|
A general equilibrium approach is required to understand why similar political institutions may produce different policy outcomes in different countries.
Chapter 6 provides some perspective on the connections between the policy-
making process (PMP) and policy outcomes. It shows how some key aspects of the policymaking process (such as whether congresses have policymaking capabilities, whether parties are institutionalized and programmatic, or whether judiciaries are independent) can have an important impact on the quality of public policies. While chapter 6 emphasizes that these aspects are not independent of one another, it does not go all the way in terms of showing the full dynamic of policymaking processes as they function in general equilibrium, and the impact they have on public policies.
Rather than focusing on some particular aspects of the policymaking process, this chapter offers a glimpse of how the process works in a few countries. The chapter is largely based on the country studies from the Political Institutions, Policymaking Processes, and Policy Outcomes project of the IDB’s Latin American Research Network. This project initially included ten countries in Latin America (Argentina, Brazil, Chile, Colombia, Ecuador, Mexico, Paraguay, Peru, Uruguay, and Venezuela) but was later extended to incorporate three more (Costa Rica, Guatemala, and Jamaica). In the interest of brevity, this chapter will discuss five of them.
The first is Chile. This choice is obvious, given that this country ranks at the top in each of the features of public policies discussed in Chapter 2. The other four countries were chosen to illustrate a number of important issues. One of them is the trade-off between representativeness (or inclusiveness) and policy effectiveness. A number of countries, including Colombia, Mexico, Paraguay, and Venezuela, have moved in the direction of more inclusiveness and participation. In some cases, however, this has come at the expense of some policy effectiveness. The second section in this chapter discusses this trade-off, using the changes in Colombia’s PMP before and after the 1991 Constitution to illustrate the point.
The chapter’s third section highlights the challenges involved in policymaking within the context of very fragmented political systems. The contrasting cases of Brazil and Ecuador are used to illustrate the issue. These countries have the most fragmented party systems in the region. They also share other important characteristics. In both, the president has strong legislative powers, designed in part to compensate for his weak partisan support. In both, legislators are elected using open lists, encouraging close connections between the legislators and the voters. Yet, while Brazil has a stable democracy and reasonably good public policies, in Ecuador the quality of policies is poor and presidents find themselves involved in difficult battles for political survival. While several factors help produce these contrasting results, especially important is the ability of the president in Brazil to build and maintain a stable coalition.
These sections are complemented with a box, based on the case of Costa Rica, which discusses an alternative (and, in this case, reasonably successful) way of making policies: one in which both presidents and legislatures are comparatively weak (in terms of their constitutional powers), and an important portion of policymaking is delegated to autonomous institutions.
Policymaking in an Institutionalized Setting:
The Case of Chile1
Like any country, Chile must contend with many unfulfilled aspirations, and its social and economic outcomes, as well as the functioning of its policymaking system, have been subject to some criticism. Even though poverty fell by half from 1990 to 2003, dropping from almost 40 to 19 percent of the population, income inequality has not decreased. A large percentage of Chileans are dissatisfied with the “authoritarian enclaves” that were left over from the Pinochet government, and political participation, especially among the young, is low.2
Yet of all the countries included in this study, Chile is the one with the best policy characteristics. It has the highest value for the policy index and for most of the components of that index. These numerical measures are buttressed by the policy cases analyzed in the background study on Chile prepared for this report.3
Chile returned to democratic rule in March 1990, after almost 17 years of dictatorship. The workings of the recent Chilean polity and policymaking system present a mix of change and continuity with respect to the country’s history of democracy before its period of dictatorship.
The constitution drafted in 1980 by the military authorities and reformed in 1989 and 1991 at the time of the transition to democracy established some important characteristics of Chilean political institutions, such as a very strong presidency, an electoral system that favors the formation of two coalitions while overrepresenting the second-largest coalition, and the presence of an important number of “authoritarian enclaves.”4
The Chilean president is very powerful (see Chapter 3, Table 3.5), with near-monopoly control over the legislative agenda, and with proposal and veto powers that make him the de facto agenda-setter.
While the president is very powerful, the Chilean policymaking system is studded with veto players, initially written into the constitution by the outgoing military government to impede policy changes by subsequent elected governments. These include a bicameral congress, a comptroller general, and an independent locus of judicial power, including regular courts, a constitutional tribunal, and an electoral tribunal. Less traditional (and more contested) checks on policy formation include the presence of unelected senators in the upper chamber of congress and the relative autonomy of the armed forces.
Yet some characteristics of the Chilean polity, with respect to its institutions and its policies, are surprising. While it has a strong executive in terms of constitutional prerogatives, Chile also has the strongest congress, as evidenced by its top ranking in the Congress Capabilities Index. While its system features numerous veto players, its policies in many areas are perceived as having the greatest capacity for adaptation among the countries examined in this study. The discussion that follows addresses these “puzzles” within the more general context of the workings of the Chilean policymaking process.
The Chilean Congress in the Latin American Context
Chilean legislators do not have access to technical input of the same quality as the executive does. And as noted, the legislative powers of the Chilean president are very strong.
