Political and Institutional Determinants of the Executive's Legislative Success in Latin America

Introduction

Our objective is to report the results of an ongoing comparative research project that aims to investigate the institutional and political determinants of government performance in Latin America. Our dependent variable is the legislative success of the executive. Studies on government performance in parliamentary countries usually focus on the stability of the government measured by its duration. The literature on presidential performance, in turn, has turned its attention to the fate of bills sponsored by the government.

These different approaches perhaps derive from opposite views concerning the way parliamentarism and presidentialism function, particularly when a majority government does not emerge from an election. In parliamentary systems, the executive's legislative success presumably stems from the formation of the government itself. When the executive fails, the government falls. In presidential systems, on the contrary, given the president's fixed term, legislative success depends on various dimensions of the process that follows the inauguration of the president, including the formation of coalition governments.

The legislative success of the executive in presidential systems varies greatly, as shown by the sample of countries covered here. The yearly average of presidential bills enacted ranges from zero, as in Colombia, to 98.8%, as in Mexico. On the other hand, Latin American presidential countries also vary greatly as to their institutional framework and to the political conditions under which governments rule. Therefore, they make up an exceptional array of cases to compare.

This research note is divided into three sections. In the next section, we trace an overall picture of presidential governments in the region from 1979 to 2006. In the third section, we describe the variables, present the hypotheses and discuss the results of a multivariate analysis of the executive's legislative success.

Latin American Presidential Government

Despite some attempts to single out a Latin American model of presidentialism, the political and institutional diversity in government organization among the continent's countries recommends caution in generalizing. This is transparent in our sample of twelve countries for the period 1979-2006. The sample includes all South American countries except Peru and the Guyanas, two countries in Central America (Costa Rica and Panama) and Mexico. Considering the period covered, we work with 200 year-country observations. The details are laid out in Table 1, which distinguishes between three sets of countries according to the type of government they experienced during the period analysed: one-party, coalition or both.

Table 1
Types and characteristics of government in Latin America - 1979-2006

COUNTRY	PERIOD	NO. OF YEARS	NO. OF. PRESID.	NO. OF. GOVT/COA	AVERAGE DURATION (MONTHS)	% YEARS MAJORITY GOVT.	NO. OF EFFECTIVE PARTIES
			ONE PARTY GOVERNMENTS
Costa Rica	1986-2001	16	4	4	48	50	2,3
Mexico	1989-2002	14	3	3	56	86	2,5
Sub-total		30	7	7	51	68	2,5
			COALITION GOVERNMENTS
Bolivia	1995-2000	6	2	3	24	100	4,5
Brazil	1989-2006	18	5	14	15	72	6,4
Chile	1990-2006	17	4	7	29	58	5,1
Colombia	1992-2008	17	4	7	29	65	4,4
Ecuador	1979-2002	24	8	14	21	10	5,1
Panama	1990-2002	13	3	7	22	75	3,8
Sub-total		95	26	52	22	63	4,9
			BOTH TYPES OF GOVERNMENTS
Argentina	1984-2006	23	5	13	21	48	3,0
Paraguai	1993-2002	10	3	3	40	50	2,4
Uruguai	1985-2006	22	5	5	53	91	3,1
Venezuela	1979-1998	20	5	6	40	30	2,8
Sub-total		75	18	27	34	55	2,8
Total	1979-2006	200	51	86	28	67	3,8

It is worth noting that we do not consider only the first coalition formed at the beginning of a president's term, but all coalition reshuffles by the same president. The duration of governments therefore does not necessarily coincide with the president's term. In order to identify the beginning and end of the coalitions, we employed the same criteria to define the making and breaking of governments in parliamentary systems, namely: “1) any changes in the set of parties holding cabinet membership; 2) any change in the identity of the prime minister [or the president]; and 3) any general election, whether mandated by the end of the constitutional inter-election period, or precipitated by a premature dissolution of parliament” (Müller and Strom 2000, 12). These criteria are sufficiently general to be applicable to the formation of and changes in governments in the presidential system, with slight modifications that do not affect comparability with parliamentary countries.

The first aspect of Table 1 worth stressing is the high incidence of coalition governments: in six of the twelve countries, only coalition governments were formed, while in another four coalition governments existed for at least some of the time. In three of these countries, Argentina, Uruguay and Paraguay, the formation of coalitions was a direct consequence of the transformation of their party systems. This means that 67% of the presidents who did not obtain a majority formed coalitions, i.e., they sought to increase their parliamentary support bases. These results allow us to question the widely held thesis that “minority presidents do not have incentives to form coalitions”.

