SciELO - Scientific Electronic Library Online

vol.29 número3El faccionalismo en el Partido Socialista de Chile (1990-2006): características y efectos políticos en sus procesos de toma de decisiónMidiendo la libertad civil: Una evaluación de base de datos estandarizados índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados




Links relacionados


Revista de ciencia política (Santiago)

versión On-line ISSN 0718-090X

Rev. cienc. polít. (Santiago) v.29 n.3 Santiago  2009 

REVISTA DE CIENCIA POLÍTICA/VOLUMEN 29 / N° 3 / 2009 / 693 - 720



Veto players in presidential regimes: institutional variables and policy change*

Jugadores con veto en regímenes presidenciales: Variables institucionales y cambio en las políticas públicas




1 University of Pittsburgh (,
Universidad de los Andes


This paper develops a computational model to explore the effect of veto players in presidential regimes. We discuss the relationship between the legislative powers of the president and the configuration of veto players, and analyze how the number of legislative parties and their cohesion affect policy stability we develop a computational model to estimate the simultaneous impact of these factors in a conventional two-dimensional policy space. The results suggest that party fragmentation has mixed implications while decree authority is a key factor explaining policy stability. The impact of most institutional variables is conditional on the position of the players.

Key words: Veto Players, Presidential Regimes, Legislative Parties, Party Fragmentation, Legislative Power.


Este estudio presenta un modelo computacional para analizar el efecto de los jugadores con poder de veto en regímenes presidenciales. Exploramos la relación entre los poderes legislativos del presidente y la configuración de jugadores con veto, y analizamos deque manera el número de partidos legislativos y su cohesion afectan la estabilidad de las políticas públicas. A partir de estas ideas desarrollamos una simulación para evaluar el impacto conjunto de estos factores en un espacio bidimensional. Los resultados sugieren que la fragmentación partidaria tiene efectos mixtos, mientras que el poder de decreto es un factor clave para explicar la estabilidad de las políticas. El impacto de las variables institucionales está en general condicionado por la ubicación de los actores políticos.

Palabras clave: Jugadores con poder de veto, regímenes presidenciales, Partidos legislativos, Fragmentación partidaria, Poderes legislativos.

This paper expands the theory of veto players into the realm of presidential regimes. We bridge two strands in the study of comparative institutions: the work on veto players in parliamentary systems (Tsebelis 1999; Tsebelis 2002) and the analysis of executive-legislative relations under presidentialism (Mainwaring and Shugart 1997b; Shugart and Carey 1992). Early studies of veto players in presidential systems (Cameron 2000; Krehbiel 1998) have not always shared the same theoretical assumptions or even the same language used by students of parliamentarism, and only recently some systematic attempts to bridge this gap have taken place (Alemán and Tsebelis 2002; Ames 2001). At the same time, a burgeoning comparative literature on presidentialism has generated important conclusions about policy stability and policy change, many times borrowing (but not always formalizing) the assumptions of the veto-player theory.

The emerging study of veto players in presidential regimes has confronted two major problems. On one hand, soft rational choice analyses of presidential regimes have tended to adopt the veto player model as a useful heuristic without paying much attention to its underlying spatial assumptions. As a result, as we show below, the literature has generated intuitive but contradictory hypotheses about the effects of some institutions. On the other hand, formal models of veto players have tended to address the impact of specific institutions in isolation, without pursuing a "general equilibrium" approach in which institutional effects are modeled simultaneously The empirical literature, however, suggests that institutional variables interact (and sometimes cancel each other) in order to shape policy outcomes.

We seek to address these problems by developing a computational model of veto players in presidential regimes. In the first part of the paper we introduce the basic concepts and outline the setup for the analysis. In the following sections we revise the literature and develop a series of systematic hypotheses about presidential systems. Section two explores the relationship between the legislative powers of the president and the configuration of veto players. Section three discusses how the number of legislative parties and their infernal cohesion affect policy stability. In the fourth section, we use the computational model to estimate the simultaneous impact of these factors in a two-dimensional model of policy stability. The analysis does not intend to provide an empirical test of the hypotheses, but rather to establish to what extent different predictions about institutional effects are consistent with the standard veto player assumptions usually invoked to justify them. The simulation suggests that decree authority must be deemed as a critical factor affecting policy stability in presidential regimes, that the effective number of parties may have inconsistent effects, and that the impact of most other institutional variables is conditional on the spatial location of the key players.


Veto players are political actors whose approval is necessary but not sufficient to alter the policy status-quo. They should be distinguished from decisive players, whose approval is sufficient but not necessary to change existing policies, and from dictators, whose approval is both necessary and sufficient to impose new policies (Strom and Swindle 2002). Multidimensional models of veto players seek to predict the probability of policy change rather than the specific policies that will emerge in equilibrium. The theory assumes that any proposal representing an improvement over the status-quo for all veto players will not face resistance and thus will defeat the existing equilibrium. The set of policy positions thus capable of defeating an existing policy is known as the winset of the status-quo, W(SQ). The larger W(SQ), the greater the probability of policy change and henee the lower policy stability (Tsebelis 1995). Foliowing this standard approach, we model the size of the winset (the proportion of policy space able to defeat the status-quo) as our dependent variable.

In order to explore the role of veto players in presidential regimes, we construct two-dimensional policy landscapes represented by a 50x50 lattice. Political actors in our models (whether individuáis or institutions) lócate themselves in one of the 2,500 discrete policy positions and seek to minimize the distance between their ideal points and the policy implemented. Figure 1 illustrates a simple case of executive-legislative relations in this policy space. A president with veto power is located to the right and to the north of the status-quo, while the House and the Senate are located to the left and south. Let us cali the preferred-to set ViSQ) to the set of policy positions contained within the indifference curve of any given actor i. In the absence ofdecisive actors or dictators, the winset of the status-quo is the intersection of the preferred-to sets of all veto players (represented by the gray area in Figure 1). Any proposal within W(SQ) defeats the status-quo, while policies outside the winset are by definition unenforceable.