Nonetheless, the Chilean Congress is the strongest congress in Latin America in terms of its role in the policymaking process. As shown in Table 3.6 in Chapter 3, Chile’s Congress has one of the highest levels of technical specialization in the region (through its system of policy committees). Chilean legislators are fairly well educated, and they have long careers in congress. Consequently, their levels of technical expertise are high by Latin American standards. Even though public opinion of the congress is low in absolute terms in Chile (as it is in the whole region), it is the second highest in Latin America.5 A seat in the lower chamber (and even more so, in the senate) is a high-profile and desirable position for Chilean politicians.
Despite the strength of the executive and the fact that some important negotiations within and among parties do not necessarily take place in congress, the Chilean Congress is an important political and policymaking arena. Crucial political and policy matters are debated openly and later enforced in the national legislature.6 The level of debate and transparency of the Chilean Congress is quite high.
The Making of Policy in Chile 7
The Chilean president is constitutionally very powerful. Yet the three presidents who have served since democracy has been restored (Patricio Aylwin, Eduardo Frei, and Ricardo Lagos) have exercised that power in a relatively careful and consensual manner. The Chilean president is, undoubtedly, the agenda-setter in the policymaking process, and has enough tools at his disposal to be able, on occasion, to exert pressure to get his preferred policies through the PMP. Many important policies are developed primarily within the cabinet (with the assistance of technically capable and politically adroit ministers). There is a practice of negotiation and agreement that operates in several (usually sequential) stages.8 Since Chile has relatively strong parties and party identities, the president initially tries to develop consensus for his policies inside his own party, and then within his own coalition, usually through negotiation with the leaders of the other parties in the coalition. Then interactions with the opposition take place, mostly through open forums such as congress.
Technical input enters the policymaking process at multiple nodes. The Chilean cabinet and bureaucracy are very solid by Latin American standards. Chile also has several well-established and reasonably well-staffed think tanks, with institutionalized links to different political parties and coalitions.
In a system with many veto players and a president with strong agenda-setting powers, policies are difficult to pass, but they are passed through negotiations, policy concessions, and, on occasion, distribution of a few particularistic benefits. Once policies are passed, any bargains struck during these negotiations are very stable, and policy is very credible (as seen in this study’s indicators). This very policy stability makes policies a very strong currency in political exchanges. Moreover, the institutionalized nature of parties makes them important actors for the inter-temporal enforcement of these negotiations, minimizing transaction costs and associated distortions.
The Party System
Political parties in Chile are currently moderate, pragmatic yet programmatic, and strongly institutionalized. The tradition of three ideological blocs, left–center–right, has been maintained, but with a substantial degree of convergence. For instance, socialist parties still receive their historical share of electoral support, even though their Leftist ideological orientation has moderated considerably. Since 1990 there have been six parties with congressional representation, organized into two national coalitions, the Concertación (Center-Left), and the Alianza (Right), which formed around the 1988 plebiscite called to decide whether Augusto Pinochet should remain as president.
The workings of the Chilean party system are influenced by a peculiar type of proportional-representation electoral system, with just two seats elected in each district (“binominal”). The lists (coalitions/parties) that receive the two highest shares of votes each win one of the two seats available in each district. Only if the first-place list wins by a ratio of more than two to one do both seats for the district go to the list that won the most votes.
This electoral system reduces the number of relevant actors by encouraging parties to coalesce. Together with the requirement that coalitions can be formed only at the national level (and thus are binding in every electoral district), it strengthens the national leadership of parties. There is a strong incentive to coalesce at the district level, since when the top vote-getter gets at least twice as many votes as its rival, it obtains all (100 percent) of the seats. If a list comes in second, even with only slightly more than half of the votes of the winning list, it gets half (50 percent) of the seats being contested. The provision that coalitions are binding at the national level leads parties to form encompassing national coalitions. Since it is difficult to form coalitions that secure more than two-thirds of the votes in each district, and it is relatively easy to secure one-third of the vote share, the most likely outcome is the formation of two national coalitions.
Any individual party would pay a high price for leaving a coalition. This has been important for keeping both coalitions in Chile united for many years, even though they include parties with different platforms on several issues and political leaders with strong personal rivalries. Despite publicized bickering within coalitions, both coalitions have remained united because of pressure exerted by the congressional members on their respective party leaderships. Congressional members know that their chances for reelection would be jeopardized should the coalitions break up.
Continuous party negotiations within coalitions to decide which candidates will be nominated to the coalition’s list in each district strengthen the parties’ national leadership (although nationally endorsed candidates must be appealing to their local districts). The legislative electoral system does not have term limits. This, together with the need for candidates with strong local support, encourages politicians to seek long legislative careers. In Chile, 75 percent of congressional members are renominated and about 60 percent are reelected. This reelection rate is one of the highest in Latin America.
The binominal system reduces the number of relevant actors to a few parties organized into two encompassing, stable coalitions. It strengthens the party leadership, but at the same time it encourages politicians to respond to their constituencies and have long legislative careers. Finally, given how difficult it would be for one coalition to obtain twice the vote share of the other in any given district, under the binominal system congressional representation for each coalition hovers at around 50 percent of the members of each chamber.9
Summary
The salient features of the Chilean PMP are a party system characterized by two long-lived coalitions, a powerful executive with de facto control over the agenda, a relatively independent judiciary, a bureaucracy that is relatively free from corruption (compared with OECD countries), and a series of veto points in the policymaking process that permit adversely affected actors to block policy change.10 The small number of actors, which interact repeatedly, as well as the predictability of policy implementation and law enforcement, leads to a policymaking process in which transaction costs are low and inter-temporal political exchanges are credible.