This remark leads us to another thesis, also widely held, which is also the inspiration for proposals to reduce the number of parties: “multiparty systems make the formation of majority governments difficult”. It has been demonstrated that, at least theoretically, this relationship is not a necessary one. It does not sustain itself empirically either. One example from our sample is Costa Rica, where one-party governments obtained parliamentary majorities for 50% of the period analysed, whilst having minority governments, though with a share of the seats above 40%, during the other half of the time. On the other hand, five of the six multiparty countries formed majority governments most of the time. The only exception is Ecuador, which had minority governments for 90% of the period.

The thesis claiming that multiparty systems will face difficulties is accompanied by the following corollary: “the higher the level of party fragmentation, the more difficult the formation of majority governments”, which is also hard to sustain in light of the data shown on Table 2. Brazil, with the highest level of party fragmentation in the region, had majority governments 72% of the time in its first 18 years of democracy, while Venezuela, with an advisable number of effective parties, 2.8, only formed majorities for 30% of the period considered here. As we will see below, the fragmentation thesis equally fails when submitted to a multivariate analysis.

Table 2
Executive's legislative success - Latin America - 1979-2006

COUNTRIES	% AVERAGE BILLS PASSED	NO. OF YEARS	% MINIMUM	% MAXIMUM	STANDARD DEVIATION
		ONE-PARTY GOVERNMENTS
Costa Rica	49.3%	16	11.6	94.5	24.5
Mexico	94.2%	14	81.8	98.8	6.8
Sub-total	70.2%	30	11.6	98.8	29.1
		COALITION GOVERNMENTS
Bolivia	69.8%	6	48.4	86.7	14.5
Brazil	84.9%	18	70.3	92.8	5.7
Chile	72.6%	17	16.7	91.1	19.0
Colombia	42.5%	17	0.0	84.9	27.3
Ecuador	41.8%	24	10.7	65.2	16.1
Panama	74.5%	13	55.0	92.3	10.4
Sub-total	61.8%	95	0.0	92.8	24.5
		BOTH TYPES OF GOVERNMENT
Argentina	59.3%	23	12.8	80.2	12.7
Paraguay	74.5%	10	60.3	90.9	10.6
Uruguay	49.4%	22	28.0	68.0	13.5
Venezuela	63.5%	20	37.8	89.5	20.1
Sub-total	59.5%	75	12.7	90.9	16.8
Total period	62.2%	200	0.0	98.8	22.9

Determinants of the executive's Legislative Success

In this section, we conduct a multivariate analysis to examine the impact of political and institutional variables on the executive's legislative success. We draw on a linear regression model, using the ordinary least squares (OLS) procedure.

the dependent variable

Our dependent variable is the rate of bills sent by the executive that were enacted. Table 2 provides the main information regarding this variable for each country.

As one can see, there is ample variation in the yearly success of the executive. Many countries display rates above 90%, but in Colombia the executive did not have any bills approved in 1995, 1998 and 2002. It is worth noting, on the other hand, that the countries with the highest and most stable averages, Mexico and Brazil, organize their government differently. Moreover, four countries with coalition governments display higher success rates than the Costa Rican one-party government.

The independent variables

Our analysis includes only political and institutional variables. We believe it is plausible to maintain constant contextual factors that may put pressure on governments’ legislative agenda, such as economic and fiscal problems, social inequality, poverty, inflation and the international environment.

The independent variables can be grouped into three sets. The first relates to the powers of the president: political powers referring to the president's support among the parties and institutional powers comprising constitutional legislative prerogatives. The second set of variables includes the characteristics of the government and/or its coalition. The third focuses on the characteristics of the party system, and is an attempt to capture the bargaining environment in which the negotiations between the executive and the legislature take place (Laver and Schofield 1990).

President's institutional powers

The president's institutional powers include agenda-setting powers and veto powers. The first one is an index composed of 16 different prerogatives constitutionally conferred to the president to set the legislative agenda. Although different statuses may grant legislative powers to the president, we considered only constitutionally-assured prerogatives, mainly because the constitutional status is by itself an indication of the importance attributed to the provision, making their comparison more reliable.

Agenda-setting powers

In order to measure the agenda powers of the president, we created an index comprising sixteen constitutional prerogatives grouped into five dimensions: 1. Constitutional and delegated decree authority, 2. Budget powers, 3. Exclusive legislative initiative, 4. Urgency request for bill consideration and, 5. Right to introduce constitutional amendments.

