Power Asymmetries And Dem Journal Version

Power Asymmetries and Post-Pact Stability: Revisiting and Updating Venezuela’s Pacts

Javier Corrales

Assistant Professor Department of Political Science Amherst College P.O. Box 5000 Amherst, MA 01002 413-542-2164 [email protected]

May 2004

Words: 9248 2

Power Asymmetries and Post-Pact Stability: Revisiting and Updating Venezuela’s Pacts

Abstract

Foundational pacts, otherwise known as constitutions, are more likely to emerge and yield stability under conditions of “reduced asymmetry,” i.e., when incumbents are strong, and opposition parties, not that much weaker. When the opposition is too weak

(i.e., large asymmetry), pacts may emerge, but they will displease the opposition, increasing the chance of instability. When the opposition is too strong (negative asymmetry), pacts may fail to emerge, precluding regime renewal. I illustrate these points by examining pact-making in Venezuela since 1947. This finding challenges theories that democratization is contingent on maximizing the power of the democratizing force, whichever that may be. Instead, stability-yielding pacts may depend on equalizing the distribution of power between the opposition and incumbent, with a slight advantage for the latter.

May 11, 2004 When do political pacts yield compliance? What makes incumbent and opposition forces pact with one another and adhere to such pacts? I define pacts broadly as any agreement of compromise signed, at a minimum, by an incumbent and an opposition political force. They can be as simple as a document of understanding about procedures or policies, or they can be as encompassing as a national constitution. In this paper, I will focus on “foundational pacts,” i.e., pacts signed at the inauguration of a new democratic regime and which typically yield a new constitution.

Foundational pacts are quintessential functional institutions: actors draft and sign pacts for the purpose of restraining their counterparts, mitigating their own political insecurity, and achieving a modus vivendi. Pacts are antidotes to polarization and mutual suspicion. As such, they help pave the way for democratization.1 Yet, intentions do not always come true. Hard as they may try, actors may fail to reach an agreement, or if they reach an agreement, pacts may fail to mitigate political insecurity, leading to post-pact instability.

What determines these outcomes? Pacts can take multiple forms, occur in very diverse political settings, and involve diverse actors. Yet there are two simple distinctions in pact-making that can explain their propensity to yield stability or instability. The first distinction has to do with the balance of forces among signatories.

The other distinction is the extent to which the pact lowers the cost of being in the opposition. This paper argues that foundational pacts in which the balance of power among signatories is deeply asymmetrical and whose provisions do not lower the cost of being in the opposition are most likely to fail. They either do not get signed or, if signed, they do not satisfy the opposition. 2

While balance of forces helps explain the incidence as well as the content of pacts, the content, in turn, helps explain whether the opposition will be “satisfied” or not with the signed pact. Herein lies a key to the sustainability of pacts. A satisfied opposition is necessary for post-pact stability. An unsatisfied opposition, on the other hand, foretells instability, but only if other conditions hold. By influencing the content, the balance of forces at the moment of pacting can influence post-pact stability.

I will illustrate these arguments by revisiting the paradigmatic case of Venezuela.

Ever since Levine’s2 study of Venezuela’s pact-guided transition to democracy in 1958, the case of Venezuela became an obligatory reference among all students of comparative democratization. Scholars marveled at the durability of these foundational agreements in a region plagued by instability and dictatorship. Today, scholars marvel instead at the decay of political stability in Venezuela since 1989. And yet, despite widespread familiarity with the Venezuelan case, many non-Venezuelanist political scientists may still be unaware that there have been, not one, but four major episodes of pact-making in

Venezuela’s democratic history, each with different outcomes:

1) the Constitution of 1947, which failed to deter the opposition from

overthrowing the democratic regime in 1948, leading to a decade-long

authoritarian regime;

2) the pacts of 1958,3 followed by the Constitution of 1961, which had two

post-pact stages: first, cooperation among the leading parties and conflict

between them and the smaller parties (1961 and 1968), and second, political

peace (1968-1978); 3 3) the attempt by Congress to revise the Constitution, starting in 1989 and

abruptly shelved in 1992; and

4) the Constitution of 1999, which has failed to prevent polarization and

instability.

Revisiting the Venezuelan case offers an opportunity to study variation in post- pact stability within the same country, and often, by the same actors. Furthermore, it allows to test the effects of constitutions on political stability in a society without major ethnic-political cleavage.4 Although each of these episodes has generated sufficient scholarship attention in its own right, few of these works attempt to compare all these different episodes with one another.5 The real challenge is to generate a theory of pact- making that would account for this variation and be applicable to other cases as well.

This paper seeks to do that.

In offering a theory of pact-making, this paper synthesizes and, in some cases, refines different schools of thought in the literature on pacts. This is a rich literature that has been influenced by rational-choice, legal, historic-institutionalist and structuralist approaches. My argument borrows from each. It is informed by rational choice in its attention to the strategic interaction between incumbent and opposition forces, and how electoral rules affect that interaction. It is informed by legal constitutionalism in its attention to the content of pacts and how it shapes the post-pact response of the opposition. It is informed by historic-institutionalism in its attention to how existing institutions provide alliance opportunities for actors who are unsatisfied with the content of pacts. And it is informed by structuralism in its attention to economic performance, although I place less emphasis on economics and class dynamics than most structuralist

Venezuela specialists.6 4

I. Asymmetry of Power and Extra Help

Arguments about enduring pact-making are predicated on three basic claims.

First, the politics of foundational pact-making depends on the strategic interaction of two sides among the elites—namely, incumbents (henceforth I) and the opposition

(henceforth O).7 Second, pact-making depends on the prevalence of soft-liners on each side.8 Soft-liners are actors who prefer to deal with their opponent through peaceful, rule-bound competition rather than through open confrontation or simply noncooperation.9 And third, successful pact-making depends on “mutual guarantees.”

Each actor must be willing to offer guarantees that it will not threaten the “vital interests” of their counterparts.10

Yet, this picture is incomplete. First, even in societies without major ethnic fragmentation and conflict, the participants involve not two, but three actors at a minimum: I, the large opposition party (henceforth LOP), and the small-size opposition force (henceforth SOF). As I will argue later, the role of SOF is crucial for understanding post-pact stability.

Second, the prevalence of soft-liners and mutual guarantees is not enough either.

Other factors matter, perhaps more, for pacts to yield compliance. The first crucial factor is the balance of forces between I and O. Following Colomer11 and Bermeo12, I argue that under conditions of high asymmetry (e.g., if I is very strong and O is very weak), pacts will fail, if they occur at all. As Olson13 argues, democracy emerges when there is a

“broadly equal dispersion of power that makes it imprudent for any leader or group to attempt to overpower the other.” The “overthrowers” cannot have the power to make 5 themselves autocrats. I has no incentive to offer guarantees to O because it does not fear it, and O is not strong enough to compel I to negotiate.

On the other hand, under reduced asymmetry (i.e., if both I and O are strong), the chance of pact-making increases, although it is not guaranteed. Pact-making is not guaranteed because a strong I may still conclude that it ought to repress O, and a strong

O might feel that it has no reason to abdicate and negotiate with I. Stuck on these hard- line positions, parties won’t generate pacts.14 Thus, low asymmetry in the balance of forces between I and O must be accompanied by soft-liner positions on each side.

And yet, O cannot be entirely moderate.15 If it doesn’t pressure I sufficiently with mobilizations, denunciations, and attacks, I will not have an incentive to negotiate.16

O’s dilemma is that it must offer a combination of some hard-line resistance (to force I to negotiate) and also some degree of moderation (to offer incentives for I to negotiate). It cannot be one or the other; it must be both. This is a difficult game to play, to say the least. It typically splits O.

Once I and O make the decision to negotiate, the next important variable is the balance of forces between them. Specifically, it is important that the distance between I and O not be large. Two types of asymmetries matter at this point. The first is what I would call “inter-institutional asymmetry:” the distance between O’s position at the negotiating table and its position at some other important political institution in the country, such as Congress. O might be politically strong in the nation, but if its representation at the negotiating table does not reflect that, there is a risk that O will deem the process of pact-making as unfair. 6

The other type of asymmetry that matters, perhaps even more, is what I would call “table asymmetry:” the distance between I and O at the negotiating table itself. High table asymmetry is fatal for post-pact stability. The reason is that, with few seats, O lacks the necessary bargaining leverage to extract favorable concessions from I.