Chile seems to be on a path of institutional and policy consolidation. The initial democratic governments have maintained the core of the economic reforms undertaken during the military dictatorship, while steadily (albeit slowly, according to some views) advancing on the social and democratic front. These steps have taken place according to a style of policymaking that is much more consensual and institutionalized than that of other Latin American countries.
The Trade-Off between Inclusiveness and Effectiveness:
The Case of Colombia
After several decades in which most Latin American countries frequently switched between democratic and military governments, the countries of the region gradually returned to democratic rule in the 1980s. Democracies, however, come in different shapes and sizes. While some are fairly inclusive, others exclude certain parties or sectors of the population, either explicitly or because the electoral and constitutional rules prevent adequate representation of minorities. While some produce decisive governments that are able to adopt and implement their policies effectively, others tend to experience gridlock and suffer serious governability problems. It is quite obvious that both inclusiveness (or representativeness) and effectiveness are desirable traits. The problem is that sometimes there are trade-offs between these two characteristics.
Take, for example, the impact of electoral rules. A system of proportional representation with large district magnitudes will score high on representativeness, since the proportion of seats going to each party will be closely matched to the proportion of votes corresponding to each of them. In the limit of perfect proportionality, the number of parties would be large, and no sizable group of voters would be left unrepresented. Yet, unless the conditions are in place for the government to be able to form and sustain a majority coalition, multi-party systems are much more prone to gridlock, and even democratic interruptions. At the other end of the spectrum, plurality systems leave some groups without representation, but are much more likely to produce majority governments that can effectively implement their policy agendas.
In recent decades, Latin American countries have tended to move toward greater inclusiveness. This is obvious in the cases of countries that have switched from dictatorships to democracy. But the movement is broader than that. In Peru, in addition to the return to democratic rule, the 1980s were characterized by a significant expansion of the franchise. Even countries that have been under democratic rule for many decades moved toward greater inclusiveness. In Mexico, the end of Partido Revolucionario Institucional (PRI) hegemony led to significant changes in terms of the representation of political minorities. Together with the adoption of elections for governors, this provided new opportunities for politicians to launch political careers outside of the control of the PRI party leaders. In Venezuela, discussed in more detail in Chapter 11, the introduction of elections for governors had a similar impact. It broke the strong control on political careers that had been firmly exercised by the national leaders of the two traditional parties, AD (Acción Democrática) and COPEI (Comité de Organización Política Electoral Independiente), and ended the collusive agreement between these parties that had been in place since the Pacto de Punto Fijo.11 In Colombia, the adoption of the 1991 Constitution has also represented a movement in the direction of more inclusiveness, in a political system in which the Left had traditionally been excluded and had resorted to non-democratic means of expression (guerrilla movements).
Rather than covering each of the cases, this section focuses on just one of them—Colombia—to illustrate the delicate balance (or trade-off) between inclusiveness and policy effectiveness, as well as the quest to find a satisfying position on these two dimensions through institutional change.
The 1991 Constitution and Colombia’s PMP
One key building block of the policymaking process in Colombia before the 1991 Constitution was the Frente Nacional, an agreement between the two traditional parties (Liberal and Conservador) to share power, with parties alternating in the presidency, and strict parity in key policymaking arenas such as congress, the cabinet, courts, governors, and mayors.12 While the Frente Nacional was formally in effect between 1958 and 1974, several of the key features of the Frente extended well into the late 1980s.
During this extended period of the Frente Nacional, policymaking in Colombia was quite effective. Growth performance was strong, averaging 4.7 percent between 1950 and 1990. Fiscal deficits were low, and fiscal policy played a stabilizing role, aided by mechanisms such as the coffee stabilization fund. The perception in the region was that Colombia had avoided the populist tendencies that were prevalent in Latin America at the time.13
In the policymaking process that operated before 1991, the president was institutionally very powerful, as was the finance ministry, typically headed by a well-respected technocrat. The rules of the political game constrained the role of congress in economic policy and enhanced the decision making capacity of the government. The Frente Nacional coalition, sustained through cabinet and governorship appointments, ensured ample majorities in congress. While the structure of committees provided some incentives for legislators to specialize and gain policy expertise, electoral rules provided incentives for them to focus more on their constituents and less on national policymaking. Thus to some extent legislators focused on getting support for their local projects (in the form of auxilios parlamentarios) and for the most part delegated national policymaking—particularly macroeconomic policy—to the executive, which was able to rely on a bureaucracy that was relatively strong, at least by Latin American standards.14 But even in cases in which the president could not count on the support of the legislature, he had ample powers to legislate by decree, through the use of economic emergencies or the declaration of the state of siege. As will be discussed in Chapter 8, these powers were instrumental in the passing of some important tax reforms during this period. The Supreme Court did exercise some control over the use of decrees, overturning about 25 percent of them, but was not as independent and active as today’s Constitutional Court.
