SciELO - Scientific Electronic Library Online

 
vol.24 número2Redibujando el mapa electoral chileno: incidencia de factores socioeconómicos y género en las urnasCalidad de la representación política y tamaño de las circunscripciones electorales: una comparación de las asambleas panameñas de 1945 y 1999 índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados

Revista

Articulo

Indicadores

Links relacionados

Compartir


Revista de ciencia política (Santiago)

versión On-line ISSN 0718-090X

Rev. cienc. polít. (Santiago) v.24 n.2 Santiago  2004

http://dx.doi.org/10.4067/S0718-090X2004000200004 

 

Revista de Ciencia Política / Volumen XXIV / Nº 2 / 2004 / 67-93

ARTÍCULOS

Political Institutions, Policymaking Processes, and Policy Outcomes in Paraguay,

1954-2003*

José Molinas

Instituto Desarrollo, Paraguay

Aníbal Pérez Liñán

University of Pittsburgh, Estados Unidos

Sebastián Saiegh

University of Pittsburgh, Estados Unidos


 

Resumen

Este artículo caracteriza la evolución del proceso de elaboración de políticas (PMP) en Paraguay durante el período 1954-2003. El siguiente trabajo ofrece una perspectiva general del PMP durante la dictadura de Alfredo Stroessner (1954-89) y explora el contexto institucional resultante a partir de 1989. Tras analizar la fragmentación del partido Colorado, se describen los patrones característicos de la elaboración de políticas surgidas con la Constitución de 1992. Se sostiene como hipótesis que la presencia de un amplio número de jugadores con capacidad de veto ha dificultado el cambio y que los legisladores tienden a promover políticas particularistas. Para poner a prueba tal supuesto, utiliza una base de datos que contiene los proyectos de ley ingresados al Congreso desde abril de 1992. Las conclusiones sugieren que el proceso político paraguayo puede ser flexible para generar beneficios particularistas y rígido a la hora de realizar amplias transformaciones de tipo regulatorio o redistributivo.

Abstract

This paper characterizes the evolution of Paraguay's policymaking process (PMP) between 1954 and 2003. We present an overview of the PMP under the rule of Alfredo Stroessner (1954-89) and explore the institutional setting emerging after 1989. We discuss how the Colorado Party progressively broke up into several factions and characterize the distinctive patterns of policymaking that emerged after the adoption of the 1992 Constitution. We hypothesize that the presence of a large number of veto players has made policy change more difficult and that legislators are inclined to pursue particularistic policies. In order to test those hypotheses we rely on a database containing virtually every bill introduced in Congress since April of 1992. Our conclusions suggest that the current Paraguayan PMP may be flexible for the provision of particularistic benefits, but rigid for the approval of broad regulatory or redistributive policies.

PALABRAS CLAVE • Paraguay • Veto-players • Policymaking • Parties • Legislative Politics


 

INTRODUCTION

The average growth rate of the Paraguayan economy between 1996 and 2000 has been -1.9%, and per capita GDP fell 6.1% in 2002 with respect to its 1995 level. Despite the country's deteriorating macroeconomic conditions no major reforms had been implemented in recent years. Following the 1995-1997 financial crisis, fiscal deficits have steadily increased from 0.3% of GDP in 1995 to 4.3% of GDP five years later. Economic reforms intended to stabilize the economy also came to a standstill in the legislative stage. For example, in November, 2002 the Minister of Finance resigned in protest for the congressional delays to authorize a stand by agreement with the IMF. The fact that most reform attempts had been stalled rather than implemented and shortly after overturned, or implemented but in a very inefficient way, reveals that the country's levels of policy adaptability are very low.

Why has it been so difficult to modify public policies in Paraguay? Is this rigidity a problem for all policy areas or only for some contested policy dimensions? This article characterizes the Paraguayan policymaking process (PMP) between 1954 and 2003. We identify four distinctive periods in the evolution of the Paraguayan PMP: the "golden age" of the Stroessner dictatorship (1954-1981), the late Stronismo (1982-1989), the transitional Rodríguez regime (1989-1993), and current period marked by democratic institutions (1993-present).

In the following section we present a stylized model of the PMP under the rule of Alfredo Stroessner (1954-89). Section Three provides an overview of the 1989-93 transitional regime. We discuss how the dictatorship's policymaking rules, that initially remained in place, allowed for an initial period of fast policy change. In the following years, however, these rules changed in order to allow for the democratization process and the emergence of new players. The Colorado Party, a "granitic" organization under the dictatorship, progressively broke up into an "archipelago" of factions (Arditi, 1993). In Section Four we characterize the distinctive patterns of the policymaking process that emerged after the adoption of the 1992 constitution, identify the new institutions and players, and develop some hypotheses about the ways in which these factors are expected to shape the underlying rules of the new PMP (Spiller and Tommasi, 2003). In order to test those hypotheses we rely on a database containing virtually every bill introduced in congress since April 1992. Our conclusions suggest that the current Paraguayan PMP may be flexible for the provision of particularistic benefits, but rigid for the approval of broad regulatory or redistributive policies.

The study of the Paraguayan case offers valuable theoretical insights regarding two issues. First, in recent years an emerging literature has emphasized that the presence of multiple veto players (political actors whose consent is necessary but not sufficient to alter the policy status-quo) reinforces policy rigidities and prevents policy change (Tsebelis, 2002). We show that, to some extent, this has been the case in Paraguay: the democratization process started in 1989 multiplied the number of veto players and thus made policy change more difficult. However, we also show that concentration of power by a policy dictator (a single player whose consent is necessary and sufficient to alter policies) does not guarantee policy adaptability: when the dictator is able to transfer the costs of failure to somebody else (as the Stroessner regime did in the 1980s), he or she has little incentives to correct inefficient policies. Second, there is uncertainty in the political economy literature with regards to the relationship between the dimensions of policy adaptability-rigidity and public-private regardedness. Publicly-regarded policies tend to generate public goods, while privately-regarded policies tend to generate private goods or benefits for narrowly targeted sectors (Cox and McCubbins, 2001). We show that these two dimensions are in fact orthogonal: while levels of policy adaptability in Paraguay varied considerably over time, private-regardedness has been constant, as the system has been historically marked by patronage and clientelism.

I. THE POLICY MAKING PROCESS UNDER STROESSNER (1954-1989)

On May 4, 1954, after seven years of unstable one-party rule, a military coup removed President Federico Chaves from office. The leader of the coup, Gen. Alfredo Stroessner, soon gained control over the Executive Branch, the Colorado Party, and the Army, establishing a regime that would last for almost 35 years. Stroessner held office longer than any other ruler in Paraguayan history (Lambert, 1997) and by 1989 approximately 75% of the population had grown up under the Stroessner regime characterized by fear, repression, and co-optation, having no experience on democratic rules of the political game nationwide.

This section outlines the structure and evolution of the policymaking process (PMP) during the Stronismo. We argue that the basic structure of the PMP remained generally constant over this period: the Chief Executive remained a dictator not only in the political sense of the term, but also in the technical sense of his approval being both necessary and sufficient to implement policy change. We also claim that this concentration of power fostered policy adaptability during the 1960s and the 1970s. However, when several conditions changed in the 1980s, these changes were not successfully addressed by the Stroessner regime; rather policy rigidity ensued.

II. THE PLAYERS AND THE "RULES"

 

1. The Dictator

For more than thirty four years, Gen. Alfredo Stroessner managed a "cooperative" arrangement among state officials in all three branches of government and the bureaucracy, the Colorado Party, and the armed forces. This so-called "trilogy" of power (State-Party-Army) was based on a single premise: Stroessner would continue in office for the foreseeable future.1 Stroessner's re-election was secured through several mechanisms. First, repression against political dissidents was backed up by a "permanent" (i.e., renewed every 90 days) State of Siege and by ambiguous regulations, like Law 209, that penalized "fostering hatred among Paraguayans" as a criminal offense. Second, the system established a culture of fear due to the arrest, torture, and death or exile of selected opposition leaders. This selective repression, combined with a widespread informal network of political spies, was internalized by the population with the result that repression on a large scale was not always necessary given the distrust and suspicion that dominated the country. Third, the Colorado Party was strengthened nationwide, rewarding membership with economic and political incentives under the patronage system that tied people to the regime at all social levels. Fourth, after two elections with Stroessner as the only candidate (in 1954 and 1958), there was a tightly controlled multi-party system with limited political activity granted to selected opposition parties (1962-1989).