High table asymmetry is a huge handicap for O because O has a particularly ambitious agenda, perhaps more so than I’s.17 O needs the pact to offer not just protection (or minimum guarantees), but also, provisions that make it easier for it to leave its opposition status.18 O does not just want to survive, it wants to thrive. Minimal guarantees are thus not enough. O needs extra help in the form of wide presence in the country’s institutional landscape, or at the very least, rules that do not penalize it for staying in the opposition. Thus O is interested in huge gains and huge restraints on I.

This is the paradox of pact-making. O needs the most out of the pact, and yet, it is the weaker party to the negotiation. O must come to the negotiating table with the highest bargaining leverage possible.

In short, the balance of forces, both inter-institutionally and at the negotiating table, shapes the incidence and content of the pact.19 The latter, in turn, shapes how pleased O will come out of the negotiation. This raises yet another issue: the selection rule used to determine representation at the bargaining table.

One can imagine two possible selection rules: favorable and unfavorable to O. A favorable selection rule is one that yields a level of representation that accurately reflects, or even exaggerates, the true strength of O territorially (e.g., across different subnational government units) and institutionally (e.g., in parliament). An unfavorable rule is one that yields the opposite—under-representation of O. The specifics of the rule are not 7 important. What matters is the outcome, i.e., whether the rule inflates or deflates the power of O.

Under favorable electoral rules, negotiations are likely to produce a content that will please LOPs. LOPs will have the bargaining leverage to extract subsidies from I.

Pleased LOPs increase the chance of post-pact stability. Cooperation between I and

LOPs to defend the new regime is more likely. The problem will occur with SOFs. Even if selection rules are favorable, SOFs will always have low representation. With minimal bargaining leverage, SOFs will not extract as good a deal as LOPs. Thus, the same pact that will please both I and LOPs may not please SOFs at all.

Under unfavorable selection rules, neither LOPs nor SOFs will come to the negotiating table with sufficient strength. LOPs might find themselves trapped. They cannot prevent the negotiation from favoring I, which tempts them to walk away, but they might decide to stick to the process due to a prior verbal commitment. LOPs probably spent a great deal of time asking for a pact, and they may not want to appear hypocritical by walking away. Consequently, LOPs may end up signing the pact, but grudgingly, aware that they are getting a raw deal.

In sum, I propose the following propositions to explain the incidence and content of pacts:

1. Foundational pacts are more likely to yield stability under reduced asymmetry, in

which O is strong, but I is stronger. If O is weak relative to I, it will lack the

bargaining leverage to pressure I to sign a pact of self-restraint. If I is too weak

relative to O, and thus uncertain about its electoral fate, it will shy away from

pact-making. 8

2. Rules to decide representation at the negotiating table can either mitigate or

exacerbate this asymmetry. If rules reduce the asymmetry between I and LOPs ,

then LOPs will have a higher chance to meet their ambitious agenda (subsidies),

and thus walk away satisfied. SOFs, on the other hand, will be unsatisfied, since

they will not obtain as much as LOPs.

3. If rules are unfair, diminishing the degree of representation of LOPs, a pact might

still be signed, but it will be excessively biased toward I. Both LOPs and SOFs

will be unsatisfied. Post-pact instability is likely.

My focus differs from the now-more frequent works that seek to explain the impact of formal rules on the president’s power, the legislature, and policy-making.20 I focus instead on the effect on a lesser-emphasized actor—O —and its prospects on complying with the new regime. Furthermore, my take on asymmetry departs somewhat from Przeworski’s21 focus on information and balance. For Przeworski, the key issue in determining the outcome of pacts is whether the balance of forces between I and O is even or uneven, and whether this balance is known to signatories. If the balance is uneven, Przeworski predicts, as I do, an agreement will emerge that “ratifies” this relation of forces, and is thus unfavorable to O. If the balance is even, he argues, the outcome of the negotiations is unpredictable (“no equilibrium” exists). I contend, instead, that one can be more predictive than this by looking at variations in asymmetry. If asymmetry is large, the outcome is a flawed agreement that will displease O. If asymmetry is low, the outcome is a stable, compliable pact. If asymmetry is unfavorable to I, pacts are unlikely to occur. 9 II. Evidence

One of the most significant differences among Venezuela’s foundational pacts of

1947, 1958-1961, and 1999 was the balance of forces among the signatories. The measure of asymmetry that I use is the numerical difference between the size of I and O, so that the higher the difference, the higher the asymmetry. A negative value indicates a pro-O asymmetry. Asymmetry was very high in 1946, even higher in 1999, and low in

1958-61. And when asymmetry was very pro-O (i.e., negative), circa 1992, an initiative to change the constitution was aborted by orders of I.

A. High Asymmetry: The Constitution of 1947

The fundamental problem of the democratic experiment of 1945-1948, the so- called Trienio, was the enormous power asymmetry between I and O. After removing

General Medina Angarita from office for his refusal to hold elections, Venezuela’s leading party, Acción Democrática (AD) enjoyed overwhelming popular support.

Consequently, AD saw no need to accommodate O. To take advantage of this popularity,

AD called for a universal, direct vote for the 1946 Constituent Assembly and received 78 percent of the votes. Table asymmetry was a formidable 57 points (see Table 1). It led to an incredibly skewed representation at the negotiating table: 137 delegates for AD and twenty-three for the opposition (nineteen delegates for COPEI and two each for the URD and the PCV22—an even larger asymmetry of seventy-one points. All five seats of the directorship (directiva) of the Constituent Assembly were filled with AD representatives.23

The result was a Constitution heavily biased toward I at the expense of O. For instance, the Constitution gave the president the right to “order the preventive detention” 10 of persons who there is reason to believe are implicated in plans to overthrow the

Government by coup or violence. The Constitution did not even recognize the right to form political parties or to protest (“manifestar”). Presidents also obtained the right to appoint all governors. O strongly disliked these provisions,24 as well as all provisions pertaining to the Church, education, the formation of the Council of Economics, and agrarian reform.25

O’s level of discontent with the 1947 Constitution, and with AD in general, was manifest from the start. A number of delegates recorded their discord. The two URD delegates quit. Copei delegates signed the Constitution, but also made it clear that they did so “with reservations,” condemning the process as illegitimate and denouncing AD’s

“totalitarian tendencies.”26 With this level of discontent, the Trienio regime was likely to collapse, as in fact it did just a year later.

B. Low Asymmetry: The Constitution of 1961

The balance of forces during Venezuela’s next attempt at pact-making was entirely different. By 1958, one of the main lessons learned by AD in the 1950s was not so much that democracy was valuable, but that it was dangerous to alienate O. This is because O was rightly deemed to be large—not large enough to defeat I electorally, but large enough to unseat it from power.27 This reduced asymmetry between I and O was confirmed in the 1958 elections. The difference in votes between I and LOPs was a tiny

–1.64 (Table 1). AD approached pact making in the late 1950s, including the constitutional process of 1962, with an attitude of respect toward this new symmetry, a true turn-around from the 1940s. 11 AD was so respectful of this new symmetry that it chose a selection rule for constitutional delegates that was extremely favorable to O. Rather than rely on proportional representation based on direct vote (which would have given AD a strong advantage), I and O agreed to choose delegates based on congressional representation, where the LOPs were strongly represented and SOFs were slightly overrepresented (in relation to the results for the presidential election). Raúl Leoni of AD and Rafael Caldera of COPEI decided that the constitutional commission would be composed of eight representatives from AD, four from COPEI, four from URD, three from PCV, and three independents.28 O obtained 65 percent of seats in the Constitutional Commission, a pro-

O asymmetry of –27.2 (Table 1). This distribution of seats at the negotiating table was even more favorable to O than the distribution of seats in congress, where I had an advantage of 9.7 points. In short, table asymmetries were highly favorable to O.

The result was a Constitution designed to prevent single-party hegemony.

Article 113 explicitly stated that “electoral bodies shall be composed in such a way that no political party or group predominates,” which was achieved by proportional allocation of ministers and the provision of state funding for the opposition parties. Such measures precluded the development of an overbearing majority party, lowering the cost of losing and increasing incentives for opposition parties to uphold the pact. While SOFs failed to make as many gains as LOPs, their gains were not trivial either. Nevertheless, SOFs were still underrepresented, and most of their leftist demands did not make it into the

Constitution. They were the unsatisfied actor of the 1958-61 foundational pacts.