This policymaking process is reflected in the quality of public policies discussed in Chapter 6. Before 1991, Colombia ranked in the top five countries in the region in terms of stability, adaptability, coordination and coherence, and efficiency. If there was one weak spot in Colombia’s PMP, it was linked to deficits in political participation. The Frente was an agreement of a collusive nature, which ensured cooperation among those included, but excluded important sectors of the population, notably the Left. Political participation at the subnational level was also lacking. Governors and mayors were appointed rather than elected, and on occasion had weak roots in the jurisdictions in which they held office. Additionally, this period was characterized by widespread use of clientelistic practices as a way to gather and maintain political support. The most obvious example was the discretionary and often arbitrary use by individual legislators of the resources they obtained through auxilios parlamentarios. Both factors—lack of participation and clientelistic practices—help explain the poor performance of the country in terms of public-regardedness, a dimension on which Colombia ranked near the bottom, in contrast to its relatively high rankings on most of the other dimensions.
Demands for further political participation at both the subnational and national levels were important factors in the transition to the new constitution. The process of decentralization, which had led to the election of mayors in 1988, coupled with the emergence of new sources of revenue associated with discoveries of oil, made it necessary to redefine the way the political and economic pie was going to be shared. The exclusion of the Left from regular channels of political participation had led to guerrilla activity and escalating violence, involving the drug cartels as well, which culminated with the assassination of three presidential candidates in the 1989 electoral campaign, including the likely winner, Luis Carlos Galán. The idea of constitutional reform gained political support, at a time of political unrest when the incorporation of the insurgent groups into the political system was seen as a priority. This led to the adoption of a new constitution.
The 1991 Constitution introduced very important changes into the policymaking process. While the president is by no means weak compared to those in other systems in the region (see Chapter 3), the 1991 Constitution limited the power of the executive in a number of ways. It introduced the election of governors (which the president had previously appointed) and endowed them with significant fiscal resources. It reduced the discretion of the president (and congress) on a number of important policy issues: in some cases, such as pensions or intergovernmental transfers, because they were “hard-wired” into the constitution; in others, such as monetary and exchange rate policies, because they were left within the orbit of a newly independent central bank. The new constitution gave congress a more active role in policymaking by curtailing the ability of the executive to legislate by decree, and by making it easier for the legislature to overrule a presidential veto. It gave the judiciary a more active role, by creating a Constitutional Court endowed with ex ante constitutional review powers and appointment procedures that encouraged judicial independence.
 The constitution also introduced reforms that weakened the partisan powers of the president. Rules for the election of presidents were modified from plurality to majority run-off elections, a system that encourages independents to run in the first round, and tends to reduce the president’s legislative contingent. In addition, parties themselves became more fragmented. While electoral rules for congress had traditionally produced incentives for the fragmentation of political parties into a large number of very small and fairly independent factions, changes to campaign finance rules channeling public funding to factions rather than parties, as well as changes in the ballot structure, further weakened the ability of presidents and central party leaders to discipline their own legislative contingents (see Box 7.1).15 In summary, not only did the constitution limit the legislative powers of the president, but his capacity to do the things that remained within his power was curtailed as well.
As a result of the changes introduced by the 1991 Constitution, the PMP in Colombia experienced significant alterations. Congress became increasingly involved in national policy discussions, introducing significant changes into legislation proposed by the executive. As the example of tax policies in Chapter 8 will make clear, passing legislation has become more costly. The executive must make more concessions, and more rewards need to be distributed in order to pass legislation, as the number of “sponsors” involved in each bill has substantially increased.16 Even after legislation passes through congress, it can still be derailed by the Constitutional Court, which not only has made it more difficult for the executive to bypass congress (associated with its role as enforcer) but has also more generally become a more active—some observers say overly active—player in the policymaking game. Subnational governments have additionally become more important players in some aspects of policymaking, including social policies. This has sometimes had important implications for macroeconomic policies, as several departments and municipalities have incurred excessive debt and had to be bailed out by the national government.
 In summary, the number of key players in the national policymaking game has increased, and cooperation among players has become more difficult to achieve. While these developments are positive in that they generate stronger checks and balances on the discretion of the executive, they can also have a negative impact on policy effectiveness. In fact, a few of the key features of public policies more directly linked with effectiveness (stability, efficiency, coordination, and coherence) declined after 1991 (see Figure 7.1).17 On the other hand, the country achieved substantial improvements in terms of public-regardedness (which is still low, however). This improvement can be traced to at least four factors: the increase in checks and balances; increased participation, at both the national and subnational levels; more control on the clientelistic practices of politicians, particularly legislators, through the elimination of auxilios parlamentarios;18 and more control over the discretion of the executive to distribute subsidies and credit to certain sectors of the population, as an independent central bank is subject to restrictions on lending to the private sector.
Colombia has not been the only country moving in the direction of more participation and inclusiveness. Other countries moving in that direction include Mexico and Paraguay.19 In all these cases, the improvement in terms of inclusiveness has occurred alongside some declines in policy effectiveness (see Figure 7.2). However, the rate at which countries have sacrificed effectiveness in the quest for more inclusiveness differs greatly across countries. The challenge is to try to increase inclusiveness and participation without losing effectiveness in the process.