Under this system, Stroessner was elected to office in eight opportunities with 98.4 percent of the vote in 1954; 97.3% in 1958, 90.6% in 1963, 70.9% in 1968; 83.6% in 1973, 90.0% in 1978, 90.1% in 1983, and 88.6% in 1988 (Lambert 1997:21). According to the electoral law in place since the late 1950s, the winning party in a "competitive election" was granted 67% of the seats in Congress. Congressmen were able to be re-elected indefinitely. Following the legal precedent of previous charters, the 1967 Constitution gave Stroessner great powers over Congress and the judiciary. The President was able to dissolve Congress without restrictions and Congress had no ability to convene itself. Congress was not an important arena of policymaking because: a) the initiative for all major policymaking was expected from the executive, and b) legislators had no real veto power. The role of Congress -as well as the judiciary- was merely to provide a democratic façade and was therefore expected to be supportive of the executive's policy initiatives. Hence, policymaking was essentially decided by Stroessner and key collaborators from the Colorado Party and the Armed Forces that surrounded him.

2. The Colorado Party

Stroessner transformed the Colorado Party from a party deeply divided along factional lines into a highly united and vertically organized political instrument (Lambert, 1997:5-7). He purged the party of opposition, converted it into an instrument to legitimize his regime, and used this mass_based, agrarian party to mobilize support and to repress the opposition.

Between 1954 and 1966, Stroessner consolidated his own control over the Colorado Party by selectively purging internal party opponents (Lambert, 1997). As a result of these moves, Stroessner imposed a "granitic unity" on the party under his control through repression and expulsion of the internal opposition and through the promotion of the politicians loyal to him. The Colorado Party became a unified, mass-based party with extensive control of the media, patronage resources, and extensive grassroots support.

The Colorado Party developed a nationwide network of seccionales (local party offices) that mobilized support for the regime and represented the "eyes and ears" of the government, employing thousands of spies to report antigovernment sentiment among the population (Lambert, 1997). Unopposed elections of centrally nominated candidates at the seccionales guaranteed tight control over the party. Moreover, the seccionales administered the political patronage system.

Membership in the party was mandatory for all state employees (teachers, members of the armed forces, judges, etc). Patronage, dispersed mainly through the Party, tied people of all social levels to the regime. By 1989, State employees accounted for 9% of the labor force (Nickson, 1997). The Colorado Party was partially funded by compulsory deductions from public employees' salaries. In exchange for public employment, public employees attended meetings and parades, donated part of their income to the Party (automatically deducted from their paychecks), and even bought a compulsory subscription to the Colorado newspaper Patria (Arditi, 1993: 164). Senior party members received lucrative contracts for state projects, holdings in public companies, and other opportunities to become successful businessmen (Lambert, 1997: 7).

3. The Armed Forces

The armed forces did not govern in their own right, but represented a key component of the trilogy of power during Stroessner's regime (integrated by the State, the Colorado Party, and the Armed Forces). The alliance between the Colorado Party and the military was consolidated in 1947 (seven years before Stroessner took over) after a short but brutal civil war that depleted the military from 90% of its officer corps (Lambert, 1997:5). After the civil war, the armed forces were re-organized under Colorado supervision.

Stroessner, however, strengthened this alliance through mandatory membership of military officers in the Colorado Party, and by unifying the command in his position of President, Commander in Chief of the Armed Forces, and Honorary President of the Colorado Party. Throughout the 1950s Stroessner brought the military under his own control. While disloyalty was punished, loyalty was rewarded with promotion and economic privileges. Military officials had access to land and lucrative positions in State monopolies, as well as a free hand in illicit business like widespread contraband.

Given this consolidation of power, the government, the Colorado Party, and the Armed Forces had long time horizons during the 1960s and the 1970s. The incentive structure was stable, and the partnership remained cohesive and strong. The main enforcer of the inter-temporal deals was Stroessner himself and the punishment for disloyal behavior was the main deterrence to defection from Stroessner's policymaking game.

III. PUBLIC POLICIES IN PARAGUAY, 1954-1989

This system generated a dictator, not only in the literal sense but also in the technical sense -a player whose acquiescence is both necessary and sufficient to alter the policy status-quo. To sustain this arrangement, two lasting deals were enforced: the first one allowed major military leaders to control illegal rent-seeking activities (smuggling operations, drug trafficking) and the second one allowed the Colorado politicians to control the distribution of patronage positions in the public administration. Within this general framework, specific intertemporal policy agreements were viable but for the most part irrelevant, because politicians and military officers had delegated the policymaking function to the executive.

We argue that policies in Paraguay were characterized by their low public regardedness throughout the whole 1954-1989 period. Policies were aimed to benefit mainly the loyal members of the Colorado Party and the armed forces. We also claim that the Paraguayan PMP fostered adaptable public policies between1954-1981, but tended toward much more rigid policies in the1982-89 period. Cooperative arrangements among key stakeholders of the Stroessner regime were easier to sustain during the first sub-period because of two conditions: (i) a long time horizon for the regime based on Stroessner leadership, and (ii) the absence of acute recessionary conditions. When these conditions were challenged, the ability of the regime to mobilize resources for policy change declined, and the regime lost much of its capacity to adapt to exogenous shocks.

1. Adaptable Policies, 1954-81

Within the first years of the regime, Stroessner reversed some of the nationalistic policies of the 1940s (for instance, the monopoly of internal beef trade controlled by the Paraguayan Meat Corporation, Copacar). In 1957, a major IMF-sponsored stabilization program involved the deregulation of trade, the reduction of the monetary base, cuts in public expenditures, and a wage freeze. The ensuing stand-by loan in 1957-61 and the recovery of foreign credit allowed for an extensive road construction program (Campos and Canese, 1987: 27).

In the late 1960s, the distribution of public land and subsidized agricultural credit promoted the emergence of a local agricultural business sector related to the regime. This sector took advantage of the boom in soybean and cotton production during the following decade. At the same time, the construction of the Itaipú dam in the 1970s encouraged the emergence of a large construction sector, from less that 22 family-owned construction companies in the early 1970s to some 250 corporations after the Itaipú period (Borda, 1993: 76).

In fact, the Stroessner regime showed a significant capacity to reap the benefits offered by long-term economic opportunities. This tendency was certainly reinforced by the capacity to impose intertemporal policy deals. For instance, the development toward the Brazilian border in order to diversify a dependent pattern of economic relationships with Argentina required large investments and an active and stable foreign policy. This large-scale movement toward the Eastern border demanding coordination of land distribution, internal rural migration, and massive construction, among others, was possible due to the intertemporal "cooperation" of the key actors (the government, the Party and the armed forces). The adaptation of the development model to allow for an increasing integration with Brazil would have been unlikely under short lived governments like the ones that characterized the Post Chaco war period (1936-1954). In this 18 year period there were 12 different presidents and political volatility prevented the adaptation to changing economic environments.

During the 1956-1981 period the government distributed more than 88,000 farms covering 7.4 million hectares in 48 colonies (Lynn Ground, 1984). This land distribution represented 59% of the existing farms in 1956 and 35% of the existing farms in 1981. During the 1960s and the 1970s, Paraguay built roads, silos, and most important, jointly with Brazil the Itaipú hydroelectric dam, the biggest in the world. The long term growth strategy turned out to be effective. During the 1960s real GDP growth was of 4.2%. During the 1970s, Paraguay had one of the highest growth rates in the region with real GDP increasing at 8% over the decade.

2. Rigid Policies, 1982-89

In contrast to the 1954-1981 period, by the mid-1980s the "shadow of the future" was becoming increasingly shorter. The main reason was the dictator's aging and his (perceived) health deterioration, particularly since 1987 (Abente Brun, 1993: 147). Since the enforcer of intertemporal agreements was Stroessner himself, the problem of succession was a big one. In contrast to less institutionalized neo-patrimonial regimes in which power was easily transferred to the dictator's son after his death (Nicaragua in 1956, Haiti in 1971), the Militantes (the faction that sought a dynastic solution to the succession problem) confronted autonomous groups within the party and the armed forces.

