars have offered different answers about the conditions under which legislators choose to strategically cooperate with or disregard the president's preferences in congress. More specifically, there is no consensus among the political scientists who study legislative behavior. On the other hand, there are also many theories about how presidents lead in congress. These theories offer differing explanations regarding which determinants (or independent variables) best explain the legislator's behavior inside the congress.
Each scholar, and each differing approach, has helped to elucidate the complex set of reasons that explain legislators’ pattern of voting; specifically, why some presidents experience grater success than others and why presidents face more problems in approving certain issues than others. However, most of the literature has presented limited and partial explanations and has depicted an incomplete picture, by privileging one variable that they select as the most important to explain the legislator's voting behavior. In other words, these researchers have only used a single variable to explain this complex and certainly multivariate phenomenon.
With particular regard to the Brazilian political system, many studies have emerged from the 1988 Constitution. That document established a constitutionally strong presidential regime with an extremely factionalized multiparty system in Congress. In this institutional environment, it is essential that a president builds a post-electoral legislative coalition. The present review synthesizes the literature that has been published on this topic since 1988. 1 Because it would be impossible to critically analyze every single contribution regarding the Brazilian multiparty presidential system, 2 we decided to select a wide range of contributions on this intriguing subject.
Abranches’ 1988 work was the first to deal with the Brazilian issue in this way. It was he who first used the term “coalitional presidentialism” to describe the Brazilian political system.
Also known as coalitional presidentialism.
Among the papers that gave a general overview of the system, Mainwaring (1990) compared the old institutionalism in Latin American politics with fresh contributions about presidentialism and democratic stability, and proposed a research agenda that emphasizes the conditions when executive power strengthened. Power (2010) noted that a research agenda that mainly looked at aspects related to the electoral system was skeptical about the future of democracy in Brazil, while the research on the internal functioning of legislature rules and the constitutional powers held by the executive reached an opposite conclusion. Recently, Raile, Pereira, and Power (2011) and Chaisty, Cheesman, and Power (2014) clarified the options available for presidents in what all these authors have termed the “executive toolbox”.
Our main objective is to review why these studies’ reservations about the system were unfounded. Section one covers studies that were critical of the future of Brazil's political system, while section two presents the main features of coalition building in presidential regimes. Next, we analyze the tool used to keep the coalition working sustainably, followed by an analysis of political institutional design. Finally, we present a section dealing with the role played by party leadership in executive–legislative relations.
The first wave of post-1988 studies to focus on presidentialism
Abranches (1988) was the first to take the unique Brazilian political system of organizing its executive power with coalitions and compare it with consolidated democracies around the world. He concluded that the main difference was the kind of government. In Abranches’ view, the main problem with Brazil's system lay in its emphasis on the government's performance.
Linz (1990) argued in favor of parliamentary regimes, noting the contradictions in presidential systems that create a position with huge political powers – the president – and institutions responsible for limiting those powers (auditing courts, for example). In Linz's opinion, another problem is the internal conflict of the president: acting as a politician inside a party and also being the executive chief of a nation are mutually exclusive options. In addition, the fixed term is an impediment for quick solutions in the case of a corruption crisis involving the president because it is more difficult to implement an impeachment process for a president than to dissolve a parliamentary cabinet.
Scholars also have concerns regarding the Brazilian electoral rules, which combine an open list and proportional representation. It has been argued that this combination allows citizens to select their candidates instead of parties. They can base their selections on candidates’ personal qualities, their activities, and personal records. This provides a strong incentive for candidates to develop direct links with their constituency groups rather than to mediate such relations through political parties. These constituency groups may include local government, local business elites, and professional groups, among others. Thus, the personification of the vote is highly influenced by the way that citizens elect individual legislators (Mainwaring and Scully 1995; Mainwaring and Shugart 1997; Haggard 1995; Haggard and Kaufman 1995; Lima Junior 1993; Ames 1995a and 1995b; Lamounier 1994).
Papers from the 1990s focused on how individual behavior incentives – such as open-list proportional representation, 3 incumbents with guaranteed re-election rights, 4 the possibility of having more candidates than there are contested seats, and the right of a member of congress (MC) to change from one party to another one without any penalty 5 – have created system weaknesses, spawning catch-all parties with consequent legislative disciplinary problems (Mainwaring 1991 and 1997; Mainwaring and Pérez-Liñán 1997).
Problems associated with open list proportional representation were extensively debated in Ames (2002a, 2002b).
The possessor of a particular political position had direct access to the party list in the upcoming elections. This rule was suspended by the Supreme Court in 2002.
