HISTORICAL CONTEXT: ELSTER’S WAVES

According to Elster (Reference Elster1995), there were seven waves of constitution making worldwide until the early 1990s. The first wave came with the new and novel constitutions elaborated by the American and French revolutions in the eighteenth century, which subsequently came to the new Latin American republics. The second wave swept through Europe following the revolutions of 1848. The third wave came after World War I, with newly created states writing their own constitution (e.g., Poland and Czechoslovakia). In the fourth wave, Germany, Italy, and Japan approved constitutions under pressure for democratization from the victorious allied forces that defeated them in World War II. The fifth wave emerged from the breakup of the European colonial empires, beginning in India and Pakistan in the 1940s and gradually through Africa and the Caribbean in the 1960s. The sixth wave is a result of the restoration of democracy in Southern Europe (i.e., Greece, Portugal, and Spain) and Latin America (i.e., Ecuador and Brazil). The seventh wave broke out in Eastern Europe in the 1990s with the introduction of many new constitutions written after the end of the Cold War.

This overview demonstrates that the writing of new constitutions has been observed frequently in exceptional, disruptive contexts such as decolonization, military coups, and regime changes. Most of these constitutions have in common that citizens were not engaged in their elaboration (Brazil in 1988 is one of the few exceptions), and they have no regular means to directly influence them once in force. An exception in both senses is the Swiss constitution, in which citizens can activate a total replacement by signature collection—something that has never happened—and the few cases in which citizens can introduce constitutional amendments.

Constitutional replacements have been infrequent in stable democratic settings. Negretto (Reference Negretto2021) identified only 25 cases between 1900 and 2015. This low frequency is explained by a narrow definition of what a replacement is (e.g., the Argentinian process of 1994 is considered an amendment, not a replacement). Moreover, a strict definition of a democratic setting was required—that is, to be included in the sample, a new constitution must have been adopted at least five years after the founding election. This was expected to provide a clear replacement of the institutions of the authoritarian regime, excluding cases such as Spain in 1978. Finally, and particularly relevant for this argument, there is a legal explanation for this low frequency: typically, a constitution in force either does not regulate its replacement (e.g., the 1980s Chilean constitution) or it puts in place obstacles that make change difficult.

After the publication of Elster’s (Reference Elster1995) book, there were prominent new cases that shaped what can be considered an eighth wave. This was based on the expectation of including “direct” citizen participation and being “inclusive” (i.e., 13 of the 25 cases included by Negretto occurred in the 1990s)Footnote ² in which the constitutions approved of the so-called political revolutions in Latin America. These included Venezuela in 1999, Ecuador in 2007, and Bolivia in 2009 and also the Colombian and Ecuadorian constitutions of 1991 and 1998, respectively. Other constitutions that resulted from the “Arab Spring” (e.g., Tunisia) can be included—all of them controversial (Welp and Soto Reference Welp and Soto2019).

For those scholars who explain the growing dissatisfaction with representative democracies due to the perception that politicians are part of the elite pursuing their own interests,Footnote ³ the introduction of citizen participation is a solution. Despite having diverse and massive participation, the prominent cases of Iceland and Chile failed to provide legitimacy of the new constitutions or approval and implementation. Furthermore, past experiences demonstrate that there are different forms of engagement. When political parties are rooted in society, informal channels could communicate and legitimate the elite’s decision, which should have been the case of Spain in 1978 (García Guitián Reference García Guitián, Reuchamps and Welp2023). Therefore, it is key to consider that (1) legitimacy is not granted but instead built into the process; and (2) there is a need to surpass a superficial concept of participation or deliberation to understand how it is shaped and how it connects to the decision-making process.

FEATURES OF PARTICIPATORY CONSTITUTION-MAKING PROCESSES

Institutions of citizen participation refer to an array of forms that can be complementary. Five dimensions help in understanding the diversity of participatory institutions: (1) timing: when participation is activated (i.e., before, during, and/or after); (2) subject: who can participate (i.e., self-appointed, sortation of citizens (i.e., participants are sorted by lots in assemblies), invited, and/or elected); (3) place: where it takes place (i.e., virtual platforms, assemblies, and commissions); (4) format: how participation is expressed (i.e., signing to submit proposals, deliberating to produce a draft or on a draft prepared by a convention, and/or voting on issues); and (5) outcomes: with which results (i.e., not specified, consultative, or binding). Ordinary citizens can participate by campaigning in elections to become members of the constituent assembly (as happened in Chile), and they can vote to launch a constitutional amendment (as the law states in Switzerland) and/or to ratify the final draft (as in Spain in 1978). Nonpartisan actors can deliberate to produce proposals (as in Ireland in 2018) or on drafts prepared by a convention (as in Ethiopia in 1991–1995).

