A failed but useful constitution-making process: How Bachelet’s process contributed to constitution-making in Chile

Abstract

This article shows how failed constitutional proposals may contribute to future constitution-making processes by exploring the relationship between the recently failed Chilean constitution-making process (2019–22) and the previous unsuccessful one led by former President Michelle Bachelet (2015–17). Comparative constitutional scholars are yet to fully understand how constitutional failures of this kind can take place, and Bachelet’s process has not received the attention it should. This article fills that gap by showing how both processes were driven by shared principles initially set by Bachelet. It also shows how those principles may serve as a blueprint for future constitutional changes in Chile. Bachelet had campaigned on the basis that any constitutional replacement attempt should be participatory, institutional and democratic – all ideas that have remained popular in Chile’s political landscape. Those ideas have served the purpose of both reducing transaction costs among constitutional negotiators and securing large compromises in polarized political scenarios.

I. Introduction

Comparative constitutional scholars have not paid much attention to failed constitution-making processes. The Icelandic¹ and Chilean experiences² are relevant exceptions, but exceptions nonetheless. Further theorization is still pending.³ It may be easy to assume that failed constitution-making processes are unlikely to contribute to future agendas of constitutional change but, as I will show in this article, this assumption is a mistake. We cannot understand, for example, the French IVth Republic’s Constitution without understanding the reasons behind the failed referendum of May 1946. This article is a first attempt to fill a gap in the literature, and it focuses on one specific issue that needs more attention: whether the prevailing political ideas in a failed constitution-making process can still influence a broader constitutional change agenda. I will argue that they can, and I will use the Chilean case as an example to illustrate this argument.

In October 2020, most Chilean voters supported a fully elected, gender-balanced Constitutional Convention that was going to propose a new Constitution. However, this historic opportunity to complete a constitutional replacement in democracy⁴ came with two background goals that had influenced the constituent-making debate for the last decade: to avoid the Bolivarian paradigm of constitution-making (closely associated with an unlimited and fully sovereign Constituent Assembly) and to put an end to the institutional and symbolic legacy of the Pinochet dictatorship.⁵ The first of these factors explains why the Chilean constituent process of 2019–22 was highly procedurally regulated by the constitutional amendment that authorized the constituent process.⁶ The process was supposed to be pursued within established and accepted channels of reform that did not entail a drastic break from existing institutions. In other words, the process was ‘institutional’.

I claim that this institutional dimension of the process was not new for Chilean politicians and scholars who participated actively in the design of the 2019–22 process. This dimension became popular due to the political narratives that justified former President Michelle Bachelet’s constituent process (2015–17). Bachelet herself had advocated that any constitution-making process should observe three core principles: an institutional, democratic and participatory constitution-making process – hereinafter, Bachelet’s trilogy. This article shows how Bachelet’s trilogy remained a politically feasible and constitutionally desirable solution that could help the parties find common ground to unlock the constitutional framework after the October 2019 social demonstrations pushed them to achieve a deal. By doing this, the parties could enrich Chile’s constitutional tradition by identifying a way to manage the tension between the need to maintain a channel of institutional continuity and respond to demands for constitutional change. In that way, parties could set up the new bases for the new constitution-making process. Despite the failure of the Constitutional Convention in having the citizens approve its proposal, I claim that Bachelet’s trilogy is likely to remain a political narrative that could still help to find common ground in negotiating constitutional reforms. It is sufficiently abstract to adapt to many types of procedures and sufficiently dense to reject extreme solutions.

Comparative scholars should take note of Bachelet’s trilogy because it shows how two failed constitution-making processes can still contribute to develop long-term constitutional change solutions by building common political narratives that can then be used by rival political actors when bargaining about alternative paths for constitutional reform. This is no minor contribution to a failed constitutional change project.