C. Negative Asymmetry: The Failed Constitutional Reform of 1992

Pact-making depends on the relative position, not only of O, but also of I. If I is weak relative to LOPs, then pact-making is improbable. According to Przeworski, this is 12 the situation that most seriously impairs pact-making—uncertainty about electoral prospects. This explains the 1992 decision by I to abort an initiative to revise the constitution.

The political context for the rise of a new pact was optimal in 1989. Specifically, asymmetry was low: O was strong, but I was much stronger. Although Copei suffered an unexpected defeat in the 1988 presidential elections, its performance was stronger than in the previous election of 1983 (see Table 1). Furthermore, the asymmetry between I and O shrank even further in the 1989 election for mayors. O was strong enough to demand a new pact, and I was still in a strong enough to accede.

As predicted, legislators from COPEI, Venezuela’s LOPs, created a legislative commission to study the possibility of reforming the constitution, maybe even drafting a new one.29 Copei’s president, Rafael Caldera, was appointed president of the Bicameral

Commission for the Revision of the Constitution. By 1992, the commission had created a huge agenda for change, including a proposal for a Constituent Assembly. The commission wanted a new constitution calling for fairly significant changes: require the internal democratization of parties, introduce the recall referendum, create a prime minister and an ombusdsman, reform term limits of elected officials, etc. These measures were meant to help O.30

But by 1992, the asymmetry of forces had changed dramatically: the low asymmetry of 1989 gave way to a negative, pro-O asymmetry. The electoral decline of I resulted from a series of political shocks: disagreement within and among party elites about economic policy, two coup attempts, frequent and precipitous decline in the popularity of the president. AD was perfectly aware that its prospects for the 1992 mid- 13 term and 1993 presidential elections were dim. Aware of its decline in relation to O, I lost all interest in pact-making. Predictably, AD decided in 1992 not to go forward with constitutional change.31

D. Extreme Asymmetry: The Constitution of 1999 (The Fifth Republic)

The 1999 Constitutional process, on the other hand, was characterized by a pro-I asymmetry that was even more extreme than during the Trienio. Both table and inter- institutional asymmetries were huge. A new constitution emerged, but one that intensely displeased O.

President Hugo Chávez ran for office in 1998 largely on a platform to “refound”

Venezuela’s democracy and change the Constitution. After the December 1998 presidential elections, in which Chávez did well (56.2 percent of the vote), and the April

1999 referendum to change the constitution, in which the “yes” vote did even better

(71.8 percent of the vote), Chávez realized the magnitude of his popularity across the voting electorate (whether it was widespread across the entire electorate is hard to gauge because abstention rates were high). He needed to take advantage of this formidable honeymoon. Chávez acted like AD in 1946: feeling no need to accommodate O, he did little to accommodate it. He chose a selection rule for constitutional delegates based not on parliamentary representation, where asymmetry was pro-O (-7.6 in Deputies, Table 1), but on direct popular elections.

Although the total votes obtained by opposition and independent candidates reached approximately 30 percent, only six made it into the Constitutional Assembly.32

The result was a dramatic asymmetry of 88.6, the largest distance between I and O ever in Venezuela—what Blanco33 calls an excessive “ventajismo” for the president. 14

Asymmetry was huge not just between I and O at the negotiating table, but also between O’s real strength across the country and its representation at the table. In 1999,

O parties were in trouble, but they were not decimated. In the November 1998

Congressional elections, which took place only nine months prior to the elections to the

Constituent assembly, O parties received 53.8 percent of the seats in the House of

Deputies. O parties also obtained fourteen of twenty-three governorships.

Given these table and inter-institutional asymmetries, O arrived at the constitutional assembly feeling cheated. O could do nothing to deter I from drafting a self-serving document. The result was a consititution heavily biased toward I to the detriment of O. 34 Although the constitution guarantees enormous civil rights, it undermines the rights of organized opposition. State funding for parties, for instance, was banned (de facto, reserving state funding only for I). More laws can be created by referenda—thus bypassing party input. Congressional (and thus party) control of the military was abolished.35 The judicial council, the mechanism through which parties control the judiciary, was also abolished. The president obtained a six-year term in office, with the possibility of reelection, whereas legislators obtained a five-year term. A longer term in office for the president aggravated the cost of being in the opposition.

Some responsibilities were transferred from the legislature to the Presidency (e.g., voting three times against the President allows the President to dissolve Congress).

Additionally, the 1999 Constitution reversed the process of decentralization underway since 1989 and which was so dear to O. The elimination of the Senate meant that no branch of government was left to voice state interests, where LOPs had strong representation. The stress on direct participation rather than parties meant that traditional campaigning techniques, in which Os had an advantage, would no longer be useful.36 15 Strengthening the power of the Executive vis-à-vis Congress and the provinces, by definition, hurt O, whose stronghold was precisely those arenas. The costs and obstacles of being in the opposition had never been higher in Venezuela’s democratic history. O parties were in disarray after the 1998 elections, but the provisions of the 1999

Constitution made a recovery difficult.

Today, many O leaders argue that the 1999 Constitution was directly written by the President. This is not true. I delegates were able to introduce initiatives that did not exactly match the preferences of President Chávez, and which in some occasions, surprised him. However, it is true that all the initiatives that the Executive strongly wanted were approved by the Assembly, including changing the name of the country, expanding the presidential term in office and permitting reelection, abolishing the Senate, and providing greater autonomy and political rights to the Armed Forces.

The most pro-I initiative that emerged from this process was the Executive’s

March 1999 decision to treat the Constituent Assembly as an “originating” (originaria) body. The government defined an “originating” body as one that enjoys all the rights to

“transform the state and create a new juridical order,” not just to reform the constitution.37

Several jurists and O leaders condemned this as a dangerous and excessive concession of powers to the Assembly. They wanted instead the existing political bodies (“poderes constituídos”) to limit the actions of the Constituent Assembly.38

O’s nightmare about the steam-rolling effects a pro-I “originating” Constituent

Assembly materialized quickly. In August 1999, the Assembly seized the parliament building (the Capitolio), thus disactivating congress. In September, the Assembly began to destitute federal judges, in part because the Supreme Court expressed some reservations about the extraordinary powers being assumed by the Constituent Assembly. 16

The president of the Supreme Court, Cecilia Sosa, resigned, arguing that the courts preferred to “commit suicide” rather than accept assassination.39 In November, the directorate of the Assembly approved a decree restricting the extent of the debate, in order to rush the signing of the constitutions, thereby foreclosing the possibility of introducing amendments and corrections.40 In December, a week after the electorate approved the new Constitution, the Constituent Assembly decreed the “Public Power

Transition Regime” (Régimen de Transición del Poder Público), whereby the existing

Congress, the state legislatures, the authorities in the Supreme Court of Justice, Attorney

General, the National Comptroller, and the National Electoral Council were disbanded.

The Assembly proceeded to appoint “provisional authorities” for these posts, including a twenty-one-person provisional legislature (the Congresillo). The newly appointed chief magistrate of the Supreme Court, Iván Rincón, was one of the previous justices to rule that the Constituent assembly was “originaria.”

Endowing the constituent assembly with so much sway was Chávez’s “self- coup,” a term applied to Peru’s President Alberto Fujimori when in 1992 he disbanded the leading political institutions at the federal level capable of holding him accountable.

Chávez accomplished the same by way of the Constitutional Assembly.41 O leaders were aghast.

To summarize, asymmetry is a critical variable in explaining the incidence and content of pacts. Under extreme asymmetry, pact-making will disproportionately favor I, displeasing O. Under low asymmetry, pact-making is more likely to favor O. Under negative asymmetry, I will do all it can to avoid pacts. The Venezuelan cases show that the emergence of pacts, and thus political liberalization, occurs under low asymmetry, i.e., when both O and I are politically strong. 17 III. From Unhappiness to Unrest

What happens to unhappy actors? Do they turn against the pact (i.e., create regime instability) or resign themselves to their lot (lend stability)? SOFs will be unsatisfied with any form of pact-making. LOPs, however, will be unsatisfied only under conditions of high table asymmetry. Whether these actors turn actively disloyal to the regime is not exclusively a function of levels of unhappiness. Other variables play a role, namely:

1) The post-pact behavior of I. After a pro-I pact is signed, I can either take

advantage of these favorable provisions to expand its power and weaken O,

or instead, take steps to integrate O. In short, I can become “overbearing” or

“accommodating.” An overbearing I is destabilizing because it further

infuriates O.42

2) Available allies. For unsatisfied actors to turn into active challengers, they

need international and domestic allies. Allies provide moral, political and

financial support. Without allies, unsatisfied actors remain isolated and

weak, rendering them unmotivated to fight, or, if they decide to fight, easy to

defeat.