The comparative literature has conventionally classified constitutional law-making powers accorded to the executive as proactive and reactive (Shugart and Carey 1992; Shugart and Mainwaring 1997). Proactive legislative powers allow presidents to unilaterally "establish, or attempt to establish a new status quo" (Shugart and Haggard 2001, 72) while reactive powers allow them to defend the status quo from any changes attempted by the legislature. The evidence of the impact of presidential powers on policy stability is somewhat mixed. Johnson and Crisp (2003) claimed that constitutional decree authority was not significantly related to the adoption of neoliberal reforms, while Nielson (2003) showed that stronger legislative powers favored the adoption of lower tariffs in middle-income presidential democracies (Nielson 2003). Empirical conclusions about institutional designs often reflect the specific policy areas analyzed, but taken together, these works suggest that the effect of presidential powers may be conditional not only on the nature of the policies at stake, but also on the broader institutional context. In turn, Cox and Morgenstern have argued that presidential powers play an important role when the president is politically weak but less so when the president has vast support in congress (Cox and Morgenstern 2002).

a.      Proactive Powers: Redefining the Winset

We contend that constitutional decree authority (CDA) reduces policy stability by transforming the president into a decisive player (Carey and Shugart 1998)1 . Figure 1 above illustrates how variance in constitutional decree authority may affect the dependent variable. A president with strong decree powers, for instance, would be able to move the status-quo northeast to any position within his preferred-to set (including his own ideal point), disregarding the preference of the two chambers.2

Thus, a complete definition of the winset of the status-quo must include not only the intersection of the preferences of all veto players but also, when they exist, the unión of the preferred-to sets of all decisive players (or just the preferred-to set of the dictator). Formally:

Where T is the preferred-to set of the dictator, Dj is the preferred-to set of the j-th decisive player, and VK is the preferred-to set of the k-th veto player-any of those sets is empty when the respective type of player does not exist. In democratic presidential regimes, W(SQ) is normally the union of three sets: (1) the policies that the president can implement autonomously via executive decrees, (2) the policies that congress can implement overriding any presidential veto, and (3) the policies in which both veto players agree. To illustrate this point, consider two examples. First, imagine a system in which the president lacks any major proactive powers (is not allowed to issue decrees) and in which he or she has weak reactive powers (the unicameral congress can override a package veto by simple majority). In this situation: (1) no policies can be implemented unilaterally by the executive, (2) a majority of legislators can alter the existing law, and (3) it is irrelevant whether the president agrees with the legislators because and the same majority that passes the bill in congress is sufficient to override a veto. Therefore, the winset will be defined as the set of alternative policies in which a majority of legislators agree.

Now, assume that the president has strong decree powers (decrees have immediate forcé), and that he or she is endowed with strong reactive powers (congress can only reject a package veto with two-thirds of the votes). In this case (1) the executive can alter the status-quo unilaterally and the new policy will become effective immediately; (2) congress can reject this change and adopt any alternative by simple majority, but (3) the president can veto the congressional alternative and sustain the veto with support of one-third of the legislators plus one. Therefore, the status quo can be replaced by any of the policies preferred by the president (which can be implemented by decree), by any of the policies supported by two thirds of the legislators (even against the president's will), or by any of the policies supported by a simple majority of legislators and the president. Note that with two decisive players in this example, policy outcomes could be quite unstable. Policies preferred by the president but supported by less than one-third of the legislators may be implemented by decree but may not last in place for long. And even policies supported by a two-third majority in congress could be soon undone by a new presidential decree.

b.      Reactive Powers

Package veto. Most presidents are allowed to exert package veto on a bill approved by congress. They have the option of either signing the bill, accepting the new status quo, or rejecting the bill completely, allowing the reversionary point to prevail (in the absence of a veto override). This is the typical veto player identified by Tsebelis (1995) and depicted in Figure 1. Package veto therefore increases policy stability simply by transforming the president into a veto player (Cameron 2000; Groseclose and McCarty 2001; Tngberman and Yao 1991; Kiewiet and McCubbins 1988; McCarty and Poole 1995; Rohde and Simón 1985).

Item veto. Under partial veto, a president confronting a proposal dealing with several policy dimensions can selectively accept some changes in the status quo while rejecting others. Thus, the reversionary point (provided that the veto is not overridden, a point we consider below) is defined by the location of the congressional bill in the approved dimensions and the location of the status quo in the vetoed ones. This reduces the legislators' capacity to log-roll with the executive, for it is now difficult to trade policy gains in some dimensions for policy gains in others (Baldez and Carey 1999,37-38).

We contend that line-item veto increases policy stability by reducing the size of the president's preferred-to set. Figure 2 illustrates this situation in a two-dimensional space where the president's ideal point P is located to the left and South of the status quo SQ. Because the president now has the capacity to consider (approve or reject) each dimension independently, his indifference curve is no longer circular as in the case of package veto, but it is shaped as a rectangle with center P and the status quo SQ as one of its corners (Schap 1986).

Veto override. The most lenient the conditions for a veto override, the lower policy stability. If congress has the capacity to override a presidential veto by simple majority, the executive is virtually deprived of any effective veto authority (it may delay but not block policy implementation) and the congressional majority becomes a decisive player. If a qualified majority is required to exert this power, the president remains a veto player-but any policy position within the preferred-to set of the supermajority can defeat the status-quo even without the executive's consent (Krehbiel 1998). It follows that the larger the q-majority required by an assembly to override an executive veto, the greater the stability of the status-quo.3


So far we have treated the legislature as a unified actor (or as two individual chambers) without taking into account its internal structure as a collective player. The location of a legislative chamber in the policy space is, however, the product of the aggregation of the preferences of N individual legislators through a particular voting procedure. In this section we extend the model to incorporate the role of individual legislators, parties, and party cohesion into the analysis. We also address the effect of the so-called "partisan powers" of the president on policy stability.

a.      Individual Legislators

Following the standard assumptions of spatial modeling, we assume that the more distant individual legislators are from the status-quo, the greater their incentives to enforce policy change. It is harder to make equally clear predictions about the effects of the distance among individual legislators. On one hand, the greater the distance among legislators, the greater the radius of the "yolk" of the collective veto player, and -under simple majority rule- the lower policy stability (Tsebelis 2002, 48). On the other hand, greater distance among individual legislators probably indicates a greater likelihood that the status-quo is Pareto-optimal (Le., belongs to the "core" of the system as a whole) and thus likely to be stable. According to Scott Mainwaring, "as ideological distance in the party system increases, the president's capacity to implement major reforms decreases" (Mainwaring 1999,286).