In each one of these dimensions, the characteristics of each constitutional prerogative was converted into a dummy variable with the score 1 referring to its presence in the country's constitution. The following prerogatives were thus considered:

• Constitutional Decree Authority (CDA)

• CDA is immediately effective as policy

• CDA is valid indefinitely (does not require legislative action)

• CDA is not restricted to substantive policy area

• Delegated Decree Authority (DDA)

• DDA is immediately effective as policy

• DDA is valid indefinitely (does not require legislative action)

• Executive has exclusive initiative regarding new expenditures in the budget law

• Restrictions on the legislature's ability to amend the budget in specific policy areas

• Restrictions on the legislature's ability to increase expenditure in the budget

• Adoption of the executive budget proposal if the legislature does not approve the budget on the regular schedule

• Executive's exclusive initiative on administrative matters

• Executive's exclusive initiative on fiscal matters

• Executive's exclusive initiative on other matters

• Executive's right to request urgency on bills

• Executive's right to introduce constitutional amendments

These sixteen prerogatives were summed up so as to compose the agenda power index. No weight was assigned because, when necessary as in the case of decrees, their different characteristics were counted separately, as one can see above. Cronbach's α test of consistency and reliability of the agenda power index was carried out, obtaining 0.77. We also performed a factor analysis which showed that the index is practically one-dimensional, only two of the sixteen prerogatives listed above represented a second dimension.

Veto powers

Only the partial veto or item veto was taken into account, since in all constitutions considered the presidents had total veto prerogatives. The strength of the “partial veto” was measured according to an index which took into account the varying conditions for its approval: 0 = no veto; 1 = Plurality or no quorum specification; 2 = Majority chamber quorum; 3 = Majority of chamber members; 4 = 3/5 chamber quorum; 5 = 3/5 of chamber members; 6 = 2/3 of chamber quorum; 7 = 2/3 of chamber members; 8 = No override.

President's political powers

• Honeymoon period: first year of a president's term

• President's party share of seats in the lower house

• Centrality of president's party – absolute difference between the president's party and the centre of the ideological position of the parties

• Membership of the biggest party in the government coalition

Characteristics of the government/executive coalition

• Type of government (coalition or one-party)

• Number (absolute) of parties in the government/coalition

• Legislative status (majority/minority) of the government/coalition

• Share of seats – % of seats held by coalition parties

• Percentage of partisan ministers: percentage of ministers who participate in the government as representatives or with the authorization of their parties

• “Cabinet coalescence”: proportionality between parties’ share of ministries and their share of seats in the lower house – Amorim Neto's (2002) formula:

  $$Gabinet=1-\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$2$}\right.{\Sigma }_{i=1}^n\left(\left|Si-Mi\right|\right)$$

• Ideological dispersion of coalition parties – absolute difference between the most extreme parties

Characteristics of the party system

• Number of parliamentary parties in the lower house (NPP)

• Effective number of parties (ENP) – Laakso and Taagepera's (1979) formula:

  $$N=1/{\displaystyle \sum {\left(\text{Proportion of seats}perparty\right)}^2}$$

• Party fragmentation – Rae's (1967) formula:

  $$F=1/{\displaystyle \sum {\left(\text{Proportion of seats}perparty\right)}^2}$$

• The average values for these variables are shown on Table 3.

Table 3
Political and institutional powers of the president, coalitions and presidential success in Latin America - 1979-2006