3) Economic performance. A sharp economic deterioration after the pact

expands the number of economic losers, and thus, available allies for O.

However, the reverse is not true. Good economic conditions might not be

enough to offset the destabilizing effects of an overbearing I and an ally-

supported O.43 18

The following section illustrates these propositions by looking at post-pact stability in the three cases of pact-making.

A. The Collapse of the Trienio

Two conditions for post-pact collapse were present in the Trienio. First, I turned overbearing. As a ruling party, AD persisted in its attempt to monopolize most political spaces, not unlike the PRI in Mexico and Peronism in Argentina in the late 1940s. AD’s aggressive policy of secularization, land reform, and de-militarization (one of President

Gallegos’ last policies before being overthrown was to reduce military spending and the presence of the military in his cabinet) alienated not just Copei, but also the traditional corporatist groups: the Church, the landowners, and the military.

Second, and as a result of the previous point, allies were available. All the aggrieved groups (the church, the landowners, and the military) sided with Copei. By

November 1948, the Patriotic Military Union (UPM), a group of junior officers, approached Gallegos demanding that Copei be given a share of political authority and that Betancourt be sent to exile—a sign of LOP-military alliance. Gallegos refused, and he was overthrown.

B. The Punto Fijo Regime: First Two Decades

In contrast to the Trienio, none of the conditions leading to pact collapse were present during the first two decades of the Punto Fijo regime, except I ’s somewhat overbearing attitude toward SOFs (President Leoni’s counterinsurgency policies, 1964-

68). Predictably, only SOFs rebelled, but only temporarily. In time SOFs lost allies, and more decisively, I became more accommodating. 19 In the first years of the Punto Fijo regime, the economy expanded (see Figure

1). But more important than good economics, the main stabilizing force was AD’s attitude of accommodation toward O, the Church and the Military. Consequently, almost no major political actor had reasons to turn against the regime, with the exception of

SOFs—the one unsatisfied actor from the foundational documents of 1959-61. Their turn to violence was predictable, especially given the availability of an international ally—the post-1959 revolutionary government of Cuba.44 However, international support did not compensate for the dearth of domestic allies available to SOFs. Furthermore, President

Caldera in 1969 launched a rapprochement with Cuba, thus ending Cuban support for

SOFs. The SOF-led insurrection crumbled by the mid 1960s.

By 1968, all the factors were aligning to yield post-pact stability. LOPs not only continued to feel included, but one of them, Copei, actually won the presidency twice (in

1968 and 1978). And most important, SOFs found ways to incorporate themselves into the Punto Fijo regime.

The story of SOF incorporation after the 1960s is worth retelling since recent scholarship on Venezuela, which has focused on the suffocating effects of partyarchy, has developed some amnesia about it. There is no question that the top political offices in

Venezuela were off-limit to SOFs (e.g., the Presidency, management of SOEs, leadership positions in labor federations, advisory boards to the president). These were reserved for the large parties,45 and large interest groups.46 But there were a number of secondary institutions that became quite accessible to SOFs, and which in fact, SOFs made their home (e.g., small parties, small unions, small neighborhood associations, some media venues, even the military, albeit not at the highest levels). Landmarks in SOF incorporation were the founding of the Movimiento al Socialismo (MAS) by former 20 guerrilla leader Teodoro Petkoff in 1968, and the legalization of the Communist Party and the Movimiento de Izquierda Revolucionaria in 1969 and 1973, respectively.47

One crucial Venezuelan institution that SOFs were able to colonize was the university system, which, under Punto Fijo became one of the largest, most resourceful and autonomous institutions in Venezuela. A good indicator of the importance of this institution is the level of resources devoted to it. In the 1970s, Venezuela conducted one of the largest expansions in spending on higher education in the region (Figure 3).

Between 1969 and 1974, public university enrollment expanded by a phenomenal 72 percent, compared to a 38 percent and a 23 percent expansion in secondary and primary education.48 Even in the economically-depressed mid 1990s, Venezuela still devoted 6.8 percent of its national budget to higher education—the highest in Latin America, whose average was 3.4 percent.49 Universities became state-funded, autonomous “offsprings” of the regime that SOFs could capture, as in fact they did. Venezuelan leftists, including former insurgents, gravitated toward the university system, as either students or, more comfortably, faculty enjoying a generous living-wage and handsome pensions.

The combination of economic growth (see Figure 2) and I-sponsored institutional openings after the 1960s contributed to a change of attitude by SOFs—from insurrection to integration-seeking. At times, SOF leaders even became players at the highest levels of politics in the Punto Fijo Republic: the number of presidential candidates doubled from 1968 to1973, mostly with candidates from the left. The marginal left was not governing, but it was not homeless in the 1970s.

C. The Fifth Republic (1999-present): 21 The politics of the Fifth Republic closely resemble the politics of the Trienio: intense polarization between I and O, with the latter seeking to unseat I. But there is one crucial difference: O has failed to unseat I, despite having come close twice (during the mass-rebellion-tuned-into-a-coup of April 11, 2002 and during the business-labor strike of December 2002 and January 2003). To understand these similarities and differences, it helps to focus, once again, on the three variables that shape the post-pact interaction between I and O. But first, let me explain why I find problematic some conventional accounts of instability under this regime.

The first conventional explanation, inspired by Marxism, sees instability as the result of a classic socioeconomic conflict between the haves (the elites in power prior to

Chávez) and the have-nots (represented by Chávez). The Chávez’s administration is trying to protect economic losers and non-whites from the perils of market economics and globalization.50 Such a “popular” government, the argument goes, is bound to generate resistance from the previous elites, hence the instability. The second explanation, frequent in op-ed pieces published by Chávez’s detractors, focuses on a presumed change in political regime. Accordingly, Venezuela under Chávez switched from an imperfect democracy to an imperfect dictatorship characterized by an unchecked

“cult of personality”51 and a foreign policy of befriending anti-democratic regimes and political movements.52 Polarization is seen as the result of civil society rising against an authoritarian state.

The problem with the class-conflict explanation is that it exaggerates the extent to which Chávez’s social spending is “pro-poor,” to use the World Bank’s term for policies that effectively deliver social assistance to the needy, rather than merely opportunistic. 53 Furthermore, the class-conflict theory cannot explain the profile of

Chávez’s opposition: it is multitudinary in size (rather than minoritarian), encompasses 22 diverse income groups (not just the privileged), and contains different ideological currents, including former chavistas. The problem with the second explanation is that it exaggerates the degree of regime change. Undoubtedly, Chávez has undermined mechanisms of accountability, has engaged in human rights violations, and has packed political institutions with loyalists. Yet, these democratic infractions are not much worse than infractions committed in other Latin American countries recently (e.g., Mexico in the 1980s, Peru in the 1990s), where instability was less pronounced. Venezuela’s democraticness has declined, but its deterioration is not proportionate with the degree of unrest.

To understand the polarization of the Fifth Republic, it is more useful instead to focus on the factors that mediate the relationship between O and I. First, the new regime started with an unhappy, and thus, potentially disloyal O, and an overbearing, and potentially authoritarian I. Each actor was aware of each other’s mood. Each actor faced high incentives to act preemptively against the other. However, O lacked institutional means to restraint the president, whereas I enjoyed formidable institutional powers. I acted quickly to take advantage of such powers, leading to the second condition conducive to instability: an extremely overbearing I.54

Feeling that he had a chance to forever decimate O, Chávez in 2000 became fixated on obtaining control of any remaining autonomous institutions. Chávez began by targeting the parties (the anti-party bent of the 1999 Constitution, Chávez’s 1999 decision to count his first year-and-a-half in office starting in 2000, rather than early 1999, thus adding almost two years to his first term in office). The government then targeted the

Church and education institutions (e.g., the conflict over education reform in early 2001), the public universities (e.g., the effort to replace autonomous deans), the labor unions 23 (e.g., the government’s tinkering with the internal elections of the main labor union), the press (e.g., the proposed bill to regulate newspaper content), the oil sector (e.g., the illegal firing of approximately 18,000 workers and managers who participated in the

December 2002-January 2003 strike), the private sector (e.g., the establishment of exchange rate controls in 2003, which allows the government to implement a policy of

“no dollars for the coup-plotters”), and the country’s technocracy (Chávez has changed the president of PDVSA five times, and sacked 85 percent of the staff of PDVSA’s finance department).55 The government is thus replicating—in fact, exaggerating—the very same behavior that it deplored of AD and COPEI prior to the Fifth Republic: suffocating civil society, squandering resources, and conspiring against technocracy.