Similarly, it is hard to anticípate the effects created by the size of the legislature. Consider for simplicity a unicameral legislature (or committee) composed by N members. The number of individual legislators has two potentially opposite effects on policy stability. On one hand, Tsebelis (2002,48) has conjectured that an increase in the number of individuáis reduces the "yolk" of the collective veto player, increasing its m-cohesion and thus constraining policy change.4 On the other hand, an increase in the number of legislators reduces the effective majority required to alter the status-quo. By effective majority we refer to the actual proportion of votes necessary to achieve a simple majority in a collective body, given its size. In a committee of two, a majority of the vote can only be achieved with a hundred percent of the votes. In a committee of three, a simple majority is reached with two-thirds of the members. In a body of a hundred, the effective majority drops to 51 percent. As the size of the chamber goes to infinity, the effective majority asymptotically approaches the fifty percent threshold. Against the previous hypothesis, this pattern suggests that the larger the size of the chamber, the easier it is to form a majority to alter the status-quo.

Figure 3 illustrates the overall decline of the effective majority as a function of N. The effective majority threshold is not only asymptotic on N; it also displays a cycling pattern by which an increase in N from an odd to an even number locally raises the threshold. An odd number of legislators allows for the formation of a majority with fifty percent of the votes plus "half" of a legislator. With an even number of legislators, a minimal winning coalition can only be achieved with literally fifty percent of the votes plus one. For example, if the size of a committee expands from three to four, the number of members needed to form a majority grows from two to three, so that 2/3<3/4.

b.      Political Parties

For the most part, cross-national comparative studies do not focus on individual legislators (who are hard to pin-down) but on legislative parties. It is now common wisdom that an increase in the number of legislative parties breeds greater policy stability. Because a majority party is less likely to exist in a fragmented legislature and the costs of collective action are presumably greater, the ability of any group to challenge the status-quo is expected to decline (Cox and McCubbins 2001; Johnson and Crisp 2003; Mainwaring 1993; Stein, Talvi, and Grisanti 1999,111).

In order to model the impact of legislative parties on policy stability, we introduce two additional concepts, the (raw) number of parties in congress (J) and the legislative weight (w) or share of the seats of each party. Given a legislature with N seats where s= is the number of seats held by the j-th party, wj = sj/N. The effective number of parties, conventionally measured by the Laakso-Taagepera index, is defined as ENP=1/Σ (wj2) (Laakso and Taagepera 1979). A reduction in the effective number of parties may be achieved by reducing J while holding the relative size of all parties constant (consolidation), or by increasing the relative weight of one or few parties while keeping J constant (concentration). We discuss the effect of these two possibilities on policy stability below.

Multipartyism. The connection between multipartyism and policy stability normally assumes a causal chain with two distinct links. The first one relates the effective number of parties to the formation of legislative majorities, and the second one links the presence of legislative majorities to the prospects of policy change.

First, consider the connection between the number of parties and the presence of legislative majorities. Given a number J of partisan players in the legislature, the maximum possible value for the ENP index is J (when every party has N/J seats) and the minimum approaches 1.0 (when every party has one seat with the exception of a hegemonic party controlling the remaining N-J+1 seats).5 Within this range, there are two values of interest. The first one is the maximum ENP score under which a majority party is guaranteed to exist in the legislature. Below this value (roughly, ENP=2), a reduction in the effective number of parties does not increase the probability of observing a partisan majority (which is already guaranteed). The second value of interest is the maximum ENP under which a majority party is possible. Above this value, any further increase in the effective number of parties does not reduce the probability of observing a partisan majority (which is already zero).

We refer to the guaranteed majority threshold as ENPGM. For any effective number of parties smaller than or equal to ENPGM a partisan majority is assured to exist. The guaranteed majority threshold is set by the marginal case in which the largest legislative party controls a bare majority of the seats and the second largest part controls all the remaining seats (except for one seat left to each of the remaining J-2 parties). This figure is equivalent to the effective number of parties closest to, but smaller than, 2.0 given parameters J and N. For any N and J, the guaranteed majority threshold is therefore:

Where c= 1 when N is even and c=0.5 when N is odd.

The second relevant threshold is the maximum effective number of parties under which the presence of a majority party is possible. This threshold corresponds to the ENP when the largest party controls fifty percent plus one seat in the chamber and all other parties have an equal number of seats. Let us cali this value ENPPM, so that

Where c = 1 when N is even, and c=0.5 when N is odd.6

These two formulas allow us to specify the relationship between the number of parties and the presence of legislative majorities. Any legislature with an effective number of parties smaller or equal to ENPGM is guaranteed to have a majority party, any legislature with a number between ENPGM and ENPPM may or may not have a majority party, and any legislature with an effective number greater than ENPPM is guaranteed not to have a majority party. Figure 4 illustrates this relationship in a hypothetical legislature of one hundred members. It follows that any increase in effective number of parties below ENPGM or above ENPPM will not reduce the probability of a majority party, in the first case because the probability is fixed at one and in the second case because the probability is zero. It is therefore true that the greater the effective number of parties, the lower the probability of finding a majority party, but this probability is fixed at the extremes, falling from one to zero within interval [ENPGM, ENPPM].

A second, related issue is whether a partisan majority guarantees policy change. We shall ignore for the moment the issue of party cohesion and focus instead on the spatial location of the majority party. We expect policy instability when the majority is distant from the status-quo and policy stability if the majority agrees with, and is willing to preserve, the existing policies. Therefore, we must conclude that policy instability will be high when ENP ≤ ENPGM and the largest party is distant from the status-quo, but very low if ENP > ENPPM or the largest party is cióse to the status-quo.

Party weights. The effective number of parties also declines as one -or few- parties tend to concéntrate more seats. This effect is particularly important within the interval [ENPGM, ENPPM] because in this case legislative majorities are not guaranteed and can only be achieved if one party acquires enough weight in the legislature.

We distinguish three situations created by the relative weight of the largest party in congress. When the largest party is relatively small, its acquiescence is not necessary to alter the status-quo and an alternative coalition may challenge the existing policy. When the largest party has enough weight to become a necessary partner in any winning coalition, it becomes a partisan veto player within the legislature, because the body cannot propose any changes to the status-quo without taking into account the party's position. Finally, if the largest party concentrates a majority of the seats in the chamber, it becomes a dictator within the legislature and a partisan veto player in its relation with the president. The majority party is now able to make decisions on behalf of the collective body without taking into account the other parties' ideal points. This sequence suggests that the relationship between partisan weight and policy stability is not linear. With small parties, policy stability may be low because múltiple coalitions may challenge the status-quo. With an "internal" veto player, policy stability will tend to increase because a single party becomes the formateur of any winning coalition. Finally, when the largest party achieves a majority, policy stability will abruptly decline because the party, in the absence of a presidential veto, will be able to enforce substantive policy change alone.