Consistent with the prediction of the folk theorem in game theory, this intertemporal cooperation was more difficult to sustain when the end of the game was approaching with less uncertainty. In fact, the end of the Stroessner regime was ultimately a consequence of a struggle for succession (Martini and Lezcano, 1997).

The twilight of Stroessner's regime came also at a time when economic conditions became increasingly harsher. Following the conclusion of the Itaipú dam, the Paraguayan economy entered a deep recession that lasted well into the 1980s. At the same time, the debt crisis hurt two of Paraguay's main trade partners, Brazil and Argentina. Paraguay's GDP declined for two consecutive years, leaving production 4 % lower in 1983 than in 1981. Growth resumed in 1984, but the average rate over 1984-86 was well below the growth of population (World Bank, 1992).

The sectors that suffered most after 1982 -construction, and commerce and finance-were those that had grown most rapidly in the 1970s boom. By now, the support of these sectors had become very important for the Stroessner regime. Instead of forcing the domestic economy into conformity with the external conditions, the government tried to avert the post-Itaipu recession by continuing with its investment and spending programs. These policies only piled a financial disequilibrium on top of the recession. The economy still stagnated, and inflation accelerated. To finance its increased spending, the government sharply increased its external borrowing. Medium- and long-term debt (including arrears) rose from 15 % of GDP in 1981, to 62 percent in 1987. Moreover, the external funds were largely used on unprofitable projects, which added to the macroeconomic instability and eventually led to the suspension of disbursements from several creditors (World Bank, 1992).

3. Low Public Regardedness, 1954-89

Stronismo provides a good example of the independence between policy adaptability and public-regardedness. Even during its "golden" years, membership in the Colorado party was a must to conduct businesses with the government, and an extensive patronage system was administered by a nationwide network of seccionales. According to Fogel, the resources invested in the hydroelectric dams (76% corresponding to Itaipú and 24% to Yasyretá) were "channelled by a reduced group of civilian and military followers of the dictator nucleated in few companies (...) and ultimately devoted to speculation, the development of financial enterprises, and the acquisition of ranches; therefore the powerful economic groups integrated themselves to new economic sectors" (Fogel, 1993: 16).

Campos and Canese (1987, 88-104) identified a series of "inconvenient" public investments that began in the second half of the 1970s. Among them were the ACEPAR steel plant, the Aeropuerto del Este near President Stroessner City (today Ciudad del Este), the bridge over the Paraguay river at the level of Concepción City, the purchase of military helicopters and vessels, the construction of large hospitals, and the Chaco telecommunications program. The authors claimed that these projects shared a common pattern characterized by a dubious need for the venture, an oversized plan, the impossibility of recovering the investment, the evaluation of the proposal by interested parties, the resort to foreign loans, the avoidance of public biddings, and the systematic recourse to foreign suppliers. In trying to explain these "white elefants," the authors concluded that "the large magnitude of the public sector expenditures (...) leads to the conclusion that this is one of the most important ways to generate a surplus that is concentrated by a small number of people" (Campos and Canese, 1987: 82).

The lack of public regardedness also reinforced the policy rigidities of the 1980s. As an example of this problem, the recession of the 1980s did not lead to a reduction in clientelism or patronage politics. Quite on the contrary, public employment -which had grown at an average 4.5 % year in 1975-80- grew at a 4.8 % rate in 1980-84, and at a 5.6 % rate in 1984-87 (Campos and Canese, 1987: 65). The total number of public jobs expanded from 82 thousand in 1980 to 119 thousand in 1989 (Campos and Canese, 1990: table 4). In 1982, prior to the 1983 general election, the total wage bill in the public sector (excluding state enterprises) grew by 19 %, just at a time when the economy was decelerating. As a result, the share of the public wages on total public consumption grew from 58 % in 1981 to 68 % in the following years.

IV. THE POST-DICTATORIAL PERIOD (1989-1992)

1. The End of Party Unity

On February 3, 1989, a coup led by General Andrés Rodríguez drove Alfredo Stroessner out of power and initiated the transition to democracy. During this transitional period, the policymaking process initially operated under the existing rules -but now under a new leadership. It was not until 1992 that the adoption of a new Constitution re-structured the underlying rules of the game.

Following the "standard procedure" in place since 1954, Gen. Rodríguez was nominated as the Colorado candidate for the May 1989 presidential election and won with 74 % of the vote. However, the collapse of the Old Regime had three consequences. In the short-run, the arrival of a new coalition to the López palace, combined with a PMP that remained centered on the executive, allowed for a new period of rapid policy change. In the medium-run, the internal struggles within the new coalition progressively eroded the president's control over the Colorado Party and established the seed of factionalism discussed below. In the long run, increasing respect for civil liberties and the adoption of a new Constitution in 1992 facilitated the incorporation of new players into the policymaking process-and thus multiplied the number of veto points.

The ensuing atomization of the Colorado Party took place in five stages, starting with the election of a Constituent Assembly in late 1991 and leading to the splinter of the ANR in 2003:

1) Although the Colorado Party won a majority of the seats in the 1992 Constituent Assembly, the presence of internal factions created room for the formation of strategic coalitions with opposition forces. For instance, the Colorado leaders from the hinterland (the "Bancada Campesina") managed to establish the constitutional bases for the decentralization process (Barreda and Costafreda, 2002: 80), and the faction led by Luis M. Argaña was able to impose a constitutional ban on Rodríguez's presidential re-election.

2) In response to this move the commander of the Cavalry and President Rodríguez's right-hand, Gen. Lino Oviedo, blocked Argaña's presidential ambitions. In late 1992, when Luis M. Argaña and Juan Carlos Wasmosy confronted each other in the presidential primary, Oviedo intervened in the vote-counting process to ensure the defeat of Argaña. This move placed Gen. Oviedo in a highly influential position during the new Wasmosy administration after 1993.

3) In 1996, the conflicts within the Colorado Party entered a new stage. The insistence of Gen. Oviedo to encroach in the political process eventually led to a showdown with President Wasmosy.

In April 1996, Wasmosy ordered the retirement of his military ally and Oviedo responded with a failed insurrection (Ayala Bogarín and Costa, 1996). This action ultimately led to Oviedo's arrest and justified his proscription in the 1998 general election, even though the General had emerged as the favorite candidate in the Colorado primary, defeating Luis M. Argaña.

4) Because Oviedo was under arrest, his running mate Raúl Cubas Grau was placed as the official Colorado candidate for 1998. For legal reasons, Luis M. Argaña became his vice-president. President Cubas's decision to release Oviedo from prison immediately after he took office in August of 1998 created a new confrontation with the Argaña faction and an impeachment threat from Congress. There was some speculation that Congress would remove President Cubas and install Argaña as the new chief executive, when the Vice-President was shot in March of 1999. The killing of Argaña triggered a wave of protests (known as the "Paraguayan March") that ended with the resignation of Cubas and the installation of Luis González Macchi as interim president.

5) Over the next three years, the Oviedista faction reciprocated by attempting to unseat President González Macchi several times. A failed military coup took place in May 2000 and, in the midst of several corruption scandals, the Oviedistas and the Liberal Party attempted to impeach González Macchi at least three times.2 Still in exile and banned from running in the presidential election, Gen. Oviedo ordered the transformation of his Colorado faction into a new party, Unace (Unión Nacional de Colorados Eticos), for the 2003 race.3

The "triumvirate" of the Colorado party, the government, and the military that had characterized the Stroessner regime was significantly weakened by this transition process. Internal factionalism became the norm in the party, and traditional Colorado-military relations were eroded. In this context, new regulations instituted party primaries, banned the practice of deducting contributions to the Colorado Party from public sector pay checks (EIU, 1990, 1:20), and prevented party affiliation among military officers (EIU, 1990, 2:20).