COUNTRIES	% SEATS PRES. PARTY	% SEATS GOV. COAL.	AGENDA POWER INDEX	NO. OF PARLIAM. PARTIES	% PARTISAN MINISTERS	NO. OF PARTIES GOV. COAL.	COALESCENCE INDEX	IDEOL. DISPERSION GOV. COAL.	CENTRALITY OF THE PRES.'S PARTY	VETO POWER INDEX	PARTY SYSTEM FRAGMENT	% EXECUTIVE BILLS ENACTED
						ONE-PARTY GOVTS.
Costa Rica	49.6	49.6	0	5.3	97.3	1.0	0.75
	0.0	0.84	7	0.56	49
Mexico	54.0	54.0	0	5.0	92.6	1.0	0.75	0.0	1.0	6	0.59	94
						COALITION GOVTS.
Bolivia	32.3	64.5	2	7.5	70.2	3.6	0.72	2.2	1.37	6	0.77	70
Brazil	14.8	59.9	10.7	16.7	60.1	4.6	0.51	1.2	0,43	3	0.82	85
Chile	24.9	56.5	7.6	7.6	90.2	4.4	0.77	1.5	0,78	6	0.80	72
Colombia	29.8	65.5	9.3	14.7	71.9	2.4	0.70	1.2	0.67	3	0.69	42
Ecuador	15.0	30.9	5.3	11.6	64.7	2.5	0.53	0.5	1.39	7	0.79	42
Panama	26.1	45.9	5	9.9	82.9	2.9	0.74	0.8	0,60	7	0.73	38
						BOTH TYPES OF GOVT.
Argentina	43.4	48.8	3.6	19.6	95.4	1.5	0.74	0.2	0.51	6	0.65	59
Paraguay	51.8	69.4	3	3	95.5	1.8	0.80	0.4	0.81	2	0.57	64
Uruguay	37.9	63.3	8	4	98.0	2.1	0.76	0.6	0.72	3	0.67	74
Venezuela	37.5	44.7	5	8.3	65.9	1.3	0.71	0.1	1.06	6	0.59	49
Average	33.7	52.6	5.4	10	81	2.4	0,69	0.6	0.81	5.2	0.69	62.2
Number of cases	200	200	200	200	200	200	200	169	185	200	200	200

Hypotheses

We expect that the executive will tend to be more successful in approving its legislative agenda during the “honeymoon” period. This expectation is based on an intuition: in a system where the president is elected, often by a majority, it is natural to expect a cooperative congress immediately following the election. Several studies on Latin America and elsewhere have confirmed this outcome (Alemán and Navia 2009; Saez and Montero 2007; Altman 2008), regardless of the political and institutional context.

The literature is also consensual regarding the effects of other political variables. Our forecasts concerning the variables related to the president's political power follow conventional wisdom. The importance of the size of the president's party to his/her success cannot be denied. When the presidential party is not large, one expects the participation of the largest party in the governing coalition to have a positive effect on the government's legislative success. Lastly, as suggested by the median voter theory, the president's party's proximity to the centre of the ideological spectrum tends to imply higher approval rates of his/her legislative agenda. In all of these cases, we make the assumption — a necessary one in comparative studies dealing with a large number of cases — that the party acts as a unitary actor, even though intra-party politics is an important factor in the functioning of governments, particularly multiparty governments, as Laver and Schofield (1990) point out.

However, there is theoretical and empirical disagreement when it comes to the effects of the institutional factors. Theoretically, the power to propose does influence outcomes (Baron and Ferejohn 1989). In the legislative decision-making process, as Cox (2002) claims, the two dimensions of agenda-setting powers, i.e., the power to put bills on or to keep bills off the legislative agenda and the power to protect legislative proposals, shape legislative outcomes. The literature on Latin America, in contrast, tends to associate the president's agenda-setting powers to lack of political support or conflict with the legislature. For instance, Shugart and Carey (1992, 165) assert that presidentialism is more prone to conflict and instability when presidents are institutionally powerful. On the other hand, for Cox and Morgenstern (2002, 450-1), politically weak presidents resort more frequently to their constitutional powers and tend to push the limits of such unilateral actions in “constitutionally provocative ways”. An opposite view is put forth by Huber in his study of the Fifth Republic in France. A general conclusion from his studies is that agenda-setting powers can be considered instruments to manage the government's majority, particularly in coalition governments, rather than an expression of “vertical conflict” between the executive and the legislature (Huber 1996; 1998). Along this reasoning, and relying on previous studies on Brazil, we expect to find a positive effect of agenda powers on the executive's legislative success (Figueiredo and Limongi 2000; 2007).

We also expect a positive relationship regarding the strength of the partial veto, i.e., the harder it is for the legislature to override the president's vetoes, the more the latter will succeed in having his/her bills passed.