Figure 5 illustrates this pattern by looking at three parties (A, B, and C) located in our two-dimensional policy landscape. The parties are fixed in their policy positions, but their relative weights (in a legislature of a hundred members) are allowed to change. In Figure 5.1, the weights are balanced, and the winset of the status-quo represents 2.5 percent of the contested policy space. In 5.2, Party A concentrates half of the seats without gaining a majority. The largest party serves as an "attractor" for W(SQ), which now falls to 1.6 percent of the policy space. Finally, in 5.3 the largest party achieves 60 percent of the seats, and the size of the winset abruptly expands to cover 13.6 percent of the policy space.

This significant change is achieved without altering the preferences of parties in congress and without increasing the raw number of parties (J).

c.      Party Cohesion

For the most part, the comparative literature on presidentialism has linked low party cohesion to policy paralysis (Cox and McCubbins 2001,39; Mainwaring 1999, chapter 10; Mustapic 2002,27). Empirical studies have also related low party discipline to the legislators' incentives to cultívate their "personal vote" and a low motivation to provide public goods (Ames 2001; Cain, Ferejohn, and Fiorina 1987; Carey and Shugart 1995; Nielson 2003). Against this standard view, Tsebelis argued that low party discipline makes it "difficult to identify partisan veto players in presidential systems," concluding that "the lower the party cohesion, the lower is the policy stability" (Tsebelis 2002,84-85; see also Ames 2001, 14, 205). In turn, Scott Mainwaring noted that party defections may work for or against the status-quo, and claimed that "as party discipline decreases, the president's capacity to implement reforms fluctuates in unpredictable ways" (Mainwaring 1999,286,289-91).

We contend that the effect of party cohesion on policy stability is not independent from the size and location of the partisan majorities. Presumably, if the majority party in congress is far away from the status-quo, greater party cohesion will mean lower policy stability. Because of this reason, students of presidentialism have usually assumed that large, cohesive ruling parties facilítate the implementation of economic reform policies initiated by executive while uncohesive parties hinder the provision of public goods (Cox and McCubbins 2001). However, if the majority party identifies with the status-quo, greater party cohesion will impose greater policy stability (Johnson and Crisp 2003,131). Discussing economic reform in Venezuela, Javier Corrales warned that "strong parties can pose serious dangers for reform sustainability. The decision of a strong ruling party to sabotage its own administration could very well be one of the most destabilizing developments to take place in any political system" (Corrales 2002, 37).

Figure 5 above suggests how to conceptualize party cohesion from the point of view of our spatial model. High party cohesion (as measured by the Rice índex, for instance) simply means that a large proportion of the members of one party adopt the same policy position when they vote (Cox and McCubbins 1993; McCarty, Poole, and Rosenthal 2001; Rohde 1991; Snyder and Groseclose 2000). Cohesion may result from 1) party leaders providing selective incentives to ensure that backbenchers do not deviate from the party's position; 2) party leaders adjusting the party's policy position to approximate the members' ideal points (conditional party leadership); or 3) party members simply sharing the same ideal point (ideological cohesion) (Aldrich 1995; Ames 2002; Cox and McCubbins 1993; Eaton 2002; Jones 2002; Krehbiel 1993; Mainwaring and Pérez-Liñán 1997; Nacif 2002; Rohde 1991). From the point of view of our model, all three processes lead to the same outcome: members of the same party converge (cluster) around a single ideal point in the policy space when they reveal their preferences. In Figure 5, all three parties are perfectly cohesive. Uncohesive parties would have shown some non-trivial degree of dispersion, allowing their legislators to "spread-out" around the party's ideal point.

d.      Partisan Powers of the President

A similar logic applies to what Mainwaring and Shugart (1997) called the "partisan powers" of the president (Le., the size of the president's party weighted by its cohesion). Many authors have discussed partisan powers as a key factor promoting policy innovation or explained paralysis as a consequence of weak partisan powers (Archer and Shugart 1997; Crisp 1997; Mainwaring and Shugart 1997a). But this is true only to the extent that the president is distant from the status-quo. Otherwise, the partisan powers of the president will essentially guarantee greater stability (Casar 2002). A similar debate has taken place regarding the gridlock effects of divided government (Cutler 1988; Krehbiel 1998; Mayhew 1991; Sundquist 1988).

The literature has also emphasized that strong partisan powers may offset the independent impact of presidential proactive powers (discussed above) on policy stability. When the president controls a strong legislative majority, the use of unilateral decree powers becomes less necessary to achieve policy reforms (Cox and Morgenstern 2002). Paradoxically, presidential decrees may be easier to issue and to sustain in this context. Congressional majorities may be more willing to delégate decree powers to the executive if they trust the president as an agent (Carey and Shugart 1998), and CDA may be more likely to produce equilibrium outcomes when the president is closer to the median legislator (Negretto 2004). In both cases, the greater viability of presidential decrees reflects the fact that the president is likely to rule within the legislature's preferred-to set.7

Table 1 below condenses the previous discussion into sixteen hypotheses about veto players in presidential regimes. For consistency, the dependent variable is cast as the probability of policy change-or more precisely, the expected size of W(SQ) relative to the total policy space. Four of these hypotheses deal with presidential powers; five, with the number and location of the individual legislators; four, with the number of parties and their cohesion, and the remaining three with the partisan powers of the president.

The table suggests that the veto-player approach can be a convoluted heuristic for the study of presidential regimes even before we attempt to account for empirical problems of measurement and cross-national comparison. In some cases, hypotheses about the same variable present contrasting predictions. In others, they suggest complex interactions between variables. It is not only that some of these hypotheses challenge any straightforward connection between conventional institutional variables and policy stability, but also that some of them potentially offset others. For example, if a president has constitutional decree authority and line-item veto, shall we expect more or less policy stability than if the president has none of these powers? We address this problem in the following section.