2. Towards a Plural PMP

Since 1989, Paraguay has been undergoing slow but steady processes of democratization and decentralization. Freedom of speech and association have been granted to the people; elections have become free and progressively cleaner; a new constitution guaranteed a more balanced relationship among the executive, congress and judiciary systems; advances have been made toward the renovation of the judiciary system and many judges appointed for political reasons under the dictatorial regime have been removed.

The recovery of political rights by Paraguayan citizens was reflected in the electoral participation of new political forces. During the transitional period, the emergence of independent candidates and new political forces challenged the two-party system that had characterized Paraguayan politics. Independent candidates ran for positions in the National Assembly in late 1991. A new political party, the middle-class Partido Encuentro Nacional (PEN) achieved 23% of the total votes for president in the 1993 election. Additional parties (Oviedo's Unace, the Movimiento Patria Querida, and País Solidario) emerged in the early 2000s.

With the unfolding of the democratic transition, the opposition represented by the Liberal Party (PLRA) and the new political forces increasingly gained access to the policymaking process. The Executive Branch for the most part has remained under Colorado control (exceptions to this rule have been the capture of the Vice-President's office by the Liberals in the 2000 election, and the incorporation of the PEN to the cabinet in 1999-2003), but the adoption of proportional representation has made the congressional arena increasingly plural. Figure 1 depicts the transformation of Paraguay from a hegemonic party system into a multiparty system, using the effective number of parties index (Laakso and Taagepera, 1979).4 Before 1963, when Stroessner allowed token opposition into Congress, the effective number of parties was 1.0. Between 1963 and 1992, the average effective number of parties was 1.9. Over the last decade, however, multipartyism has emerged in the Paraguayan Congress, with currently 3.2 effective parties in the Chamber of Deputies.

FIGURE 1: Effective Number of Parties in the Chamber of Deputies, 1954-2004

3. Public Policies during the Transition

Between 1989 and 1993, the transitional regime preserved the executive-centered policymaking process under the command of a new chief executive. This PMP structure generated two results: the continuation of previous public policies (i.e., policy stability) along the dimensions in which the policy preferences of the old and the new regime coincided, and rapid policy change along the dimensions in which the preferences of the old and new regimes diverged.

In some key aspects, the policy preferences of the Stroessner and the Rodríguez administrations proved divergent, and an initial flurry of policy reform took place in the early 1990s. After a long period of post-Itaipú rigidity, economic policies during the transitional period (1989-1993) targeted the deregulation of the economy. This process included exchange rate liberalization, substantial tariff reductions, deregulation of agricultural prices, and deregulation of interest rates (Análisis del Mes, December 1989, 1990, 1991). The Rodríguez and Wasmosy administrations also privatized a few medium-sized public companies: LAPSA airlines (Líneas Aéras Paraguayas), the merchant fleet (Flota Mercante del Estado), the ACEPAR steel company (Aceros del Paraguay), and the Paraguayan Administration of Sugar-Cane Sprits (APAL).

Other policy dimensions, like the land-tenure policies, constituted less of a priority for the Rodríguez administration and the implementation of reforms was slow and incomplete. However, in the context of increasing public liberties, policy rigidity was confronted by the re-emergence of the peasant movement. Peasant mobilization was another signal of the growing pluralism in the policymaking process.

Although some public policies moved from rigidity to adaptability in the early 1990s, low public regardedness remained a constant outer feature. Changes in the judiciary system were slow; in 1991 government resources were still used to finance Colorado party in elections (Análisis del Mes, December, 1991); high-ranked military officers were still quite active in party politics until 1996 (Análisis del Mes, December, 1994; December, 1995; December, 1996); there was still violent repressions of peasant demonstrations, (Análisis del Mes, December, 1994) and corruption was not seriously attacked. For example, in 1990 the executive vetoed a law targeted to create a National Commission of Investigation of Corruption against the Public Sector; high government officials under the dictatorship were given soft punishment for corruption and human right violations (most sentences ranged from 2-5 years of prison), and accusations of corruption in the armed forces were not seriously investigated.

V. THE EMERGENCE OF A NEW POLICY MAKING PROCESS (1993-2003)

Economic policies in the early 1990s mainly focused on tax reforms, the balance of payments liberalization, financial liberalization, price deregulation, and the privatization of a few medium-sized public enterprises. However, most policy reforms took place in the period 1989-1992, and only marginal changes have occurred afterwards. Many areas of reform identified as crucial by key stakeholders have been stalled. Among those delayed reforms are the privatization of public enterprises in the telecommunications, safe water, and railroad sectors; the reform of social security and the health system, a civil service reform, the reform of the state-owned banks, the implementation of policies to promote rural land markets, and the modernization of the public sector management, including the strengthening of the regulatory system in the financial sector.

In this section we seek to characterize the emerging PMP during the period following the establishment of the new Constitution. This period includes the Wasmosy (1993-1998), Cubas Grau (1998-1999), and González Macchi (1999-2003) administrations, and the first semester of the Duarte Frutos government (August-December, 2003). In the first sub-section we show that the number of veto players and the rules of the PMP have changed considerably since 1993. In the second sub-section we discuss how the new conditions affect the outer features of public policies.

VI. THE NEW PLAYERS

In this section, we provide a brief discussion of the workings of the Paraguayan policymaking process following the adoption of the 1992 Constitution. We first describe the institutional players according to the new Constitution, and then we attempt to map the rules of the PMP. Because the underlying rules of the new PMP are in many ways still unchartered, we rely on detailed bill-initiation data to identify emerging regularities in the process.

1. The Executive

The Executive Branch has been considerably weakened by the 1992 Constitution. The Constitutional Assembly established that the president and vice-president would be elected by plurality for five-year terms with no re-election. It also deprived the president of the power to dissolve Congress, and endowed the executive with relatively weak "proactive" and "reactive" powers.5 The Paraguayan president can enforce policies unilaterally only if Congress fails to address an urgent bill within sixty days6, and his line-item veto can be overridden by an absolute majority of Congress.

2. Congress

Deputies and Senators are elected in concurrent elections every five years and can be reelected indefinitely. The Paraguayan Chamber of Deputies has 80 members who are elected from 18 districts for five-year terms. The deputies are elected from closed party lists using the d'Hondt divisor form of proportional representation in relatively small districts (the average district magnitude is 4.4). In theory, the closed-list system creates little incentives to cultivate the "personal vote" (Carey and Shugart, 1995; Hallerberg and Marier, 2004) but, as we show below, the legally-mandated party primaries and the relatively small districts have encouraged particularistic politics in the lower chamber. The Senate is composed of 45 members elected from closed party lists in a nation-wide electoral district. Thus in contrast to the deputies, who have clear incentives to promote departmental interests, senators are not representatives of local elites but rather visible party figures at the national level. In other respects, the Paraguayan bicameral system is highly congruent -meaning that the partisan composition of the House and the Senate is usually quite similar (Lijphart, 1999; Llanos, 2002).

Although congressional re-election is not banned by the constitution, the electoral context marked by competitive primaries, increasing party factionalism, and new political parties has created greater uncertainty about the political survival of legislators. In 1998, 54 % of the deputies were re_elected a figure comparable to Chile's lower house (Morgenstern, 2002: 416). Five years later, only 21 % of the deputies remained in their seats, a situation that resembled the turnover of Argentina's "amateur" legislators (Jones et al. 2002).7 The decline was consistent across all major parties.

3. Political Parties

Historically, partisan politics in Paraguay has centered on the competition between two nineteenth-century organizations, the Colorados (or National Republican Association, ANR) and the Liberals (nowadays called Authentic Radical Liberals, PLRA). Although the ANR has remained the largest party (winning 67% of the seats in the lower chamber in 1989, 50% in 1993, 56% in 1998, and 46% in 2003), the Liberals were able to capture the vice-presidency in the 2000 election (scheduled to fill_in the vacant position after the vice president was killed in 1999). Smaller middle-class parties (Encuentro Nacional, País Solidario, and Patria Querida) and a Colorado splinter (Unace) have been able to capture a few seats in Congress.