We also expect a positive effect with respect to the following features of the government or governing coalitions: majority status, high percentage of seats in the legislature and low ideological dispersion of the parties that make up the coalition. Conversely, we expect the number of parties in the coalition to have a negative effect. All of this assuming the party is a unitary actor and, as do classical coalition theories, taking into account the transaction costs and the role of conflicts of interest in the formation of coalitions (Axelrod 1970; De Swaan 1974). Recent studies have innovated by stressing the importance of two features of government coalitions in presidential systems. First, the cabinet's level of partisanship, i.e., to what extent ministers are chosen according to and as representatives of their parties and; second, the level of proportionality between the representation of a party in the cabinet and the number of seats it holds in the legislative or cabinet “coalescence index” (Amorim Neto 2002; 2006). Given the president's institutional prerogative of choosing his/her ministers, these indexes can be taken as an indicator of the level of cohesion between the executive and the legislative branches in presidential coalitions. The parties’ agreement to participate in the cabinet and the distribution of cabinet positions according to the parties’ legislative strength would result in disciplined support to the government legislative agenda. One can then expect that the higher the percentage of partisan ministers and the higher the proportionality between the representation of a party in the cabinet and the number of seats it holds in the legislature, the greater the executive's success

The third set of variables aims to capture the effect of the institutional context in which bargaining between the branches of government takes place, employing the characteristics of the party system as the main indicators. The argument that presidentialism and the multiparty system are a difficult combination is broadly accepted (Mainwaring 1993 Mainwaring and Shugart 1997; Jones 1995). But empirical results challenge this argument (Chasquetti 2001; Deheza 1998). The conception that the larger the number of parties, the higher the costs of governing can also be theoretically questioned. The number of parties in itself does not determine the cost of bargaining between the branches of government. This depends on the positioning of parties in the ideological spectrum. If the number of parties is large but the ideological difference between them is small, mere addition may not make a difference. On the other hand, if the parties are small in number yet ideologically polarized, it may be harder to garner majority support. Unfortunately, it has not been possible to fully obtain the required information in order to identify the ideological positioning of parties. Therefore we have used the absolute number of parties with parliamentary representation and have adopted different measures of party fragmentation/concentration: the number of effective parties and party factionalism. Our expectation is that party fragmentation by itself has no effect on the success of the government. Table 4 summarizes the hypotheses:

Table 4
Expected effects of independent variables

Independent Variables	Dependent Variable
President's political and institutional powers
Honeymoon period	+
President's party share of seats	+
Centrality of president's party	+
Agenda power	+
Veto power	+
Characteristics of the government
One-party government	+
Majority status government	+
Share of seats - Lower house	+
Number of parties in the government/coalition	-
Largest party in the coalition	+
Percentage of partisan ministers	+
Proportionality between parties' share of ministries
and seats in the lower house	+
Ideological dispersion of coalition parties	-
Characteristics of the party system
Number of parliamentary parties	No
Party fragmentation	No

Analysis and results

It is not surprising that some of the variables within and among the three sets of institutional and political factors singled out here are highly correlated. For this reason, the regression models presented below excludes some of them. Obviously, since both are measures of concentration, party fragmentation and effective number of parties are correlated, and both are also strongly correlated with the president's share of seats in the lower house. We decided to keep the party fragmentation variable only. Coalition, as one would expect, is correlated to the number of parties in the coalition. In this case, we kept the number of parties in the coalition because it provides more information than the dummy for coalition. For the same reason we excluded the dummy for majority government, since we have another variable with information for the share of the coalition seats. Finally, the ideological distance among the coalition parties was excluded due to its high correlation with the number of parties in the coalition.

After the collinearity tests, we ran the Model 1 regression, the results of which are displayed on Table 5. The two institutional variables related to the powers of the president – the president's agenda and veto powers – are significant at different levels. Nevertheless, contrary to our hypothesis, their effect on presidential legislative success is negative. These results differ from the positive effect found in the most systematic comparative study to date of presidential success in Latin America (Alcântara and Montero 2008). It is worth noting, though, that the latter differs from our research in two aspects. First, in addition to constitutional provisions, it also took into account agenda powers included in congressional internal regulations. Second, presidential success was not considered on a yearly basis as it is here, but rather on the average of presidents’ terms.

Table 5
Determinants of legislative success of the Executive (OLS)

President's political and institutional powers	Model 1	Model 2	Model 3
Constant	63.152***	52.393*	96.788***
Honeymoon period	.045	.067	-
Centrality of president's party	.010	.074	-
Largest party in the coalition	.041	.222***	.214***
Agenda power	-.417***	-.117	-.289***
Parcial Veto	-.223	.333	-
Characteristics of the government
Share of seats - Lower house	-.191	-.234**	-.270***
Number of parties in the government/coalition	.389***	.057	-
Percentage of partisan ministers	-.119	.030	-
Proportionality between parties' share of ministries and seats in the lower house	.253**	.038	-
Characteristic of the party system
Number of parliamentary parties	.043	.029	-
Party fragmentation	.020	.092	-
Countries
Argentina	-	-.534***	-.330***
Bolivia	-	-.211***	-
Brazil	-	.113	.277***
Chile	-	-.214	-
Colombia	-	-.406*	-.292***
Costa Rica	-	-.622***	-.499***
Ecuador	-	-.756***	-.515***
Panama	-	-.241*	-
Paraguay	-	.028	-
Uruguay	-	-.332	-.238***
Venezuela	-	-.170	-
R2	.125	.568	.528
F	2.237**	9.686***	21.976***
N	184	184	184

• Statistical significance:
• ^***
  1%;
• ^**
  5%;
• ^*
  10%.