In order to "weight" these hypotheses against each other, we developed a computational model of the policy landscape. The system randomly locates the president and the legislators in a two-dimensional policy space and endows the agents with a constitution. The use of a computer simulation has two major advantages for our purposes. In contrast to formal models anchored in analytical solutions, our approach is able to assess the validity of several "conjectures" that the literature has presented as stochastic properties of the interaction between veto players (Tsebelis 2002, chapter 2). At the same time, and in contrast to empirical models, the simulation setting allows us to grow artificial legislatures under a broad range of constitutional rules, irrespective of considerations about limited variance across cases, constitutional path-dependence in individual country histories, and small-N samples (Epstein and Axtell 1996). The model allows us to take a set of assumptions formulated by the veto-players theory, incorporate the insights of comparative presidential studies, use an experimental procedure to genérate data, and analyze the emerging information inductively (Axelrod 1997, 24; Gilbert and Troitzsch 1999, 25). Thus, although the simulation approach does not provide an empirical test of the hypotheses discussed above, it addresses in an experimental setting the two key issues that motívate this paper: a) to what extent different intuitive hypotheses about the role of institutions are consistent with the spatial assumptions of the theory, and b) to what extent the isolated effect of some institutional designs is to be expected after we control for other institutional arrangements.

Procedure. The computation follows five simple steps. First, a status-quo policy and J parties (for 2 ≤ J ≤ 10) are randomly located in the 50x50 policy lattice. Each party is assigned a different level of cohesion, Cj, ranging between zero and one.8 Second, N legislators (5 ≤ N ≤ 200) are randomly distributed among the existing parties and assigned ideal point coordinates in each dimension according to a normal distribution with mean equal to their party's position and standard deviation (1-Cj)*6.25.9 The varying size of the legislature is intended to capture a whole range of collective veto players: powerful committees, small legislatures, and regular size chambers. Because there has been an extensive treatment of bicameralism in the literature (Binder 1999; Bottom et al. 2000; Hammond and Miller 1987; Llanos 2002; Miller, Hammond, and Kile 1996; Tsebelis and Money 1997), and in order to simplify the computation, we assume a single collective veto player. Third, a president is randomly assigned one of the previously created parties and is located at that party's ideal point (i.e. as its leader).10 Fourth, the system is endowed with a constitution that determines: a) whether the president has constitutional decree authority or not; b) whether the president has package veto, item veto, or no veto power at all; and c) whether such veto can be overridden with a simple majority, two-thirds, or three-fourths of the votes in the chamber.

Based on this information, fifth, the system estimates the preferred-to set of the president and all legislators, every possible majority (and super-majority) in the legislature, the relative size of each party, the effective number of parties, and ultimately (given the constitutional arrangements) the size of the winset, measured as the proportion of the total policy space that defeats the status-quo. Note that the institutional features of the system are taken jointly to determine the size of the winset, controlling for the strategic interactions among the players.

We run this procedure five thousand times, generating a dataset with ampie variance in the number, location, and nature of the players, as well as in the constitutional settings. Qnly the size of the policy lattice was preserved constant to allow for a consistent metric of the dependent variable. As an example, Figure 6 illustrates the policy landscape for observation number 487, showing the location of the president and the legislators, the density of the legislative support for different policy positions, the preferences of a simple majority and a supermajority of legislators, and the preferred-to set of the president under package veto. Graph 6.4 presents the emerging configuration of the winset, a very narrow segment of the policy space (14/2500) determined by the override majority and the overlap of the president's and the congressional majority's preferred-to sets.

Results. The variables captured in our dataset are described in Table 2. Table 3 shows different OLS models in which W(SQ) is the dependent variable and the institutional conditions presented in Table 1 are the predictors.11 The large number of simulated observations allows us to ignore the idiosyncratic effects created by specific spatial configurations and focus on the overall impact of the selected variables. The simulation results suggest that, according to the logic ofthe veto-player theory, we should expect constitutional decree authority and veto override by simple majorities to have a significant negative impact on policy stability. In both cases, a "decisive player" is born, but the impact of these institutions can be substantively different because, in contrast to congress, the executive faces no problem of collective action. In our tests, CDA increased the size of the winset between 11 and 17 percent, and simple majority override, by about 5 percent. Against hypotheses 2 and 3, the impact of reactive powers on policy stability was feeble.

The number of legislators (a variable of some theoretical relevance but rightly ignored by empirical studies) had no significant impact on the dependent variable. Not surprisingly, the greater the distance between the mean legislator and the status-quo, the greater the size of the winset. The average distance between legislators has a somewhat intuitive effect. In spite of the larger "yolk" of the collective veto player, greater separation between legislators tends to create greater policy stability-presumably because the status-quo is more likely to be in equilibrium.

The simulation challenged the conventional wisdom that a larger effective number of parties breeds greater policy stability-even after we controlled for instances of "impossible majorities". It also suggested that guaranteed majorities can significantly reduce policy stability as the largest party gains distance from SQ. Against all expectations, the share of seats of the largest party had no clear influence on the capacity of the legislature to enforce policy change (Model 5), and party cohesion was relevant only in interaction with its distance from the status-quo (in Model 6, but not in Model 8).

Although the partisan powers of the president had the hypothesized effects in Model 7 (expanding the size of the winset as the president gains distance from SQ), in the fully specified model the distance of the president's party from the status-quo, more than its size or cohesion, emerged as the strongest predictor of the propensity for policy change. The interaction between CDA and partisan powers showed the expected sign, suggesting that decree authority becomes less relevant for presidents with strong legislative support.


Table 4 displays the predicted size of the winset (based on Model 8) for four ideal-types of presidentialism: strong presidentialism (when the executive has both strong constitutional and partisan powers), weak presidentialism (when it lacks both types of power), "efficient" presidentialism (relatively weak constitutional powers but strong partisan powers), and "inefficient" presidentialism (strong formal powers but weak partisan powers) (Shugart and Carey 1992). For each ideal type we estimated the expected value of W(SQ) when the president is relatively cióse to (10 spatial units) and distant from (40 spatial units) the status-quo. The estimations suggest that, paradoxically, under the spatial assumptions of the veto-player model, the policy preferences of the key players are more relevant for shaping policy stability than the institutional rules themselves.

The results of our computational model have direct relevance for empirical studies of presidential institutions. First, the significant role played by policy preferences underscores the need to intégrate behavioral and institutional data in order to understand the policymaking process. Most large-N studies of public policy under presidentialism have modeled the effects of institutional variables (particularly, the effective number of parties) without capturing the distance of the key players from the status-quo. There are several practical reasons for this omission: the "status-quo" may be hard to identify and the position of key players in the policy space may be hard to locate-even for major political parties operating in a simple left-right continuum. As a result, there is little comparative information on policy positions in presidential regimes (as exceptions, see Alcántara Sáez 2002; Coppedge 1998). However, our analysis suggests that the effect of some key institutional variables is conditional on spatial factors. The presence of legislative majorities, party cohesion, and the partisan powers of the president may have substantially different effects when the president or the congressional majorities are cióse to the status-quo than when they are distant from it. This suggests that greater efforts should be made to produce reliable comparative indicators of spatial positions across countries and over time.

Second, empirical studies have produced inconsistent assessments of the impact of proactive powers on policy stability, sometimes suggesting that presidential powers have limited relevance for policy change (Johnson and Crisp 2003) and sometimes suggesting that they are critical (Ferreira Rubio and Goretti 1998; Nielson 2003). This may be in part the result of the nature of different policy areas covered by empirical studies. Our estimations demónstrate that, under the assumptions of the theory, constitutional decree authority (for those countries and policy areas in which is allowed) and line item-veto should be regarded as relevant institutional factors shaping policy stability, even after other institutional fea tures are accounted for. These findings suggest that studies of policymaking in presidential regimes should play systematic attention to those issues. Repeated empirical results indicating that CDA or partial vetoes are virtually irrelevant for policy stability would seriously question the assumptions of veto player models and forcé a reevaluation of the theory.

A potential problem with this approach is that the veto-players model imposes assumptions that may be too restrictive to guide empirical studies of presidentialism. For instance, different policy dimensions may opérate under different institutional rules (e.g., decrees may be allowed for some issues but not others). An extended version of our computational model could simúlate cases with different institutional rules operating in the two dimensions (e.g., the president could have decree power along X but not Y). In addition, the assumption that politicians are policy-maximizers (more than vote- or office-maximizers) permeates the heuristic. Our computational model allows for technical extensions to expand the analysis. A two-level or "nested" extension would map the distribution of voters into the policy space, forcing individual legislators to trade off their own policy preferences for constituency support. In a multi-district version of this model, individual legislators would respond to different underlying spatial distributions of voters in their respective districts (Saiegh 2002). A three-level extension of the model would also incorporate coalition formation (Altman 2000; Amorim Neto 2002), forcing legislators to trade their own preferences for greater closeness to the president. These extensions would shed light on critical questions about comparative presidentialism that students of institutions are just starting to explore.



*      Authors' names are in alphabetical order. An earlier version of this paper was presented at the 2003 Annual Meeting of the American Political Science Association. We are indebted to Thomas Bräuninger, Mark P. Jones, George Tsebelis, and two anonymous reviewers for their valuable comments.

1      Following Carey and Shugart (1998) we distinguish constitutional decree authority (CDA) from decree authority delegated by congress (DDA). Under spatial assumptions, delegation of decree powers simply indicates closeness between the president and congress (in the sense that congress considers the president a trusted agent). We therefore ignore DDA in most of the discussion below.

2      Our stylized representation of CDA assumes that: a) executive decrees are immediately effective as policy, and b) there is a "rule of tacit approval" by which the new status-quo remains in place unless the assembly explicitly rejects the decree. On these assumptions (and their institutional variations) see Carey and Shugart (1998) and Negretto (2004).

3      In fact, the threshold for an override can be thought of as a measure of how much of a veto player the president is. At the limit, when only a simple majority is required in congress to override the executive's veto, the president is deprived of any substantial veto-playerness.

4      Simultaneously an increase in the number of individual players presumably expands the yolk for qualified majorities (q-yolk), which in turn reduces the q-cohesion, and increases policy stability (Tsebelis 2002, 51-55).

5      These extremes are not merely hypothetical. The first case is illustrated by Colombia, 1958-74, when Liberals and Conservatives constitutionally shared an equal number of seats in the legislature (ENP=2.00) (Hartlyn 1988); the second, by the Dominican Republic, 1947-52, when the Dominican Party controlled 43 out of 45 seats in the lower chamber and allowed the two "opposition" parties to have one seat each (ENP=1.09) (Galíndez 1973).

6      An approximation to this value irrespective of N is 4(J-1)/J.

7      Dependence of decrees on legislative support is greater when presidents lack veto powers and when decrees require the explicit approval of congress to become laws (Negretto 2004).

8      Although the veto players theory developed by Tsebelis (2002) does not delve with parties, most comparative studies of policy-making in presidential systems address hypotheses regarding the number (and position) of parties in the legislature.

9      The maximum dispersion of 6.25 is arbitrary It essentially means that a highly uncohesive party will have 97.7% of its members spread within a 25x25 "quadrant" of the policy space. In contrast, a cohesive party with a standard deviation of 1.25 would have 97.7% of its members within a 5x5 policy area.

10       In the real world, the distribution of policy preferences may not be random: presidents may be more centrally located than legislators, for instance, because they are selected by a nation-wide electorate. However, such distribution may depend on múltiple factors that vary across countries: the policy dimensions at stake, the underlying distribution of voters' preferences in the policy space, whether legislative and presidential elections are concurrent, the rate of change in the distribution of voters' preferences between elections, and the nature of the electoral districts for legislative elections, among other factors. Given the number of possible determinants and the range of empirical variance across cases, assuming a random location for presidents and parties over a large number of simulation triáis is more prudent than imposing an arbitrary set of assumptions constraining their spatial location.

11      Tobit models with upper and lower censoring generated virtually identical results.


Alcántara Sáez, Manuel. 2002. "El programa de los partidos políticos latinoamericanos". PostData - Revista de Reflexión y Análisis Político (8): 116-158.         [ Links ]

Aldrich, John H. 1995. Why Parties? The Origin and Transformation of Party Politics in America. Chicago: University of Chicago Press.         [ Links ]

Alemán, Eduardo, and George Tsebelis. 2002. "Agenda Control in Latin American Presidential Democracies". Presented at the Annual Meeting of the American Political Science Association, August 29-September 1.         [ Links ]

Altman, David. 2000. "The Politics of Coalition Formation and Survival in Multi-Party Presidential Democracies - The Case of Uruguay, 1989-1999". Party Politics 6 (3): 259-283.         [ Links ]

Ames, Barry. 2001. The Deadlock of Democracy in Brazil. Ann Arbor: The University of Michigan Press.         [ Links ]

Ames, Barry. 2002. "Party Discipline in the Chamber of Deputies". In Legislative Politics in Latin America, ed. S. Morgenstern and B. Nacif. Cambridge: Cambridge University Press, 185-221.         [ Links ]

Amorim Neto, Octavio. 2002. "Presidential Cabinets, Electoral Cycles, and Coalition Discipline in Brazil". In Legislative Politics in Latin America, ed. S. Morgenstern and B. Nacif. Cambridge: Cambridge University Press, 48-78.         [ Links ]

Archer, Ronald P, and Matthew S. Shugart. 1997. "The Unrealized Potential of Presidential Dominance in Colombia". In Presidentialism and Democracy in Latin America, ed. S. Mainwaring and M. S. Shugart. Cambridge: Cambridge University Press, 110-159.         [ Links ]

Axelrod, Robert M. 1997. "Advancing the Art of Simulation in the Social Sciences". In Simulating Social Phenomena, ed. R. Conté, R. Hegselmann and P. Terna. Berlin: Springer-Verlag, 2140.         [ Links ]

Baldez, Lisa, and John M. Carey. 1999. "Presidential Agenda Control and Spending Policy: Lessons from General Pinochet's Constitution". American journal of Political Science 43 (1): 29-55.         [ Links ]

Binder, Sarah A. 1999. "The Dynamics of Legislative Gridlock, 1947-96". The American Political Science Review 93 (3): 519-533.         [ Links ]

Bottom, William P, Cheryl L. Eavey, Gary Miller, and Jennifer Nicoll Victor. 2000. "The Institutional Effect on Majority Rule Instability: Bicameralism in Spatial Policy Decisions". American journal of Political Science 44 (3): 523-540.         [ Links ]

Cain, Bruce, John Ferejohn, and Morris Fiorina. 1987. The Personal Vote: Constituency Service and Electoral Independence. Cambridge: Harvard University Press.         [ Links ]

Cameron, Charles M. 2000. Veto Bargaining: Presidents and the Politics of Negative Power. Cambridge: Cambridge University Press.         [ Links ]

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 ]

Carey, John M., and Matthew S. Shugart. 1998. "Calling Out the Tanks or Filling Out the Forms?" In Executive Decree Authority, ed. J. M. Carey and M. S. Shugart. Cambridge: Cambridge University Press, 1-29.         [ Links ]

Casar, Ma. Amparo. 2002. "Executive-Legislative Relations: The Case of Mexico (1946-1997)". In Legislative Politics in Latin America, ed. S. Morgenstern and B. Nacif. Cambridge: Cambridge University Press, 114-144.         [ Links ]

Coppedge, Michael. 1998. "The Dynamic Diversity of Latin American Party Systems". Party Politics 4 (4): 547-568.         [ Links ]

Corrales, Javier. 2002. Presidents Without Parties - The Politics of Economic Reform in Argentina and Venezuela in the 1990s. University Park: The Pennsylvania State University Press.         [ Links ]

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

Cox, Gary W., and Matthew D. McCubbins. 1993. Legislative Leviathan: Party Government in the House. Berkeley: University of California Press.         [ Links ]

Cox, Gary W., and Scott Morgenstern. 2002. "Epilogue: Latin America's Reactive Assemblies and Proactive Presidents". In Legislative Politics in Latin America, ed. S. Morgenstern and B. Nacif. Cambridge:Cambridge University Press, 446-468.        [ Links ]

Crisp, Brian. 1997. "Presidential Behavior in a System with Strong Parties: Venezuela, 1958-1995". In Presidentialism and Democracy in Latin America, ed. S. Mainwaring and M. S. Shugart. Cambridge: Cambridge University Press, 160-198.         [ Links ]

Cutler, Lloyd. 1988. "Some Reflections about Divided Government". Presidential Studies Ouarterly 17:485-492.         [ Links ]

Eaton, Kent. 2002. "Fiscal Policy Making in the Argentine Legislature". In Legislative Politics in Latin America, ed. S. Morgenstern and B. Nacif. Cambridge: Cambridge University Press, 287-314.         [ Links ]

Epstein, Joshua M., and Robert Axtell. 1996. Growing Artificial Societies - Social Sciencefrom the Bottom Up. Washington, DC: Brookings Institution Press and The MIT Press.         [ Links ]

Ferreira Rubio, Delia, and Matteo Goretti. 1998. "When the President Governs Alone: The Decretazo in Argentina, 1989-93". In Executive Decree Authority, ed. J. M. Carey and M. S. Shugart. Cambridge: Cambridge University Press, 33-61.         [ Links ]

Galíndez, Jesús de. 1973. The Era of Trujillo. Edited by R. H. Fitzgibbon. Tucson: University of Arizona Press.         [ Links ]

Gilbert, Nigel, and Klaus G. Troitzsch. 1999. Simulation for the Social Scientist. Buckingham: Open University Press.         [ Links ]

Groseclose, Timothy, and Nolan M. McCarty 2001. "The Politics of Blame: Bargaining Before an Audience". American Journal of Political Science 45 (1): 100-119.        [ Links ]

Hammond, Thomas H., and Gary Miller. 1987. "The Core of the Constitution". American Political Science Review 81 (4): 1155-1174.         [ Links ]

Hartlyn, Jonathan. 1988. The Politics of Coalition Rule in Colombia. New York: Cambridge University Press.         [ Links ]

Ingberman, Daniel E., and Dennis A. Yao. 1991. "Presidential Commitment and the Veto". American Journal of Political Science 35 (2): 357-389.         [ Links ]

Johnson, Gregg B., and Brian Crisp. 2003. "Mandates, Powers, and Policies". American journal of Political Science 47(1): 128-142.         [ Links ]

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

Kiewiet, D. Roderick, and Matthew D. McCubbins. 1988. "Presidential Influence on Congressional Appropiations Decisions". American journal of Political Science 32 (3): 713-736.         [ Links ]

Krehbiel, Keith. 1993. "Where's the Party?" British Journal of Political Science 23 (2): 235-266.         [ Links ]

Krehbiel, Keith. 1998. Pivotal Politics: A Theory of U.S. Lawmaking. Chicago: The University of Chicago Press.         [ Links ]

