Term Limits Manipulation across Latin America – and what Constitutional Design could do about it
The recent decision of the Supreme Court of Honduras – which essentially excised that country’s strict one-term limit from its constitution – has raised hackles from scholars concerned that it will open the door to would-be authoritarians seeking power. Indeed, the temptation for political leaders to extend their terms has preoccupied many countries and scholars in Latin America at different periods in history. Presidents in particular have an obvious incentive to extend their periods in office in order to consolidate power. The failure of presidents to cede power can lead to a significant amount of democratic erosion.
At the same time, the specificity of term limits has made them increasingly difficult for Latin American executives to disregard. Presidents who simply ignore limits on their mandate send a very loud signal to domestic and international constituencies of their defiance of the constitution. Mechanisms like the OAS “democracy clause,” which allows for possible suspension of member states for an “unconstitutional alteration that seriously impairs the democratic order” of a member state, might be activated in such circumstances.
As a result of these cross pressures, recent Latin American attempts to alter term limits have utilized the mechanisms of constitutional change – constitutional amendment or replacement. In other words, rather than simply ignoring constitutional term limits, presidents have friendly legislatures, courts, or voters alter or erase those limits. These efforts thus represent classic forms of “abusive constitutionalism” – the use of the tools of constitutional change to undermine the democratic order.
The key question is whether, and under what conditions, constitutional design can combat these threats. Five recent cases of presidential manipulation of term limits in Latin America – Nicaragua, Venezuela, Honduras, Colombia, and Ecuador – suggest that constitutional design is unlikely to offer a simple answer. Sophisticated tools of constitutional theory, such as the unconstitutional constitutional amendment doctrine, have proven effective in one key case (Colombia), but ineffective in others. In the extreme cases of Nicaragua and Honduras, courts acting at the behest of politicians used the unconstitutional constitutional amendment to abolish rather than protect term limits. (The very recent case of Guyana, where a court also eliminated constitutional term limits by holding that the wrong process of constitutional change had been used to insert them, is not treated in depth here). And in four out of the five cases, defenses failed and the term limits were (or are in the process of being) swept away. It is thus difficult to be overly optimistic, but the existing cases offer lessons and perhaps some modest improvements in the techniques of constitutional design.
The rest of this brief essay first discusses the background of all five cases. Second, I argue that constitutional designers will usually need to count on poorly-functioning and low-capacity, rather than high-capacity and heroic, courts when term limits problems come up. Third, I note that this has implications for constitutional design, and that scholars may be able to help develop best practices for constitutionally protecting term limits. Finally, I close by speculating about whether the emergence of modest regional standards – particularly against the existence of no term limits on executives whatsoever – would be useful and feasible as a starting point.
The Five Cases
First, the one successful case of Colombia: the popular, outsider President Alvaro Uribe won election in 2002 and gained significant popularity through his perceived successes in combatting guerrilla groups. He used his control of Congress to pass a constitutional amendment allowing a second consecutive term in office. This required approval by a simple majority of Congress in one term and an absolute majority in a second, thus amending Colombia’s fairly long-standing provision allowing only one presidential term. This amendment was challenged in front of the Colombian Constitutional Court as a “substitution of the constitution” or unconstitutional constitutional amendment, as well as for procedural defects, but a majority of the Court upheld the amendment, and Uribe won reelection in 2006.
In advance of the next presidential election of 2010, Uribe’s allies in Congress again sought to amend the constitution, in the form of a proposed referendum, in order to allow a third consecutive term. This amendment was also challenged in front of the Constitutional Court, which this time struck down the proposed referendum on both procedural grounds and as a substitution of the constitution. The Court’s reasoning focused on the fact that three consecutive terms would allow the president to dominate all of the various institutions – the Attorney General, Prosecutor, National Ombudsman, Central Bank, and judiciary, for example – that were supposed to be independent of the executive by constitutional design. It also emphasized that two term presidencies were fairly common internationally, but three term presidencies were considerably less common because of concerns about the consolidation of executive power and the erosion of democracy.
Now, for the four unsuccessful cases:
Venezuela: President Hugo Chavez won election in 1998 and quickly moved to consolidate power by writing a new constitution. The Constituent Assembly, elected via an electoral rule selected by Chavez, was more than 90 percent dominated by Chavez supporters. It drafted a constitution over about two months that strengthened presidential power but also included some interesting elements of popular participation, including recall provisions, popular participation in the selection of judges, and referenda. The new constitution lengthened presidential terms from five to six years and allowed presidents to be immediately reelected one time.