Second, Venezuela’s economic performance under Chávez has been catastrophic

(Figures 1). Political polarization and inept policies have taken a heavy toll on the economy. Between 1999 and 2002, GDP per capita declined an average of 3.4 percent each year.

Third, the number of possible allies for O has expanded, partly because of the economic contraction. The middle class, which was originally conflicted vis-à-vis

Chávez in 1999, has now turned decidedly against him. And low-income groups, which overwhelmingly supported Chávez in 1999, now overwhelmingly reject Chávez.56 The evenness of these opinions across classes makes it impossible to explain Venezuela’s political unrest as a class conflict.

Conditions are optimal for a repeat of the Trienio—a quick overthrow of I. Yet,

O has failed to unseat the government. Why? The answer has to do with the composition of alliances: I is not as isolated as it was in the Trienio, and O is not as unified. 24

There is no question that I has suffered two major defections: the non-radical left, which has since abandoned Chávez, finding his regime too whimsical and militaristic, and second, parts of the military,57 who see Chávez replicating, rather than changing, the Punto Fijo practice of arbitrary state favoritism. Yet, despite these defections, I preserves a core of support: the radical left, some sectors of low-income groups, organized (armed) urban action groups, and more important, parts of the military.

Unlike AD in the Trienio, Chávez has not alienated the entire military apparatus. By

2002, the military was split in three: government loyalists, defectors,, and the neutral.

This split became obvious during the April 11-13, 2002 crisis. First, there was an historically enormous civil society march against Chávez. Although armed chavistas fired against protesters, killing seventeen, top military officials refused to follow

Chávez’s order to repress the protesters, siding instead with the protesters and pressuring

Chávez to resign. The next day, a new government headed by Pedro Carmona, leader of

Fedecámaras, appointed itself in office with the explicit support of labor groups, the military, and broad sectors of society. But Carmona’s first measure in office— predictably, the abrogation of the Constitution of 1999—was followed by the dissolution of Parliament and several governorships, and the detention of some chavistas. These anti-I measures alienated his moderate allies. In the meantime, some chavistas (fewer in number than the anti-chavistas two days before) took some streets of Caracas, surrounded the presidential palace, and demanded the return of Chávez. These protesters received the explicit support of some military officers still loyal to Chávez. The officials who backed Carmona two days prior refused to repress and reinstated Chávez.

For some, this was an example of a betrayed popular uprising; for others, it was a coup (or rather, a double coup). Perhaps it was both, or more accurately, it was the 25 former before it became the latter. But most important, it’s evidence of the split of the military in Venezuela. As long as I retains support from some (even if not all the) military officials and radical civilians, it can defend itself in office.

Second, O is internally divided, suffering from the typical dilemma that plagues

O when confronting a semi-authoritarian president, i.e., whether to take a hard-line or a soft-line approach against I. The hard-liners support extra-constitutional means to unseat

I; soft-liners are holding them back. But even among soft-liners, divisions run deep.

Some prefer negotiations with I, others prefer mass mobilizations, others prefer legislative blockage, others prefer waiting until scheduled elections. These divisions within O, unheard of during the Trienio, works toward I’s advantage.

The anti-I coalition is suffering not just a strategy dilemma, but also a coordination dilemma. The multiplicity of groups under O has given rise to a multiplicity of preferences. What makes O particularly strong—namely, its broad multi-class, multi- institutional composition—is also its Achilles’ heel because it makes agreement difficult.

Multiple preferences, in and of itself, is not a lethal problem—in fact, it is the staple of democratic politics. But in healthy democracies, political parties aggregate and conciliate these multiple preferences. As Corrales argues, opposition to an authoritarian regime is more fruitful if it is coordinated by national-level parties,58 which is precisely what is lacking in the Fifth Republic.59 O is large, but balkanized.

In short, I has suffered defections, but retains the core support of a minority of radical civilians and military loyalists. O has expanded its ranks, but it lacks strong parties to coordinate its actions. The result is an impasse in which I is strong enough to survive but not strong enough to destroy O, and O is strong to block some of I’s authoritarian outbursts, but insufficiently coordinated to prevail. 26

In traditional Latin American politics, this impasse between I and O has been recurrent. It was usually solved through militarization: either I called in the military to repress, or O called in the military to stage a coup. This is what began to happen on April

11, 2002. The novelty was that Chávez (and Carmona) discovered that military allies were unwilling to repress, and protesters discovered that the international community was intolerant of military coups. Each actor retracted. That the military option has been ruled out in Venezuela has been good for democracy. But the cost has been to leave Venezuela in the hole of an impasse between I and O from which there seems to be no easy escape.

V. Venezuela and Latin America Compared

The argument that power asymmetries influence post-pact stability can be extended to other cases. Table 2 shows asymmetries prior, during and immediately following the eight constitutional processes that took place in Latin America since 1987.

Venezuela is clearly an extreme case.

Table 2 shows that most constitutions occurred in a context of pre-existing pro-I or low negative asymmetry, which is consistent with this paper’s argument about the conditions that lead to pact-making. The only cases of negative asymmetry (Ecuador,

Peru and Venezuela) exhibited exceptional circumstances. In Ecuador, for instance, I called for a constitutional reform as a way to placate the political turmoil that followed the impeachment of President Abadalá Bucaram and as a distraction from implementing necessary economic reforms. In Peru, I offered the constitutional assembly after carrying out a self-coup and bowing to international pressures. In Venezuela, I rushed the constitution during the honeymoon, when there were assurances of high popularity. 27 Table 2 also shows two types of constitutional processes: subsidizers

(negative table asymmetry) and non-subsidizers (positive table asymmetry). Among the three subsidizers, Colombia stands out as the most generous, with a dramatic reversal of pro-I asymmetry prior to the constitutional process to a pro-O asymmetry during the constitutional process. In Colombia after 1991, as in Venezuela after 1958, it has been the SOFs, rather than the LOPs, that have become insubordinate, fueled by the drug trade.

Among the non-subsidizers, Venezuela 1999 is by far the most extreme. It went from a significantly pro-O asymmetry prior to the constitution to a formidable pro-I asymmetry—the exact opposite of Colombia, except to a higher degree. The only case that comes close is Peru, but even in Peru, the reversal of asymmetries is less pronounced. In the other cases, the pro-I asymmetries during the constitutional processes were low. In Brazil, for instance, the constitutional asymmetry did not depart much from the pre-existing asymmetry. In Nicaragua and Paraguay, the pro-I asymmetry actually declined in relation to the period prior to the constitutional assembly.

In short, the Venezuelan constitutional process stands apart. In Latin America’s constitutional assemblies since 1987, table asymmetry favored O, improved in favor of

O, or was not excessively biased in favor of I. Peru and more significantly Venezuela are the only two cases in which asymmetry was significantly pro-I and significantly different from the pre-constitutional period.

Table 2 also offers some clues about post-constitutional stability. Although many factors contribute, this paper argued that an important ingredient for post-pact stability is whether O emerges satisfied with the pact, which in turn depends on table 28 asymmetry. In the subsidizing cases, this paper hypothesizes that LOPs emerge satisfied.

O will turn confrontational only if it achieves a significant power advantage vis-à-vis I in the election immediately following the constitutional assembly. In this situation, O enjoys enviable power resources: those granted by the pro-O constitution and those obtained by O’s control of key institutions. O is capable of blocking the actions of I through institutional means, typically through non-cooperation in Congress. This explains the case of Ecuador, the most unstable country of the subsidizers, experiencing severe Executive-Legislative gridlock in the two administrations that have followed the constitution (Mahuad and Gutiérrez).

In the non-subsidizing cases, this paper hypothesized that LOPs emerge relatively unhappy with the new regime. But as argued, unhappiness is not sufficient to render LOPs confrontational. Other factors matter such as I’s attitude, whether O recovers or obtains new allies, and economic conditions. In the case of Brazil after the pact (the Collor administration of 1990-1992), these conditions materialized. O’s reaction was predictable: it used institutional means to block I’s initiative and forced I to resign. In contrast, in Nicaragua, where pro-I asymmetries declined during the constitutional process, conditions were ripe for the amelioration of instability, culminating with the peaceful, pact-respecting transfer of power from the Sandinistas to

UNO in 1990.