On the other hand, the negative effect of agenda powers seems to confirm the argument that they reflect a more conflictive relationship between the executive and the legislative branches. In fact, however, the mere existence of these powers does not mean they are used, as attested by the cases of Chile (Siavelis 2000) and Uruguay, where they are rarely employed. Moreover, differences in constitutional provisions directly affect the amount of bills entering in the calculation of the rates of success. Taking Brazil and Argentina as examples, we see that in the latter, constitutional decrees do not require congressional approval to become permanent. They are not counted as laws but as decrees. In Brazil, on the contrary, the issuing of decrees (medidas provisórias) requires congressional approval, and they are enacted as ordinary laws and added to the total number of laws enacted. Therefore, in Brazil presidential rates of success comprise bills initiated as constitutional decrees, while in Argentine it does not. Future analyses must take these differences into account.

The two other significant variables in Model 1 are related to the characteristics of the government. The number of parties in the government has a positive effect, contrary to the thesis of transaction costs. It is plausible to think that the costs involved in the negotiation of many parties in the coalition could be overcome by an increase in the share of seats in the legislature. This however is not the case here, since the share of seats is not significant and the sign of this variable is negative.

It is important to observe, though, that as expected, even if at a lower level of significance, the proportionality between parties’ share of ministries and seats in the lower house has a positive effect on success. This reinforces, even if not very strongly, studies on the role of cabinet's characteristics on government performance (Amorim Neto 2006).

Model 2, in the second column of the Table 5, is the result of a regression taking into account unobserved, fixed effects. As we can see when we include fixed effects for the countries, the model becomes more robust and the results change considerably. Unobservable variables negatively affect the success rates of Argentina, Bolivia, Costa Rica and Ecuador, and at a lower level Panama, with Mexico as the reference-country.

The most significant variable in this model turns out to be the presence of the largest party in the government/coalition. This result supports our hypothesis and is important because it challenges the “difficult combination” hypothesis, i.e. that multipartism and presidentialism are incompatible. The irrelevance of the number of parliamentary parties and party fragmentation in both models further reinforce this result. In other words, the degree of fragmentation of the party system and the number of parties obtaining parliamentary seats matter little, unless the largest party is in the government (regardless of whether or not it is also the president's party). Probably this has to do with the ideological position of the parties, as we have argued before. This claim, however, still needs to be substantiated by more reliable information on party ideology.

Another change regarding the previous model is that the share of seats becomes significant, if at a lower level, but maintains the negative sign. This result seems to be contradictory to the importance of having the largest party in the government/coalition. However, it should be noted that share of seats is a continuous variable and that consequently its negative sign may indicate that minimal winning and not super-majority coalitions influence the executive's legislative success. This is a speculation worthy of further exploration.

Finally, after testing a series of models, we arrived at the model with the best fit to the data (Model 3). The results found in Model 2 regarding the presence of the largest party in the government/coalition and the coalition's share of seats do not change, except for a slight increase of the level of significance of the later. Nevertheless, the president's agenda power is significant and negative, as in Model 1, and its values increase. Unobserved variables negatively affect the presidents’ success in Argentina, Costa Rica and Ecuador, as in Model 2, but also in Uruguay and Colombia. Brazil is the only country where the unobservable effects are positive.

We have seen that the absolute number of parliamentary parties and the level of party fragmentation do not seem to play the role usually attributed to them in terms of government performance. It is highly likely that this result could be explained by the distribution of party preferences, as noted above. In other words, countries with higher levels of fragmentation are not the most polarized; therefore, the existing fragmentation is not reflected in the government's legislative success. The positive effect of the number of parties in the coalition may also be a consequence of parties’ low ideological polarization level, which would diminish transaction costs and get translated, on the contrary, into support. Such support would not require the largest possible number of seats, simply a number sufficient to approve the executive's bills.