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 ]

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

Mainwaring, Scott. 1993. "Presidentialism, Multipartism, and Democracy - the Difficult Combination". Comparative Political Studies 26 (2): 198-228.         [ Links ]

Mainwaring, Scott P. 1999. Rethinking Party Systems in the Third Wave of Tierno cratization. The Case of Brazil. Stanford, California: Stanford University Press.         [ Links ]

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

Mainwaring, Scott, and Matthew S. Shugart. 1997a. "Conclusion: Presidentialism and the Party System". In Presidentialism and Democracy in Latin America, ed. S. Mainwaring and M. S. Shugart. Cambridge: Cambridge University Press, 394-439.         [ Links ]

Mainwaring, Scott, and Matthew S. Shugart, eds. 1997b. Presidentialism and Democracy in Latin America. Cambridge: Cambridge University Press.        [ Links ]

Mayhew, David. 1991. Divided We Govern: Party Control, Lawmaking, and Investigations, 1946-1990. New Haven: Yale University Press.         [ Links ]

McCarty, Nolan M., and Keith T. Poole. 1995. "Veto Power and Legislation: An Empirical Analysis of Executive and Legislative Bargaining from 1961 to 1986". Journal of Law, Economics, and Organizations 11:282-312.         [ Links ]

McCarty, Nolan M., Keith T. Poole, and Howard Rosenthal. 2001. "The Hunt for Party Discipline in Congress". American Political Science Review 95 (3): 673-687.        [ Links ]

Miller, Gary, Thomas H. Hammond, and Charles Kile. 1996. "Bicameralism and the Core: An Experimental Test". Legislative Studies Ouarterly 21 (1): 83-103.         [ Links ]

Mustapic, Ana María. 2002. "Oscillating Relations: President and Congress in Argentina". In Legislative Politks in Latin America, ed. S. Morgenstern and B. Nacif. Cambridge: Cambridge University Press, 2347.         [ Links ]

Nacif, Benito. 2002. "Understanding Party Discipline in the Mexican Chamber of Deputies: The Centralized Party Model". In Legislative Politks in Latin America, ed. S. Morgenstern and B. Nacif. Cambridge: Cambridge University Press, 254-284.         [ Links ]

Negretto, Gabriel L. 2004. "Government Capacities and Policy Making by Decree in Latin America: The Cases of Brazil and Argentina". Comparative Political Studies (Forthcoming).         [ Links ]

Nielson, Daniel L. 2003. "Supplying Trade Reform: Political Institutions and Liberalization in Middle-Income Presidential Democracies". American journal of Political Science 47 (3): 470-491.         [ Links ]

Rohde, David, and Dennis Simon. 1985. "Presidential Vetoes and Congressional Response: A Study of Institutional Conflict". American journal of Political Science 29 (3): 397-427.         [ Links ]

Rohde, David W. 1991. Parties and Leaders in the Postreform House. Chicago: University of Chicago Press.         [ Links ]

Saiegh, Sebastian. 2002. "Government Defeat: Voting Coalitions, Constituency Influence, and Legislative Success". Unpublished Manuscript. New York University.         [ Links ]

Schap, David. 1986. "Executive Veto and Informational Strategy: A Structure-Induced Equilibrium Analysis". American journal of Political Science 30 (4): 755-770.         [ Links ]

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

Shugart, Matthew S., and Stephan Haggard. 2001. "Institutions and Public Policy in Presidential Systems". In Presidents, Parliaments, and Policy, ed. S. Haggard and M. D. McCubbins. Cambridge: Cambridge University Press, 64-102.         [ Links ]

Shugart, Matthew S., and Scott Mainwaring. 1997. "Presidentialism and Democracy in Latin America: Rethinking the Terms of the Debate". In Presidentialism and Democracy in Latin America, ed. S. Mainwaring and M. S. Shugart. Cambridge: Cambridge University Press, 12-54.         [ Links ]

Snyder, James M., and Timothy Groseclose. 2000. "Estimating Party Influence in Congressional Roll-Call Voting". American journal of Political Science 44 (2): 193-211.         [ Links ]

Stein, Ernesto, Ernesto Talvi, and Alejandro Grisanti. 1999. "Institutional Arrangements and Fiscal Performance: The Latin American Experience". In Fiscal Institutions and Fiscal Performance, ed. J. M. Poterba and J. v. Hagen. Chicago: The University of Chicago Press, 103-133.         [ Links ]

Strom, Kaare, and Stephen M. Swindle. 2002. "Strategic Parliamentary Dissolution". American Political Science Review 96 (3): 575-92.         [ Links ]

Sundquist, James L. 1988. "Needed: A Political Theory for the New Era of Coalition Government in the United States". Political Science Ouarterly 103 (4): 613-635.         [ Links ]

Tsebelis, George. 1995. "Decision Making in Political Systems: Veto Players in Presidentialism, Parliamentarism, Multicameralism, and Multipartyism". British journal of Political Science 25 (3): 289-325.         [ Links ]

Tsebelis, George. 1999. "Veto Players and Law Production in Parliamentary Democracies: An Empirical Analysis". American Political Science Review 93 (3): 591-608.        [ Links ]

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

Tsebelis, George, and Jeannette Money 1997. Bicameralism. Cambridge: Cambridge University Press.        [ Links ]


Aníbal Pérez Liñán es profesor de ciencia política en la Universidad de Pittsburgh e investigador afiliado al Centro de Estudios Latinoamericanos de dicha universidad. Su trabajo se centra en cuestiones de democratización, diseño institucional y estabilidad política. Es autor de juicio político al presidente y nueva inestabilidad política en América Latina (Fondo de Cultura Económica, 2009).

Juan Carlos Rodríguez Raga es profesor de ciencia política de la Universidad de los Andes y director del Observatorio de la Democracia de dicho departamento. Su trabajo de investigación gira alrededor del estudio comparado de las instituciones políticas, incluyendo los sistemas electorales, las legislaturas y, más recientemente, las cortes. Es coautor de "Cetisismo Otimista. A reforma eleitoral colombiana de 2003" (En C. U. Santander y N. Penteado, eds., Os processos eleitorais na América Latina. Brasilia, Editorial LGE) y de "Ordenando el caos. Elecciones legislativas y reforma política en Colombia" (Revista de Ciencia Política. 26(1): 138-151).

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