Political Parties, Legislatures,
and Presidents

This chapter is the first of three that examine the role of a range of actors in the policymaking process. It focuses on a group of actors that are central to policymaking in democratic systems: political parties, the legislature, and the president. These form the inner core of actors, with important roles in democratic representation, in framing the policy agenda, and in formulating, adopting, and implementing policies. Clearly, their roles in these stages of the process are not exclusive. They share these roles with other actors with a formal presence in the constitution, such as the cabinet, the judiciary, and the bureaucracy, as well as with a wide range of actors with more informal roles, such as business groups, unions, the media, social movements, and international organizations.

The operation of the democratic system and the characteristics of policy outcomes are importantly influenced by the interactions among factors related to the nature of the political party system, the structure and functioning of the legislature, and the constraints and incentives facing presidents. Chapter 3 begins by highlighting the importance of examining the interactions among these key actors and the rules of the game.

Key Actors and Their Interactions

A useful point of departure relates to the interaction between two actors, the president and the legislature, that play a leading role in the game of policymaking as it is played in one of the most important arenas, congress. The discussion begins by looking at the ability of the president to pass his policy agenda through congress. It is natural to begin by centering the analysis on the role of presidents for a number of reasons. First, presidents have in fact acted as agenda-setters during most of the recent reforms undertaken in Latin America. Second, the rules of legislative engagement in most Latin American countries formally endow presidents with prerogatives that help them lead the policymaking process. Third, because they are elected by nationwide constituencies, presidents tend to have more “encompassing” interests in national public goods than legislators. (Presidents may also, however, have incentives that may cause them to stray from the public interest, a matter returned to below.)

The inability of presidents to pass their agendas has been at the heart of a raging debate on the merits of presidential democracies versus parliamentary systems, which fuse the selection and governing responsibilities of the executive and legislative branches. Scholars have expressed concern about the capacity of presidents to govern under divided government: that is, when their party does not control the legislature.1 The problem is much more than an academic concern. Latin America has experienced several interruptions of democratic governance that have resulted, at least in part, from the inability of presidents and legislatures to agree on policies.

The debate in political science regarding the problems of presidential democracies under conditions of divided government has been cast mainly in terms of its impact on political outcomes: in particular, on the stability of the democratic regime itself. Although this is a very important concern, this report takes a different focus: the impact on the quality of public policies. In particular, the ability of the president to pass his agenda is a very important determinant of one of the key features of public policies emphasized in this report: policy adaptability.

What determines the ability of the president to pass his agenda? In presidential democracies, one key ingredient is the share of seats controlled by the governing party. If the president’s party controls a majority of legislative seats in congress, then it will be easier for him to pass his agenda, assuming he can retain the support of his own legislators. As a result, public policies are likely to be adaptable, given that the transaction costs associated with policy change are likely to be relatively low. The size of the president’s contingent, in turn, depends on the nature of the party system and electoral rules, among other things.

What if the president’s party does not control the legislature? In that case, presidents have alternative ways of advancing their agendas. They can form a coalition based on the stable support of other parties, for instance. Or they can use different bargaining chips to persuade legislators to support particular legislative measures on an ad hoc basis. These rewards include appointments to political offices, policy concessions or changes that benefit legislators’ constituencies or parties’ bases of support, local investment projects, budgetary transfers, and/or public employment and governmental contracts. These options are not mutually exclusive, as even in the case of stable coalitions, presidents need to use different inducements such as cabinet positions or policy concessions to form coalitions, or to keep them together.

Coalitions also vary in their degree of formality, cohesion, and durability. Some are very strong and long-lasting, as in Chile, where coalitions function in important respects as parties and have provided the government with a reliable majority in the lower house (see Chapter 7 for a detailed discussion of policymaking in Chile). Some coalitions, such as those in Ecuador, are much more informal, ad hoc, and short-lived. They are often based on exchanges of various types of immediate and targeted rewards for short-term legislative support. As a consequence, in Ecuador coalitions have not provided presidents with a stable base of legislative support and thus have undermined policy adaptability and stability. Between these two extremes lie coalitions, such as those in Brazil and Uruguay, that may be formalized through participation of multiple parties in the cabinet and/or through agreements on core policies. On occasion, these coalitions do not last the entire presidential term or do not provide as firm a guarantee of support.

The stability of coalitions varies across countries and across administrations, and it depends at least in part on the extent to which there is a match between what the president is able to offer legislators and what legislators want. What a president can offer depends on factors such as his appointment powers, his agenda-setting powers, and his budgetary powers. What legislators want (whether policy concessions, local investment projects, or transfers for their constituents) depends on factors shaping their incentives, including electoral rules. For example, where there is a strong electoral connection between legislators and voters (that is, where voters have a great degree of influence on who gets elected), legislators may want local investment projects for their communities. Such is the case in Brazil, where the distribution of such projects has been an important inducement used by presidents to gather political support.2

The potential of such exchanges to sustain coalition support also depends on the existence of commitment mechanisms that ensure that the president and parties (or individual legislators) honor their promises. In Paraguay, for example, the absence of commitment mechanisms has led to a different kind of reward: public employment. Not surprisingly, the quality of the bureaucracy in Paraguay is greatly undermined by the use of public employment as a patronage resource.3

One thing that should be clear from this discussion is that different institutional configurations may help or hinder a president in passing his agenda. Passing the president’s agenda is possible within a two-party system with majority rule; under a system of strong and stable coalitions, as in Chile; or even under very fragmented party systems, if and when there is a good match between what legislators or legislative parties need and what the president can commit to deliver.

However, when the political system fails to provide the president and his administration with the ability to pass the president’s agenda, the consequences may be very serious, even leading to interruptions of democratic governance (see Chapter 6).4 And such a failure may also lead to a lack of adaptability of policies, implying an inability to move away from failed policies, to adjust in response to shocks, and/or to pass welfare-enhancing reforms, even if these have been clearly identified.

While the inability of presidents to pass their agendas can have harmful political and policy consequences, an institutional setting that allows the president to pass his agenda is not a sufficient condition for good public policies. The discussion that follows highlights two problems that may arise, even when the president has the tools to get what he wants. To tie the discussion to this study’s general framework, the discussion links each of these problems to one of the features of public policies introduced in Chapter 2.

Public-regardedness. Are public policies better if the president gets what he wants? Not necessarily. Because presidents are elected by national constituencies, they tend to have more encompassing interests in national public goods than legislators, who tend to favor their local constituencies. However, presidents do not always act like the idealized social planners described in economics textbooks, who seek to maximize social welfare. While they care about the public interest, presidents, like any politician, also have personal and political ambitions.