II. Bachelet’s constituent process (2015–17)

Although many provisions in the Chilean Constitution could be traced back to either older constitutional documents (1828, 1833 and 1925) or the several amendments enacted during post-authoritarian era, the current version of Chile’s Constitution is mostly the result of a political process initiated when the Pinochet dictatorship imposed the 1980 constitutional framework,⁷ manipulating a plebiscite that lacked democratic credentials.⁸ This imposition imprinted a hardly redeemable sin of origin on the Constitution.⁹

Despite the major amendments of 1989 and 2005,¹⁰ the second of which was presented by former President Lagos as a ‘new Constitution’ that ended Pinochet’s ‘authoritarian enclaves’,¹¹ new efforts towards reform began in 2006.¹² In 2008, the centre-left presidential candidate and former President (1994–2000), Eduardo Frei Ruiz-Tagle, a Christian Democrat, campaigned to replace the Constitution. A long-standing left-wing agenda became mainstream. However, the proposals of that time typically established that the new Constitution had to be drafted in Congress or by a committee of experts, not by a specialized elected constitutional assembly. In those years, that type of assembly was widely perceived as a non-institutional mechanism, far from the historical (and often subject to controversy) ‘Chilean exceptionalism’ when considering institutional development in Latin America.

Among the myriad reasons that explained the new scenario, it has rarely been highlighted that the 2005 reforms were too little and came too late. Most of these reforms were already negotiated in March and April of 1989, in a bipartisan roundtable that took place right after the 1988 referendum that ended with Pinochet’s dictatorship. For example, the Concertación coalition (a centre-left alliance that had opposed the Pinochet regime) and Renovación Nacional (RN, the center-right party of the later President Piñera) had unsuccessfully agreed to amend key controversial institutions such as the binomial system to elect legislators – which was designed by Pinochet’s government to enhance the veto power of the rightwing parties¹³ – and the supermajority requirement to approve and modify the organic laws – a set of particularly relevant laws enacted by the Junta with the aim of preserving crucial policies of the regime.¹⁴ Those reforms were approved, respectively, only in 2015 and 2022. Even though influential later, the Concertación–RN agreement did not have a real impact on the (official and final) 1989 amendment due to the final veto power of the Junta.¹⁵ For the next decades and until recently, the centre-left criticized the lack of commitment of the centre-right (and RN in particular) to fulfil the 1989 agreements during the post-authoritarian era. Under this scenario, President Bachelet’s impressive victory in the 2013 presidential run-off elections (ballotage) made it plausible that during her second administration (2014–18), the promise to replace the existing Constitution with an entirely new constitutional text could be fulfilled.¹⁶

The constitution-making process was officially launched in a television address to the nation in October 2015. It comprised four phases and was supposed to be completed by 2017. The first phase was a civic education preparatory stage aimed at instructing citizens on the basic concepts and components of a constitution and a constitution-making process. The second phase was a participatory stage composed of individual consultations and local, provincial and regional meetings, which was overseen by a bipartisan fifteen-member Citizen Council of Observers (CCO, Consejo Ciudadano de Observadores), handpicked by Bachelet. With the help of a Systematization Committee, the future constitution-making organ could use the inputs of the participatory stage by examining a consolidated document, including the proposals made by the constitution-making body. The third phase of Bachelet’s process consisted of submitting and approving a constitutional amendment aimed at opening a constitutional replacement process within the established channels of reform. The amendment was supposed to be approved by a two-thirds majority of Congress, and it was going to have two parts: first, the establishment of a special procedure to elaborate a new Constitution; and second, the presentation of a constitutional proposal to be discussed by the constituent body approved by Congress.¹⁷ Finally, the discussion and elaboration of a proposal for a new constitution would be undertaken by the specific body chosen by congressional decision, which should be ratified in a referendum.¹⁸

Therefore, it should come as no surprise that Bachelet’s political and constitutional strategy came to the three sensibly chosen ideas mentioned above: ‘institutional’, ‘democratic’ and ‘participatory’, a carefully designed trilogy.¹⁹ Although the ‘participatory’ component was the most innovative, both from a comparative²⁰ and a domestic perspective,²¹ the ‘institutional’ component was the most critical because of Chile’s aversion to a Bolivarian constitution-making process²² and the need for a post-sovereign limited strategy of constitutional replacement.²³ The institutional element also helped to respond to the right-wing parties main fears, which still translated into a strong and permanent opposition to Bachelet’s process.²⁴ However, that same institutional component successfully ensured that the fears of the right could be addressed during the negotiations that later set up the 2020–22 constitution-making process.