Subsequent referendums demonstrated a clear intent to abolish term limits so as to allow Chavez to remain in power indefinitely. A first effort to eliminate presidential term limits in 2007 was bundled with a broader set of constitutional reforms intended to deepen Venezuela’s transformation into a socialist country; the referendum was narrowly defeated. A second effort in 2009 was more narrowly focused on eliminating term limits for the president as well as governors, mayors, and national and regional legislators: this time the referendum was successful with 54 percent of positive votes. Despite a tiered system of constitutional change that might have provided some protection, the Venezuelan Supreme Tribunal (controlled by Chavez) did not play any meaningful role in limiting the efforts at constitutional change. Similar to the Ecuadorian Constitution, discussed in more detail below, the Venezuelan constitution distinguishes “amendments” from “constitutional reform” and a “constituent assembly,” establishing three different procedures for these three devices. According to the constitution, an amendment is not supposed to “alter the fundamental structure” of the Constitution; while a constitutional reform is not supposed to “modify [its] structure and fundamental principles.” By implication, only a constituent assembly can be used for changes that modify the fundamental structure. Nonetheless, the 2007 amendment was carried out as a “constitutional reform” while the 2009 attempt was carried out through the “amendment” procedure. The latter process was strongly contested by the opposition but the Supreme Court upheld it. Venezuela thus now has no presidential term limit, a provision that has survived the death of Chavez.
Ecuador: President Correa followed a somewhat similar strategy to that of Chavez – he won election in 2006 and replaced the Constitution in 2008, in a process controlled by his supporters. The 2008 Constitution originally created four year terms and allowed only one reelection; an early reform changed the constitution to allow three consecutive terms. In 2014, President Correa and his allies sought the elimination of all term limits on the country. The Ecuadorian constitution creates a tiered system of constitutional change, distinguishing amendment, partial reform, and the Constituent Assembly, and creating a different device for each procedure. Amendment may be carried out either by referendum, solicited by the president or eight percent of registered voters, or two-thirds vote in the unicameral Congress. Partial reform must be approved by a majority in Congress and then by a referendum. A Constituent Assembly must be called by majority approval in a popular consultation activated by either the president, two-thirds vote in Congress, or twelve percent of registered voters.
The text states that amendment of the constitution may not be used to “alter its fundamental structure, or the character and constitutive elements of the state, … establish restrictions on rights and guarantees, or … modify the system of constitutional reform.” Partial reform may not be used to “restrict constitutional rights or guarantees, or modify the constitutional reform procedure.” There are no explicit textual limits on the competence of an elected Constituent Assembly. The text also states explicitly that the Constitutional Court has the power to determine which route is the correct one for any particular change.
The debate in Ecuador was thus about which of these procedures needed to be pursued in order to eliminate term limits – the president sought to utilize the amendment mechanism, while the opposition argued that at least the partial reform procedure needed to be used. They argued that a constitutional elimination of term limits at least “altered [the] fundamental structure” of the constitution, and likely also “restrict[ed] the constitutional rights and guarantees” of the opposition. Politically, elements of the opposition noted that while Correa was personally extremely popular, the elimination of term limits was not, so they hoped to use the referendum as a way to build resistance. The Constitutional Court, in late 2013, nonetheless held that the amendment route could be used, and therefore that Congress could pass the change unilaterally. Given Correa’s domination of the Assembly, approval of the amendment appears to be a certainty.
Nicaragua and Honduras: The courts in Venezuela and Ecuador were ineffective in limiting proposals to eliminate term limits despite potentially deployable constitutional defenses; courts in Nicaragua and Honduras played an even more pernicious role, essentially deploying constitutional defenses themselves to sweep term limits away. In October 2009, after President Daniel Ortega had sought the elimination of term limits but lacked the support in Congress needed for a constitutional amendment to do so, the Supreme Court held that article 147 of the constitution prohibiting presidential reelection for more than one additional term was itself “inapplicable.” The Court reasoned that the term limit discriminated against the president and limited his political rights. Ortega was subsequently reelected to a third term, and in 2014 the Congress (now controlled by Ortega’s supporters) amended the constitution to remove the term limits, clearly allowing Ortega to run for a fourth term.
The Honduran case is even more bizarre and dramatic. The 1982 Honduran Constitution established a one term limit for presidents, made this term limit completely unamendable by any means, and removed from office any official who “attempted” to change the no-reelection rule. In 2009, this provision played a major role in the coup that removed President Manuel Zelaya. Opponents argued that the coup was legal because Zelaya was attempting to hold a non-binding public consultation on whether to replace the constitution, thus “attempting” to change the no-reelection rule. The coup was nonetheless widely condemned.