Opportunity extinguished by the Supreme Court in 2007.
A few incentives for representatives to follow party directives include whether the party leadership list access control is weak, whether the vote is nominal and not on the list, and whether there is a high proportion of candidates in relation to the magnitude of the district. All of these features weaken the bargain between Brazilian legislative parties and the executive (Shugart and Carey 1992; Carey and Shugart 1995; Shugart and Mainwaring 1997).
As noted earlier, many authors have attested that electoral rules resulting in multiparty systems and the impossibility of a majority government would lead this kind of presidential system to failure. However, this was not observed in Brazil or in other Latin American countries. Nonetheless, as Pereira and Melo (2012) claimed, multiparty presidentialism appears to be here for good. They argued that multiparty presidential regimes must be seen as special cases that do not function like parliamentary political systems or like two-party presidential models. In other words, the literature does not fully grasp the operational nuances of coalition-based presidential regimes. Scholars have been misestimating the outcomes of these regimes because they have been using theoretical and analytical tools designed to analyze either European multiparty parliamentary regimes or the American two-party presidential system. It is only recently that scholars have started to examine the relevance of coalitions in multiparty presidential regimes. This newer research has addressed topics such as coalition voting discipline (Amorim Neto 2002), the relationship between coalition type and legislative success (Cheibub 2007), the flexibility that executive appointment powers provide presidents (Martinez-Gallardo 2005), the impact of cabinet formation on presidential survival in times of crisis (Negretto 2006), the relationship between presidential policymaking strategies and cabinet formation (Amorim Neto 2004), and the use of ‘pork’ and other tradable political currencies for political bargaining and coalition management (Raile, Pereira, and Power 2011).
Primary presidential support from MCs stems from offering jobs in the federal bureaucracy. Since we can observe cohesiveness inside the group of patronage recipients created, support for the president will be effective (Figueiredo and Limongi 1999). A new cabinet is formed whenever a new president is inaugurated, when there is a change in its party composition, or when more than five percent of the ministers are changed (Amorim Neto 1994).
Figueiredo, Salles, and Vieira (2010) noted that 67 percent of presidents without an electoral majority in Latin America, including Brazil, had built government coalitions using their cabinets. For Raile, Pereira, and Power (2011), the cabinet is a way of coalition building, with the presidential party's share inside the House and the president's popularity being inversely proportional to the number of departments (or ministries) given to other parties.
Comparing the 1946–1964 period with the post-1985 years in Brazil it is concluded that, in both periods, the bigger the job offering inside the federal bureaucracy, the higher the legislative discipline (Amorim Neto 1994; Amorim Neto and Santos 2001; Amorim Neto, Cox, and McCubbins 2003; Figueiredo 2007).
Vis-à-vis cabinet proportionality, the ratio between departments offered to a certain party and its share inside the coalition is called the coalescence degree. Empirically, coalescence has a positive legislative submission relationship; it also negatively correlates with the number of decrees 6 issued by the president, with the supposition that weak executive–legislative relations that are a by-product of a low coalescence degree stimulate a higher number of executive decrees (Amorim Neto 2000 and 2002; Amorim Neto and Tafner 2002; Amorim Neto, Cox, and McCubbins 2003). 7
Known as
Research about cabinet-building in Europe (Amorim Neto and Strom 2006; Amorim Neto and Samuels 2010) showed that the share of independent ministers inside the cabinet is a positive function of the following: electoral volatility; semi-presidential regimes; minority governments; and from the president's legislative powers, being negatively related with congress fragmentation.
The studies considered here, mainly based on econometric analysis, have shown the process of budget amendment from proposition through to execution as being vital for executive–legislative relations. In this scenario, representatives can amend the annual budget sent every year by the president to congress for ‘pork barrel’ projects. Executive bargaining power over this process is expressed first in authorizing the money, and second, in effectively giving it to deputies as patronage.
Budget amendments can influence deputies’ electoral ambitions. Ames’ explanation of the spatial patterns in Brazil's 1990 House elections found that candidates in 1989 and 1990 had looked to strongholds in vulnerable cities to solve their electoral weaknesses by offering patronage (1995a). Mayoral candidates who had previous experience as representatives allocated more patronage to the city where they were running for local executive. Those who tried a higher-level position, but also had earlier been representatives, allocated more money for their states than deputies who did not run for those jobs, although incumbents seeking re-election had a similar performance in terms of implementing their budget amendments to those who were running for higher office (Samuels 2002; Leoni, Pereira, and Rennó 2003). In the 1998 House elections, patronage had a more positive electoral effect than legislative propositions (Pereira and Rennó 2002 and 2003; Pereira and Mueller 2003).