The results regarding the role of the institutional powers are inconclusive. If they do not back up our expectation of a positive effect they do not rule it out either. In sum, more work to investigate the conditions under which they may have a positive or a negative effect on governments’ legislative success is still necessary. Not only do we need to increase the amount and improve the quality of the information used, we must also employ other statistical tools and analytical strategies or perhaps even adopt other research designs.

• 1
  See Calvo (2007), Alemán and Calvo (2008; 2010), Saiegh (2008). Altman (2008) and Amorim Neto (2006) focus on the duration or stability of coalitions.
• 2
  We have been unable to obtain sufficient information on Peru.
• 3
  Data was drawn from the following sources: Keesing's Record of World Events; Observatório Eleitoral Latinoamericano; Centro de Estudos Latinoamericanos da Universidade de Georgetown; Rulers; Latin American Weekly Report; Elections Results Archive; International IDEA; Banco de Dados - Facultad de Ciencias Sociales de la UdelaR - Universidad de La Republica; Datos de Opinión, Elites Parlamentarias Latinoamericanas, Universidad de Salamanca; the national Congresses of Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Ecuador, Mexico, Panama, Paraguay, Uruguay and Venezuela; Coppedge and Michael (1995), The Dynamic Diversity of Latin American Party Systems. Various colleagues helped us with data and feedback. We are particularly grateful to André Mejía Acosta, David Altman, Kenneth Benoit, José Antonio Cheibub, Simone Diniz, Mercedes Montero, Octavio Amorim Neto, Monica Pachón, Anibal Perez-Liñán, Sebastian Saiegh, Luciana Santana, Sergio Toro, Nina Wiesehomeier.
• 4
  This percentage is higher than that found by Cheibub, Przeworski and Saiegh (2004). However, these authors only consider the first coalition formed, while here we consider every coalition change.
• 5
  The matrix of the arguments is provided by Linz (1994). For an examination of the arguments and additional detailed references, see Cheibub (2007, 49).
• 6
  See arguments and demonstrations for a much larger sample, which includes parliamentarist countries, in Cheibub, Przeworski and Saiegh (2004, 575-6).
• 7
  This index is adapted from Altman (2008).
• 8
  This is the name given to this index by Amorim Neto (2002, 53), who introduced it in the analysis of Latin American presidentialism.
• 9
  Aleman and Tsbelis (2002) also raised this hypothesis, but did not test it.
• 10
  The information about Uruguay was provided by Daniel Chasquetti.
• 11
  In fact the president's party percentage of seats is not relevant for the president's legislative success.