Relative to Brazil and Ecuador, Venezuela’s post-constitutional O was angrier with the regime because the pro-I table asymmetry was bolder. But Venezuela’s O does not enjoy the same institutional means to challenge I as its counterparts in Brazil and

Ecuador—it remained minoritarian in parliament and it operates under a heavily pro-I 29 constitution. Consequently, it has relied on non-institutional means to challenge I, which explains its predilection for street protests. It has also had to rely on extra-partisan allies, which have not been difficult to find precisely because of I’s overbearing behavior and the economy’s overall decline. In contrast to Peru, where an overbearing I

(Fujimori) generated formidable economic growth from 1993 to 1999, Venezuela’s overbearing I has only produced economic ruin (until 2003), which explains the massive exodus from I toward O in 2001-2003. I has also bullied other corporatist groups (the church, university students, labor unions, and members of the military). All of this explains why O has an anti-constitutional streak (the April 2002 Carmona government), has a formidable size (a majority of Venezuelans), and is so divided (encompassing actors with different ideological backgrounds forming a coalition of LOPs and SOFs ).

********

Added 8/23/04: Riker and Weimer 1995 in Banks and Hanushek’s Modern

Political Economy discuss that credible property rights are necessary for investment. What makes property rights credible? Separation of power. In

Venezuela, separation of powers eroded in 1980s and 1990s (democratic decay), but greater societal activism increased (democratic revival). Elections of 1998 worked: society rose and yielded an opportunity to fix the democratic decay—hence the rise of Chavez and the const assembly. But then, the state hijacked the process, and undermined the democratic decay (separation of powers). This was destabilizing.

VI. Conclusion 30

The literature on pacts has created the image of pacts as medicines with adverse side effects. Pacts can cure intolerance among political opponents, thereby paving the way for democratization, but with a side effect—their conservative bias. Pacts empower the already-powerful, restrict contestation, foreclose policy choices, and exclude non- mainstream political groups, all of which puts breaks on democratic deepening.Hartlyn

1998:108-111 in Ellner 12; Karl and Schmitt60.

My paper suggests that this side effect has been overstated. An understudied characteristic of successful pacts is the extent to which the weaker actor (O), rather than the powerful one (I), is subsidized. Pacts signed in a context of high power asymmetry between I and O are more likely to fail. Undeniably, I must gain from pacts, but O must gain more. Subsidies for O must come in the form of favorable rules for the selection of representatives at the negotiating table, and more important, stipulations in the content of the pact that benefit O more than I. The former is a precondition for the latter, and the latter is a precondition for O’s satisfaction with the pact. Successful pacts are noteworthy not so much because of their conservatism, but because of their generosity toward the weaker forces.

Successful pacts thus require a combination of auspicious historic-institutional conditions (low asymmetry among signatories), but also strategy (an effort on the part of

I to tame majoritarinism by choosing an electoral formula that is favorable to O).

Perhaps one reason that the literature has focused so much on the conservative bias of pacts is that this literature seldom examines failed pacts. Successful pacts are compared to successful pacts only. Karl and Schmitter explore variations in modes of transitions, but they compare pact-based modes with other modes (“reform,” “imposition,” and

“revolution”) rather than with failed pacts. This inattention to variation—a 31 methodological trap that this paper has avoided by looking at completed and aborted pacts in Venezuela—blinds scholars to the variable that most strongly distinguishes successful pacts from unsuccessful ones, namely generosity toward non-dominant groups.

Another reason for the inattention to the generosity of pacts toward O is that most studies on pacts were written several years, sometimes several decades, after the signing of the studied pact. By then, the political force that was weak at the moment of signing had probably turned into a powerful group—maybe even a ruling party. Scholars thus observe two dominant groups that gain far more than the SOFs, leading them to conclude that the pact was conservative. But to conclude that both I and O gain from the pact is already an implicit recognition of the higher generosity of pacts. It is this generosity that permits the weaker force to match the status of the dominant group.

Subsidies, naturally, can spoil the beneficiary. Successful pacts, therefore, carry the risk of eventual decay since they are intrinsically subsidy-intensive. Successful pacts provide so much extra help to non-dominant actors that is easy for such beneficiaries to develop political vices over time. In Venezuela, both AD and COPEI received enormous subsidies even when each was in the opposition, and this, in many ways, protected them from undergoing necessary adaptation to new economic conditions. Founding pacts are susceptible to what Horowitz calls “retrogression,” the tendency for those who gain to alter the course of the new regime to their favor.61 Unless pacts are continuously renewed, and subsidies lessened over time, successful pacts can be expected to eventually generate decay among their main beneficiaries. 32

Table 1: Venezuela’s Constitutional Processes: Asymmetry Among I, LOPs and SOFs

I LOP SOFs Asymmetry Source s I - (LOPs +SOPs) Trienio (High Asymmetry) 1946 Constitutional Assembly (% of votes) 78.43 13.2 7.88 57.35 Corrales 1946 Delegates to Const. A. (% of seats) 85.63 11.88 2.5 71.25 Kornblith 1947 Congress (% of votes) 70.38 20.28 8.15 41.95 Corrales 1947 President (% of votes) 74.47 22.4 3.12 48.95 Corrales

Punto Fijo Pact (Low Asymmetry+ Favorable Election Rule) 1958 Presidential elections (% of votes) 49.18 45.85 4.97 -1.64 CNE 1958 Deputies (% of seats) 54.9 39.9 5.3 9.7 CNE 1961 Constitutional Commission (% of seats) 36.36 36.36 27.2 -27.2 Kornblith 1963 Presidential elections (% of votes) 32.81 37.7 29.49 -34.38 CNE

Failed Constitutional Reform, 1989-1992 (From Positive to Negative Asymmetry) 1983 Presidential Elections (% of votes) 57.7 35.1 22.6 Corrales 1988 Presidential Elections (% of votes) 52.7 40.0 12.7 Corrales Molina and 1988 Elections for Parliament (% of votes) 43.2 41.2 15.6 -13.6 Pérez 1998 1989 Elections for Mayors (% of votes) 39.7 48.3 11.9 -20.5 OCEI 1992 Elections for Mayors (% of votes) 32.3 32.3 14.1 -14.1 OCEI 1992 Elections for Governors (% of votes) 31.1 52.4 16.5 -37.8 OCEI

Fifth Republic (Extremely High Asymmetry) 1998 Deputies (November) (% of seats) 46.2 53.2 0.6 -7.6 CNE 1998 Presidential (December) (% of votes) 56.2 39.97 2.82 13.41 CNE 1999 Delegates to Const Assem. (% of seats) 93.13 4.5 88.63 Penfold 1999 Referendum to approve Constitution 71.78 28.22 43.56 CNE 2000 Presidential elections (% of votes) 59.76 37.52 2.72 19.52 CNE 2000 Assembly elections (% of seats) 60 21.82 18.18 20 Georgetown 2004 Referendum 58.26 41.74 16.52 Georgetown Notes: I: AD in Trienio, Punto Fijo, and in 1989-92; MVR+MAS+PPT in Fifth Republic LOP: COPEI in Trienio; COPEI+URD in Punto Fijo; COPEI in 1988 Presidential Elections; COPEI + MAS in 1989-92 Governor and Mayoral Elections; AD+PRVZL+COPEI in Fifth Republic SOF: URD+PCV in Trienio, PCV in Punto Fijo; LCR+ORA+NGD+Other in 1992 Governor Elections LCR+CONVER+APERT+IRENE+OFM+RENOV (parties with less than 1 percent of the vote not listed). 1989 Governor election data not available.