In Colombia, recent reforms have passed through congress that seek to improve effectiveness, hopefully without compromising inclusiveness and participation to any significant extent. One change involves the reelection of the president, a trend that has been present in many Latin American countries. Probably the most important one, from the perspective of this study, relates to changes in the electoral rules for the legislature. Political parties can now present only a single party list in each electoral district. In addition, a threshold was introduced (equivalent to 2 percent of the national electorate) that precludes small local movements from participating in elections. These changes should help reduce fragmentation and increase party discipline. Thus they should enhance the role of parties as encompassing institutions interested in policymaking at the national level, where inter-temporal bargains can be reached. They should also help facilitate the passage of some legislation through congress.
Policymaking in Fragmented Political Systems:
The Contrasting Cases of Brazil and Ecuador20
One of the main concerns about the workings of presidential democracies is the potential for governability problems. In contrast to his counterpart in parliamentary democracies, the president in a presidential democracy is not guaranteed a winning coalition that will allow him to pass his agenda through the legislature. Lack of support for the president in the legislature has been associated with difficulties in obtaining approval of welfare-enhancing reforms and in adapting to shocks, and even with an increased likelihood of constitutional interruptions. These concerns are particularly relevant in the case of fragmented political systems, in which the party of the president typically does not hold a majority of the seats in the legislature. In these cases, policy adaptability and, more generally, democratic governability will depend on the ability of the government (and of the president in particular) to form and maintain a winning coalition.
This section looks at policymaking in Brazil and Ecuador. These two countries have the most fragmented party systems in Latin America. As a result, in both cases the presidential party’s contingent in the legislature is among the lowest in the region (see Figure 7.3).21
Yet Brazil and Ecuador differ considerably with regard to the quality of their public policies. According to the policy index developed in Chapter 6, Brazil appears in the group with relatively good policies, while Ecuador appears in the group in which the quality of policies is low. In fact, Brazil ranks above Ecuador in each of the features of public policy presented in  Table 6.1 (see also Figure 7.4). Furthermore, democracy in Brazil is stronger and more stable than in Ecuador, where none of the last three popularly elected presidents (Abdalá Bucaram, Jamil Mahuad, and Lucio Gutiérrez) has been able to complete his term.
In addition to their political fragmentation, political institutions in Brazil and Ecuador share other important features. In both countries, the president has been endowed with unusually strong constitutional powers. Legislators in both countries are elected using similar electoral rules: open-list proportional representation systems, which provide them with incentives to cater to their geographical constituencies. Given their important common elements, how can the contrasting political and economic policy outcomes of Brazil and Ecuador be explained? The answer does not lie in any single factor.
To some extent, differences in policy outcomes may be associated with substantial differences in the quality of the institutions that Chapter 6 identified as essential for good policymaking. While Brazil has a congress with relatively good policymaking capabilities, a strong bureaucracy, and an independent judiciary, Ecuador lacks all of these things. While in Brazil the president’s main focus (at least in recent times) has been on maintaining macroeconomic stability, the president of Ecuador on occasion becomes engaged in an uphill battle for political survival, which prevents him from focusing on the long run.
In addition, there is another important difference between the policymaking process in Brazil and Ecuador. While the president in Brazil seems to have the tools he needs to engage in political exchanges with legislators and sustain a coalition in congress, in Ecuador coalitions are very fickle and tend to collapse as the presidential term progresses and elections draw near. The rest of this section explores each of these factors in more detail.22
Policymaking Capabilities and the Role of Congress
Brazil and Ecuador differ substantially with regard to the policymaking capabilities of their legislatures and the role they play in the PMP, in spite of important similarities in the electoral rules for the legislature.
Brazil
Legislators in Brazil are elected for a period of four years under an open-list, proportional representation system. Having an open list means that voters can vote for an individual candidate within the party list and help shape which candidates get elected. This electoral system typically provides incentives for legislators to respond according to the preferences and demands of their districts, rather than respond to the wishes of party leaders. Not surprisingly, their reelection chances depend crucially on their ability to deliver on those demands.
However, in Brazil these decentralizing electoral forces are compensated for by other features that lead to greater party discipline. Party leaders have the power to appoint and substitute members of the legislative committees, which play a relatively important role in policymaking. Perhaps more importantly, interactions between the executive and legislators are not handled individually, but through the mediation of party leaders, who act as brokers in the exchange. Thus, by controlling the access of individual legislators to the benefits that may affect their chances for reelection, party leaders can get individual legislators to vote according the party’s preferences, resulting in a relatively high degree of party discipline. This is why, in spite of “decentralizing” electoral rules, political parties in Brazil tend to be strong and cohesive within the legislative arena, increasing the scope for cooperative inter-temporal exchanges.23
A number of other factors contribute to the policymaking capability of congress in Brazil. Reelection rates for legislators are relatively high, at least in regional perspective. Nearly 70 percent of legislators in the lower chamber seek reelection, and about 70 percent of those who run are reelected. Thus close to 50 percent of legislators have prior legislative experience, which contributes to an accumulation of policy expertise, as well as a longer-term focus.
