text new page (beta)
English (pdf)
Article in xml format
Article references
Automatic translation
Send this article by e-mail
Cited by SciELO 
Similars in
SciELO 
Permalink
Introduction
In order to lay a due foundation for that separate and distinct exercise of the different powers of government, which to a certain extent is admitted on all hands to be essential to the preservation of liberty, it is evident that each department should have a will of its own; and consequently should be so constituted that the members of each should have as little agency as possible in the appointment of the members of the others. But the great security against a gradual concentration of the several powers in the same department consists in giving to those who administer each department the necessary constitutional means and personal motives to resist encroachments of the others. The provision for defense must in this, as in all other cases, be made commensurate to the danger of attack.
Federalist No. 51
The connection between political institutions and policy production is probably one of the most important in Political Science, not only because focusing on the impact that public policy has in the life of citizens may be vast but also because the status quo of society can always be changed by the institutions that are able to alter the life of their constituents. Institutions are important because the way rules are organized directly affects how decisions are made, and because the decision-making process may make the drafting of public policy more or less difficult.
Literature (Linz & Mainwaring, 1990; 1993; Shugart & Carey, 1992) has seen the presidential system as difficult to maintain, especially when combined with a multi-party system, which would increase the probability of a gridlock between the Executive and Legislative powers. Bipartisan systems, in this sense, would be less likely to have a political crisis and institutional problems regarding the drafting of public policy. This paper’s hypothesis is that there is no significant difference between two-party and multi-party systems regarding conflict and the drafting of public policy, and that success rates are similar, which is central to the argument made by Figueiredo and Limongi (1999) and others. It is still necessary to discuss how Presidentialism works, and to note that even though its characteristics may vary, cooperation and conflict will happen no matter how the political system is designed.
Many of the multi-party presidential political systems that have emerged in Africa, Latin America, and the former Soviet Union have proven to be stable. This is largely because the presidents have been able to act as Prime Ministers (Colomer & Negretto, 2005), forming coalitions between parties to secure a legislative majority and, consequently, the (at least partial) success of their legislative agendas. According to Cheibub (2002), the existence of a majority in a congress, whether unicameral or bicameral, happens in more than half the cases in the world due to party fragmentation.
Institutionally, Yates and Whitford (2005) have observed that presidents are not fully able to set the agenda or at least draw attention to all the issues they would like in it. This can be explained by the limits of presidential control, since there are different forces and phenomena like other relevant political actors who have their own agendas and priorities. Yet, they are direct representatives of the entire nation, so they cannot act unilaterally. The establishment of a system of checks and balances, which also counts on the presence/agency of other actors, is part of the institutional foundations of its agenda. In this paper, this is represented by the presence of vetoes in legislation proposed by the Executive power.
Part of the difficulty in the policy-making process lies in the multiple instances in which a bill can be vetoed. According Ames (2001), party leaders can secure the cooperation of its representatives, since many political actors have veto power over the process. As defined by Cheibub (2006), veto power is the ability to object to or refuse to sign a bill. While the former means the power to object to a portion of the bill (a partial veto), the latter is the complete refusal of a bill, a complete or total veto. Presidents with partial veto power do not have an all-or-nothing choice, but they are presented with more ways to influence legislation and hence are more powerful. When the president vetoes a bill, either partially or completely, it is often sent back to the legislature, which is given the opportunity to disagree with the president’s desire to change it, maintaining its original position and overriding the presidential veto. For a veto to be overridden, a larger legislative majority is required than the one needed for the law to pass.
This paper is divided into five parts. In the first, it considers how presidential power and public policy drafting are related in Brazil and the United States. Even though both countries have a presidential system, Brazil is characterized by the centrality of the Executive in the drafting of bills and setting of the agenda. Multi-party political systems are more likely to have internal conflicts and therefore less ability to govern. The countries compared here have similar types of governments: eight years of liberal administrations and eight years of more conservative ones. The second part of the article deals with the methodology used, and how the authors analyzed the data. The third gathers information on presidential success rate in welfare policy in Brazil and in the United States, considering two similar yet different political systems: bipartisan and multi-party Presidentialism. The fourth section navigates empirically through the president’s ability to veto in these countries. In conclusion, it shows that the bipartisan system of the United States displays more conflict when it comes to welfare bills than Brazil, making it possible to reject the thesis that multi-party systems are more likely to deadlock.
Presidential Power and Public Policy
The influence of the United States Constitution can be seen in countless other constitutions that arose during the nineteenth and twentieth centuries, especially in Latin America and Eastern Europe (Ackerman, 2000). Cheibub, Elkins and Ginsburg (2010), seeking to answer the question of whether it is possible to say that there is a Latin American Presidentialism, identify several characteristics of the American Constitution that have been used by Latin American countries, such as federalism, judicial oversight of the idea of separation of powers, and Presidentialism as a system of government. In the case of the United States, such a system is defined by the popular election of the head of the Executive, the lack of legislative authorization to remain in office, the impossibility of dissolving Congress, the absence of explicit legislative powers, the presence of emergency powers, the possibility of vetoing legislation initiated by Congress, etc.
However, it is obvious that not all those characteristics were used directly and without adaptation. In Latin American countries, Cheibub, Elkins, and Ginsburg (2010) verified that the Executive branch also has the power to initiate legislation, which represents a different type of Presidentialism. This power, which can be seen in a positive light, is evidenced by being granted emergency powers, issuing decrees and submitting of constitutional amendments and legislation.
The Brazilian system is closely linked to the American one. The chambers that make up the Legislative power are elected by the people, and the federalist option entails the existence of a representative chamber of the States of the Federation. In the United States, the Senate is the review house of the National Congress. Terms are fixed, not subjected to the approval of the head of the Executive but directly linked to the popular vote.
Checks and balances allow for interdependence between powers, for example, in the drafting of rules. In these cases, the Executive branch can propose legislation and holds the powers of sanction and veto. On the other hand, the Legislature has the power to amend and reject the Executive’s proposal. The veto power prevents unintended changes or projects of parliamentary initiative. Nonetheless, the veto can be rejected not only by the President, but also by the National Congress, by the absolute majority of its members, or the law may be promulgated by the Head of the Senate if the president does not veto it within constitutional terms.
Presidential Power and Public Policy in Brazil
Presidents in Latin America have extensive policy-making powers, and are able to influence the legislative agenda, control resources, appoint and dismiss officials, and see to electoral demands (Inácio & Llanos, 2015). There is no agreement among Brazilian political scientists to whether the extent to which this presidential power can impact the decision-making process and facilitate the approval of the Executive Power’s agenda (Santos, Pérez-Linãn & Garcia Monteiro, 2014). Abranches (1988) points to the problems of Brazil’s institutional design. These include the president’s extensive powers, the need for legislative support to facilitate governability, and the fragmentation of Brazilian parties coupled with the heterogeneity of society and political personalism that leads to low or nonexistent party discipline in Congress. The aforementioned decision paralysis is the focus of the discussion by numerous academics like Lamounier (1994), who believes that the institutional arrangement is aimed at blockading and not at decision-making; Mainwaring (1991, 1997), who focuses on the existence of multiple veto points; Mainwaring and Shugart (1997), who argue that the Executive’s constitutional prerogatives and the nature of the party system influence the Executive’s ability to implement its agenda; and Shugart and Carey (1992), who believe hybrid regimes are more likely to fail than “pure” systems.
However, this perspective is not shared by a second group of authors, such as Figueiredo and Limongi (1995, 1997, 2001), Santos (1999) and Diniz (1997). For Figueiredo and Limongi (2001), the 1988 Constitution creates a decision-making process that is institutionally sustained by the centralization of the president’s agendas and the existence of party leaders in Congress who also centralize the work and ensure the formation of stable coalitions and a majority, allowing the president’s agenda to be sustained. In this way, the Executive’s exclusive legislative powers, which are a provisional measure, are intended to protect the parliamentary basis of unpopular issues, as well as to solve bargaining problems within the coalition.
One aspect of Brazilian Presidentialism that is pointed out by all these authors is that the Executive branch has the power to propose legislation and therefore draft public policy. The way the Presidentialist system is designed determines how decision-making power is divided among domestic actors. For Milner (1997), the balance of power between the Executive and the Legislature depends on which one controls certain elements of the legislative process. Considering the constitutional prerogatives that the Executive has in relation to foreign policy, Shugart and Carey (1992) point out that the Executive can control the legislative process through the exclusivity of initiating certain legislation, which in Brazil is exemplified by the possibility of deciding over budget and financial matters since the 1988 Constitution.
Figueiredo and Limongi (1999) show that the 1988 Constitution maintained the vast legislative prerogatives of the president of the Republic instituted by the military, as well as the legislative recourses for party leaders in the National Congress. Party leaders channel and represent the party’s opinion to the Executive, facilitating the bargain between the two. This process explains, as per Figueiredo and Limongi, the high level of discipline within the National Congress. Since the president can cut off access to economic, bureaucratic or information resources, the costs are too high for a parliamentarian to act impulsively.
The key to explaining the domination of the Executive branch over the legislative process lies in its agenda power. Examples include patronage and the ability to determine what will be voted and when, by means of urgently urgent or urgent requests (the urgent request gives Congress 45 days to vote on a certain matter, often limiting the proposal to a specific committee and forcing the floor to vote, thereby limiting the possibility of amendments by parliamentarians).1
Another example is that the 1988 Constitution confers on the Executive Branch the exclusivity of the budgetary initiative, tax, and administrative matters. According to Figueiredo and Limongi (1999) and Silva and Araújo (2013), the Executive has control of legal production in an almost imperial way, since 80% of federal laws are of the president’s formal authorship, whereas only 1.9% of parliamentary initiatives became laws during the period between 1990 and 2006.
In addition, the Executive branch has reactive powers, like that of total or partial veto, which allow the president to adjust legislation (Shugart and Carey, 1992). In Brazil, the veto process must take place within 30 days from the date one of the Chambers receives it (Amorim Neto, 2000; Tsebelis and Aléman, 2005), although this view is only formalistic, since Congress has the power to postpone the deliberation of vetoes.
Presidential Power and Public Policy in the United States
In the United States, as Whittington and Carpenter (2003, p. 495) show, the narrative of legislative dominance is implicit. Much of the president’s power in the policy-making process comes from the Constitution. According to Grant (1997), the president’s roles are head of the Executive, Head of State, Chief Diplomat, Commander-in-Chief of the Armed Forces,
Head of Party, and Chief Legislator. These roles outline the president’s actions in the Executive, on foreign affairs, as a party leader and in his relations with the Legislative power. By delegating the legislating duty to Congress, the Constitution says little about the president’s role in legislating. This role comes from Article II, Section 3, of the Constitution: “He shall from time to time give the Congress information of the State of the Union and recommend to their consideration such measures as he shall judge necessary and expedient” (Shallus, 2019). The practical effect is that the president may suggest, request, and insist that Congress enact laws he believes are needed.
In addition to State of the Union Address, the president communicates his positions through the media and in special messages, reports, and mandatory documents like annual budgets. According to Brown (1985), the president’s priorities are raised from deep beliefs and convictions, others are products of a style or structure of thought and others may stem from campaign promises, partisan platforms, or demands from influential interest groups or powerful lobbies. These actions provide Congress with something diffused and that the decentralized structure prevents legislators from obtaining for themselves: an agenda. Indeed, as Chief Legislator, the president shapes public policy.
In addition to communicating his positions, the president may call for emergency sessions of Congress. There is also a source of legislative power that is not specified in the Constitution: the president has a formal capacity to act unilaterally and legislate on his own. Presidents do this through executive orders, proclamations, executive agreements or national security directives (Moe; Howell, 1999). An executive order is a presidential directive that requires or authorizes some action within the Executive branch to reorganize its agencies, alter administrative and regulatory processes, establish policies, and affect how legislation is interpreted and implemented (Mayer, 1999).
Still, the most common practice is the president’s communication of his position to Congress. Mack et al. (2011) found that the president may avoid taking a stand on a vote when he is less likely to find support from Congress on the issue, such as in the last months of his administration or when he is running for reelection. However, presidents are more likely to find support from Congress when they effectively state their positions, and when their party is dominant. Gibbs (2009) had previously indicated that having a Congressional majority, combined with presidential popularity and the economic conditions of the country, significantly influenced the success of the president in Congress.
Bill sponsorship, which is the object of this analysis, has been a possibility in the United States since 1967 in the House and around the 1930s in the Senate. A representative may introduce a bill as a courtesy, such as legislation proposed by the president. As Whitthington and Carpenter (2003, p. 501) put it, “[l]egislative proposals emerging from the executive branch, and in particular from the White House, often come attached to a larger political effort on behalf of those proposals, which can mobilize public support for the president’s own favored issues and crowd other issues off the legislative agenda. Congress need not buckle under to presidential pressure, but Congress cannot ignore the presidential agenda without costs.”
Method of Analysis
Drawing from Mahoney (2007), comparative analyses can be used for testing theories, which is the goal of this article. When comparing similar systems, such as Brazil and the United States, one cannot go beyond medium-range theories, due to the risk of overdetermination (many variables can intervene, but their influence cannot be controlled). On the other hand, by maximizing the differences between the cases, one can identify the independent variables that do not violate the claim of homogeneity of the total population. The analysis of different countries, for example, aims to determine whether similar mechanisms and processes cause changes in different periods, places and regimes (McAdam, Tarrow & Tilly, 2001). By identifying mechanisms linking a supposed explanatory variable and an effect variable, trust in the causal relationship is increased and complex causal combinations are identified (Ragin, 2000).
To calculate the presidential success rate, this study used the percentage of the legislation supported by the president that passed in the low chambers, the party conflict rate, and, when applicable, a probit regression. To verify the intra-party conflict on welfare legislation, the index developed by Stuart Rice (1925) was used, which consists of the difference between the percentage of votes in favor and against within a party, in a given vote, ranging from 0 (situation in which half of the party voted yes and the other half voted no) to 100 (a situation in which all legislators of a party voted in the same direction, whether yes or no), and can be expressed by the following formula:
The party cohesion index is then calculated from the average of votes. This average can come from temporal and thematic clippings. To measure interparty party conflict in roll call votes we use the party conflict rate, which is calculated from the Rice index. The party conflict rate is given by the following formula:
Where:
PCr = |
Party Conflict rate |
RICE = |
Rice Index |
G = |
Government Party(ies) |
O = |
Opposition Party(ies) |
j = |
given roll call |
k = |
direction of the votes of the majority |
α= |
check factor of majority votes between the government and opposition parties (α> 0 for roll calls where the majorities of the government coalition and the opposition parties voted in the same direction and α <0 for polls in which the majority of the government and opposition parties voted in opposite directions). |
In each situation where 100 % of the governing parties voted in the same direction as 100 % of the opposition parties, the conflict rate would be equal to zero (100-100). On the other hand, in a situation where 100% of the two sides voted in opposite directions, the conflict rate would be equal to two hundred (100 + 100). From the average of the indices presented in each vote, we can calculate the party conflict rate in each subject and in each specific legislature, in addition to the results in the period as a whole. In this way, zero and two hundred would be the extreme points of consensus and polarization, respectively, that the parties manifest in each vote.
Finally, to calculate the probability of the majority control of the chamber increasing the presidential success, a probit regression was used whose main characteristic is to constrain the estimated probabilities to be between zero and one, assuming that there is an unobserved latent continuous variable Y* that determines the value of Y.
The legislation initiated by the president of Brazil, as well as their vetoes, were collected at the Centro Brasileiro de Análise e Planejamento (CEBRAP) database and they are available online.2 Unlike the constitutional powers held by the president of Brazil, the president of the United States does not have the constitutional prerogative to initiate legislation. Thus, this study has selected the welfare state votes in which the president expressed his approval. The source for determining these votes is the yearly almanac of Congressional Quarterly (CQ) and the data collected by Lewis et al. (2019). It analyzes all the public statements and messages of the president to determine what legislation he does or does not desire.
Presidential Support in Welfare Legislation
Considering the constitutional prerogative of the president of Brazil to initiate legislation and that of the president of the United States to be able to express his position on legislation initiated by legislators, this paper analyzes the success rate of the presidents of these countries on welfare legislation and the degree of conflict between government and opposition parties on welfare roll call votes. The time period studied here covers a period of 16 years in each country, from 1995 to 2010 in Brazil and from 1993 to 2008 in the United States. The purpose is to compare presidential success and legislative conflict over a period of time when a party is replaced by its main opposition party in the presidency and to compare these same indicators in countries with different party systems. Therefore, this analysis focuses on the governments of Fernando Henrique Cardoso and Luiz Inácio Lula da Silva in Brazil and Bill Clinton and George W. Bush in the United States.
Presidential Success and Conflict Rate in Brazil
Fernando Henrique Cardoso was elected president of Brazil for the first time in 1994 and reelected in 1998. His party, the PSDB (Partido da Social Democracia Brasileira), formed a coalition with the PFL (Partido da Frente Liberal), the PMDB (Partido do Movimento Democrático Brasileiro), and the PTB (Partido Trabalhista Brasileiro) in 1995, and the PPB joined in 1996 (Partido Progressista Brasileiro). By the end of Cardoso’s first term, the PTB left the coalition, and in the last year of his second term, the PFL did so too. Most of the time, this coalition accounted for more than 70 % of the members of the Brazilian Chamber of Deputies. Because of that, the opposition did not destabilize Cardozo’s success in passing the legislation.
During Cardoso’s administration (1995-2002), 239 welfare bills were passed. In his first term, he initiated 112 bills and 127 in his second term. Cardoso’s success rate was an astounding 99.16 %. Only two bills were rejected, one in each term. The first one would have revoked the articles of the Consolidation of Labor Laws on trade union organization and the second aimed to reinforce appropriations in favor of the Ministry of Education.
Since the coalition was broad and agreements were easily reached, most of those 239 bills were voted symbolically. Only 14 nominal votes on welfare occurred. Compared to the apparent consensus on Cardoso’s bills, there was a high party conflict rate between the governing coalition and the opposition in those nominal votes. During Cardoso’s first term, the party conflict rate in the Chamber was 139.4, which indicates a strong polarization. The most polarized legislation was one that sought to regulate the acquisition of products for the implementation of health measures by the Ministry of Health. In his second term, the party conflict rate decreased to 110.0, meaning the Chamber became more moderate. The cohesion rate of the PSDB (99.3) was higher than that of the main opposition party, the PT (Partido dos Trabalhadores) (93.7), although both sustained strong cohesion.
Luiz Inácio Lula da Silva was elected president in 2002 and reelected in 2006. His party, the PT, started a government coalition initially with the PTB, the PDT (Partido Democrático Trabalhista), PPS (Partido Popular Socialista) and the PV (Partido Verde). In 2004, the PMDB, the PSB (Partido Socialista Brasileiro), the PL (Partido Liberal), and the PCDOB (Partido Comunista do Brasil) joined the coalition, while the PDT left it. In 2005, the PPS and PV left the coalition and the PP (formerly PPB) joined it. In 2007, the PDT and PV returned to the coalition and the PRB (Partido Republicano Brasileiro) joined it for the first time. In 2009, the PV once again left the coalition. During his two terms, Lula da Silva had between 58.3 % and 73.5 % of the Chamber’s members on his side. As in the Cardoso government, the large majority that supported Lula in the Chamber was crucial for his success in approving welfare legislation.