Thus, the ‘institutional’ component was manifested throughout the process in several decisions involving institutional design and rules. Examples abound. First, Bachelet decided to explicitly amend Chapter XV of the existing Constitution and include a special procedure for constitutional replacement. The special procedure, approved by a two-thirds majority of sitting members of both houses of Congress, was largely considered as a consensus-building and inclusive process that would avoid a drastic legal rupture.²⁵ Second, Bachelet’s Constitutional Bill of March 2018 regulated the option of calling a specialized Constitutional Convention with the single purpose of enacting a new constitution and the obligation to dissolve itself after the purpose was achieved.²⁶ This decision aimed at avoiding the possibility of a runaway convention abusing its powers and trespassing on other institutions’ authority. Third, the process was deliberately designed to include a number of institutional actors – the Government, the CCO, the Systematization Committee, the Congress and a constituent-making body – working incrementally and inclusively. Each of them was supposed to have specific tasks and limitations in a logic of checks and balances. Even though the process was largely seen as ‘Bachelet’s process’, the design aimed to prevent an executive branch-driven process, or even the appearance of it. The core idea was to prevent the caudillo’s dominant role that Latin America has experienced in previous constitution-making processes, signalling a commitment to a competitive and democratic process.²⁷

Bachelet’s process failed at the Congress stage, but it succeeded in leaving a powerful blueprint for future constitutional change. That blueprint was based on the trilogy’s political narrative, justifying the establishment of several participatory, consensual and institutional mechanisms.

Bachelet’s constitution-making process failed due to several factors. A key issue was the exclusion of political parties from participating in the earlier stages of the process. Even President Bachelet’s supporting coalition parties were not involved, and their demands were ignored. (This mistake was also replicated in the 2021–22 Constitutional Convention, as other contributions from this symposium have shown.) Furthermore, the left heavily criticized Bachelet’s strategy.²⁸ Some argued that it was too conservative, while others possibly wanted to be consulted about the content of Bachelet’s constitutional proposal. Criticism from the left increased when Bachelet became a lame duck president, and the right-wing candidate Sebastián Piñera was elected. That moment was important because Bachelet had decided to submit her proposal to Congress after the presidential elections (November 2017) and before the effective rotation of power (March 2018). She presented the constitutional proposal during the last week of her mandate.²⁹ The constitutional draft was written in secret by officers of her administration, and it did not consider relevant citizens’ proposals systematized by the CCO.³⁰

When President Piñera took office in March 2018, he immediately announced that he was not going to continue with Bachelet’s process – only minor constitutional reforms were part of his electoral manifesto³¹ – and the Congress interrupted the process by putting the proposal aside.

III. Bachelet’s trilogy and the design of the failed Constitutional Convention (2019–22)

The 15 November Agreement (Acuerdo por la Paz Social y la Nueva Constitución)³² was the political parties’ response to the October 2019 social outburst and its demands. Large parts of the 15N Agreement and the constitutional amendment that implemented it were based on the parties’ fears. The right-wing parties feared a Socialist and neo-Bolivarian takeover that could potentially damage the market-based economy and other constitutional bases. The left-wing parties feared the right-wing parties’ long-standing opposition to demand for a new Constitution. Bachelet’s trilogy (the democratic, participatory and institutional constitution making-process) appeared as an appealing common-ground solution, which could help to find a compromise among politicians who distrusted each other.³³