The technical capacity of the legislature for policymaking purposes is further aided by a large technical staff that provides legislative support (see the discussion of knowledge actors in Chapter 5). More than 500 staff members in the two chambers provide technical support. Most of them are specialists in different areas of policy. They are highly educated and well paid, and they obtain their positions through a highly competitive selection process. These legislative support offices, developed mostly during the 1990s, have helped improve the technical level of legislative deliberations, as well as the quality of the policies that are discussed in congress. In this dimension, Brazil is at the top of the ranking of the countries in Latin America.
As a result of these factors, Brazil ranks high within the region on the index of policymaking capabilities of congress, lagging behind only Chile. These capabilities allow the Brazilian Congress to play a constructive role in policymaking, a role characterized by a workable relationship with the executive, despite the high level of political fragmentation. The mechanisms through which the executive can keep the coalition together and advance its agenda are discussed in more detail below.
Ecuador
In Ecuador, electoral rules for the legislature have been subject to a number of changes. Until 1997, Ecuador had a mixed electoral system in place for its unicameral congress. While “national” deputies were elected from a single district for a period of four years, “provincial” deputies (comprising more than 80 percent of the total) were elected from provincial districts under proportional representation, for a period of just two years. Immediate reelection was banned by the constitution until 1994. As a result, the legislative experience of the Ecuadorian legislators was by far the lowest in the region. These short tenures led to legislators with short horizons, without policymaking capabilities, and without incentives to develop them.
The midterm elections for provincial deputies had another important drawback. They typically reduced the size of the president’s legislative contingent even further, making it more difficult for the president to build a winning coalition that would allow him to pass his agenda. As an example, during the Rodrigo Borja administration (1988–92) the share of seats controlled by the ruling party, the Izquierda Democrática, dropped from 42.3 percent in the first half of the term to 19.4 percent after midterm elections. During the Sixto Durán-Ballén administration, the share corresponding to the Partido Unión Republicana fell from 15.6 percent to just 3.9 percent. More generally, the size of the coalition supporting the president in congress declined substantially throughout the presidential term, with the midterm elections accounting for an important share of the decline.
In recent years, electoral rules for the legislature have been subject to a number of changes. The ban on immediate reelection was lifted in time for the 1996 election. The electoral system was switched to one with open lists in 1997. Midterm elections were eliminated, and the tenure of provincial deputies was extended to four years in the constitutional reform of 1998. For the 2002 elections, deputies were elected only out of provincial districts. In addition, the rules for electing the president were also changed in 1998, in an attempt to provide the executive with a larger legislative contingent. The recent episodes of democratic interruptions suggest that the impact of these changes has been insufficient. The legislative contingent of Ecuadorian presidents continues to be very small; only a small portion of legislators (about 27 percent) are reelected, which means that they still tend to be focused on the short run and have weak policymaking capabilities.
As a result of scant legislative support, particularly toward the end of the presidential terms, executives have often resorted to using their considerable constitutional powers to carry out their agendas. These attempts to bypass congress have led to a highly adversarial relationship between the executive and the legislature, with the president often trying to legislate by decree, and congress threatening to impeach cabinet members.
The contrasting roles of the legislatures in Brazil and Ecuador are reflected in the rate of success of the executive in passing legislation, presented in Chapter 3. While Brazilian presidents have been able to pass 72 percent of their initiatives through congress, Ecuadorian executives have been successful only 42 percent of the time.24
Judiciaries and Bureaucracies
Brazil and Ecuador also differ substantially with regard to the strength and independence of the judiciary and the bureaucracy. As discussed in Chapter 6, both are important building blocks that contribute to good policymaking.
Brazil
In Brazil, supreme court justices are nominated by the president and confirmed by the senate for lifetime terms, but must retire at age 70. Tenures of supreme court justices have averaged over ten years, making Brazil second only to Chile within Latin America (see Figure 4.6 in Chapter 4). As a result, each president typically appoints a small number of the 11 supreme court justices. For example, during the eight years of the Cardoso administration, only three justices were appointed to the supreme court. The independence of the courts has been enhanced by the 1988 Constitution, which established that the judiciary would determine its own budget, and the courts themselves would appoint lower-court judges. All this translates into a reasonably high degree of judicial independence that, according to the World Economic Forum, places Brazil in third place in Latin America, with a score of 3.9 out of 7 (see Table 4.3 in Chapter 4).25
The judiciary in Brazil has played the important role of enforcer of the constitution, ensuring that the other branches of government do not overstep their boundaries. There have been a number of high-profile cases in which the supreme court ruled against the executive on issues that were of vital importance to the executive. One of these was the attempt by the Cardoso administration to tax retired workers, a measure that was seen as an important component of a fiscal restraint program. The measure, which was passed through congress after considerable effort, was highly controversial because it involved acquired rights and entitlements. In the end, the supreme court declared the measure unconstitutional. Alston and others discuss this and other cases that support the notion that the judiciary in Brazil has a reasonable degree of independence.26
The bureaucracy in Brazil also contributes to the high quality of public policies. In fact, according to most measures of bureaucratic quality, Brazil has the strongest bureaucracy in Latin America. This is reflected in the indices presented in Chapter 4, in which Brazil appears at the top of the rankings. Most appointments are handled through a well-institutionalized system of competitive entrance examinations. Employment conditions (tenure and competitiveness of salaries) are attractive, and there are ample resources available for training of personnel. 27 While there are a significant number of positions that are reserved as political appointments (called “DAS positions”), the number is limited by law, and in making these appointments technical criteria are still taken into consideration. 28 Brazil’s bureaucracy, which combines a high degree of autonomy with strong technical capabilities, is an important institutional actor that can constrain the executive, and at the same time contribute to policy stability and the public orientation of policies.
Ecuador
In Ecuador, the judiciary has traditionally been highly politicized. On occasion, the ability of the executive to influence the courts has even been used as a token of exchange in order to build political support for the government coalition. Before 1998, the tenure of supreme court justices was formally six years, and their terms could be renewed beyond that. In practice, however, the average tenure of supreme court justices until the mid-1990s was 2.8 years, among the shortest in Latin America. The constitutional reforms of 1997 and 1998 introduced important changes, aimed at providing the judiciary with independence from political pressures. Tenure for supreme court justices was extended from six years to lifetime terms, and the appointment of members (31, including the court’s president) was assigned to an administrative branch of the judiciary, the Consejo Nacional de la Judicatura. The constitutional reform also introduced the nine-member Constitutional Tribunal as the supreme entity for constitutional oversight.
In spite of these efforts, the judiciary in Ecuador continues to be politicized. The best example of this was the removal of 27 of the 31 supreme court justices in December 2004. That the justices would be removed was decided by pro-government deputies, with the support of the Roldosista Party, the party of ex-president Bucaram, whose trial on charges of corruption was immediately voided. Public demonstrations against Bucaram’s impunity became the trigger that eventually contributed to President Gutiérrez’s dismissal, in April 2005. This episode is yet another example demonstrating that changes to the letter of the law do not necessarily change institutions in the long run.
The bureaucracy in Ecuador presents a stark contrast to that in Brazil. Both appointments and dismissals tend to be politically motivated, and a substantial number of people enter and exit the bureaucracy with every new administration. Ecuadorian legislation as far back as the 1980s has introduced merit criteria for the appointment of civil servants, but there is a large gap between the letter of the law and established practice. The frequent constitutional interruptions experienced by Ecuador in the recent past have contributed to the lack of stability of the civil service. In terms of the indices presented in Chapter 4, Ecuador ranks 13th out of 18 countries in the combined index of civil service development presented in Figure 4.4.
Forming and Maintaining Coalitions
Box 6.2, on constitutional interruptions, clearly showed that minority governments have a substantially higher risk of democratic instability. It also showed that a fragmented party system does not need to be a problem, provided that the president is able to form and sustain a majority or near-majority coalition that can provide political support in congress. In Brazil, coalitions tend to be stable and have for the most part allowed the president to pass a significant part of his agenda during the entire presidential period. In contrast, in Ecuador coalitions are typically weak and short lived, tend to provide ad hoc support for selected initiatives, and typically break down as new elections approach. Thus, it is important to examine the mechanisms through which coalitions are built and maintained in each case.
Brazil
The president in Brazil has many instruments for building and maintaining the support of the coalition. Perhaps most important is his strong budgetary powers.29 The budget approved each year in congress typically includes a number of amendments introduced by legislators, involving small projects such as roads or sanitation infrastructure in their districts. Even after these amendments are approved, the president has discretionary power during the budget execution stage to decide which amendments get funded and which do not. This constitutes a powerful bargaining chip, which the president can use to build political support for his agenda. Legislators’ electoral success depends on getting their projects implemented in their districts, so they are often willing to make policy concessions in exchange for these investment projects.
A second tool used by the president is his power to distribute positions in the federal government. Within certain limits described above, the president can use appointments to DAS positions to reward coalition partners or their supporters, in order to help cement the coalition.
A third mechanism used to build support in congress is the power to appoint cabinet ministers. All major parties in the coalition are granted ministries, typically in proportion to their share of votes in congress, although the extent to which this resource is utilized varies from administration to administration. While the president has to “concede” certain ministries to coalition partners in exchange for their support, he still is able to dictate the major policy guidelines and holds the authority to reclaim any position at any moment.
In Brazil, these exchanges are credible because they are part of a repeated game among actors that tend to be long-term players. Presidents can be reelected, and party leaders, who act as brokers in these transactions, are long-term players, as are many individual legislators, who stand a good chance of reelection. If the government were to frequently renege on implicit promises made in the course of these exchanges, it would debase the bargaining chips it holds, and coalitions would break down.
Recently, some allegations regarding illegitimate political exchanges in Brazil have attracted considerable attention. In this context, it is important to point out that the particular transactions discussed in the last few paragraphs are legitimate ways for Brazil’s president to gather political support and pass his agenda, in the context of a very fragmented party system. While some observers may find these political transactions objectionable as a matter of principle, the transactions play a vital role in helping cement stable coalitions, and they contribute to governability. Eliminating these opportunities for political exchange would either put governability at risk, or increase the likelihood that other less legitimate and less transparent forms of exchange might take place. The case of Ecuador is a vivid example of what can happen in fragmented party systems in the absence of stable coalitions.