Source: Javier Corrales. 2004. “Power Asymmetries and Post-Pact Stability: Revisiting and Updating Venezuelan Pacts.” Amherst College, mimeo. 33

Table 2: Power Asymmetry in Latin America’s Constitutional Assemblies, 1987-2000

Country Pre-Constitution (year)4 Const. Assembly (year) Post-Constitution (year) Colombia5 20 (1990) -42.2 (1991) 8 (1994) Ecuador6 -34.14 (1996) -40 (1997) -41.66 (1998) Argentina7 -2.72 (1993) -24.2 (1994) -0.4 (1995) Brazil8 6.52 (1986) 6 (1988) -57 (1990) Nicaragua9 25.8 (1984) 9 (1987) 10.8 (1990) Peru10 -31.2 (1990) 10 (1993) 11.6 (1995) Paraguay11 33.34 (1989) 16 (1991) 0 (1993) Venezuela12 -40 (1998) 86.26 (1999) 20 (2000)

Sources: Most data taken from Base de Datos Políticos de las Américas.[online]]. Georgetown University and Organization of American States http://www.georgetown.edu/pdba/Elecdata.html. 2001, except for the following cases. For Colombia 1991: Cepeda. For Ecuador 1996: Mark Payne, Democracies in Development: Politics and Reform in Latin America. (Washington, DC: Inter-American Development Bank, 2002). For Ecuador 1997: CNN Election Watch http://us.cnn.com/world/election.watch/americas/ecuador2.html. For Argentina 1993: “Elections Results Archive.” Center on Democratic Perofrmance. Binghamton University. For Argentina 1994: Towsa, Andy, “Atlas de elecciones en Argentina.” [online] http://www.towsa.com/andy/index.html1995: CNE, “Resultados de Elecciones Presidenciales Anteriores,” [online] http://www.pjn.gov.ar/cne/cne_res95.html. For Nicaragua 1984, 1987: Susanne Jonas and Nancy Stein. “The Construction of Democracy in Nicaragua.” Latin American Perspectives, Vol. 17, No. 3. Summer 1990, 10-37. For Nicaragua 1990: “Voter Turnout: Nicaragua” International IDEA, 23 October, 2001. For Paraguay 1989: Payne. For Paraguay 1991: “Facts on File, Inc.” Facts on File World News Digest. For Venezuela 1998 and 1999: “Consejo Nacional Electoral.” [online] http://www.cne.gov.ve/.

4 Asymmetry is measured as the difference in number of seats between I and O in the House of Deputies or in the Constitutional Assembly, except Argentina 1994 and 1995, Colombia 1991, and Nicaragua 1984, for which percentage of vote, rather than seats, was used. 5 I = Partido Liberal 1990, 1991, 1994. 6 I = Partido Social Cristiano 1996, 1997. Democracia Popular 1998. 7 I = Partido Justicialista. 1993, 1994, 1995. 8 I =Brazilian Democratic Movement Party, 1986, 1988,1990. 9 I = Frente Sandinista de Liberación Nacional 1984, 1987; National Opposition Union 1990. 10 I = Fredemo 1990; Alianza Nueva Mayoría – Cambio 90 1993, 1995. 11 I =Asosiación Nacional Republicana 1989, 1991, 1993. 12 I =AD 1998; MVR-MAS-PPT 1999, 2000. 34

Figure 1

GDP per capita growth in different Post-Pact periods

g 1945-1949 n 0.05 a

h 1959-1963 C

n 1988-1992 e 0 c r 1999-2003 e 0 1 2 3 4 5 6 P

Source: World Development Indicators Online. 35 Figure 3

Spending on Higher Education

Venezuela

Uruguay

Mexico

Ecuador Circa 1994 Circa 1985 Circa 1975 Colombia 1965

Brazil

Bolivia

Argentina

0 5 10 15 20 25 30 35 40 45 Percent of Total Education Spending

Source: UNESCO (various years). 36 1 Andrew Reynolds, ed. The Architecture of Democracy: Constitutional Design, Conflict Management, and Democracy

(Oxford University Press, 2002). Adam Przeworski, Democracy and the market: political and economic reforms in

Eastern Europe and Latin America. (New York: Cambridge University Press, 1991). Terry Lynn Karl and Philippe

Schmitter, “Modes of Transition in Southern and Eastern Europe and South and Central America,” International Social

Science Journal, 128 (1991), pp. 269-284. Terry Karl, “Dilemmas of Democratization in Latin America,” Comparative

Politics, 23 (October 1990), 1-23. Giuseppe Di Palma. To Craft Democracies: an essay on democratic transitions.

(Berkeley: University of California Press, 1990). Guillermo O’Donnell and Philippe Schmitter, Transitions from

Authoritarian Rule: Tentative Conclusions about Uncertain Democracies (Baltimore, Maryland: Johns Hopkins

University Press, 1986).

2 Daniel Levine, Conflict and Political Change in Venezuela (Princeton, NJ: Princeton University Press, 1973).

3 In addition to the Punto Fijo Pact of 1958, signed by the three leading parties, there were two other pacts signed during this period: the “Labor-Business Rapprochement Pact,” which normalized relations in the workplace, and the

“Declaration of Principles and Basic Program of Government,” which committed the political parties to draft and defend a new constitution and to a new model of state-based economic development, see Margarita López Maya, Luis

Gómez Calcaño, and Thaís Maingon, De punto fijo al pacto social : desarrollo y hegemonía en Venezuela, 1958-1985,

(Caracas: Fondo Editorial Acta Científica Venezolana, 1989).

4 This study thus differs methodologically from the chapters in Reynolds which focus on how constitutions mitigate conflict in ethnic group-divided societies.

5 One notable exception is Michael Coppedge, “Explaining Democratic Deterioration in Venezuela Through Nested

Inference,” paper presented at the Congreso Latinoamericano de Ciencia Política, (Universidad de Salamanca, Spain,

2002). Coppedge tries to show how the evolution of democracy in Venezuela since the 1940s defies general theories of comparative politics.

6 Kenneth Roberts, “Social Polarization and the Populist Resurgence in Venezuela,” in Steve Ellner and Daniel

Hellinger. Venezuelan Politics in the Chávez Era: Class, Polarization, and Conflict. (Boulder, Colorado: Lynne Rienner

Publishers, Inc., 2003). Steve Ellner, “Introduction: The Search for Explanations,” in Steve Ellner and Daniel Hellinger, eds., Venezuelan Politics in the Chávez Era: Class, Polarization, and Conflict, (Boulder, Colorado: Lynne Rienner

Publishers, 2003), pp. 7-25. Daniel Hellinger, “Political Overview: The Breakdown of Puntofijismo and the Rise of

Chavismo,” in Steve Ellner and Daniel Hellinger. Venezuelan Politics in the Chávez Era: Class, Polarization, and

Conflict. (Boulder, Colorado: Lynne Rienner Publishers, Inc., 2003) Medina, Medófilo and Margarita López Maya.

Venezuela, confrontación social y polarización politica, (Bogota: Ediciones Aurora, 2003). Terry Karl, The Paradox of

Plenty: Oil Booms and Petro States, (Berkeley, CA: University of California Press, 1997).

7 Przeworski. 8 O’Donnell and Schmitter.

9 Juan J. Linz, The Breakdown of Democratic Regimes, (Baltimore, Maryland: The Johns Hopkins University Press,

1978).

10 Karl and Schmitter. Karl “Dilemmas of Democratization in Latin America.”

11 Josep Maria Colomer, “Strategies and Outcomes in Eastern Europe,” Journal of Democracy, 6 (1995), 74-85.

12 Nancy Bermeo, “Myths of Moderation: Confrontation and Conflict During Democratic Transitions,” Comparative

Politics, 29 (April 1997), 305-322.

13 Mancur Olson, Power and Prosperity, (New York: Basic Books, 2000), pp. 31.

14 Przeworski.

15 Bermeo, “Myths of Moderation.”

16 Charles Tilly, Coercion, capital, and European States, AD 990-1992, (Cambridge, MA: Blackwell, 1992).

17 On how asymmetry of preferences emerge from different sizes of players, see Donald Horowitz, “Constitutional

Design: Proposals Versus Processes,” in Reynolds, p.27.

18 Ljiphart 1992:208.

19 Octavio Amorim Neto, “Presidential Cabinets, Electoral Cycles, and Coalition Discipline in Brazil, “in Scott

Morgenstern and Benito Nacif, eds., Legislative Politics in Latin America, (Cambridge: Cambridge University Press,

2002) applies a similar argument to Executive-legislative relations under fragmented political party systems. When the representation of parties in the cabinet is proportional to their presence in parliament, what he calls “cabinet coalescence,” the legislature is more likely to cooperate with the Executive.

20 For a recent example, see Stephan Haggard and Mathew D. McCubbins, eds. Presidents, Parliaments and Policy

(New York: Cambridge University Press, 2001).

21 Przeworski, p. 82-83.

22 Miriam Kornblith, “The Politics of Constitution-Making: Constitutions and Democracy in Venezuela,” Journal of

Latin American Studies, 23 (February 1991), p. 67.