What determines the president’s incentives? Among other factors, the nature of the party system may play an important role. For example, if the workings of political parties are dominated by clientelistic politics, presidents and other politicians will care more about generating the resources needed to maintain the clientelistic system that is the basis of their political support, and less about whether policies are “good” or “bad” from the standpoint of the longer-term public interest. If the president is thought of as a social planner, anything that stands in the way of passing his agenda will reduce the quality of policy outcomes. Once one departs from the naive social planner view, then checks on one’s ability to pass one’s agenda will be seen to involve trade-offs. These checks can reduce the extent to which the president caters to narrow interests, but they may lead to less decisiveness and less adaptability.

Like the president, members of congress can also seek to derive private or narrow benefits from their public role in the policymaking process. Various types of institutional checks and balances, including those between the executive and legislative branches, can limit the scope for all players involved to cater to narrow interests.5 Additional actors that can enforce more public-regarding policy choices and policy implementation include an independent and impartial judiciary and a professional and politically neutral civil service. The role of these actors is considered in Chapter 4.

Policy stability. The ability of the president to pass his agenda may also lead to policy instability under some circumstances. Consider a hypothetical country with a two-party system. Each of the parties has stable support of close to half of the electorate. As a result, they alternate in power. Both parties are programmatic, so they compete primarily on the basis of distinctions in their policy ideas and proposals, which tend to be stable over time. Electoral rules tend to produce majority governments and disciplined parties, so policies are adaptable. Some analysts would consider this to be a favorable scenario for presidential democracies. Yet even this rosy scenario of stable parties with stable preferences does not guarantee a very important feature of public policies: policy stability.

If, for example, one of the parties is on the Right and the other on the Left in ideological terms, policies could shift from one extreme to the other every time there is a change in administration.6 Trade openness and privatization could be followed by increased protection and nationalization of assets; primary education could shift from a focus on public provision to private; and so on. The inefficiencies associated with policy volatility are obvious. While those on the Right and the Left may disagree on their preferred policy options, they would probably agree that cycling between extremes represents the worst possible scenario. The question is how to make sure that the parties “compromise,” engaging in inter-temporal agreements to adopt an intermediate policy that is acceptable to both, rather than adopting their preferred one-sided option every time they gain access to power.

Now that it has been established how important it is to examine the interactions among actors and their underlying characteristics to understand the policymaking process, the rest of the chapter will focus in greater detail on each of these central political actors. The purpose is to highlight the potential roles that the actors play in the policymaking process and what characteristics of the actors may affect the extent and nature of their role, as well as to describe some of the characteristics of these actors across the Latin American region.

Political Parties 7

Political parties are key players in the policymaking process, as they are in the workings of a democratic system more generally. An indispensable element of democracy is the holding of regular electoral contests to allocate control of governmental positions and legislative seats. Parties recruit candidates for such offices, mobilize electoral support, and define their policy aspirations, should they win a place in the government. Apart from their direct roles in elections, parties play other key roles in forming governments, organizing the work of the legislature, and articulating and aggregating citizen interests and preferences.8

More related to the topic of this report, the policymaking process in a country is strongly influenced by the structure and organization of political parties, both directly and indirectly. For instance, in some countries, parties—even when they are out of power—are important actors in defining and articulating broad policy programs and can engage effectively in public policy debates, on occasion with the help of affiliated think tanks (see Chapter 5). Characteristics of the party system also affect the policymaking process somewhat more indirectly: influencing, for example, the workability of executive-legislative relations; the possibilities for coordination in congress; and/or the incentives of elected officials to cater to narrower or broader sets of societal interests.

Not only do the characteristics of the party system interact with one another, they also interact with other institutions and actors, such as the presidency, the legislature, and the judiciary. Thus the expected effects of party system (and other institutional) characteristics highlighted in this part of the report may not be observed in all countries, given that the particular impact of each institutional feature depends upon its interaction with other institutional characteristics of the country.

In the discussion that follows, several party system characteristics and their expected impact on the policymaking process are considered: their degree of institutionalization, degree of fragmentation, and degree of nationalization.

Party System Institutionalization

A first important characteristic of party systems is their level of institutionalization. Party systems can be considered institutionalized when:

• Patterns of inter-party competition are relatively stable.
• Parties have reasonably strong links with organized societal interests, and citizens tend to be attached to particular parties.
• Parties and elections are perceived as decisive in determining who governs.
• Party organizations are well developed and influential in shaping the policy directions and leadership of the party.9

Institutionalized party systems are more likely to promote greater policy consistency over time and a greater potential for inter-temporal agreements. This is because, from one election to the next, the partisan identity and relative political importance of the main players in the policymaking game in the executive and the congress are less likely to change dramatically. Moreover, because parties play a strong role in political recruitment, and political elites strive to promote and protect the value of the party label (which implies maintaining relatively consistent policy positions over time), institutionalized parties are less likely to change their basic policy stances very often. In addition, it is more likely that elected presidents will enjoy some degree of partisan support in the legislature and/or will be better able to build somewhat stable party-based coalitions.

Institutionalized party systems tend to be programmatic. That is, they compete and derive support on the basis of differences in their policy orientations and achievements. But institutionalized parties—parties with relatively durable bases of political support—can also be clientelistic. In such cases, they compete for and obtain support based on the distribution of selective benefits to voters (such as public sector jobs, governmental contracts, cash, or meals) and are judged by voters primarily on their ability to deliver these benefits.

Parties that are institutionalized and programmatic are likely to improve democratic accountability by enhancing citizens’ ability to select parties and candidates that best match their policy preferences. Incentives should also be stronger for investments in public goods, such as education and well-functioning public utilities, and for policy reforms aimed at advancing broader public interests.

By contrast, policymaking in clientelistic systems is likely to be constrained by the need to maintain the parties’ clientelistic system of support. Targeted government spending on public works is likely to be favored over investments in more wide-reaching public goods. Implementing policy reforms that can undermine the pillars of the clientelistic system may necessitate making large side payments to clients that have been favored by the party in the past.

To gauge the extent and dimensions of party system institutionalization, this study used an aggregate index of party system institutionalization developed by Mark Jones.10

• The stability of inter-party competition is measured on the basis of measures of the volatility in party shares of votes and legislative seats between recent elections.
• The extensiveness of parties’ roots in society and the legitimacy of parties and elections are measured on the basis of responses to questions in the Latinobarometer public opinion survey and a survey of legislators.11
• The strength of party organizations is gauged based on a measure of the age of significant parties and responses to a question from the survey of legislators, asking them to rate the permanence and strength of their party’s organization.12

To analyze the extent to which party systems are programmatic, this study used another index constructed by Mark Jones. This index attempts to capture the extent to which parties develop distinct approaches to policy and the extent to which citizens are aware of these differences. But it does so indirectly, by examining legislator and citizen perceptions of differences between parties in Left-Right ideological terms, rather than directly observing whether parties’ policy programs and positions adhere consistently to distinct ideological perspectives.13

Figure 3.1 plots the countries according to their scores on the two aggregate indices. According to these measures, parties are relatively institutionalized and programmatic in Uruguay, Nicaragua, El Salvador, and Chile.14 They are weakly institutionalized and least programmatic in Guatemala, Ecuador, and Peru.