Ecuador
If the president in Brazil can overcome the weaknesses associated with party fragmentation by forming a stable coalition that allows him to pass his agenda, the question is, why can the president in Ecuador not do the same? The contrasting results are puzzling given that, like his Brazilian counterpart, the Ecuadorian president is also endowed with a number of resources—pork-barrel projects, cabinet positions, positions in the bureaucracy, policy concessions, contracts—that he can offer in exchange for support for his agenda. As usual, there are no single-factor explanations to account for the inability of presidents in Ecuador to maintain a stable coalition.
It is worthwhile to begin by examining the three types of mechanisms identified for the case of Brazil and attempting to understand why they do not seem to contribute to the building of stable coalitions in the case of Ecuador. The first of these was the exchange of projects for political support. In Brazil, the president can deliver these projects, because of his discretionary power over the budget, and legislators value this, since it contributes to their reelection. In Ecuador, as in Brazil, legislators are elected from open lists and can be reelected, so they should have similar incentives to deliver projects to their districts. However, presidents in Ecuador do not have discretionary power over the budget. They previously had the power to make discretionary use of off-budget allocations, but lost this power in 1995.30 Even before that date, these types of exchanges would not have been very useful, since at the time legislators could not be reelected and were elected from closed lists. Thus, they had a weaker electoral connection with their voters, and they did not have the kind of incentives to deliver investment projects to their communities that Brazilian legislators have.31
The second bargaining chip discussed in the case of Brazil was public employment. While the DAS positions discussed in Brazil are very prestigious and well paid, this is not the case for civil service positions in Ecuador. In addition, while these positions in Brazil may last up to eight years in case of reelection of the president, in Ecuador both entry to and exit from the civil service tends to be highly political, and there is a great deal of turnover (particularly in the case of political appointees) every time there is a change in administration. The absence of immediate presidential reelection further shortens the expected length of these appointments. In the case of Ecuador, the result is less prestigious and shorter-term positions, which are likely to be less attractive. Naturally, as the end of the presidential term draws near, this loses most of its value as a token of exchange.
The third bargaining chip discussed in the case of Brazil was cabinet ministries. As can be seen in Figure 4.1, Ecuador is the country with the greatest degree of cabinet instability among the 12 Latin American countries for which data are available.32 The study of the PMP in Ecuador presents evidence showing that, out of 292 ministers that occupied cabinet positions between 1979 and 2002, 61 percent did not finish the presidential term, and 90 percent of these either resigned or were dismissed by the executive.33 In other words, tenure of ministers is short and uncertain, and the president, who awards these positions, can also take them away when he needs to build a coalition along a different policy dimension. This reduces the credibility of the reward, and thus the attractiveness of cabinet positions for coalition partners, as well as the value the president can expect to receive in exchange for a cabinet position. Interestingly, the background study of Ecuador’s PMP conducted for this report also states that Ecuador has one of the lowest shares of partisan ministers in Latin America. It is likely that short tenure and the lack of credibility surrounding the reward may be contributing to this result.
But there is another important component to the explanation of why coalition partners do not find cabinet positions attractive, and more importantly, why coalitions in Ecuador are unstable. There is an embedded perception in the Ecuadorian political culture, reinforced by the electoral calendar, that collaborating with a government (or being a “gobiernista”) can be a politically costly move, especially if this involves the support of unpopular adjustment policies. This anti-government attitude—and the liability of being associated with the government—increases when the president’s job approval ratings decrease over time. With decreasing levels of presidential popularity, potential coalition partners often prefer to engage in what have been called “ghost coalitions”: secret agreements whereby party leaders agree to collaborate with the government on a narrow set of ad hoc policy issues, but avoid—and sometimes publicly deny—any long-term commitment that may affect their own electoral chances.34 These exchanges are made possible by the absence of roll calls in the legislature. In this way, parties can obtain the benefits of coalition membership, without having to pay the political costs of being associated with the government.
The cost of being a part of the government coalition increases as the term of the presidency advances, new elections draw near, and party leaders need to position themselves for the elections. At the same time, the benefits of being in the coalition decline as the presidential term approaches its end, since positions in the bureaucracy, and even ministries, become shorter in expected tenure and thus less valuable. It is therefore not surprising that the size of the coalition declines substantially over the presidential term (see Figure 7.5).
As should be clear from the above discussion, none of the political mechanisms that seem to work in Brazil as a way of cementing stable coalitions appear to be working in the case of Ecuador. These mechanisms are further hindered by the relatively short tenure of the political actors that participate in them. All these factors contribute to explaining the relative instability of democratic governance in Ecuador.
The contrasting cases of Brazil and Ecuador also illustrate the need for a general equilibrium view of political institutions and policymaking processes. Focusing on just a few institutional dimensions, such as the extent of presidential powers, the electoral rules for the legislature, and the degree of party fragmentation, may lead to the wrong conclusions, however important these dimensions may be. Policymaking processes are very complex, and understanding them requires attention to details such as presidential budgetary discretion and roll calls in the legislature, which may affect in an important way the nature of the political transactions among the relevant political actors. This point is further illustrated by the discussion of the case of Costa Rica, in which a key element of the policymaking process is the delegation of important policymaking responsibilities to autonomous institutions (Box 7.2)
|