In April 2015, the Constitutional Chamber of the Supreme Court issued a decision holding that all three parts of the no-reelection rule (the rule itself, its unamendable nature, and the “attempt” clause) were unconstitutional pieces of the constitution, and thus held them inapplicable. The Court’s reasoning focused on what it saw as international human rights law: it held that these provisions limited the right to free speech and of the “free choice” of voters. Beneath the surface, the Court was likely acting at the behest of National Party incumbents, who were integral in the removal of Zelaya and who illegally removed and replaced four of five members of the Chamber in late 2012. Honduras now apparently has no presidential term limit.
“Problematic” Constitutional Courts
The Colombian Constitutional Court, in issuing the second reelection decision that stopped Uribe’s third term, is often theorized as a heroic court that helped prevent substantial democratic erosion in the country. Indeed, both the Colombian Court itself and the broader political context are clearly unusual, as the other four cases show. The Colombian Court is a high capacity institution that was able to create a “substitution of the constitution” doctrine despite the absence of explicit, textual tiers of amendment in the constitution. The design of the Court, particularly the judicial selection mechanism, may have helped some. The nine judges (who serve eight-year non-renewable terms) are selected by the Congress from lists of three candidates proposed by three different institutions: the president, Supreme Court, and Council of State (each of whom control the lists for three seats on the court). The fragmentation of the selection mechanism made the Court relatively difficult for even a strong president like Uribe to pack. Moreover, the political context proved important. Uribe was an extremely popular president, but his legislative support depended on a loose coalition, and since he won election as a political outsider he lacked the backing of a durable, cohesive political party or movement. Key political actors who had supported Uribe acquiesced in the decision and left him with little choice but to do likewise.
The other cases show what is likely the more typical context of a “problematic” constitutional court with low independence and low capacity. All four courts issued rulings that were almost certainly incorrect, and had serious negative consequences. Courts in both Venezuela and Ecuador should have held that constitutional changes to eliminate term limits were on a higher tier because they altered or modified the “fundamental structure.” But neither court did so. The courts in Nicaragua and Honduras issued bizarre rulings deploying the “unconstitutional constitutional amendment” doctrine (or something like it) to themselves eliminate term limits. The arguments in these two decisions are extremely weak and clearly politically motivated. The Honduran Supreme Court, for example, swept away the entire term limit provision despite having arguments that aimed only at the “intent to change” clause, and it seemed untroubled by the fact that it was utilizing the unconstitutional constitutional amendment doctrine against an original part of the 1982 constitution, not a later amendment.
Thus, theories of constitutional design that rely on heroic or even competent courts in these context are likely to fail. Constitutional designers must figure out regulations of term limit provisions that will work under sub-optimal conditions.
Constitutional Design Under Bad Conditions
Using this lens, the different experiences with “failure” all suggest potential improvements in constitutional design. Constitutional tiering is a broad term that encompasses a range of practices: designers sometimes protect a specific provision or section with a more demanding procedure, while in other cases they protect a much more abstract concept like the “fundamental structure” of the state or “constitutional rights.” Both Venezuela and Ecuador embody the latter approach. Both the specific and abstract approaches have advantages: the former provides clear, specific protection to particular constitutional articles, while the latter is more flexible, essentially providing textual support for a judge-made unconstitutional constitutional amendment doctrine.
But the two approaches are not mutually exclusive. In both countries, the flexible system of tiering based on whether a given change infringed on the “fundamental structure” could and should have been supplemented with specific reinforcement of the term limits provision. A specific protection of term limits, requiring any change to the term limit to be passed via a more demanding procedure, would have been more difficult for problematic courts – or international actors – to ignore. The broader lesson is that specificity may be an advantage when dealing with “bad” constitutional courts.
Honduras, of course, embodies a strong form of the more specific method of tiering: the term limit, along with the provision making it unamendable and certain other provisions, were specifically singled out in the text and made completely unamendable. This ultimately provided no protection. One might argue that the provision was actually too strong. First, as Richard Albert has argued, complete unamendability may be a counter-productive, as it provides no route short of legal revolution for actors to achieve their goals. A more realistic provision might have aimed to make change of the term limit difficult – both time-consuming and requiring a substantial supermajority as well as perhaps a referendum – but not impossible. A requirement that constitutional change of this type be carried out over several years, or take effect only after an intervening election, may be particularly useful as a roadblock to authoritarian projects which does not completely frustrate the possibility of constitutional change. Second, the “intent to change” provision in the Honduran constitution probably went too far. This provision caused serious problems during the Zelaya episode and gave the Constitutional Chamber at least some ammunition for striking down the term limit by arguing that the provision limited freedom of speech and political debate. The term limit itself, and a provision making it difficult and time-consuming to change, would have been sufficient.