At the same time, the "granitic unity" of the parties during the Stroessner era has given way to factionalism and increasing internal competition in the form of primary elections to define the candidacies and the composition of party conventions. Figure 2 below displays the evolution of the Rice index of party unity for the Colorado and the Liberal parties in the Chamber of Deputies over the last nine years. The Rice index ranges from 0, when the party is evenly split on any legislative vote, to 100, when all the members of the party vote together (Rice, 1925).8 Because the Chamber of Deputies, in contrast to the Senate, collects roll-call votes on a regular basis, we were able to estimate the cohesion scores using 1409 controversial votes since 1995.9

FIGURE 2: Rice Index of Party Unity (ANR and PLRA, by Year)

Figure 2 suggests that the cohesion of Paraguayan parties at the present is quite low. At the peak of the factional confrontation in 1999, the Colorado party reached an abysmal unity score of just 30 points. The average ANR score during the González Macchi administration was 43 points. In contrast to the ANR, the PLRA has shown increasing cohesion over time, but the Rice index is not particularly high (75 on average for all years and 81 during the best year). In a comparative study, John Carey found average Rice scores of 91 points for Mexican parties, 81 for Chilean parties, 79 for Peruvian and Uruguayan parties, 75 for Brazilian parties, and 88 for Argentine parties (Carey, 2002). In his study of the Argentine Chamber of Deputies, Mark Jones found that the two largest parties (PJ and UCR) consistently had scores above 90 points between 1989 and 1997 (Jones, 2002).10

A long historical tradition of clientelism has encouraged factional leaders to build their own private clienteles in order to compete for the control of their parties.11 The legally mandated primaries have compounded the problem, triggering a competitive drive for enrolling new party members under the banner of each faction. Voters many times enroll to participate in the primaries in response to selective incentives provided by party bosses, but they are far from being "true" partisans. As an indication of this trend, former Senator Gonzalo Quintana has noted that: "The PLRA entered the race for the 1989 national election with some 70,000 members and obtained 240,000 votes. In the 2003 election, we entered the race with more than 600,000 affiliates and we could not reach 400,000 votes. (…) People do not join the party anymore; they are just hired by the candidates to register in the electoral roster for sole purpose of the primaries."

VII. THE NEW "RULES" OF THE POLICYMAKING PROCESS

Given the historical background described above and the institutional structure created by the 1992 Constitution, we anticipate the emergence of eight basic characteristics in the new Paraguayan policymaking process:

1. Given the tradition of strong presidentialism, we anticipate that the president will remain the initiator of all "relevant" policies, even though his constitutional powers have been limited.

2. Legislators confronted with legally-mandated primaries and with more competitive general elections are expected to initiate particularistic bills for credit-claiming purposes and major bills for position-taking purposes. Given the declining re-election rates and the tradition of a weak legislature, we do not anticipate legislators to be interested in grand policymaking.

3. The increasing fragmentation of the ANR has eroded the party's capacity to play the role of a "legislative dictator." Therefore, we hypothesize that the capacity of successive presidents to enforce policy change has declined following the Rodríguez administration.

4. Because of the re-election ban on the executive, the capacity of the president to enforce policy change in key areas is further expected to decline as the end of the term approaches.

5. Points 1 to 4 suggest that the ability of Congress to operate as a veto player has increased. Thus, the number of controversial bills is expected to grow over time.

6. Controversy over bills (and thus a greater role of veto players) is expected to concentrate on comprehensive (nation-wide, region-wide, or sector-wide) policies of regulatory or redistributive intent. In contrast, particularistic policies (distributive bills with a local focus) of low visibility and low marginal cost will be less likely to generate frictions within the legislature or in executive-legislative relations.

7. Given 5 and 6, it is possible that the Paraguayan system may provide relative flexibility for the provision of particularistic (privately-regarded) policies, but relative rigidity for the provision of comprehensive regulatory or redistributive policies.

1. Empirical Evidence

In order to map the emerging Paraguayan policymaking process and to test these expectations, we built a database with all the legislative bills (proyectos de ley) that received final treatment in Congress between 1993 and 2003. By final treatment we mean that: a) the bill was approved and signed by the president; b) the bill was rejected by Congress or vetoed by the executive; or c) the bill was treated for the last time at some point between 1993 and 2003 but the legislators made no final decision (i.e., the bill was withdrawn, archived, or is pending). The database covers 4576 bills, virtually every bill introduced in Congress since April 1992 until December, 2003.

2. Policy Initiation

Who initiates the policymaking process under the new system? Table 1 compares the rate of bill initiation by branch and by administration. On average, the executive branch has initiated about one third of the bills during the democratic era. There is no clear indication of a secular trend in terms of a growing or declining role of the president as policy initiator. The somewhat lower scores for the Rodríguez and Duarte administrations are presumably artifacts of the truncated sample (we are only dealing with the last months of the Rodríguez administration, when the president was already a lame duck, and just a few months of the Duarte administration).

We expected the type and relevance of the policies initiated by the executive to differ from those initiated by Congress. In order to test this idea, we classified the content of the bills according to three standard criteria in the literature. The first one is the intent of the policy (distribution of benefits to constituencies, social redistribution of resources, or regulation of social and economic activities) according to the typology outlined by Lowi (1964). The second criterion is the scope of the bill (national, local, sectoral, individual) as described by Taylor-Robinson and Diaz (1999). Our final criterion identifies the key policy areas for structural reform according to Lora and Panizza (2002, 2003).

TABLE 1: Initiation of Congressional Bills, by Administration (1992-2003)

Table 2 presents the distribution of bills by initiator, according to Lowi's typology. Against our initial expectations, the executive does not dominate the initiation of redistributive and regulatory policy, and it is not less active than the legislators in the promotion of distributive policy. The president clearly dominates the initiation of policies in the "others" category (a residual category that includes all foreign policy issues).

TABLE 2: Bill Initiation, by Policy Goal (1992-2003)

There is, however, a clear division of labor in terms of the scope of the policies initiated by different branches, as shown in Table 3. The executive dominates the formulation of nation-wide policies and policies related to the public sector (constitutionally, the executive has gatekeeping power over the initiation of the budget). In contrast, Congress dominates the production of individual-level and local bills, and more surprisingly, the production of bills targeted to the private sector.

Interestingly enough, the lower house dominates the production of bills related to local and regional issues -suggesting that the electoral system gives deputies a clear incentive to represent the interests of their districts. However, in spite of being elected in a nation-wide district, senators have displayed a significant interest in individual-level bills, while deputies also showed some interest in sectoral bills.

TABLE 3: Bill Initiation, by Policy Scope (1992-2003)

Tables 2 and 3 thus suggest that congress members have a greater propensity to initiate particularistic bills. We define as particularistic any policy characterized by i) a distributive purpose and ii) a limited scope (individual or municipal range). Particularistic policy is normally related to the distribution of pork and has a distinctively low level of public-regardedness (Cox and McCubbins, 2001). Under this definition, we find that only 7% of all the bills initiated by the executive during the period under study can be classified as particularistic, as opposed to 50% of all the bills initiated by the Chamber of Deputies, and 55% of all the bills initiated by the Senate.

We could not find any evidence of executive dominance in the key policy arenas for structural reforms. During the period under study, the executive branch initiated 6 of the 13 (42%) bills that explicitly focused on trade and tariffs, 30 of the 70 (43%) bills centered on the financial system, 14 of the 66 (21%) bills related to the tax system, 6 of the 31 (19%) bills dealing with labor policy, and only one of the 12 (8%) bills dealing with privatization policies. This does not necessarily mean that the median legislator initiated the most critical policies adopted in Paraguay. It is likely that many legislators initiate bills mainly for "position taking" purposes. In the following sub-section we show that the bills initiated by the legislature are less likely to pass.

We found some evidence of an electoral cycle driving bill initiation. Figure 3 shows the average number of bills introduced in Congress and the number of months pending to the next general election. Although generally speaking, bill initiation is not driven by the electoral cycle (downward spikes simply reflect the congressional summer recess), the share of particularistic bills (as percentage of the total bills initiated) grows by about 10 percent points during the last 12 months prior to a general election (sig. at .01 level), even after we control for the effects of the summer recess.