23 Andrés Eloy Blanco, President; Ambrosio Oropeza, First Vice President; Augusto Malavé Villalba, Second Vice

President; Miguel Toro Alayón, Secretary; Octavio Lepage, Undersecretary. Rodolfo Jose Cárdenas, COPEI en la

Constituyente: La tentación totalitaria de Acción Democrática, (1987), p. 24.

24 New York Times, July 6, 1947.

25 Kornblith, “The Politics of Constitution-Making,” p.68. Levine.

26 Kornblith, “The Politics of Constitution-Making,” p. 69, 73. Cárdenas.

27 Javier Corrales, “Strong Societies, Weak Parties: Regime Change in Cuba and Venezuela in the 1950s and Today,”

Latin American Politics and Society, 43 (Summer 2001). 28 Kornblith, “The Politics of Constitution-Making,”, p. 72.

29 Venezuela in 1989 was suffering from a case of hyper-reformism. Economically, the government of Carlos Andrés

Pérez launched the most profound package of neoliberal reforms ever in the history of Venezuela. Politically, enormous pressure mounted in the 1980s to force political parties to become more democratic. In 1989, Venezuela allowed the direct election of governors and mayors. The massive riots of February 1989 convinced legislators that further political reforms were necessary to placate this discontent. For a summary of this reform spirit, see Javier

Corrales, Presidents Without Parties: the politics of economic reform in Argentina and Venezuela in the 1990s,

(University Park, PA: Pennsylvania State University, 2002). The decision by Copei to launch a process of constitutional reform was an effort to jump on this reformist bandwagon of 1989.

30 Ricardo Combellas. Venezuela: Crisis política y reforma constitucional. (Caracas: Universidad Central de Venezuela,

1993). Miriam Kornblith, “La reforma constitucional en Venezuela: el debate actual” presented at Nuevas tendencies politico constitucionales en latinoamérica, (Oñati, España, 1995).

31 Revista SIC, September-October 1992, p. 381. My account differs from Kornblith, “La reforma constitutcional…” in that I emphasize negative asymmetry whereas Kornblith emphasizes opposition from various interest groups, including the media.

32 Michael Coppedge. “Venezuela: Popular Sovereignty versus Liberal Democracy,” in Jorge Domínguez and Michael

Shifter, Constructing Democratic Governance in Latin America. (Baltimore, Maryland: Johns Hopkins University

Press, 2003). Brian Crisp and Gregg Johnson, “De instituciones que restringen a instituciones ausentes.” In

Carrasquero, et al. Venezuela en transición: elecciones y democracia. 1998-2000, (Caracas, Venezuela: CDB

Publicaciones, 2001). Thaís Maingon, Carmen Pérez Baralt, and Heinz R. Sonntag. “La batalla por una nueva

Constitución para Venezuela,” Revista Mexicana de Sociología, 62 (Octubre-Diciembre 2000), 91-124. Michael

Penfold, “Constituent Assembly in Venezuela: First Report,” (Atlanta, GA: Carter Center, mimeo).

33 Carlos Blanco. Revolución y desilución. La Venezuela de Hugo Chávez, (Madrid: Catarata, 2003), p. 235.

34 Blanco, pp. 250-254; Maingon, Pérez Baralt, and Sonntag, p. 118

35 Deborah Norden, “Democracy in Uniform: Chávez and the Venezuelan Armed Forces,” in Steven Ellner and Daniel

Hellinger, Venezuelan Politics in the Chávez Era: Class, Polarization, and Conflict, (Boulder, Colorado: Lynne Rienner

Publishers, 2003), p. 100.

36 Angel Alvarez. “State Reform Before and After Chávez’s Election,” in Steven Ellner and Daniel Hellinger,

Venezuelan Politics in the Chávez Era: Class, Polarization, and Conflict, (Boulder, Colorado: Lynne Rienner

Publishers, 2003), p. 152.

37 Ricardo Combellas, “El proceso Constituyente y la Constitución de 1999” (mimeo), 2003, p. 16.

38 Maingon, Pérez Baralt, and Sonntag, p. 102. 39 Combellas, “El proceso Constituyente.”

40 Ibid.

41 Coppedge, “Popular Sovereignty versus Liberal Democracy.”

42 On the impact of the “founding leadership’s” behavior on democratic survival, see Larry Diamond, Jonathan Hartlyn, and Juan J. Linz, “Introduction: Politics, Society, and Democracy in Latin America,” in Larry Diamond, Jonathan

Hartlyn, Juan Linz and Seymour Martin Lipset, eds., Democracy in Developing Countries: Latin America, (Boulder,

Colorado: Lynne Rienner Publishers, 1999), pp. 44-48.

43 On how economic performance affects democratic survival see Larry Diamond, et al.; Adam Przeworski, Michael E.

Alvarez, José Antonio Cheibub and Fernando Limongi, Democracy and Development: Political Institutions and Well-

Being in the World, 1950-1990, (Cambridge: Cambridge University Press, 2000). Stephen Haggard and Robert R.

Kaufman, The Political Economy of Democratic Transitions, (Princeton, NJ: Princeton University Press, 1995). Karen

L. Remmer, “Elections and Economics in Contemporary Latin America,” in Carol Wise and Riordan Roett, eds., Post-

Stabilization Politics in Latin America: Competition, Transition, Collapse, (Washington, DC: Brookings Institution

Press, 2003).

44 Jorge I. Domínguez, To Make a World Safe For Revolution: Cuba’s Foreign Policy. (Cambridge, MA: Harvard

University Press, 1989), pp. 119, 141.

45 Michael Coppedge, Strong Parties and Lame Ducks, (Stanford: Stanford University Press, 1994). Deborah Norden,

“Party Relations and Democracy in Latin America,” Party Politics, 4 (1988), 423-443.

46 Brian F. Crisp, Democratic Institutional Design: The Powers and Incentives of Venezuelan Politicians and Interest

Groups, (Stanford: Stanford University Press, 2000).

47 Janet Kelly and Carlos A. Romero, Rethinking a Relationship: The United States and Venezuela, (New York, NY:

Routledge, 2002), pp. 28-30.

48 Orlando Albornoz, “Higher Education and the Politics of Development in Venezuela,” Journal of Interamerican

Studies and World Affairs, 19 (August 1977), pp. 291-314.

49 Claudio De Moura Castro and Daniel Levy, Myth, Reality, and Reform, (New York: Inter-American Development

Bank, 2000).

50 Gregory Wilpert, “Collision in Venezuela,” New Left Review, 21 (May-June 2003), pp. 101-116. Richard Gott, In the

Shadow of the Liberator: Hugo Chávez and the Transformation of Venezuela, (New York, NY: Verso, 2000). Margarita

López-Maya y Luis Lander, “Refounding the Republic” NACLA Report on the Americas 33, (May/June 2000)

51 Mark Falcoff, “Viva Chávez!” May 21, http://www.aei.org/news/filter.,newsID.12909/news_detail.asp 2001. 52 Linda Robinson, “Terror Close to Home.” US News and World Report. (October 6, 2003) pp. 20-22, 24. Demetrio

Boersner, “La dimensión internacional de la crisis en Venezuela.” Ponencia en el simposio “Venezuela en la

Encrucijada,” Berlín, 2003.

53 World Development Report 2004: Making Services Work for Poor People, (Washington, DC: The World Bank and

Oxford University Press, 2003).

54 Kurt Weyland, “Whill Chávez Lose His Luster?” Foreign Affairs, (November-December 2001).

55 The Economist, August 2, 2003.

56 In February 1999, a few weeks after Chávez took office, 94 percent of Venezuela’s in the lower income bracket, 92 percent of middle income groups, and 91 percent of upper middle income groups supported Chávez. By July 2003, these numbers dropped to 39 percent, 30 percent and 14 percent, respectively (Based on Datanálisis, “¿Cómo evalúa usted la labor del Presidente HChF por el bienestar del país?,” Public Opinion Poll, PowerPoint Presentation Caracas,

2003).

57 Norden, “Democracy in Uniform,” pp. 106-109.

58 Corrales, “Strong Societies, Weak Parties.”.

59 Omar Encarnación, “Venezuela’s ‘Civil Society Coup’” World Policy Journal, 19 (Summer 2002), pp. 38-49.

60 Frances Hagopian, “Democracy by Undemocratic Means? Elites, Political Pacts, and Regime Transition in Brazil,”

Comparative Political Studies, 23, (July 1990), pp. 147-66.

61 Horowitz, p. 35.