164 welfare bills were presented by Lula da Silva to the Chamber of Deputies: 97 in his first term and 67 in his second. His success rate was 100%. There were only 11 nominal votes; most of the legislation was voted symbolically. In his first term, the party conflict rate in those nominal votes was 91.8, whereas in his second term it was 94.3. The most polarized bill sought to provide formal recognition of trade unions in the Consolidation of Labor Laws (CLT). During the Lula da Silva administration, PT cohesion (89.5) remained above PSDB cohesion (83.2), although in the case of both parties it was lower than during Cardoso’s administration. Lula da Silva’s coalition included more parties than Cardoso’s; the latter gained the support of a higher percentage of deputies, but the opposition seems not to have obstructed Lula as much as it did Cardoso. According to Figueiredo and Limongi (1999), the Executive has an extremely high legislative approval rate. On average, around 80 % of legislative proposals initiated by the Executive Power were approved between 1989 and 1998. It is also noteworthy that the content of the bills introduced by the Executive and Legislative branches tends to be different. Most of the laws sent by the Executive in matters of welfare involve budgetary laws for the area, corroborating this analysis.
Presidential Success and Conflict Rate in the United States
The president of the United States, in addition to not having the constitutional prerogative to initiate legislation in Congress, faces another major challenge in securing support for his legislative positions. Unlike the Brazilian case, the legislature of the House of Representatives is renewed every two years. This means that if the President is poorly evaluated by the public opinion, the latter may, during the presidential term, strengthen the members of the opposition party in the House and thus make it difficult to approve legislation supported by the president.
Elected for the first time in 1992 and reelected in 1996, Bill Clinton faced an unfavorable institutional context to gain support from the House for his positions. Among the four legislatures in the House of Representatives during President Clinton’s two terms (1993-2001), the Democratic Party only had the support of a majority in the House of Representatives during the first; in the last three legislatures, the majority of the House was Republican. In that sense, while the 103rd legislature can be characterized as a party government case, the 104th, 105th, and 106th legislatures can be characterized as divided government.
During the two Clinton terms, 42 welfare state votes were supported by him in the House of Representatives, 34 in his first term and 8 in his second. In all of them, Clinton supported 12 votes related to labor and wages, 11 to health, 7 to education, 7 to social inclusion, and 5 to housing. Clinton got support from the House in 32 of the 42 votes. This means a presidential success rate of 76.2 %; only 10 votes supported by the president were rejected. Clinton did not get the House’s approval in votes aimed at securing union rights and political rights for federal officials, removing restrictions on federal funding for abortion programs, helping children whose parents were denied social benefits, including provisions for children of non-citizens to receive food stamps, introducing a system to measure student performance annually, regulating the system of adoption and custody of indigenous children, and providing incentives to housing programs.
The data seem to indicate that there was strong partisan consensus in the House, although the legislature had been controlled most of the time by Republicans. However, the party conflict rate shows a significant difference between Clinton’s two terms: whereas in the first term the average vote showed moderate polarization (129.8), in the second term it showed a strong consensus (69.4). Considering control of the legislatures as a benchmark, the party conflict rate decreased from 130.5 in the 103rd legislature (when the Democratic party had the majority control) to 106.0 during the 104th-106th legislatures (when the Republican party had the majority control). This is because the cohesion of Republicans increased from 50.4 in the 103rd legislature to 77.1 in the 104th to 106th legislature, while in the case of the Democrats it decreased from 92.3 to 78.3 in the same period.
Indeed, Democrats and Republicans seem to have been strongly affected by the change in control of the House. In the Democrat case, the decrease in cohesion can be interpreted as a shift toward the political center, in order to attract the support of the Republican majority. On the other hand, Republicans, a minority before, took advantage of their takeover of the House to advance their political agenda. This was the first time the party had run the House since the 83rd Legislature (1953-1955).
Although narrow, the Republican majority elected in the 104th legislature has been singled out as exceptionally efficient in advancing a Republican agenda (Jaenicke, 1995, 1998; Foley & Owens, 1996; Owens, 1997). In a study that sought to test the hypothesis that legislators’ behavior is motivated by factors other than their personal interests and the moderate positions of voters in their districts, Crespin et al. (2006) find evidence that a substantial part of the Republican party took on more conservative positions than their constituents in the 104th legislature, and this effect lasted until the 107th legislature. In addition to these effects, Garand and Burke (2006) conclude that the resumption of control of the House by Republicans almost completely changed the pattern of partisan behavior in introducing bills. Analyzing the 102nd-105th legislatures, they point out that members of the majority party showed consistently higher levels of legislative initiative than members of the minority party.
Unlike Clinton, George W. Bush had a Republican majority in the House of Representatives for most of his two terms (2001-2009). Republicans dominated the House from the 107th to the 109th legislature. Only in the 110th legislature did the Democrats regain control of the House. It is precisely the opposite situation to Clinton’s. Was this condition enough to further increase presidential success in social policy votes? The short answer is no. Bush’s presidential success rate was 75.9 percent, close to Clinton’s. Bush supported 29 social welfare state votes, 16 in the first term and 13 in the second. Of these, 13 were related to health, 5 to social inclusion, 4 to education, 4 to work and wages, 2 to food programs and 1 to housing.
Only seven votes that the president supported were rejected in the two terms. They are the “No Child Left Behind Act” of 2001, the “Improving Head Start Act,” the “Alternative Pluripotent Stem Cell Therapies Enhancement Act,” and legislation seeking to provide benefits for individuals with injuries resulting from a smallpox vaccine and other purposes, to strike section 726 of the bill regarding the importation of prescription drugs, to prohibit the use of funds by the NLRB to recognize any labor organization not certified by the NLRB as the exclusive representative of employees, and to apply the one-to-one replacement requirement for units demolished under a HOPE VI grant only to units occupied prior to demolition.
Although Bush’s presidential success rate was very close to Clinton’s, the party conflict rate between government and opposition was slightly higher during the Bush administration. In the first (139.8) and the second (131.6) Bush terms, welfare state votes were marked by moderate polarization. Using control of the majority in the House as a point of reference, the party conflict rate remains in the same range. When it had a Republican majority, from the 107th to the 109th legislature, the conflict rate was 136.6, and when it had a Democrat one in the 110th legislature, the index was 133.8. Republican cohesion decreased from 91.4 when Republicans had a majority to 70.7 when the Democrats regained control. In turn, the cohesion of Democrats in the same period increased from 82.1 to 92.1. Thus, as we saw in the Clinton administration, the realignment of the majority increased the cohesion of the opposition party and decreased cohesion in the party that held the Executive.
Considering that there are elements suggesting a relationship between majority control of the House and presidential success in welfare state votes in both the Clinton and Bush administrations, a probit regression considering the former as an independent variable and the latter as a dependent variable was conducted. The 71 votes supported by presidents Clinton and Bush were jointly analyzed.
Before presenting the estimated results we examine the goodness of fit measures in the model for the House of Representatives. To test the null hypothesis that all coefficients except the constant are equal to zero, the probit model gives the value of the log-likelihood chi-squared. It corresponds to the F-test in the linear regression model. The log-likelihood chi-squared value of the model is 15.12 and is significant at the 0001 level, which allows for the rejection of the null hypothesis. The potential of the model correctly predicts presidential success to a certain extent. The model was accurate 76.06 % of the time.
Table 1: Presidential Success Model in the U.S. House of Representatives
| Independent Variable | Coefficient | T-statistic (z) | P>|z| |
|---|---|---|---|
| Majority control of the House | 1.34 | 3.74 | 0.000 |
| Constant | 1.70 | 0.00 | 1.000 |
Log-likelihood: -31.5, N= 71
The analysis of the estimates of the β coefficients in the probit model is quite different from the OLS model. Because the estimates of the β coefficients cannot directly be interpreted as marginal effects on the dependent variable Y, its derivatives have been computed. The presidential success rate was 41.1 % higher when the president’s party had control of the majority than when it did not. Together with the data discussed above, the results of the probit regression model reinforce the importance of majority control for presidential success in welfare state votes. In addition, they confirm the findings of Gibbs (2009) and Mack et al. (2011).
Presidential Vetoes in Welfare Legislation
Veto Players and Public Policy in Presidential Systems
A veto player is an individual or collective actor whose agreement is required to make the decision to change policy. According Strom and Swindle (2002), they are not to be confused with decisive actors, whose approval is sufficient but not necessary to change thestatus quo, and from dictators, who need to approve new policies. According to Tsebelis (1997), in presidential systems, another category of veto player appears: the parties that form the government coalition, or partisan veto players. However, the agreement of partisan veto players is not necessary for policy change, given that related parties can bypass each other or be pitted against each other.
It is interesting to note that being able to veto a bill does not mean power over another player. A simple example is that not all presidents have veto power, and even if they do, their veto can almost always be overturned by a certain majority of members of the legislature. As the number of veto players increases, policy stability does not decrease, i.e. even though a change of the status quo does not become easier, it does not necessarily become more difficult.
Presidential Vetoes in the United States (1993 - 2008)
President Clinton used his veto power on 37 occasions over his two terms in office, 6 of which were welfare legislation. There were no attempts to override three of these vetoes in Congress. They are the “Teamwork for Employees and Managers Act” of 1995, the “District of Columbia Student Opportunity Scholarship Act” of 1997, and the “Education Savings and School Excellence Act” of 1998. In the Teamwork Act veto message, Clinton argued that this legislation “would undermine the system of collective bargaining that has served this country so well for many decades [...] and abolishes protections that ensure independent and democratic representation in the workplace”. On the District of Columbia Student Act, Clinton wrote that this bill was “fundamentally misguided and a disservice to those children” and pointed out that “the way to improve education for all our children is to increase standards, accountability, and choice within the public schools.” Clinton used this same justification to veto the Education Savings Act.
The House attempted to override the last three vetoes. Clinton vetoed the “Marriage Tax Relief Reconciliation Act” of 2000 because although he supported tax cuts, this cut in particular would not benefit the working class. The House tried to override it, but failed. The last two bills vetoed were about the partial-birth abortion legislation. Clinton argued he supported “the decision in Roe v. Wade protecting a woman’s right to choose. Consistent with that decision, I have long opposed late-term abortions, and I continue to do so except in those instances necessary to save the life of a woman or prevent serious harm to her health.” The House succeeded in overriding the veto on both occasions. On the other hand, the Senate failed to override it two times.
According to Conley (2004), the use of vetoes was a strategy of President Clinton to face a divided government. During Clinton’s second term (1997-2001) he issued 102 public veto threats. Many of the legislation vetoed concerned elements of the Republican agenda, causing considerable controversy in Congress.
George W. Bush vetoed four bills on welfare and the House attempted to override all of them. Three of these attempts failed. Two of them were about the “Children’s Health Insurance Program Reauthorization Act.” Bush stated in his veto message that this Act “would still shift SCHIP away from its original purpose by covering adults. It would still result in government health care for approximately 2 million children who already have private health care coverage.” The third legislation was the “Departments of Labor, Health and Human Services, and Education, and Related Agencies Appropriations Act” of 2008. Bush argued that the bill “does not sufficiently fund programs that are delivering positive outcomes.”
However, Congress succeeded in overriding one of Bush’s vetoes and the bill became a law: the “Medicare Improvements for Patients and Providers Act” of 2008. Bush considered this Act fiscally irresponsible. He argued “it would harm beneficiaries by taking private health plan options away from them” and “would undermine the Medicare prescription drug program.” The Congress overrode it in the House and Senate, justified by the rising number of uninsured Americans in the face of an economic crisis. The Chart below summarizes the 10 vetoes in welfare legislation during the Clinton and Bush Administrations.
Chart 1: Presidential Vetoes in Welfare Legislation and Congress Response
| President-Term | Legislation | Description | Attempt To Override | Final Status |
|---|---|---|---|---|
| Clinton-1 | H.R.1833 | Banning partial birth abortions | Yes/Failed | Vetoed |
| Clinton-1 | H.R.743 | Teamwork for Employees and Managers Act | No | Vetoed |
| Clinton-2 | H.R.1122 | Banning partial birth abortions | Yes/Failed | Vetoed |
| Clinton-2 | S.1502 | District of Columbia Student Opportunity Scholarship Act | No | Vetoed |
| Clinton-2 | H.R.2646 | Education Savings and School Excellence Act | No | Vetoed |
| Clinton-2 | H.R.4810 | Marriage Tax Relief Reconciliation Act | Yes/Failed | Vetoed |
| Bush-2 | H.R.976 | Children’s Health Insurance Program Reauthorization Act | Yes/Failed | Vetoed |
| Bush-2 | H.R.3043 | Departments of Labor, Health, and Human Services, and Education, and Related Agencies Appropriations Act | Yes/Failed | Vetoed |
| Bush-2 | H.R.3963 | Children’s Health Insurance Program Reauthorization Act | Yes/Failed | Vetoed |
| Bush-2 | H.R.6331 | Medicare Improvements for Patients and Providers Act | Yes/Succeeded | Enacted as Law |
As shown in the chart above, there were seven attempts to override the presidential vetoes, out of ten possible ones. This indicates that Congress uses this option to defy the president, but not recurrently. All the attempts took place when the opposition party held the majority of the House, reinforcing the conclusion that the importance of holding a majority in the legislative is important for presidential success.
Presidential Vetoes in Brazil (1995 - 2010)
Unlike the United States, Brazil experienced a higher number of vetoes to legislation introduced by the president. There were 91 cases of vetoes for this analysis. Of these, 54 were made by the Executive power in welfare bills introduced by the Legislative power and 37 by the Legislative power in bills introduced by the Executive power. This distinction, which does not exist in the American political system, needed to be made in Brazil to extend the analysis of the conflict between the two powers. Since it is not possible to identify them all individually due to their sheer number, a categorization was made to try to understand if there is a connection between the subjects of the bills and a specific Power. The vetoes were classified according to which power introduced and vetoed them.
There is a clear balance in bills vetoed by the Executive and Legislative regarding the environment, justice, and labor. However, when it comes to bills concerning health, there is a strong presence of legislative vetoes. For example, PL.2242/96 was vetoed and later archived due to its unconstitutionality. Likewise, most housing vetoes were made by the Executive on bills introduced by the Legislative. No conclusions can be drawn from just the topics of the bills, but the information is nonetheless enriching.
As for the number of vetoes, most were made regarding labor (22 total), education (19 total) and welfare and social promotion (17 total). These policy areas typically divide political parties and, according to Cameron (2000), are more likely to be vetoed than other topics. Graph 1 summarizes the information:
Source: CEBRAP Legislative Database.
Graph 1: Vetoes in Brazil (1995 - 2010) in Welfare Policy per Category
Of the 37 bills introduced by the Executive that were vetoed, 28 had the veto appreciated. These 28 are partial vetoes, of which 12 were PL (bills), 15 were MPV (executive orders) and 1 PLP (budgetary bills). Unappreciated3 partial and total vetoes do not interfere in our analysis since they were not voted in Congress, but they are included in Table 2. There is little interest of both powers in vetoing budgetary bills. Since in Brazil the budget is mostly controlled by the Executive with little space for negotiation, as stipulated by the law, there would be no need to do so.
Table 2: Number of Vetoes of Bills Introduced by the Executive and Legislative Powers
| Legislative | Sustained partial | Unappreciated partial | Sustained Total | Unappreciated Total | Executive | Sustained partial | Unappreciated partial | Sustained Total | Unappreciated Total |
|---|---|---|---|---|---|---|---|---|---|
| PL | 28 | 4 | 12 | 9 | PL | 12 | 3 | 0 | 0 |
| MPV | 0 | 0 | 0 | 0 | MPV | 15 | 4 | 0 | 0 |
| PLP | 0 | 1 | 0 | 0 | PLP | 1 | 2 | 0 | 0 |
| Total | 2 | 5 | 12 | 9 | Total | 28 | 9 | 0 | 0 |
Source: CEBRAP Legislative Database
Interestingly, most vetoes made by the Legislative power were sustained and 30 % (12) of the vetoes made to laws introduced by the Executive as PLS (bills) were total and 70 % were partial vetoes (28). As for the Executive vetoes to legislation introduced by the Legislative power, there were no total vetoes, neither sustained nor unappreciated. All vetoes, totaling 12, were partial.
Conclusions
The hypothesis presented by this paper has turned out to be correct. Multi-party and bipartisan systems have produced similar results in the two countries concerning presidential success and government-opposition conflict on welfare legislation. However, there are some limitations to the findings presented here. It is important to bear in mind that this work does not apply to a generalization of the Brazilian or American systems, but it is a significant contribution to understand how the decision-making process results are similar, due to the high rates of legislation approval.
Brazilian presidents had almost all the welfare legislation on welfare they initiated approved. Lula da Silva achieved a 100 % success rate, while Cardoso had 99.16 %. American presidents were not as successful as their Brazilian counterparts, but their success rate on welfare legislation was also high: Clinton achieved 76.2 %, while Bush got 75.9 %, almost the same amount. That being said, it is possible to conclude that presidential success remained within the same range in both the countries, although there is a significant difference between Brazil and the United States.
This difference is related to how much the president’s coalition controls the lower chambers. It is important to note that the fact that the Brazilian president can veto legislation may be minimizing the effects of the multiparty system in the country, which, in the American case, might have the opposite effect. While in Brazil vetoing may have a centrifugal effect, in the United States it might cause a centripetal one. At the same time, in Brazil, presidents have managed to form broad coalitions in the chamber, gaining the support of about 70 % of its members; in the United States, presidents can rely only on their own parties. In the period studied here, the American president never had more than 60 % of support in the House. The support of the majority of both Brazil and U.S. chambers was fundamental to presidential success on welfare legislation. In Brazil, this meant absolute control of the president over the chamber. In the United States, the presidential success rate was 41.1 % higher when the president’s party had control of the majority than when it did not.
Concerning the conflict between government and opposition parties in both countries, there are important differences between the two countries. In the Brazilian case, the polarization on welfare roll call votes was higher during Cardoso’s administration than Lula’s, probably partially due to the high level of cohesion in the opposition parties, especially the PT. In the United States, polarization was moderate most of the time, except for the legislature when Republicans took over the majority of the House. Therefore, Democrats and Republicans seemed to be strongly affected by the change in control of the House, which also affected party cohesion.
Another source of conflict among governmental branches is related to the president’s prerogative to veto legislation. A larger amount of legislation on welfare initiated by the legislative or even the executive4 was vetoed in Brazil than in the United States. However, the conflict seems to be greater in the United States. All of the president’s vetoes were maintained in Brazil, whereas in the United States one veto was overridden. Additionally, the low chamber attempted to override 70 % of the presidential vetoes in the U.S., all of them taking place when the opposition party had control of the majority of the House.
Although one would expect an underlying conflict between government and opposition parties in the lower chambers of Brazil and the United States to account for the high rate of presidential success, it is surprising that the party conflict rate on roll call votes and the conflict over the presidential vetoes proved to be more intense in the United States than in Brazil. This conclusion conflicts with the understanding of some authors, who claim that bipartisan systems tend to be more moderate than multi-party systems. On the other hand, it reinforces the findings of literature that has perceived the American party system as more polarized in recent decades.
Finally, all the findings of this study point toward the importance of a majority in the Legislative branch. For the president’s party, it guarantees presidential success. For the opposition party or parties, it allows them to impose an agenda that defies the president’s positions. Future research ought to deepen the qualitative analysis of the ideological positions of legislators and political parties in relation to welfare legislation and analyze the motivations and interests of political actors in relation to presidential vetoes.