The trilogy was relevant at least in two moments: first, when legislators and leaders of the political parties negotiated the 15N Agreement; and second, when a fourteen-member bipartisan technical committee appointed by the parties worked on the draft of the constitutional amendment that was going to implement the agreement.³⁴ That committee unanimously elaborated the amendment’s draft in fifteen days, and included a special procedure to elaborate a new Constitution by adding a new section³⁵ to the existing Constitution’s Chapter XV.³⁶

The right-wing parties had demanded the institutional component of the constitution-making process to be strengthened, and the technical committee established several substantive and procedural limits to the powers and functioning of the constitution-making body.³⁷ This body was carefully named a ‘Constitutional Convention’ instead of a ‘Constituent Assembly’ for symbolically important reasons: the idea was to reject the establishment of a of a sovereign assembly with unlimited powers based on the constituent power theory. That way, a rejection of the neo-Bolivarian model was obvious. The Convention was supposed to have an explicit mandate with demanding limitations and scope, including a rigid timeframe and the obligation to dissolve itself automatically. It was supposed to be institutional.

The rules stated that the Convention’s constitutional proposal must preserve the republican form of government and the democratic regime, guarantee compliance with final court rulings and respect all ratified international treaties. Infringement to procedural limits could be presented to a panel composed of randomly selected Supreme Court justices – the left demanded that the Constitutional Court be left outside the process. The constitutional proposal was supposed to be ratified 60 days after the end of the Convention’s work by means of another referendum – the ‘exit plebiscite’. If the ‘Approval’ vote won, the new Constitution would have replaced the current Constitution after the new document got published in the Official Gazette, and the new constitutional provisions were supposed to be implemented gradually, according to the temporary provisions that the Convention was also supposed to pass. Had the Convention failed to write a constitutional proposal within its mandatory functioning period, or had citizens rejected the Convention’s proposal on the ‘exit plebiscite’, the Convention would have been dissolved and the current Constitution would have remained in place. Moreover, the constitutional amendment regulating the Convention’s procedure needed to be supplemented by additional procedural rules, including the voting procedure. To avoid the establishment of a runaway Convention, the amendment also established that the Convention was required to approve its procedures and the norms of the new Constitution by decisions supported by two-thirds of the Convention’s members.

Many of features mentioned above were taken directly from Bachelet’s proposal or inspired by it: the language of a ‘Constitutional Convention’, the limited scope of the Convention’s powers and the need to dissolve de Convention automatically after it had released a constitutional proposal, among others, are good examples that illustrate how the ideas of Bachelet’s process were present, and helped to strengthen the ‘institutional’ component that mostly interested the parties from the right-wing coalition.

The left-wing parties also had essential demands for the constitution-making process. First, they secured the existence of mechanisms guaranteeing citizens’ direct participation in the process, one of Bachelet’s most attractive and innovative processes. The idea of taking the participatory dimension from Bachelet’s process was to find a way to channel the frustration expressed by the massive social outburst. Some members of the technical committee even suggested imitating Bachelet’s local town hall meetings and self-convoked citizens’ roundtables. The final agreement was more moderate, though heavily inspired by the need to identify new channels for citizen participation. The second demand from the left-wing parties was to have the Convention write the constitutional proposal from scratch – that is, without the existence of default rules and without privileging the constitutional status quo in any meaningful way. Third, the Convention was not supposed to approve the constitutional proposal as a whole, but rather vote on the proposed provisions ‘article by article’. That way, the left ensured that the right-wing parties would not have the power to boycott the constitution-making process at the final stage.

The procedures described above were insufficient to guarantee a coherent constitutional proposal because the risk of approving inconsistent provisions could not easily be controlled. The parties had failed to secure a vote ‘chapter by chapter’ (in Chilean law, chapters are broader than articles), or a requirement that the Convention have a final vote on the constitutional proposal as a whole. The establishment of a strong harmonization committee was proposed but rejected.³⁸ These last features reflected the way the left-wing parties distrusted the right-wing parties. If they were going to design a constitution-making process, it was essential to secure that the existing Constitution was going to be left behind. The plan was to seize a favorable moment in which public opinion was particularly inclined to put an end to the ‘Pinochet’s Constitution’.³⁹

Finally, both the left-wing and right-wing parties agreed on designing a constitution-making process that could respect the ‘democratic’ dimension, thus also resembling the democratic principle of Bachelet’s trilogy. They did that by having the Constitutional Convention popularly elected. They reproduced the electoral rules of the lower chamber of Congress for electing the members of the Constitutional Convention, and used them as the basis for the new process.

Bachelet’s process was not imitated by the designers of the Constitutional Convention; however, the influence of Bachelet’s trilogy was notorious and sometimes even explicit. It served as a relevant – although not exclusive – blueprint for the 2019–22 constitutional process.

IV. Lessons from the Chilean experience

Bachelet’s trilogy has been very influential in Chile, as it was decisive in the design of the failed Constitutional Convention. Even though this has not been discussed in this article, Bachelet’s trilogy has also been influential in the design of the 2023 constitution-making process.⁴⁰ Bachelet’s process should receive more attention from comparative scholars for several reasons. First, Bachelet’s trilogy offers a plausible and influential alternative to the sovereign constituent assembly model and partly resembled Arato’s ‘multi-track’ constitutionalism⁴¹ and the post-sovereign model of constitution-making.⁴² That way, Bachelet’s process is a useful example of how Latin American constitution-makers do not need to follow the path of the neo-Bolivarian constitutions, largely modelled on the paradigm of sovereign assemblies. This is not to say that Bachelet’s process was perfect – it did fail to produce a new constitution, after all. Nevertheless, Bachelet’s process should be seeing as an invitation to reflect on how to strengthen and implement the post-sovereign model, carefully considering the institutional design problems that could have affected the final outcome. It should also be considered by those critically examining whether the influential constituent power theory is still useful.⁴³

Second, Bachelet’s process shows that even failed constitution-making processes can leave important and specific lessons for future attempts at constitutional change.⁴⁴ For the legislators who negotiated the 15-N Agreement, and for the members of the technical committee that drafted the constitutional amendment that implemented the Agreement, Bachelet’s trilogy and specific mechanisms taken from Bachelet’s process were a direct influence. Mechanisms such as the two-thirds majority rule were reactivated in ways that allowed significant reductions in the transaction costs among political parties negotiating the rules of the new constitution-making process. This is particularly notable because having consensual decisions in the context of an intense political crisis with high levels of polarization, and little time to act, is not easy. Identifying common-ground solutions, such as those offered by the Bachelet model, is particularly helpful.

Third, Bachelet’s process (as well as the 2019–22 Convention process) contributes to confirming the findings of those that claim that political parties are, and should be, a relevant component of constitution-making processes taking place in the context of a democratic regime.⁴⁵ The Chilean case shows that constitution-making processes that try to circumvent substantive political pacts between elites, with a bottom-up strategy (such as Bachelet’s process), or use the process to achieve political, economic and social transformations against political parties, elites and instituted powers (such as the Convention’s strategy) can fail. This is an important lesson for comparative scholars.

Fourth, looking at Bachelet’s process compared to the 15-N Agreement also entails assessing how trust among political elites becomes a relevant factor for constitution-making. Trust is a key component of constitution-making processes because it can help to reduce transaction costs while producing more enduring commitments among the negotiating parties.⁴⁶ This was clear in the case of Chile, where the historical mutual lack of trust between left-wing and right-wing parties regarding constitutional change conspired against achieving substantive political agreements among political elites. In this context, the way the designers of the Convention’s process used Bachelet’s trilogy showed a feasible alternative for how to build trust in unlikely scenarios.

Comparative scholars should carefully consider failed constitution-making processes such as those of Bachelet and the Convention, and identify the factors and conditions that lead to such failures. To assume that failed constitution-making processes are not a contribution to our understanding of how constitutional change takes place would be a mistake. Constitutional change can also be explained by past failures and the lessons learned from those experiences, and the Chilean experience is a useful illustration of this.