Bibliographical References

• Alemán, Eduardo, and Ernesto Calvo. 2010. Unified government, bill approval, and the Legislative weight of the president. Comparative Political Studies, 43 (4): 511-34.
• ____. 2008. “Analyzing legislative success in Latin America”: The case of democratic Argentina. In New voices in the study of democracy in Latin America, eds. Guillermo O'Donnell, Joseph S. Tulchin and Augusto Varas. Woodrow Wilson Center Reports on the Americas No. 19, October.
• Alemán, Eduardo, and George Tsebelis. 2002. “Agenda control in Latin American presidential democracies”. Paper prepared for delivery at the 2002 Annual Meeting of the American Political Science Association, August 29 to September 1, 2002.
• Altman, David 2008. “Marriages of convenience?”. Presidents, cabinets, and coalitions in presidential regimes. Paper presented at the conference New Frontiers on Institutional Research in Latin America”, Institute of Latin American Studies at GIGA. Hamburg, May 5-6.
• Amorim Neto, Octavio 2006. Presidencialismo e governabilidade nas américas. Rio de Janeiro: Editora FGV/Konrad Adenauer Stiftung.
• ___. 2002. “Presidential cabinets, electoral cycles, and coalition discipline in Brazil”. In Legislative politics in Latin America, ed. S. Morgenstern and B. Nacif, 48-78. Cambridge: Cambridge University Press.
• Axelrod, Robert 1970. Conflict of interest. A theory of divergent goals with application to politics. Chicago: Markhan Publishing Company.
• Baron, David, and John Ferejohn.1989. “The power to propose”. In Models of strategic choice in politics, ed. Peter C. Ordeshook. Ann Arbor: The University of Michigan Press.
• Calvo, Ernesto. 2007. The responsive legislatura: Public opinion and law making in a highly disciplined legislature. British Journal of Political Science, no. 37: 263-80.
• Chasquetti, Daniel 2001. “Democracia, multipartidismo y coaliciones en América Latina”: Evaluando la dificil combinacion. In Tipos de presidencialismo y coaliciones políticas en America Latina. Colección Grupos de Trabajo de CLACSO, Buenos Aires, Argentina.
• Cheibub, J. Antônio 2007. Presidentialism, parliamentarism and democracy. Cambridge: Cambridge University Press.
• Cheibub, J. Antônio, Adam Przeworski, and Sebastian Saiegh. 2004. Government coalitions and legislative success under parliamentarism and presidentialism. British Journal of Political Science 34 (4).
• Cox, Gary 2002. “On the effects of the legislative rules”. Legislative Studies Quarterly 25 (2).
• Cox, Gary, and Scott Morgenstern, 2002. “Epilogue”: Latin America's reactive assemblies and proactive presidents. In Legislative politics in Latin America, eds. S. Morgenstern and B. Nacif, 446-69. Cambridge: Cambridge University Press.
• De Swaan, Abram. 1973. Coalition theories and cabinet formation: A study of formal theories of coalition formation applied to nine european parliaments after 1918.
• DEHEZA, I. G. (1997), “Gobiernos de Coalición en el Sistema Presidencial: América del Sur”. In El presidencialismo renovado: Institucionalismo y cambio político en America Latina, ed. D. Nohlen and M. Fernández. Caracas: Nueva Sociedad.
• Figueiredo, Argelina, and Fernando Limongi. 2007. “Instituições políticas e governabilidade”: Desempenho do governo e apoio legislativo na democracia brasileira. In A democracia brasileira. balanço e perspectivas para o século 21, ed. C. Melo and M. Sáez. Belo Horizonte: Editora UFMG.
• ___. 2000. Presidential power, Legislative organization, and party behavior in Brazil. Comparative Politics, 32 (2).
• Huber, John. 1998. “Executive decree authority in France”. In Executive Decree Authority, eds. J. Carey and M. Shugart. Cambridge: Cambridge University Press.
• ___. 1996. Rationalizing Parliament. Cambridge: Cambridge University Press.
• Jones, Mark. 1995. “Electoral laws and the survival of presidential democracies”. In Legislative multipartism and presidential legislative support, ed. ___, 75-87. Notre Dame: University of Notre Dame Press.
• Laakso, Markku and Rein Taagepera. 1979. “Effective number of parties: A measure with application to West Europe”. Comparative Political Studies 12 (1).
• Laver, Michael and Norman Schofield. 1990. Multiparty government: The politics of coalition in Europe. Ann Arbor: The University of Michigan Press.
• Linz, Juan J. 1994. “Presidential or parliamentary democracy”: Does it make a difference? In The failure of presidential democracy: The case of Latin America, eds. J. J. Linz and A. Valenzuela. Baltimore: Johns Hopkins University Press.
• Llanos, Mariana and Marsteintredet, Leiv 2010. Presidential breakdown in Latin America. London, Palgrave.
• Mainwaring, Scott. 1993. “Presidentialism, multipartism and democracy”: The difficult combination. Comparative Political Studies 26 (2).
• Mainwaring, Scott, and Shugart, Matthew 1997. “Presidentialism and democracy in Latin America”: Rethinking the terms of the debate. In Presidentialism and democracy in Latin America, ed. S. Mainwaring and M. Shugart. Cambridge: Cambridge University Press.
• Marsteintredet, Leiv and Berntzen, Einar 2008. Reducing the perils of presidentialism in Latin America through presidential interruptions. Comparative Politics 41 (1). New York: The City University of New York.
• Müller, Wolfgang, and Kaare Strom. 2000. Coalition governments in Western Europe. London: Oxford University Press.
• Rae, Doulgas 1967. The political consequences of electoral laws. New Haven/London: Yale University Press.
• Saiegh, Sebastian. 2008 “Ruling by statute: Evaluating chief executives' legislative success rates”. Manuscript.
• Shugart, Matthew, and John Carey. 1992. Presidents and assemblies. Constitutional design and electoral dynamics. Cambridge: Cambridge University Press.
• Siavelis, Peter. 2000. The president and congress in postauthoritarian Chile. Institutional constraints to democratic consolidation. University Park, PA: The Pennsylvania State University Press.

Publication Dates

• Publication in this collection
  18 Sept 2023
• Date of issue
  2009

History

• Received
  Nov 2009
• Accepted
  Dec 2009