FIGURE 3: Average Number of Monthly Bills Introduced in Congress, by Closeness to Next General Election

3. Veto Players and Policy Approval

The distinctive characteristic of the 1954-89 period was the virtual absence of significant veto players. This pattern of policymaking was extended briefly into the Rodríguez administration, but it was progressively dismantled in the early 1990s. The 1992 constitution strengthened the position of Congress vis-à-vis the president and, as explained in the previous chapter, the process of democratization allowed for an increasing number of parties and internal party factions. In practical terms, this means that the Paraguayan president moved from controlling a "granitic" Colorado party with 67% of the seats in congress in 1963 to bargaining with a factionalized party with 46% of the seats in 2003.

To assess the impact of this process fragmentation on the PMP, we classified all bills according to their fate. Policy initiatives were coded as: 1) approved (signed into law); 2) rejected; 3) stalled (archived, withdrawn, or pending).12 Figure 4 shows the success rate (percentage approved) of the executive and legislative initiatives for each administration. The success rate of executive_initiated bills declined from 96% during the last 16 months of the Rodríguez administration (a significant figure considering the lame-duck position of the president) to 65% during the González Macchi administration. González Macchi's success rate is comparable to the average found in a study of 11 presidential countries (63%) and clearly lower than the rate enjoyed by other presidents with majority parties in congress (72%) (Cheibub, Przewoski, and Saiegh, 2003: 723). Figures for the Duarte administration could not be interpreted given the truncated sample (80% of the bills initiated by the executive and 93% of the bills initiated by Congress were still pending resolution by the end of 2003).

FIGURE 4: Success Rate of Legislative Bills, by Administration and Initiator (1992-2003)

Note: Rodríguez administration only covered for bills initiated between April 1992 and August 1993.

It is clear from Figure 4 that the model of executive-dominated policymaking has been increasingly challenged by the process of political fragmentation. But it is also true that fragmentation has hurt the capacity of legislators to pass their preferred policies. The rate of congress-initiated bills approved has declined from 73% during the last months of the Rodríguez administration to about 38% in the present.

Table 4 presents information on the success rate of bill initiatives by policy area. The data suggest that particularistic policies (distributive in nature and limited in scope) are the less controversial initiatives and the more likely to pass. Only the bills dealing with the public sector (budget and administrative issues) have similar rates of approval.

TABLE 4: Success Rate of Legislative Bills, by Policy Area (1992-2003)

Based on this information, Table 5 presents a logistic regression that models the simultaneous effects of institutional factors and policy types on the probability of policy change (as measured by policy approval and presidential vetoes). The dependent variable in Model 5.1 indicates whether a bill was ultimately approved (as opposed to rejected or stalled). The baseline category corresponds to a generally uncontroversial category of policy: national bills which are not distributive, redistributive, or regulatory (e.g., foreign policy issues). The results of Model 5.1 suggest that: a) individual-level policy initiatives are more likely to pass than the bills in the consensual baseline category, and municipal-level and distributive bills are equally likely to succeed, supporting the hypothesis that particularistic policy is relatively non-controversial; b) policies with a regional focus, as well as distributive and regulatory policies are more controversial than the rest; c) executive-initiated bills are more likely to be successful than congressional bills, but this advantage has eroded with successive administrations, as they faced increasing levels of political fragmentation; d) although there is no clear evidence of an electoral policy cycle, the results suggest that the capacity of the system to promote policy change declines as the president approaches the end of the term.

TABLE 5: Success and Presidential Veto of Legislative Bills, by Initiator and Policy Area

In order to illustrate the substantive significance of Model 5.1, we present two simulations (Figure 3 and Table 6). Figure 5 illustrates the effects of political fragmentation and the no-reelection clause on the approval of executive-initiated bills (for simplicity, we assume all bills to be national in scope and regulatory in nature). The simulation shows that: a) as the party system became more fragmented, every president after Andrés Rodríguez has enjoyed lower levels of policy success than his predecessor; and b) in every case, the effectiveness of presidential initiatives declined as the president turned into lame-duck.

FIGURE 5: Expected Probability of Success for Administrations (1992-2003). (Simulation assumes regulatory bill, national in scope)

Table 6 depicts the expected probability of success for bills initiated by Congress in different policy areas.13 The estimation shows that the Paraguayan policymaking process is well-equipped to deliver particularistic policy (distributive policy with an individual or local focus). Bills in this category have the highest success rate, ranging between 55 and 61%. In contrast, the capacity of the legislative process to produce regulatory or redistributive policy at the regional, sectoral, or even national level seems much weaker (16-35%). This general pattern holds even if we assume that the executive branch and not the legislators initiate the proposals. A "contemporary" president (someone with a leverage in between Cubas Grau and González Macchi), initiating a bill in his 30th month in office would face an expected probability of success ranging between 81% (for individual, distributive bills) and 35% (for regional-redistributive bills).

TABLE 6: Expected Probability of Success, by Policy Area

To verify these findings, we selected the occurrence of presidential vetoes as an alternative measure of controversy in the policy areas (Model 5.2). This measure is somewhat orthogonal to approval (a vetoed bill may still become law if Congress overrides a package veto or accepts the president's partial veto) and it is intended to measure controversy around a bill rather than its success. Our dataset contains 193 episodes of partial or total vetoes, of which information on the PMP is complete for 104 cases. Model 5.2 confirms the idea that particularistic policies are less controversial (and therefore less likely to be vetoed by the executive), while redistributive and regulatory policies are more likely to generate friction in the policymaking process.

VIII. PUBLIC POLICY UNDER THE NEW PMP

How do public policies in the current period compare to policies in previous periods? The patterns described in the previous section suggest that the new policymaking process displays both traditional and novel characteristics.

1. Stability

The Stroessner dictatorship was able to impose intertemporal policy deals until the succession crisis loomed in the political horizon. In contrast, the current PMP seems to have little ability to enforce long-term transactions, given: a) the ban on presidential reelection, b) the discord and the political realignments within the ANR, and c) the focus of the opposition in displacing the ANR from power in the medium-run.

This conclusion, however, must be qualified on three grounds. First, the current system seems to manage the production and distribution of particularistic policies. Second, the increasing number of veto players may ultimately impose a low rate of policy change in the years to come. These two conditions, however, may add rigidity (rather than stability) to the existing policies. Third, it must be noted that a straight comparison between the stability of the Stroessner policies and the uncertainty of the current policy regime may be deceiving. There is an underlying selection bias problem because at the core of the Stroessner model was the elimination of policy controversies -and precisely these are the instances in which the PMP must generate either short-term policy volatility or stable intertemporal transactions.

2. Adaptability

The Stroessner model initially generated significant levels of policy adaptability, followed by increasing rigidity in the 1980s as the privately_regarded policies ossified. In part, these levels of adaptability resulted from the capacity of the regime to marshal resources around long-term strategic projects like the "March towards the East". This capacity to marshal resources towards policy change also generated some negative effects, like inefficient patterns of public investment. It is clear that the current policymaking process for the most part lacks such capacity. Regulatory bills initiated by the executive faced declining rates of success: 87 % were approved by Congress during the last year of the Rodríguez administration, 69 % during the Wasmosy term, 60 % during the short Cubas period, and 48 % during the González Macchi administration. Low decisiveness is the product of a weakened executive coupled with legislators with little interest in policymaking. Thus, the potential for policy adaptability may be low, particularly in the areas of regulatory and redistributive policy, where controversial issues are more likely to arise.

3. Public Regardedness

Like the Stroessner system, the current PMP displays a very low level of public regardedness. This problem has adopted three forms: corruption, patronage, and pork-barrel politics. In 2002, Paraguay ranked 98 out of 102 countries in the Transparency International Corruption Perceptions Index -the lowest Latin American score. The country occupied position 140 (out of 155 ranked countries) in the Kaufmann et al. index of Graft (Kaufmann, Kraay, and Zoido-Lobatón, 1999). At the same time, the factionalization that characterizes Paraguayan politics has permeated the public bureaucracy. Public hiring practices are characterized by a high degree of arbitrariness. These hiring practices are based principally on political favors and clientelism. In a recent paper, Ugo Panizza estimates public sector wage premiums for a cross section of countries. While the average Latin American public sector employee earns roughly 4 percent more than workers with similar characteristics employed by the private sector, the wage premium for Paraguay is around 17 % (Panniza 1999). Although less visible in the media, particularistic policymaking may be a more relevant dimension of public regardedness. Particularism is the legal manifestation of the low public regardedness that permeates the system. About 38 % of all the bills introduced in congress and 40 % of the all bills passed are distributive in nature and narrow in scope. We have shown that particularistic policies take most of the legislators' efforts, they are non-controversial, and they are less likely to be vetoed by the executive.