The following studies have viewed patronage as a coalition maintenance tool that is highly valued by congressional representatives.
8
Knowing this, the president uses it as a currency to deal with the legislative coalition, as well as using the approval or execution of budget amendments to get
A different point of view in favor of the collective feature of budget amendments can be seen in Limongi and Figueiredo (2007).
Usually, dominant-concentrated elected deputies give more support to presidential bills than to congressional bills. The same trend could also be observed in MCs who receive higher amounts of money from budget amendments (Ames 1995b; Pereira and Mueller 2002). Pereira and Mueller (2004) pointed out that patronage is very low-cost. Also, the number of executed individual budget amendments in the 1990s was a direct function of the support for executive bills (Alston and Mueller 2005; Pereira and Orellana 2009).
Refining the executive support mechanism concept, Raile, Pereira, and Power (2011) stated that budget amendments also serve to attract some representatives from opposition parties, as was observed in the 2003 Pension Reform, when coalition members who already had allies inside the bureaucracy saw the execution of many amendments proposed by representatives from oppositional parties. While approving opposition amendments is not the rule, it can be used occasionally as a powerful weapon.
Many institutionalist characteristics have emerged from studies that view institutions as a kind of government, while others focus on electoral issues, or prioritize the legislative powers of the president, etc. 9
The focus relies on formal institutions, but we cannot neglect the role played by the informal ones. The latter approach has few analyses. For example, Desposato (2006), based on São Paulo and Piauí, stated that clientelism as an informal institution creates minor party cohesion and makes governance more difficult.
For instance, scholars have emphasized the institutional rules and structures that organize the legislative process itself. That is, the set of rules and internal procedures that define the level of centralization in terms of prerogatives of initiating the decision-making process (agenda setting) in the hands of deputies or in the hands of parties and/or executive. This literature has attempted to explain how institutional variables internal to the decision making process (the distribution of power inside congress) and the institutional legislative powers held by the president work as key variables in the definition of the legislator's behavior. Figueiredo and Limongi asserted:
electoral laws and lack of party control over candidacy may give politicians room for cultivating personal votes and defying party line. But individualistic behavior does not encounter a milieu to develop in Congress. The institutional powers held by the executive, on the one hand, and the centralized decision making system in the legislature, on the other, impose restrictive agendas and limit legislators’ role in policy outcome (Figueiredo and Limongi 1997a: 3).
According to Melo (1998), national constitutions reduce transactional costs when they stipulate the role played by each party in electoral and governmental processes. When politicians limit their ability to act, this could be interpreted as an attempt to avoid future irrational behavior. Cheibub, Elkins, and Ginsburg (2011) argued that the huge powers granted to Latin American executives do not occur in all cases at the expense of congress’ powers, noting that Latin American legislatures have greater surveillance powers over the executive than those in other regions.
The 1988 Brazilian Constitution gave the president huge legislative powers, such as partial veto, decree power, bill urgency requests, and exclusive initiative on budgetary issues. Nevertheless, strong presidents have not been considered dangerous for presidential democracies (Cheibub and Limongi 2010). Checks and balances, such as an independent judiciary and congressional oversight, play an important role in Brazil, and include such executive counterweights as public prosecutors (
Another interesting point is that partisan fragmentation is considered beneficial because it increases transaction costs between parties for a joint attack on judiciary and other control agencies (Melo 2009).
Regarding electoral rules, neither party fragmentation nor minor parties could affect political stability among the five biggest Brazilian political parties (Amorim Neto and Cox 1997; Cheibub 2002; Santos 2008). Also, Cheibub and Limongi (2002). Cheibub (2004) found that cooperation incentives are bigger in parliamentary regimes, but the probability of coalitions matches those of the president under certain circumstances. Therefore, it seems that party fragmentation is not a barrier for governability.
Another important threat coming from the president is his or her partial and full veto powers against congressional bills. Santos (1997) argued that current Brazilian presidents have seen their veto powers reduced; this is another institutional feature, which tries to limit some of the executive's prerogatives.
Executive decree power, which was viewed as a delegation of authority from the legislature to the executive in the 1980s National Constituent Assembly, was seen as an instrument that offered modernization and administrative action (Figueiredo and Limongi 1997b). This hypothesis was tested by Pereira, Power, and Rennó (2005), who assessed the extent to which delegation theory 11 and unilateral action 12 could be observed in Brazil from 1988 to 1998. They showed that no particular pattern among the two mentioned above were perceived. In the whole period, unilateral action could be seen; however, regarding just Cardoso's first term, delegation theory was observed thanks to the requirements of the Real Plan. 13
Increasing number of provisional decrees on situations of high presidential popularity.
Increasing number of provisional decrees in periods of low indexes of presidential approval and less congressional support.
Figueiredo and Limongi's 1995 study detailed how, after 1988, the executive proposed 88 percent of all federal laws in the country. Armijo, Faucher, and Dembinska (2006) stated that neither the propensity to political chaos nor the governance created from the sacrifice of representatives, mayors, or governors in favor of the executive can be applied to Brazil. Figueiredo and Limongi argued that all governments from Sarney to Lula in his first term were based on a cooperation system in which a strong president was supported by other political agents. 14
Party leadership is another institutional feature that affects Brazilian governability. After 1988, party delegations were observed as being very loyal to their leaders, allowing easy forecasts about future roll calls (Limongi and Figueiredo 1995; Figueiredo and Limongi 1999).
Acting by himself or herself, a representative's bargaining power against the federal executive tends to be too small. In order to get what they want, MCs must cluster in a political party with a representative to bargain with the executive on their behalf; the party leader plays this role (Limongi and Figueiredo 1998 and 2002; Figueiredo and Limongi 2000; Pereira and Mueller 2003).
In contrast to the American legislative decentralization, the Brazilian system gives wide power to political parties inside congress. The role of the leadership is important because there is no difference between representatives regarding rights in votes and other common procedures; since leadership determines nominations for important positions inside the house, a rational MC will follow his or her leader to make his or her future demands possible. Meanwhile, top-down leadership is not common due to potential party colleague rebellion. Thus, we expect cooperation between a delegation and its leader (Limongi and Figueiredo 1998).
The successful economic stabilization plan carried out by Cardoso when he was finance minister during Franco's presidency. Cardoso was subsequently elected president in 1994 and re-elected in 1998.
Another institutional design that is able to influence executive–legislative relations, but has not been studied to the same extent as others, comes from the federalist issue. There are two streams regarding federalism. The first one, represented by Carey and Reinhardt (2003), Arretche and Rodden (2004), and Cheibub, Figueiredo, and Limongi (2009), looks at the weaknesses of states and the strength of the federal executive branch acting over delegations. The other stream, led by Samuels (2000 and 2002) and Desposato (2004), argues the opposite, advocating the power of state governors at the expense of federal executive's power.
According to Figueiredo, Limongi, and Valente (1999), another important feature of a leader is his or her ability to appoint and remove colleagues from committees. The appointment to be part of committees of great importance as the Committee of Constitution, Justice and Citizenship and the Committee of Finance are very desired by several MCs, leading them to subordinate themselves to their party leader.
This paper has reviewed the Brazilian multiparty presidential system literature. Linz, Mainwaring, Shugart and Carey led a first wave of studies and developed ideas focused on presidential regimes; forecasts about governance in Brazil were skeptical. Time has shown that such skepticism was misplaced because the studies had not considered the features that could induce governability.
Other studies have attempted to understand how such an unusual system could survive with a reasonable level of stability. Amorim Neto found that Brazilian presidents build their coalitions by giving jobs in the bureaucracy to political parties.
Pereira, who is considered an authority on ‘pork barrel’ politics, showed how this tool could help a president get an agenda approved at a low cost.
Authors like Cheibub, who dealt with the theme of institutions shaping the relations between the president and the congress, observed the impact of the strong legislative powers that the 1988 Constitution granted to the executive. However, checks and balances were also observed as a way to restrain a president's power.
Another source of thought regarding governability in Brazil comes from party leadership strength with regard to their delegations. Limongi and Figueiredo have made a particularly important contribution. They systematically demonstrated that representatives are highly obedient to party leaders and vote on the floor according to the leadership's wishes.
The rich literature on presidential–congressional relations critically reviewed in this paper has identified several aspects and variables as a source of presidential support in congress. Although all of these approaches try to answer the same question – that is, what are the conditions for presidential success in the legislative arena in a multiparty institutional setting? – they do so in different ways, offering conflictual and often partial explanations. In fact, each of those approaches illuminates one facet of the inherently complex relations between president and congress. There is no doubt that these approaches have offered a great deal of insightful explanation regarding the phenomenon of the executive–legislative relationship. Nevertheless, by working with models that deal with isolated variables instead of considering the influences of all variables simultaneously, those approaches have presented analytical limitations. The contribution of this literature review, then, is to demonstrate that executive–legislative relations in multiparty presidential regimes should assume that presidential, congressional, electoral, and institutional internal variables matter in constraining the bulk of legislators’ options and coalition behavior.