Figure 3.2 shows the marked contrast between Argentina and Chile in the degree to which parties are programmatic, as measured by one component of this index: legislator perceptions of differences between parties on the Left-Right ideological scale.

Party System Fragmentation, Presidential Contingents,
and Party Discipline

Several other party system characteristics interacting with one another also have an important influence on the interaction between the executive and legislative branches. These are the level of legislative fragmentation; the size of the presidential legislative contingent; and the extent of party discipline (that is, the degree to which members of the president’s party in the legislature are responsive to the instructions of the party leadership and/or president).

Party system fragmentation and the size of presidential legislative contingents, in turn, are influenced by characteristics of the electoral system. Proportional electoral systems, in which multiple members of congress are elected per electoral district, encourage more fragmented party systems than plurality systems, in which only one legislator is elected per district. Among proportional systems, greater party system fragmentation is more likely in systems with larger districts (that is, more legislators elected per district) and those with more equitable formulas for translating the vote shares of parties into shares of legislative seats.

Other aspects of the electoral system are also relevant. Always holding legislative elections at the same time as presidential elections favors a greater concentration of votes and legislative seats in the president’s party than when such elections are held separately, or when legislative elections are held in the middle of the presidential term. The system used for electing the president can also have indirect effects on party system fragmentation. Majority run-off systems 15 tend to discourage the formation of inter-party alliances and favor the proliferation of parties, while plurality systems tend to encourage a concentration of votes in fewer parties.

The share of seats controlled by the governing party, the degree of party system fragmentation, and the cohesiveness of the governing parties have an important impact on the ability of the executive to pass its agenda. They also affect the nature of the bargaining game and the types and loci of trades that the executive engages in to gain legislative support. For instance, if the president controls a majority of seats in the legislature and can enforce discipline within this contingent, he is more likely to choose to govern through statutes (laws passed in the legislature) and to use partisan levers of influence to maintain support. But if the president’s legislative contingent is relatively small and the party system is fragmented, the executive must attempt to assemble some type of legislative coalition. These arrangements can range from being a fairly stable coalition among parties to a shifting coalition of parties and even individual legislators. In this case, the president will likely need a broader array of inducements, including offering cabinet or other governmental positions, making policy concessions, or promising favored treatment in allocating public resources and contracts. If the president expects gaining legislative support to be problematic (because stable coalitions cannot be formed), or too costly (in terms of the president’s policy agenda or in terms of public resources), then he may choose to govern unilaterally through his constitutional legislative powers (decree or agenda-setting powers) or “paraconstitutional” powers (such as discretionary powers to shift funds within the budget or use unilateral rulemaking authority extensively).16

Consistent governmental majorities and disciplined parties can promote policy adaptability. While such a scenario may promote governmental “decisiveness,”17 it may, in the absence of effective checks on executive authority, result in deficiencies in the public-regardedness of policy outcomes. In addition, policy volatility can result, if power tends to alternate among disciplined and relatively polarized parties.

On the other hand, party system fragmentation, especially if very high and combined with polarization, can complicate executive-legislative relations, making it difficult to obtain stable support for the executive agenda. This can lead to gridlock or prompt presidents to pursue unilateral strategies, which can endanger the durability of democratic institutions.

The level of fragmentation in the legislature directly influences the size of the presidential legislative contingent. It also affects the number of partners with which the president must form some type of legislative coalition or structure piecemeal alliances specific to individual pieces of legislation to implement his policy agenda. The level of legislative fragmentation in the lower house (or national assembly) is measured in Table 3.1, using an index of the effective number of parties based on the two most recent legislative elections in each country.18 Brazil, Ecuador, and Bolivia are in the fragmented extreme, with a multitude of political parties winning legislative seats. By contrast, politics in Chile, Honduras, and Nicaragua tends to be dominated by two political parties.19

Table 3.1 also shows the average percentage of seats held by the president’s party in the lower house (or national assembly) in the two most recent legislative elections, as well as an index of the proportionality of the design of the electoral system. In several countries, including Chile, Nicaragua, Paraguay (lower house), and Honduras, the president’s party has typically enjoyed a majority of the legislative seats. In others, including Argentina and Uruguay, the president’s contingent, while not reaching a majority, has been reasonably high (above 40 percent). In several countries, however, the size of the president’s legislative contingent has been very low, including Brazil, Colombia, Ecuador, and Bolivia.

From Table 3.1 it is clear that there is some correlation between party system fragmentation and the proportionality of the electoral system, but that correlation is imperfect. This is likely due to the fact that the structure of party systems is also affected by longer-term historical factors, as well as the salience of socioeconomic, ethnic, and geographic cleavages.

There is considerable debate in the scholarly literature regarding the consequences for governance in general and policymaking in particular of presidential legislative contingents that are below majority status.20 While some observers consider small presidential contingents (especially those that are around 33 percent and below) to be problematic for democratic governance, others do not. However, there is more agreement that, in instances where the president’s party lacks a majority of the seats in the legislature (or does not at least approach a majority, with at least 45 percent of the seats), the president must form some type of legislative coalition in order to be able to govern effectively. Where coalitions are not formed and sustained, governance problems are likely to emerge.21

The increasing fragmentation of party systems and the decreasing probability of single-party majorities (reflected in Table 3.1) has meant that the viability of governments increasingly depends upon the formation of some form of coalition government. Whether a stable majority government can be formed, on the basis of either a single party or a coalition of parties, is likely to be a key factor in shaping whether inter-temporal agreements among political actors are possible and whether policies are adaptable and stable over time and implemented effectively.

From 1990 to 2004, in a sample of 18 Latin American countries, only 20 percent of presidential-congressional periods22 were characterized by single-party majorities—or 36 percent, if cases in which the governing party has a near-majority are also included.23During the period, single-party majorities (or near-majorities) were common in such countries as Argentina, Costa Rica, Honduras, and Mexico. Stable, or at least relatively stable, coalition governments were common in Brazil, Chile, and Uruguay, among others. By contrast, minority governments were especially common in Ecuador and Guatemala (Table 3.2).

But it is an oversimplification to focus only on the share of seats controlled by the governing party or coalition, since the degree to which parties act as a cohesive block in the legislature varies considerably across countries and even across parties within the same party system. The ability of presidents (and the central party leadership) to enforce discipline within their party’s ranks in the legislature has an important effect on their success in advancing their legislative agendas.

The level of responsiveness of legislators to national party leaders is affected by incentives provided by the electoral system and the process of nominating candidates, as well as by other forms of reward and punishment that might be utilized by party leaders or the president. This study uses a Party Centralization Index to capture how responsive legislators are to national party leaders (Figure 3.3).24 The index centers mostly on the sources of leverage that party leaders have that stem from the electoral system. But there are other incentives, such as leaders’ control over the appointment of legislators to committees and control over the legislative agenda.

The first component of the index focuses on the extent to which national party leaders or individual candidates are responsible for determining who is able to run for office and their position on the electoral list. When national party leaders largely determine whether legislators can run for reelection and their position on the party list, then legislators have greater incentive to conform to the party line in the legislature.

A second dimension of the index focuses on core features of the electoral system, including the territorial dimension of electoral districts (national, regional, single-member, or some mixture thereof) and whether voters may choose among individual candidates instead of only among pre-selected party lists. Centralization is greater where legislators are elected in a single national district or large districts with closed party lists.

Four other factors affecting the degree of party centralization are included in the index: whether presidential and legislative elections are held concurrently; the degree of autonomy of subnational authorities (in particular, governors); the extent to which parties are internally democratic; and the extent to which presidential candidates have been selected through primaries. Parties will be more centralized, and national party leaders more powerful, when elections are concurrent; when governors are not competing with national party leaders for the control of legislators; and when the power of the national leaders to nominate candidates is not diluted by primaries or other forms of intra-party democracy.

Values for the Party Centralization Index for various Latin American countries are shown in Figure 3.3. Presidents would be expected to have the most influence on legislators from their own parties in Bolivia, Guatemala, Nicaragua, Honduras, and Costa Rica. The president should not expect especially disciplined legislators in Chile, Brazil, and Argentina, and one would expect coalitions in those countries to be formed primarily through negotiation with regional/factional leaders of political parties or with individual legislators.

Party and Party System Nationalization

A final characteristic of party systems that can affect the functioning of democracy and policy outcomes is their degree of nationalization. A nationalized party system implies that parties are generally national in scope and that they tend to speak and act with a common national orientation, rather than being divided according to regional or subnational issues and focused upon them. In highly nationalized party systems, national issues are likely to be central to legislators’ careers. Executives might have a greater ability to forge legislative coalitions centered on national issues, given the need to negotiate with only a few key national party leaders. Under conditions of weak party nationalization, the central party leadership may be less able to speak for the entire party and to deliver its legislative support.

Differences in nationalization are also likely to have public policy consequences. Where a party’s base of support is relatively constant across geographic units, it will be more likely to treat all units equally with respect to decisions in relation to such matters as transfers to subnational units, administrative reform, public investments, and subsidies. By contrast, where its support varies widely across geographic units, it will be more likely to base its decisions in part on the degree of electoral support it receives in specific geographic units. Under a nationalized party system, public policy is likely to be more oriented toward working for the national common good.

A partial approach to assessing the nationalization of party systems is to examine the distribution of the popular vote for parties across territorial jurisdictions of the country.25 Figure 3.4 shows the scores on an index of party system nationalization for Latin American countries constructed on the basis of votes in the lower house elections held closest to 2002. On the basis of this measure, parties would appear to be most nationalized in Honduras, Chile, Uruguay, and Nicaragua and least nationalized in Peru, Argentina, Venezuela, and Brazil.

Legislatures 26

Legislatures are critical institutions in the effective functioning of a democratic system and in the policymaking process. Legislatures are expected to represent the needs and wishes of citizens in policymaking; identify problems and formulate and approve laws to address them; and oversee the implementation of policies by monitoring, reviewing, and investigating government activities to ensure that they are transparent, efficient, and consistent with existing laws and regulations.

The extent and nature of the role played by legislatures in the policymaking process vary greatly from country to country. At the more proactive and constructive end of the spectrum, legislatures such as the U.S. Congress are able to develop their own legislative proposals and thus participate along with the executive in directing the policy agenda. Given their policy capabilities, such legislatures are also likely to be active and effective in overseeing policy implementation.27 At the other end, legislatures may be fairly marginal players, serving as a rubber stamp on the executive’s legislative proposals and having little capacity or willingness to scrutinize the conduct of government.28 Between these two extremes, there is a wide area in the middle where legislatures can exhibit different degrees of activity either in simply blocking much of what the executive proposes or in reformulating and/or amending executive initiatives. Among such legislatures, there can also be considerable variation in the intensity and effectiveness with which they perform the oversight role.

How the legislature plays its policymaking roles can have an important effect on the nature of policy outcomes. If the legislature is a marginal actor, this will give the executive free rein to enact policy changes that it perceives to be necessary. But the lack of legislative deliberation as policies are formulated and the weakness of oversight may mean that the policies adopted are poorly conceived in technical terms, poorly adjusted to the real needs or demands of organized interests and citizens, lacking consensus and therefore politically unsustainable, and/or inefficiently or unfairly implemented. On the other hand, legislatures that are involved more heavily in policymaking in a constructive sense can contribute to the adoption of policies that are more sustainable because they are based on a broader social and political consensus and are more carefully scrutinized in technical terms. In addition, in a constructive legislature the effective oversight of policy implementation should increase the likelihood that policies fulfill their intended objectives rather than being carried out for the benefit of particular individuals, groups, or sectors.

Legislatures with limited capacity to play a constructive role in policymaking may nonetheless be important players in the sense of obstructing or vetoing much of what the executive proposes. Such legislatures have many of the potential negative traits of more marginal legislatures in regard to policymaking, and they may also prevent the executive from advancing a positive agenda of policy reform. Given their limited capacity, such legislatures are also unlikely to play an effective role in overseeing the implementation of policies.

Against the backdrop of the region’s history of dictatorial rule and presidencialismo, scholars had tended to consider legislatures in Latin American countries to be largely irrelevant throughout much of the 20th century and not worthy of study in and of themselves. Some prominent experiences in the past two decades, such as the closing of the legislature by President Fujimori in Peru and the frequent use of decree powers by President Menem in Argentina, continued to reinforce the commonly held view that Latin American legislatures often abdicate (or are forced to abdicate) their constitutional prerogatives to the executive. But recent studies suggest that, while legislatures in the region in general may not be heavily involved in formulating and advocating policy change, they are nonetheless relevant to policy outcomes. Legislatures in some countries are active in policymaking in the sense of mainly being blunt veto players, blocking legislation proposed by the executive. Others, however, are involved in negotiating policy issues behind the scenes with the executive or in amending or reformulating executive legislative initiatives.

That legislatures in the region do not exist solely to rubber-stamp executive decisions is evident from data on the success rates of executive legislative initiatives. As is evident in Table 3.3, the rate of approval of executive initiatives varies from a low of 41 percent in Costa Rica from 1986 to 1998 to a high of 96 percent in Mexico from 1982 to 1999.29

But such raw measures are of limited value in assessing a legislature’s full influence on policymaking. Aside from proposing or killing legislation, legislatures can approve bills with extensive amendments. They can also exert influence, outside the formal legislative arena, through bilateral negotiations between legislative leaders and executive officials, as to which bills get introduced and the form that such legislation takes. In addition, the executive, not wanting to face the humiliation of a legislative defeat, can anticipate the legislature’s reaction in the way that it designs policy proposals.30 Thus the task of appraising the legislature’s policymaking role in any given country is a very difficult task that requires detailed study of individual cases.

The nature of the role that legislatures play is likely to influence the way that citizens view them. At the same time, the level of citizen trust in congress affects the likelihood that investments can be made in building its capacity. In addition, in cases in which congress has little credibility, it is likely to be less effective in representing societal interests, and the executive will have a greater incentive to seek to bypass or minimize the legislature in the policymaking process.

As seen in Table 3.4, neither the general public nor international business executives have a high degree of confidence in the congress in most countries of the region. On average over the past decade, according to the Latinobarometer, the general public has the most favorable view of congress in Uruguay, Chile, Honduras, and Costa Rica and the least favorable view in Ecuador, Bolivia, and Guatemala. By contrast, the average ratings given by business executives, as reported by the World Economic Forum, are highest in Chile and Brazil and lowest in Venezuela, Nicaragua, and Argentina. The most important differences in the views of the general public and business executives are for Venezuela and the Dominican Republic, where in each case the general public has a comparatively more favorable view than business executives.

Factors Affecting the Legislature’s Role in Policymaking

The role of the legislature in the policymaking process is shaped by a variety of factors, including the extent of its constitutional powers relative to the executive branch; the balance of partisan forces in the executive and legislative branches; the structure, organization, procedures, and technical capacities of the legislature; and the goals of its members, derived from electoral-based and party-based incentives.

Macro-Structure of the Legislature

A first characteristic that may affect the role played by the legislature in policymaking is its unicameral or bicameral structure. Nine countries in the region—Argentina, Brazil, Bolivia, Chile, Colombia, the Dominican Republic, Mexico, Paraguay, and Uruguay—have bicameral legislatures. Depending upon such factors as how the members of the two houses are elected and the balance of legislative powers between the two chambers, a bicameral legislature can provide a separate veto point in the policymaking process and also affect the extent to which territorial interests are represented. For instance, a senate elected from a single national district concurrently with the president and on a single ballot (as in Uruguay) is less likely to act as an additional veto point and does not dramatically change how territorial interests are represented. But when senators are elected separately from the president on the basis of provincial districts (as in Argentina), and representation is not tied to population, then there is a greater possibility that the upper house can become a separate veto point and accentuate the extent to which regional interests are represented in policymaking. Given the complexity that consideration of the senate adds, and the relative lack of comparative information, the analysis that follows focuses mainly on the lower house, or national assembly in the case of unicameral congresses.

Constitutional Powers

Although all the Spanish- and Portuguese-speaking Latin American countries considered in this report have adopted a basic presidential form of government similar to that of the United States, considerable differences exist with respect to the relative powers the constitution assigns to the executive and legislative branches. Several Caribbean countries and Suriname and Guyana, on the other hand, use either parliamentary or semi-presidential systems, which entail differences in the nature of the relationship between the executive and legislative branches and the expected role that legislatures play in the policy process. Box 3.1 underlines some of these differences through a discussion of the case of Jamaica.

Two fundamental characteristics of presidential systems distinguish them from parliamentary systems: the head of state is elected separately from the congress, and the terms of the president and congress are fixed. In relation to these core features, the only notable deviation among this set of countries is Bolivia, where congress has the responsibility of choosing among the leading two vote-winners in the presidential race if no candidate obtains an absolute majority in the first round.

Once one looks beyond these two defining characteristics, there are important differences among countries. The relative powers that constitutions assign to presidents and legislatures are a key factor in shaping the policymaking role of the congress.

Though presidents have the power to appoint and remove cabinet ministers in all Latin American countries,31 in some cases, such as Colombia, Peru, Uruguay, and Venezuela, legislatures also have the power to remove them through censure procedures. Given the difficulty of obtaining the majorities required for censure and the president’s full discretion in naming a successor, this power has not been used to a significant extent in most countries. However, the power of censure may still act as a constraint on the president’s discretion in controlling the composition of his cabinet.

Aside from appointment powers, constitutions grant presidents other tools with which to insert themselves into policymaking. Generally speaking, the stronger and more diverse these powers, the more constrained the legislature is likely to be in undertaking an active and effective role in policymaking and developing its capabilities.

The presidential powers that contribute to the president’s ability to unilaterally change the status quo can be referred to as proactive powers.32 These powers include decree powers, agenda-setting powers, and budgetary powers. Reactive powers, by contrast, allow the president to preserve the status quo against efforts by the legislature to change it. These powers include package veto and partial veto powers, and exclusive powers to initiate legislation in given policy areas.

In several countries the constitution grants presidents the power to enact new legislation by decree, even without the legislature first delegating this authority. This authority is applicable across most policy areas in Argentina, Brazil, and Colombia, but limited to economic matters in Ecuador and fiscal matters in Peru. Although in most cases congress has the authority to rescind the decree, this power nonetheless helps the president control the legislative agenda and obtain outcomes that would otherwise not be possible. For instance, in Brazil, the president can legislate through provisional decrees (Medidas Provisórias), which need to be ratified by the congress within 60 days to remain in effect. If a provisional decree is not acted on within the first 45 days, it is automatically sent to the top of the legislative agenda. If the congress does not approve the provisional decree in this first 60-day period, the president can reissue the provisional decree, but only once.

Constitutions also grant many presidents in the region important agenda-setting powers. For instance, presidents in Brazil, Colombia, and Uruguay can declare a legislative proposal “urgent,” thereby requiring congress to act within a set timeframe. In Uruguay, a bill becomes law if the congress does not act within the allowed timeframe. Another form of agenda-setting power, found in Brazil and Chile, is the president’s ability to convene a special legislative session in which only those initiatives set forth by the executive can be debated.

In many countries, the legislature’s role with respect to budgetary policymaking is curtailed by restrictions on the legislature’s authority and by special presidential prerogatives. For instance, in Chile the executive sets the spending limits in the budget and has the sole responsibility for estimating revenues. The legislature cannot increase expenditures for any budget item or introduce amendments that increase total spending; it can only reduce expenditures or reject them. Moreover, if congress fails to approve a budget law within 60 days, the executive’s original proposal becomes law.

The constitutions of many of the region’s countries also provide presidents with mechanisms to prevent attempts by the legislature to change the status quo policies without the president’s assent. The package veto, in which presidents can refrain from signing an entire bill approved by the legislature into law, is common in many presidential systems, including the United States. But many Latin American presidents are also given the power to reject individual items of bills approved by the legislature.

Another form of reactive power is when the president is given the exclusive authority to initiate legislation in some policy areas. For example, in Colombia, this restriction on the legislature applies to the structure of ministries, salaries of public employees, foreign exchange, external trade and tariffs, and the national debt, among other areas. Presidential legislative monopolies (that is, areas of exclusive initiative) are also fairly extensive in Brazil and Chile.

Finally, the power of presidents to put to a vote of the citizens general matters of policy or particular laws (plebiscite powers) can be valuable, even when not used, to put pressure on legislators to support the president. The magnitude of this power depends on the breadth of its applicability across policy areas and whether this power is also shared by congress.

As Table 3.5 shows, the overall legislative powers of presidents are greatest in Chile, Brazil, Ecuador, and Colombia. Proactive powers are also sizeable in Peru. Legislative powers of presidents are weakest in Nicaragua, Paraguay, Costa Rica, and Bolivia.33 While significant legislative powers give the president important levers for bargaining and shaping the legislative agenda, they usually do not substitute for the need for adequate partisan support. Decrees can be overturned, “urgent” legislative initiatives can be defeated, and vetoes can be overridden. Thus factors related to the party and electoral system are also key in shaping the legislature’s role and the nature of executive-legislative relations.

Party Dynamics and Electoral Incentives

As discussed, the likelihood that the president will be able to pass his agenda is affected by the typical size of the president’s legislative contingent, the degree of legislative fragmentation, and the extent of party centralization (or discipline). These factors are also likely to affect the role that the legislature plays in policymaking.

When presidents can count on the support of a disciplined majority party or coalition, the role of the legislature can be limited. For example, despite having relatively weak constitutional powers, Mexican presidents before 1997 dominated policymaking, since they could count on solid majorities for the governing Partido Revolucionario Institucional (PRI) in both houses of congress. Now that the currently governing Partido Acción Nacional (PAN) lacks a majority, the legislature has become a much more active player in policymaking and the success rate for executive initiatives has dropped considerably.

While a highly fragmented party system is likely to result in a more active legislature, it may tend to limit the legislature’s role to being mainly a veto player or a site of bargaining over particularistic expenditures—rather than an arena for proactive policymaking or effective oversight of the executive. Having a large number of parties, especially when they are internally factionalized, is likely to limit the possibilities for coordination over policy both within the legislature and between the executive and the legislative branches.

Differences in the extent to which parties are centralized and disciplined also entail trade-offs with respect to the legislature’s policymaking role. On the one hand, party centralization may help presidents secure support in the legislature and facilitate inter-party negotiations in the formation of governing coalitions, thus contributing to adaptability. Centralized parties that are also programmatic in orientation may encourage legislators to adopt a policy focus oriented toward national public goods, rather than a focus on the delivery of more targeted and narrow benefits.

On the other hand, high levels of party centralization are likely to limit legislators’ incentives and possibilities of responding directly to their constituents, as well as their incentives to participate independently in the policymaking process and in oversight responsibilities. Subservience to party leaders, especially when parties tend to be clientelistic, can contribute to a weak policy role for the legislature and weak incentives for legislators to invest in developing the capacities of congress. But while decentralized parties may encourage greater policy independence among legislators and more accountability of individual legislators to voters if parties are less cohesive, this can limit the ability of voters to hold representatives accountable on the basis of national policy positions and accomplishments and encourage an orientation among legislators toward satisfying narrow geographic interests.34

Election rules and the degree of centralization of candidate nomination processes can also affect legislators’ career ambitions and incentives, as well as their experience. Given the very high rates of reelection (around 90 percent) and fairly decentralized party structures of the U.S. Congress, analysts assume that legislators’ main motivating goals are to obtain reelection and to advance their careers in the legislature. By contrast, in Latin America, where the rate of reelection tends to be much lower, legislators typically have an incentive to work toward advancing a career outside the legislature (such as in national, state, or local government) and are also less experienced. Their career objectives are often furthered by satisfying party leaders rather than centering their attention on satisfying constituents’ interests and demands.

As shown in Figure 3.5, the rates of immediate reelection to the congress vary greatly across countries. In Chile and Uruguay, around 60 percent of legislators are immediately reelected; in Peru and Argentina, less than 20 percent return for a second consecutive term. In Costa Rica and Mexico (not shown in the figure), immediate reelection is not permitted, and only about 13 percent and 11 percent of legislators, respectively, are eventually reelected.

The complex interactions among various electoral rules and party system characteristics, rather than any particular factor, are what shapes the legislature’s policymaking role, as illustrated by the following examples.

In Argentina, local party leaders control the construction of the local party list. Thus legislators’ ability to pursue a legislative career independently is significantly restricted. Instead, they typically seek to continue their political career in other elective or appointive offices. As a consequence, Argentine legislators have a strong incentive to maintain a good relationship with their local party leaders. These leaders have a complex political objective: they want to maximize their party’s performance in their province, but at the same time they want to safeguard their position within the provincial party structure. The threat of challenge by popular legislators provides local party leaders with a strong incentive to reduce the national and provincial visibility of their local subordinates by rotating them among the various jobs the provincial party can offer. The electoral risks associated with nominating lesser-known candidates are mitigated by Argentina’s electoral rules, especially the use of party-supplied ballots and closed-list proportional representation. Voters tend to vote for the party list, not for the individuals on the list. A president’s ability to influence legislators of his own party thus depends in part on whether the provincial party leader supports the administration.35

In Colombia, the party system is characterized by high intra-party competition. In the last 15 years, small parties and movements have proliferated. At the national level, 45 movements/electoral lists obtained at least one seat in the lower house in the 2002 election, and more than 900 participated in the elections. The existence of high intra-party competition has weakened political leadership, ultimately fragmenting party organizations. Parties no longer have a means to control the career paths of local political leaders and candidates (see Chapter 7).

In Chile, the binominal electoral system, with two members elected per district, creates strong incentives for the formation of two electoral coalitions. Parties or electoral alliances can win the two available seats only if the winning list receives at least twice the total vote of the list that obtains the second-most votes. Given electoral incentives, legislators concerned with keeping their seats in congress know that dropping out of one of the main coalitions entails significant electoral risks. The imposition of this voting system in a country characterized by around five effective political parties has resulted in majority control of the Chamber of Deputies by the governing Concertación coalition since the return to democracy in 1989. Thus electoral system–based incentives have contributed to strong legislative support for bills initiated by the executive.

Organization of the Legislature

The legislature’s policymaking role is also affected by its organizational characteristics, which in turn are influenced by environmental factors, such as constitutional stipulations, party system dynamics, and electoral incentives.

Given the unwieldy size and lack of specialization of the full congress, if legislatures are to play an active role in shaping the content of policy and overseeing the executive, they must do so through capable committees. Most legislatures in Latin America have permanent committees with specific policy jurisdictions. Legislation is routinely referred to them before being taken up by the larger chamber. The roles of committees vary across countries depending upon procedural rules, the resources and degree of specialization of the committees and their legislative members, and broader political factors, such as the strength and organization of political parties.

The number of committees varies greatly and does not necessarily correspond to the size of the legislature. If there are a large number of committees relative to the size of the chamber, legislators may be required to serve on several committees simultaneously, which may limit their ability to concentrate their efforts and develop specialized knowledge. In addition, the efficiency of the legislature can be impaired if legislation is commonly sent to multiple committees because of overlap among committees’ policy jurisdictions.

Legislative rules also shape the size of committees, how members and committee leaders are selected, and the number of committees on which each legislator can serve. If committees are too large, this can limit their ability to function effectively. If committee memberships and leadership rotate frequently, this is likely to limit the degree of expertise that members develop and thus their policymaking effectiveness. To the extent that party leaders can exercise control over committee assignments and appointments to leadership positions, this can give them leverage in maintaining party discipline. In Brazil and Colombia, such prerogatives of party leaders to manage the committees, organize the legislative agenda, and direct public resources help impose some party discipline, despite electoral rules that allow or encourage legislator independence.36

In most Latin American countries, committee and leadership assignments are made on a partisan basis. The composition of the committees is expected to reflect the partisan composition of the legislature as a whole. Instead of allocating important committee assignments and leadership positions on the basis of seniority, as has been the practice in the U.S. Congress until recently, in most Latin American legislatures party leaders or party caucuses allocate these slots on the basis of other criteria, such as party loyalty.

The existence of an ample and competent staff to assist legislators with the tasks of administration, research and analysis, and document preparation is vital to enable committees to evaluate bills initiated by the executive and supervise policy implementation effectively. While the scope of committee staffing varies widely from one country to another, in most countries it is deficient relative to the roles assigned to the committees. For example, in El Salvador, each committee has only one technical assistant and one secretary, but they are expected to perform all three types of functions.

In a few countries, professional staff are available to assist legislative committees (and parties) with research and analysis. For example, in Brazil, a research office that has about 35 professionals assists the budget committee of the lower house. Chile has a (relatively small) legislative budget research office; several professional staff persons also advise the budget committee. In Colombia, a relatively large number of professional staff members assist the budget committee.

Grouping Legislatures According to Type

Despite the recent proliferation of research on executive-legislative relations in Latin America, comparative knowledge of how legislative institutions operate is still quite limited. Thus a detailed and empirically precise classification of Latin American legislatures in terms of their policymaking role is not possible at this time.

This section presents a tentative categorization of Latin American legislatures, based on available information, which is still far from complete. Legislatures can be grouped according to the nature of their policymaking role and the intensity with which they carry out that role.37 In part, the nature of the role is shaped by their capabilities, including the experience and qualifications of legislators, the strength and degree of specialization of committees, and the availability of professional support staff and research units. But as has been discussed, the role is also affected by electoral and party-based incentives and the balance of constitutional and partisan powers between the executive and legislative branches. In the categorization employed here, a major focus is on legislative capabilities, but a more qualitative assessment of the actual role performed by legislatures is also made.

• Relatively limited legislatures generally approve the initiatives of the executive with only minor changes and are not very active or effective in overseeing the executive.
• Reactive obstructionist legislatures are potentially more active in policymaking, but their role is primarily one of a blunt veto player, generally either blocking or approving executive initiatives. This type of legislature only rarely engages deeply in refining the technical or distributional character of the policies proposed by the executive or in actively supervising the implementation of policies.
• Reactive constructive legislatures may perform the roles that characterize reactive obstructionist legislatures, but they can significantly shape the content of policies by amending executive initiatives. They can also fulfill the oversight function somewhat effectively.
• Proactive legislatures, in addition to performing the roles of the reactive obstructionist and reactive constructive legislatures, can on occasion also take the initiative in shaping the policy agenda and developing policy proposals. No legislature in the region has persistently exhibited such proactive characteristics.

Legislatures that have more experienced legislators, well-developed committee systems, and ample support staff will tend to be more constructive and/or proactive. Legislatures with weaker capabilities will tend either to play a limited policymaking role or to be active, but only in a fairly obstructionist way rather than a constructive one.

Table 3.6 compares Latin American legislatures according to several indicators that attempt to measure some of the dimensions of legislative capabilities. Five are quantitative, objective measures. The first two assess the confidence of citizens and businesspeople in the performance of congress. The third and fourth indicators—average years of legislator experience and percentage of legislators with university education—attempt to gauge the qualifications and experience of legislators. The average number of committee memberships per legislator attempts to measure the degree of specialization of legislative committees, and thus their effectiveness. The next three measures—strength of committees, whether the legislature is a good place to build a career, and technical expertise—were constructed by Sebastián Saiegh, drawing from a variety of secondary sources, but especially from the legislator survey of the University of Salamanca.38 The final column places the legislatures into categories of “high,” “medium,” and “low” on the basis of the aggregate index that results from combining all these indicators.39 The inclusion of the subjective variables does not substantially change the overall index or the corresponding country groupings.40 However, it is important to point out that the Congress Capabilities Index should be seen only as a preliminary and imperfect effort to capture this important aspect of countries’ policymaking capabilities.

Table 3.7 then compares the assessments of legislature capabilities that arise from the previous analysis with a qualitative assessment of the actual roles played by legislatures in the region. Legislatures with greater capabilities tend to play a more constructive role in the policymaking process. But the partisan balance of power has an important role in determining whether a legislature with relatively limited capabilities is fairly inactive or active but obstructionist.

How would the characteristics of legislatures be expected to affect the key features of public policies? Relatively marginal legislatures would typically not interfere with policy adaptability. But such legislatures would be likely to generate greater policy volatility, except in cases of long-term dominance by a single party or low levels of ideological polarization. Moreover, fewer constraints against private-regarding policy outcomes would be likely, given the deficiencies in the representation of citizens in the policy process and in oversight of the executive.

Obstructionist legislatures would be likely to lessen policy adaptability, given the greater difficulty of gaining approval for executive-initiated reforms. At the same time, if a fragmented congress impedes inter-temporal bargaining to a great extent, policies might not be designed so as to be durable over time.

Reactive constructive legislatures should contribute to greater public-regardedness, given their greater capacities with respect to representation and oversight. They also might limit volatility by facilitating inter-temporal agreements and providing a check on ill-conceived policy changes.

Presidents

In Latin America, presidents play a preeminent role in setting the policy agenda and formulating policy proposals. How does the president’s central role affect policy outcomes? What is the actual experience of presidents in terms of their ability to pass their agendas and maintain their influence and prestige, and in terms of their orientation toward serving the public interest? Answers to these questions must focus on the factors constraining, enabling, and motivating the president’s choices and actions. Moreover, since the office of the president has special weight in Latin America, certain personal qualities of leadership also play a role (see discussion below and Box 2.1).

Constraints

Among the factors constraining the president’s strategic options and policy choices are his constitutional and partisan powers.

• Constitutional legislative powers, such as decree and budgetary powers, provide presidents with bargaining leverage, influence over the legislative agenda, and a means—potentially—to bypass a recalcitrant legislature.
• Non-legislative powers, such as the power to appoint cabinet officials, are important for shaping the policy effectiveness of the administration, as well as for building political support across parties and strengthening party discipline.
• Partisan powers relate to the share of seats controlled by the president’s party in congress and the cohesiveness of the governing party/parties.

In general, presidents with greater partisan and constitutional powers have more room to maneuver in developing and implementing their policy agendas. Presidents with weak partisan powers may be able to compensate for this limitation by using their constitutional powers to broaden their political support and/or increase their negotiating leverage in relation to congress. Figure 3.6 shows the relationship between the overall constitutional powers of presidents and their partisan powers, as measured by the average share of seats their party controlled in the legislature as a result of the last two elections.

In some countries, such as Chile and Argentina, presidents are fairly powerful in constitutional terms, and their parties typically control a significant share of the seats in congress. In others, such as Bolivia, presidents are relatively weak in terms of both their legislative and their partisan powers. Brazil, Colombia, and Ecuador stand out as cases in which presidents are powerful in constitutional terms but their parties’ weight in the congress is small. The reverse has typically been the case in Honduras and Paraguay.

Such traditional measures of presidential powers only imperfectly capture the actual experience of presidents. For example, despite the fact that Brazilian and Ecuadorian presidents are each strong in constitutional terms but weak in partisan terms, their experience over the past decade has been remarkably different. While in Brazil, presidents (since 1994) have tended to retain their popular standing and reasonable levels of legislative support throughout their terms of office, in Ecuador the last three elected presidents struggled to pass legislation and were unable to complete their terms of office.

Examples such as these imply that presidents face additional constraints besides those considered in Figure 3.6. Given the interaction among these various constraining factors, traditional measures of presidential power are not enough to explain such differences in outcomes and behavior. Differences in presidential incentives may be playing a role in explaining differences in outcomes as well.

Incentives

Because presidents are elected by a majority or plurality of the national vote, their electoral-based incentives are more “encompassing” than the electoral-based incentives of individual legislators, who are typically elected from smaller regional constituencies, and thus may not focus to the same extent on the provision of national public goods. Thus presidents may be more national in outlook and more encompassing in the constituent base they serve. But it is evident from the corruption scandals that have sometimes affected Latin American presidents, during or after their terms, that the ambitions of presidents are not always limited to serving the public good—and even if this remains their intention, they do not always succeed in doing so.

Beyond the goal of serving the public interest, presidents, like most politicians, are driven by personal and political goals. On the one hand, most presidents are concerned with increasing and/or maintaining their public esteem, political influence, and power, in part to increase their effectiveness as leaders, but also to increase the chance that their visions and programs will triumph over those of others. On the other hand, presidents are also driven by somewhat narrower personal and political ambitions, such as retaining their leadership over a political party, seeking reelection as president (if this is permitted by the constitution, or if the constitution is not too difficult to change), rewarding friends and allies who helped them get to the presidency and are helping to keep them there, or bolstering the popular following and power of their party for the long term.

In some institutional environments, these personal and political ambitions do not seriously interfere with the goal of serving the general public interest. A president’s desire to build his stature within a party, to strengthen the party’s base of support, or to be reelected, for instance, may be best served in some contexts mainly by maintaining or expanding the diffuse support of citizens or party supporters for his presidency. In other institutional contexts, these narrower ambitions, or the necessities of the exercise of power, may result in greater deviations from the public interest.

The extent to which parties are clientelistic, for instance, is an important constraint on public orientation. In clientistic party systems, the president may focus on generating and redirecting the public resources needed to maintain the party’s clientelistic network of support, rather than on adopting policies in the general public interest. Similarly, if parties are relatively decentralized, as for example in Argentina, then the president may be forced to negotiate with regional party leaders and distribute resources to districts where the support of such power brokers is vital to passing legislation and maintaining authority within the party.

Other types of pressures to deviate from the general public interest are likely to be generated in contexts where the party system tends to be clientelistic and highly fragmented, as in Bolivia and Ecuador. Given the large number of parties that typically need to be involved in coalition-building and the divergent and typically geographically concentrated interests the parties represent, the costs of maintaining political coalitions and societal support in terms of policy concessions to special interests, and the distribution of patronage and particularistic expenditures, can result in considerable deviations from optimal policies.

The implications of such acute constraints on presidential incentives are clear in the case of Ecuador. When asked by a reporter whether he expected to complete his term of office after the premature ouster of the two previously elected presidents, President Gustavo Noboa (2000–2002) said, “I’m like the members of Alcoholics Anonymous. I take things one day a time.”41 Clearly, if presidents in Ecuador experience such difficulty in assembling support in congress to approve their policy initiatives and must concern themselves merely with day-to-day survival, their time horizons will tend to be quite short and not conducive to reaching inter-temporal bargains with other actors for the sake of improving policy outcomes.

Numerous other institutional configurations could create presidential incentives favoring deviation from the pursuit of the public interest. While excessive political fragmentation has its own costs, an electorally dominant governing party, strong presidential powers, weakly institutionalized parties, a weak congress, and/or a politically captured supreme court can encourage presidents to put their own power or material ambitions ahead of serving the public good. In the absence of sufficient checks and balances, presidents elected at one moment with a majority or plurality of the popular vote may use this temporary authority and control of the State to further remove potential challenges to their power and distort the electoral process in their favor.

Overcoming Constraints

To some extent, presidents can overcome constraints and shape the outcomes of their administration through strategic choices, such as how to portray and order their policy priorities, whether to take issues to the public, whether to use their legislative powers or negotiate with potential opponents in the legislature, and how to use their powers to appoint cabinet ministers and other governmental officials.

Moreover, presidential power, beyond its formal functions, seems to be surrounded by what has been called institutional charisma. This is a type of influence derived from the way in which the office is perceived by others, which on occasion gives it a much greater range of action, especially in times of crisis. Institutional charisma acquires much more value in combination with interpersonal skills that can significantly increase the real power of the presidency. It is not possible to measure objectively the extent to which these traits increase presidential authority, but experience shows that their contribution to the policymaking process and the implementation of policy can be decisive.