In contrast to the Stroessner system, the current PMP displays greater inclusiveness, lower coordination, and higher decentralization of the distributive process. Therefore, the system seems to be quite flexible for the production and distribution of particularistic policies (individual pensions, public jobs, etc.). At the same time, the initiation of such bills seems to be driven by clearly electoral considerations, and the combination of legally-mandated primary races and contested general elections may increase the competition for pork among the legislators in the future.

4. Institutionalization

Maybe the most important contrast with the Stroessner model, however, is the greater potential for institutionalization of the current policymaking regime. Although the struggles within the Colorado party generated outbursts of political instability in 1996, 1999, and more recently in 2000, the absence of a "succession problem" looming in the horizon creates a more propitious context for developing a learning process about how to build intertemporal policy agreements. Thus, although the Stroessner system supplied a dominant executive capable of enforcing intertemporal deals in the medium-run, it failed to provide a policy regime with longer horizons. In contrast, the current system creates a weaker executive with little capacity to enforce pacts, but it may generate expectations for stable rules of the game in the long-run.

TABLE 7: Comparison of the Policy Making Process in Different Historical Moments

Table 7 summarizes for each of the four periods the key players, the policy initiation, the effective number of parties, the veto players, the capacity to enforce intertemporal agreements, the policy adaptability, the level of public regardedness, and the capacity to enforce rules in the long run.

Table 7 underscores two main theoretical conclusions. First, the presence of multiple veto players may be neither necessary nor sufficient to create policy rigidities. True, the current period contrasts with the "golden age" of the Stronismo and with the Rodríguez transitional regime because during these two periods executive concentration of power allowed for fast policy change. But when partisan factions converge in their policy preferences (as in the case of particularistic policies), policy change is easily achieved. At the same time, the presence of a policy dictator did not prevent policy rigidity in 1982-89. Second, the historical evidence supports the idea that policy adaptability and public-regardedness are independent dimensions of the policymaking process. Even though adaptability has varied over time, low public regardedness has remained a constant feature.

 

* Research for this paper was supported by the Inter American Development Bank through its project "Political Institutions, Policymaking Processes, and Policy Outcomes". Marcela Montero (Desarrollo) coordinated the effort to develop the database on bill initiation. We are indebted to Mariano Tommasi, Carlos Scartascini, Ernesto Stein, Pablo Spiller, and two reviewers of the Revista de Ciencia Política for their valuable comments.

1 Originally, the 1967 constitution established re-election for only one more period; but after the 1977 reform the President could be indefinitely re-elected.

2 An impeachment finally took place in early 2003, but the Senate acquitted González Macchi.

3 Lino Oviedo returned to Paraguay in late June 2004. At the time of this writing, he confronted a sentence of ten years in prison.

4 The Effective Number of Parties is an index akin to the Hirschman-Herfindahl index of market concentration (HHI) that weights the distribution of legislative seats among political parties. The formula for the index is ENP=1/Ó(p2), where p represents the proportion of seats controlled by each party in the lower chamber.

5 Consistent with Shugart and Carey (1992), Payne et al. (2002) presented the reactive powers of the Paraguayan president as "moderate" vis-à-vis the "weak" reactive powers of the Costa Rican and Honduran presidents who presumably lack partial veto powers. The fact is that articles 126-127 of the Costa Rican constitution allow the president to "amend" bills passed by Congress and return them to the assembly for reconsideration, and article 220 of the Honduran Constitution is ambiguous about the possibility of a partial veto.

6 The 1992 Constitution (art. 210) further constrains the president's proactive powers in three ways: 1) the president is not allowed to issue unilateral decrees and must instead introduce "urgent" bills in Congress (that he can legally enforce if Congress fails to act in sixty days); 2) the president is not allowed to introduce more than three urgent bills per year; and 3) Congress can lift the urgency character of a bill with two-thirds of the votes, reverting to the normal policymaking process. Because of this reason, the Paraguayan proactive powers are generally "toothless."

7 These figures are estimates of unconditional re-election, based on roll-call data. We do not know how many deputies were actually placed in the lists.

8 The formula for the index is R=|Ayes-Nays|/(Ayes+Nays)*100.

9 Controversial votes were defined as decisions in which at least 25% of the legislators voted against the winning side (Mainwaring and Pérez-Liñán 1997). The focus on controversial votes guarantees that minor legislative decisions, which are normally consensual and thus inflate party unity scores, are not given undue influence.

10 Comparison of unity scores is always difficult because roll-call votes are not equally frequent in all legislatures and researchers select votes using slightly different criteria. Instead of dropping all non-controversial votes, Carey weighted votes according to their closeness and the number of legislators absent. Jones provided "relative" unity scores, ignoring absent legislators and abstentions.

11 The practice of bitter factional disputes within the two traditional parties was interrupted in the Colorado Party by the "granitic" leadership of Alfredo Stroessner (1954-89), but it remained an essential part of the Liberal Party even during this period.

12 Some bills are archived when their content is incorporated into a larger initiative. So we understand that in few cases archived bills may have not been really stalled. Unfortunately, there is no practical way of identifying those bills.

13 Simulations are based on Model 5.1. It is assumed that bill was initiated by congress, that the administration is in the middle of the term, and that there are no intermediate elections.

REFERENCES

15 Abente Brun, Diego. 1993. "Las etapas de la transición y el camino por recorrer." In Paraguay en Transición, edited by D. Abente Brun. Caracas: Editorial Nueva Sociedad, 147-59.         [ Links ]

16 Arditi, Benjamín. 1993. "Del granito al archipiélago: el Partido Colorado sin Stroessner." In Paraguay en Transición, edited by D. Abente Brun. Caracas: Editorial Nueva Sociedad, 161-172.         [ Links ]

17 Análisis del Mes (Several Issues). Servicio de Consultoría Informativa. BASE: Asunción, Paraguay.         [ Links ]

18 Ayala Bogarín, Oscar, and José María Costa. 1996. Operación Gedeón: Los secretos de un golpe frustrado. Asunción: Editorial Don Bosco.         [ Links ]

19 Barreda, Mikel, and Andrea Costafreda. 2002. "La Transición Democrática y el Sistema Político-Institucional." In Diagnóstico Institucional de la República del Paraguay, edited by J. Prats i Català. Barcelona: IIG_PNUD, 71-126.         [ Links ]

20 Barreda, Mikel, and Andrea Costafreda. 2003. "El Sistema Político." In Libro Branco sobre la Reforma Institucional en Paraguay, edited by J. Prats i Català. Barcelona: IIG-PNUD, 11-26.         [ Links ]

21 Borda, Dionisio. 1993. "Empresariado y transición a la democracia en Paraguay." In Paraguay en Transición, edited by D. Abente Brun. Caracas: Editorial Nueva Sociedad, 69-104.         [ Links ]

22 Campos, Luis A., and Ricardo Canese. 1987. El Sector Público en Paraguay: Análisis de sus Inversiones y Empresas. Asunción: Centro Interdisciplinario de Derecho Social y Economía Políticas.         [ Links ]

23 Campos, Luis A., and Ricardo Canese. 1990. La Reestructuración Democrática del Sector Público. Asunción: CEDES-BASE.         [ Links ]

24 Carey, John M. 2002. "Getting Their Way, or Getting in the Way? Presidents and Party Unity in Legislative Voting." Presented at the American Political Science Association, Boston, August.         [ Links ]

25 Carey, John M., and M. S. Shugart. 1995. "Incentives to Cultivate a Personal Vote: a Rank Ordering of Electoral Formulas." Electoral Studies 14 (4):417-439.         [ Links ]

26 Cheibub, José A., Adam Przeworski, and Sebastian Saiegh. 2004. "Government Coalitions and Legislative Success Under Presidentialism and Parliamentarism." British Journal of Political Science (Forthcoming).         [ Links ]