The role of referendums in regulating changes to term limits is a topic that merits additional study. It is fascinating that in Ecuador the opposition’s arguments were aimed largely at forcing a referendum, which they thought they had at least an outside shot of winning; in Colombia by contrast Uribe’s supporters were confident they would have won the referendum, and in Venezuela, Chavez initially lost a referendum but returned two years later and was successful. The argument in Ecuador was perhaps overly optimistic – the evidence from Venezuela and Colombia suggests that referenda regulating term limits, and promoted by popular leaders, can become dominated by the personality of those leaders. Still, referenda might become a way for the opposition to organize and counter incumbents’ arguments in a particularly compelling way. A referendum, however, might become part of the “more difficult” amendment process, but it should be paired with other requirements like temporal limitations and super-majority requirements.
Finally, an alternative (and underexplored) approach might focus on the incentives of politicians. Absolute term limits like those in the countries under discussion here require politicians to vacate the presidential office forever once they have completed their designated term. The severity of this limit gives presidents a strong incentive to “cheat” on the constitutional bargain by any means possible. Term limits like those in Chile barring presidents from serving consecutive terms, but not prohibiting them from regaining office forever, might alter political incentives in a pro-compliance direction by lowering the stakes of leaving office.
Towards Regional Standards on Term Limits?
The analysis has focused on the domestic rather than the regional or international level. It has done so for a particular reason – the OAS in particular has recently proven reluctant to become involved in these sorts of issues. In the Honduran case, for example, the OAS (properly) suspended Honduras after the coup but showed little reaction either to Zelaya’s machinations or the recent Constitutional Chamber decision eliminating term limits. Incidents of presidents, legislatures, courts, or voters excising constitutional term limits raise difficult question of what Rosalind Dixon and Vicki Jackson call extra-territorial constitutional interpretation: they require outside actors to make an independent assessment of difficult constitutional questions, perhaps overruling domestic courts in the process.
At the very least, external actors should understand that courts can undermine rather than promote the democratic and constitutional order. Nicaragua and Honduras are both examples. The fact that a court legitimizes or carries out a given action should not automatically be a signal that the OAS should be unconcerned.
The deeper issue is that the OAS’s democracy clause is currently tied to the “constitutionality” of an attempted democratic erosion. This makes it hard, without carrying out tricky exercises of extraterritorial constitutional interpretation, for international actors to complain when domestic courts have either legitimized constitutional change as in Venezuela and Ecuador or actually caused it as in Nicaragua and Honduras. European institutions, particularly the Venice Commission intervening in the recent constitutional changes in Hungary, have moved further towards seeing that constitutional changes may be strongly anti-democratic in effect even if they are ambiguously or probably constitutional in the eyes of domestic actors. Moreover, judging the anti-democratic effect of a given change can be very tricky, because a change or package of changes that might be innocuous in one context might be very dangerous in another. Kim Lane Scheppele used the term “frankenstate” to describe the 2011 Hungarian constitutional changes where each individual element was found elsewhere in some form, but the combination of the changes was – she argued -- strongly anti-democratic.
Fortunately, the term limits issue simplifies some of these problems: changes to term limits are usually easy to observe and their anti-democratic impact can perhaps be observed more readily in isolation than other changes like those to the jurisdiction or selection mechanism of a high court. One can reasonably argue about whether one or two presidential terms is optimal, whether term limits should be absolute or merely consecutive, and about how long presidential terms should be. But the danger to democracy of unlimited or indefinite presidential term limits – which now exist in Venezuela, Honduras, and Nicaragua, and will soon exist in Ecuador – seems to greatly outweigh whatever benefit these provisions may provide. Given this, it is time for regional institutions in Latin America to say that having no presidential term limit is unacceptable.
David Landau is a Mason Ladd Professor and Associate Dean for International Programs at Florida State University College of Law. He has published on comparative constitutional law and Latin American law, focusing on socioeconomic rights, judicial role, and democratic transitions. In 2011, he served as a consultant for the Truth and Reconciliation Commission of Honduras. He has also co-authored analyses of this decision with Brian Sheppard for ICONnect (the Blog of the International Journal of Constitutional Law) and the International Edition of the New York Times.