Empirical evidence shows that referenda are activated in democratic and nondemocratic settings whereas many processes of deliberation have lacked the basic conditions to be fair. It is not surprising that deliberative processes implemented in nondemocratic contexts tend not to meet the minimum requirements to be considered open and plural; however, it is surprising that this happens in many democratic contexts as well (Welp and Soto Reference Welp and Soto2019). This brief overview demonstrates that it is essential to define standards for fair participatory and/or deliberative democratic processes with some degree of influence. Additional lessons are to be learned from recent processes developed with higher standards and in democratic contexts.

ICELAND AND CHILE IN FOCUS

Chile (2020–2022) and Iceland (2009–2013) were selected for this study because they both illustrate different dimensions of the consequences of overestimating the potentialities of including citizen participation. In Iceland, a main obstacle to succeed in approving a new constitution resulted from the lack of connection between the participatory process and the formal institutions (Bergmann Reference Bergmann, Reuchamps and Welp2023). In Chile, the debate is ongoing and there are competing arguments: media manipulation from the right (Piscopo and Siavelis Reference Piscopo and Siavelis2022); “a set of identitarian outlooks that has for too long masqueraded as radical politics” (Rojas Reference Rojas2023); and weaknesses of parties and/or ideological representation (Larraín, Negretto, and Voigt Reference Larrain, Negretto and Voigt2023), among others. My focus is on how main promoters of direct citizens’ participatory processes trusted in obtaining a more legitimate constitutions, neglecting or underestimating the complexities of building such legitimacy on a daily and an institutional basis.

In Iceland, the attempt to replace the constitution stemmed from social protests over the 2008 financial crisis and the consequent economic collapse. Massive numbers of demonstrators flooded the streets and the politicians became very unpopular. New parties replaced traditional parties. The collapse of the government in 2009 and the call for new elections facilitated what would be a temporary replacement (Olaffson Reference Olaffson, Ingimundarson, Urfalino and Erlingsdóttir2016). In 2010, a left-wing coalition came to power and addressed the demand to promote a constitutional replacement. In November, the National Forum began to function with the participation of 950 citizens chosen by lottery. The National Forum was mandated to prepare a proposal that later would be discussed by a parliamentary committee consisting of seven members nominated by the political parties; from there, it would go to Parliament. The draft then was submitted to a commission divided into party blocs, producing a debate that limited the legitimacy of the process. The next step was to convene a constituent assembly consisting of 25 elected representatives. Barely a year after the great mobilization of 2009, only 37% of registered voters went to the polls. The election presented irregularities that led the Supreme Court to invalidate it in January 2011. This decision was highly controversial, which benefited the Parliament main actor’s position, opposed to the replacement. The Parliament nominated the assembly members and, in April 2011, the constituent body began to function, deciding to open up to the citizenry for a “crowd-sourced” constitution. The constitution was approved and submitted for a referendum on October 20, 2012, with a low voter participation rate of 48%. Six questions were approved: in addition to asking about the constitution, specific aspects were consulted, including the ownership of natural resources and the introduction of direct-democracy mechanisms. However, the constitution was rejected by the traditional parties, which had regained their position in Parliament (Gylfason Reference Gylfason2014; Olaffson Reference Olaffson, Ingimundarson, Urfalino and Erlingsdóttir2016).

On October 13, 2015, President Michelle Bachelet (2006–2010 and 2014–2018) of Chile announced a schedule to change the existing constitution dating from Pinochet’s era despite several amendments (Heiss Reference Heiss2022). The process that was initiated with a participatory experience of “citizen dialogues” did not end with a constitutional replacement as expected. Demands for constitutional change reemerged in October 2019. The increase in public-transportation fares was followed by waves of protest. Despite the repression, the mobilization did not end, which forced the government to open a dialogue with the opposition parties that set the scenario for constitutional replacement. The agreement included a referendum as a first step to decide whether to change the constitution as well as the appropriate body to conduct the process. In October 2020, more than 78% of Chilean voters approved rewriting the national constitution and—in a second referendum question—opted for a directly elected Constitutional Convention equally composed of women and men (unique in the world) that guaranteed an appropriate representation of indigenous councilors. In mid-May 2021, Chileans elected 155 representatives from more than 1,300 candidates from parties, social movements, and independent candidacies. Despite the low turnout (42.5%), the election reaffirmed Chileans’ commitment to overcome the status quo. Political-party candidates on both the right and the left received so few votes that neither traditional right- nor left-wing forces were able to veto proposals on their own in a convention dominated by independent citizen candidates. The process also was open to other forms of participation. However, on September 4, 2022, the proposed new constitution was defeated: 61.9% against and 38.1% in favor, with a participation rate of 85% in the first mandatory vote.

It has been argued that the problem in Iceland was based on both the prioritization of consensus and the design of the process, which left power in the control of the traditional institutions (Olaffson Reference Olaffson, Ingimundarson, Urfalino and Erlingsdóttir2016). In other words, the National Forum that produced the draft constitution operated on a consensus basis, independently of parties. It produced poor results because there was a disconnect with the institutional framework—which partly explains the subsequent failure with traditional parties rejecting the text—whereas the search for consensus from the point of view of “ordinary citizens” produced a vague text susceptible to various interpretations. In Chile, the procedures were clearly established and agreed to; however, this process also was frustrated due to its defeat in the referendum. Of course, is not possible to isolate a single explanation; however, I do consider that the anti-partisan discourse behind both conventions led to an underestimation about the need to work not only on participatory mechanisms but also on building representation channels and connecting the process with formal and informal institutions. This led to an expectation of legitimacy simply because of the participation of independents or ordinary citizens as “the people.” To summarize, in both cases, the more “citizen-driven” nature of the process was not sufficient to reach more legitimacy due to several factors. These include the legal framework in which the process was installed (i.e., Iceland) and the limits of new forms of representation (i.e., civil-society organizations and new parties) to build legitimacy.

LESSONS LEARNED

The two cases analyzed in this study suggest that there are fallacies damaging the processes of democratic renewal expected by the promotion of citizen participation. These fallacies have been fueled by the predominance of the idea of self-determination of “the people” and the suspicion that limitations to exercise that will are imposed by the institutionalization of representation and the rule of law (García Guitián Reference García Guitián, Reuchamps and Welp2023). This led to the idealization of direct participation, which is rooted in Jean-Jacques Rousseau (i.e., the assembly of free and equal citizens as the ideal democratic model) and Hanna Pitkin (i.e., representation as the only available option of putting democracy into action because of the impossibility of implementing direct participation; in other words, representation as “the second best”), and which gave place to the three fallacies.

First, there is no people’s epistemic superiority. Nothing allows us to attribute to nonpartisan leaders or those from social movements “being the people,” acting as spokespersons of the general will and, accordingly, transcending pettiness for the benefit of the whole. A feminist or an environmental activist has a specific agenda and a party also should have a programmatic agenda. “The people” do not comprise an entelechy, and those who act as their representatives should behave according to their conditioning factors (e.g., ethnic, gender, and class). Far from being a problem, this is beneficial given that in greater descriptive representation is the foundation for greater inclusion as a necessary but not sufficient condition. Democracy is a method to organize the plurality of interest with some limits framed by the rule of law.

Second, the mechanisms that put participation and representation into action are diverse and, far from opposed, feed off on another. Participation refers to a multiplicity of formats that in no case eliminate mediation. Members of Parliament participate and represent their electorate in the same way in which a civil-society leader is expected to speak to some extent in the name of civil society. Electoral rules with their validation thresholds and requirements for decision making, the features of leadership, and even the order of speaking—to name only a few aspects—have an influence on a deliberation process and its results.

Third, the legitimacy deficit is not resolved by injecting participation because participation and representation go hand in hand. If either one is absent or severely deficient, the result will be poor. In contemporary democracies, the most widespread method of participation is in elections. Strengthening democracy requires a good design of institutional channels so that citizens can make their voices heard—for example, with popular initiatives that can be activated by collecting signatures—and high-quality representation through parties and social leaders who are recognized and valued.

The reflections in this article were inspired by the defeat of the proposed Chilean constitution, but they are not intended to point out that the overwhelming rejection can be explained by an abstract and unequivocal reason such as “the crisis of representation.” However, this article does seek to draw attention to three fallacies that lead to an underestimation of the weight and complexity of the construction of legitimacy in contemporary democratic systems.