27 Cox, Gary W., and Mathew D. McCubbins. 2001. "The Institutional Determinants of Economic Policy Outcomes." In Presidents, Parliaments, and Policy, edited by S. Haggard and M. D. McCubbins. Cambridge: Cambridge University Press, 21-63.         [ Links ]

28 EIU (Economist Intelligence Unit). (Several Issues) Country Profile: Uruguay, Paraguay, No.1. London.         [ Links ]

29 Fogel, R. 1993. "La estructura social paraguaya y su incidencia en la transición a la democracia." In Paraguay en Transición, edited by D. Abente Brun. Caracas: Editorial Nueva Sociedad, 13-30.         [ Links ]

30 Hallerberg, Mark, and Patrik Marier. 2004. "Executive Authority, the Personal Vote, and Budget Disciplinein Latin American and Caribbean Countries." American Journal of Political Science 48 (3):571-87.         [ Links ]

31 Jones, Mark P. 2002. "Explaining the High Level of Party Discipline in the Argentine Congress." In Legislative Politics in Latin America, edited by S. Morgenstern and B. Nacif. Cambridge: Cambridge University Press, 147-184         [ Links ]

32 Jones, Mark P., Sebastian Saiegh, Pablo T. Spiller, and Mariano Tommasi. 2002. "Amateur Legislators-Professional Politicians: The Consequences of Party-Centered Electoral Rules in a Federal System." American Journal of Political Science 46 (3):656-669.         [ Links ]

33 Kaufmann, Daniel, Aart Kraay, and Pablo Zoido-Lobatón. 1999. "Aggregating Governance Indicators". Washington, DC: The World Bank - Policy Research Working Paper Nº 2195.         [ Links ]

34 Laakso, Markku, and Rein Taagepera. 1979. "`Effective' Number of Parties: A Measure with Application to Western Europe." Comparative Political Studies 12 (1):3-27.         [ Links ]

35 Lambert, Peter. 1997. "The Regime of Alfredo Stroessner." In The Transition to Democracy in Paraguay, edited by Lambert and Nickson. New York: St. Martin's Press.         [ Links ]

36 Lambert, Peter, and Andrew Nickson. 1997. The Transition to Democracy in Paraguay. New York: St. Martin's Press.         [ Links ]

37 Lijphart, Arend. 1999. Patterns of Democracy - Government Forms and Performance in Thirty-Six Countries. New Haven: Yale University Press.         [ Links ]

38 Llanos, Mariana. 2002. "El Bicameralismo en América Latina." Presented at the Tercer Congreso Internacional de Latinoamericanistas en Europa (CEISAL), Amsterdam, July 3-6.         [ Links ]

39 Lewis, P. 1982. Socialism, Liberalism and Dictarship in Paraguay. New York:Praeger.         [ Links ]

        [ Links ]

41 Lora, Eduardo, and Ugo Panizza. 2003. "The Future of Structural Reform." Journal of Democracy 14 (2):123-137         [ Links ]

42 Lowi, Theodore. 1964. "American Business, Public Policy, Case-Studies, and Political Theory." World Politics 16 (July):677-715.         [ Links ]

43 Lynn Ground, Richard. 1984. "El auge y la recesión de la economía paraguaya, 1972-1983. El papel de la política económica interna". In CPES, Economía del Paraguay Contemporáneo, Vol. II. CPES. Asunción, pp. 515-574.         [ Links ]

44 Mainwaring, Scott, and Aníbal Pérez-Liñán. 1997. "Party discipline in the Brazilian Constitutional Congress." Legislative Studies Quarterly 22 (4):453-483         [ Links ]

45 Martini, C. and C.M. Lezcano. 1997. "The Armed Forces." In The Transition to Democracy in Paraguay, edited by Lambert and Nickson. New York: St. Martin's Press.         [ Links ]

46 Miranda, C. 1990 Paraguay y la Era de Stroessner. Asunción: R.P Ediciones.         [ Links ]

47 Morgenstern, Scott. 2002. "Explaining Legislative Politics in Latin America." In Legislative Politics in Latin America, edited by S. Morgenstern and B. Nacif. Cambridge: Cambridge University Press, 413-445.         [ Links ]

48 Nickson, A. 1997. "Corruption and the Transition." In The Transition to Democracy in Paraguay, edited by Lambert and Nickson. New York: St. Martin's Press.         [ Links ]

49 Panizza, Ugo. 1999. "Why Do Lazy People Make More Money? The Strange Case of the Public Sector Wage Premium." Inter-American Development Bank Working Paper No. 403.         [ Links ]

50 Payne, J. Mark, Daniel Zovatto G., Fernando Carrillo Florez, and Andrés Allamand Zavala. 2002. Democracies in Development - Politics and Reform in Latin America. New York: Inter-American Development Bank and International Institute for Democracy and Electoral Assistance.         [ Links ]

51 Plan Nacional Anticorrupcion, Asunción, 2000 (www.cisni.org.py).         [ Links ]

52 PNUD-IIG, 2003. Diagnóstico Institucional de la República del Paraguay.         [ Links ]

53 Rice, Stuart A. 1925. "The Behavior of Legislative Groups: A Method of Measurement." Political Science Quarterly 40 (1):60-72.         [ Links ]

54 Shugart, Matthew S., and John M. Carey. 1992. Presidents and Assemblies. Constitutional Design and Electoral Dynamics. Cambridge: Cambridge University Press.         [ Links ]

55 Spiller, Pablo T., and Mariano Tommasi. 2003. "The Institutional Foundations of Public Policy: A Transactions Approach with Application to Argentina." Journal of Law, Economics, and Organization 19 (2):281-306         [ Links ]

56 Taylor-Robinson, Michele M, and Chris Diaz. 1999. "Who Gets Legislation Passed in a Marginal Legislature and is the Label Marginal Legislature Still Appropriate? A Study of the Honduran Congress." Comparative Political Studies 32 (5):589-625         [ Links ]

57 Tsebelis, George. 2002. Veto Players: How Political Institutions Work. Princeton: Princeton University Press.         [ Links ]

58 World Bank, 1992. Paraguay: Country Economic Memorandum. Washington D.C.         [ Links ]

59 World Bank, 2003. Paraguay: Creando las Condiciones para un Crecimiento Sustentable. Junio 2003.         [ Links ]José Molinas es Director del Instituto Desarrollo, Asunción, Paraguay. Es Ph. D. en Economía por la Universidad de Massachussets, Amherst. Ha sido Director del Consejo Impulsor del Plan Nacional Anticorrupción, y Coordinador de la Estrategia Nacional de Reducción de Pobreza en Paraguay. Ha realizado extensivamente trabajos de consultoría para organismos multilaterales en América Latina, África y Europa del este. Sus áreas de investigación son desarrollo rural, macroeconomía y pobreza, capital social, políticas sociales y economía de la educación. Ha publicado artículos en World Development, Economía y Sociedad, y como capítulos de libros publicados por Elgar Press, Eudeba y Alfaomega, entre otros.

(E-mail: Jmolinas@desarrollo.edu.py)

Aníbal Pérez Liñán es profesor de ciencia política e investigador del Centro de Estudios Latinoamericanos en la Universidad de Pittsburgh, Pensilvania. Ha publicado artículos en la Revista SAAP y la Revista de Ciencias Sociales (Argentina), Dados (Brasil), América Latina Hoy (España), Electoral Studies, Latin American Research Review (Estados Unidos) y Democratization (Gran Bretaña), entre otras. Actualmente está completando un libro sobre la utilización del juicio político contra los presidentes en América Latina.

(E-mail: asp27+@pitt.edu)

Sebastián Saiegh (Dr.en Ciencia Politica, New York University) es profesor en la Universidad de Pittsburgh. Se ha desempeñado como consultor del Banco Interamericano de Desarollo (BID). Sus areas de investigación incluyen economía política, modelos formales de instituciones legislativas y federalismo. Ha publicado en American Journal of Political Science, British Journal of Political Science, Comparative Political Studies, y Economía. Es coeditor (junto con Mariano Tommasi) del libro La Nueva Economía Política (Eudeba, 1998).


 

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons