Blog of the International Journal of Constitutional Law
Home Developments I-CONnect Symposium on the Chilean Constitutional Referendum – The Paradox of Constitution-Making in Democratic Settings. A Tradeoff between Party Renewal and Political Representation?

I-CONnect Symposium on the Chilean Constitutional Referendum – The Paradox of Constitution-Making in Democratic Settings. A Tradeoff between Party Renewal and Political Representation?

Sergio Verdugo, IE University Law School


In 1995, Jon Elster concluded his landmark article by identifying the paradoxes of constitution-making.[1] Constitution-making processes are unlikely to occur without the urgent need that a political crisis triggers. Even though political crises are arguably necessary for constitutional replacement processes, those crises can also create unfavorable conditions for good constitution-making. In this comment, I argue that the failure of the proposed Chilean constitution of September 2022 connects to a more specific version of the Elster paradox. Democratic states attempting to change their constitutions typically face a strong push for renewing the political elites and party structure. However, the calls for political renewal also typically favor highly participatory processes requiring constitution-making bodies that can represent citizens. The problem is that, under those conditions, political representation becomes a challenging goal to achieve. Even a highly inclusive Constitutional Convention such as the Chilean Convention is exposed to this risk. Before I get to the Chilean case, I’ll briefly reflect on the dangers of constitution-making in democratic settings.

The risks of constitution-making in democratic settings

Constitution-making processes made in democratic contexts do not have obvious historical justifications. Despite Jeffersonian arguments that call for a total and periodic constitutional replacement,[2] most constitution-making processes occur not in historically predictable scenarios but in times of crisis. Post-war constitutions, transitional constitutions, and regime-change constitutions are easily understood and widely reported in the literature. Calling for a revision of all relevant political institutions while empowering a constituent organ—i.e., a convention, an assembly, a committee, a parliament—to decide which matters should be included in the constitution is a radical measure that could end up in the dismantling or the refounding of the institutional framework. We should be suspicious if calls for such a radical measure can undermine political institutions justified by a democratic principle. If there is a reasonably independent judiciary, elections are genuinely competitive, a minimum commitment to basic rights exists, and rotation in power occurs relatively frequently according to preestablished electoral rules, it would be hard to make compelling normative arguments for total constitutional change—unless one is a Jeffersonian, of course, or unless the purpose of the change is not a drastic plan for political revision but a more targeted reform agenda.

Even though political crises may give more weight to normative arguments advocating for constitutional replacement, the risks of democratic erosion that these processes may bring to political regimes may outweigh the expected benefits.[3] Populist narratives can associate their political projects with constitution-making agendas seeking to entrench a unilateral constitution and turn a ruling coalition into a hegemonic political force.[4] (But constitutions should not be unilateral or partisan documents unless one is willing to defend a situation in which the basic rules of the democratic system are permanently challenged by those who had not supported the constitution-making efforts in the first place.). Leaders advocating for constitutional change can try to get around this criticism by appealing directly to citizens and offering them a robust—though hard to enforce—bill of rights. But constitutions are no magic bullets,[5] and the democratic price for such a rich bill of rights can elevate too much.[6] To be sure, if those types of populist plans are successful in democratic settings, then it is perhaps possible that the political institutions of those settings were not working very well. Suppose this is true, and the reasons are connected to specific design flaws of the political process. In that case, it may be helpful to offer a constitutional reform plan to correct a critical weakness of the constitutional system. If a political crisis elevates, advocates of the reform plan may find arguments for trying an elected constituent assembly or a type of citizens’ convention that can allow them to achieve a symbolic gesture that can help alleviate the crisis in the short term. In the long term, the new rules are supposed—we hope—to make the political process work better. The risks of democratic erosion should be controlled, even at the expense of involving the political elites.

In these contexts, it is essential to consider that the conditions for a functioning—though flawed—democracy already exists and should not be put at risk. Constitution-making organs should thus try hard to justify each revision and should not operate with broad discretion. Suppose the judiciary and the electoral organs, for example, are independent and capable of releasing impartial decisions. In that case, constitutional replacement processes should not use their radical constituent power narratives to reboot the judiciary or revise each of the relevant rules of the judicial system. If a change is desirable, it should not harm existing institutional virtues. Otherwise, the constitution-making process may be captured by undemocratic or illiberal agendas seeking to pack courts, punish existing officers, or politicize institutions that should not be politicized—such as the Central Bank. In these cases, the constituent power narrative should not be warranted.[7]

Political parties and constitution-making

In democratic regimes, parties are supposed to reduce information costs for the voters, help discipline groups of politicians to achieve agreements and compromises, and signal the types of policies that are likely to be approved.[8] When political parties perform these functions well, political representation can work properly because voters will know the candidate’s platforms, and politicians will advocate for demands that citizens and other politicians know. They can be accountable for not pursuing those demands. On the contrary, if a set of unknown politicians are elected without specific platforms or forms of accountability—such as reelection or the power of party structures—they are likely to behave in unpredictable ways. When independents campaign on single issues without offering ideas on, for example, the form of state, reproductive rights, the way to select judges, and so on, there is no party structure behind those independents that can signal to the voters how the candidate will behave if that candidate is elected.

Voters may like a committed underdog that proposes to fight climate change or empower regional structures. They may like that candidate even more if they are campaigning against an elite they distrust. But they might not know anything else about that candidate, particularly if the information costs elevate too much. Think, e.g., of a farmer who cares about fighting climate change and votes for a single-issue candidate who offers to fight climate change. If the climate change candidate ends up advocating to undermine the farmer’s water rights and replacing them with a new structure of administrative permits, the farmer will be disappointed. Now think of the number of issues that that candidate needs to address in a constitution-making process that the farmer was unaware of during the campaign—e.g., the form of the state, reproductive rights, political regime, same-sex marriage, and so on.

The choice of using the party structure in a constitution-making process is a crucial—yet under-researched—idea that can explain the success of the process in terms of political representation. Unlike new states that need to build their party systems, existing democratic political systems have operative political parties and do not need to start designing them from scratch. As outsiders may promote a unilateral constitution-making process and reject the idea of political opposition, it makes sense to control these risks using the existing party structure. It also makes sense to use the parties to lower the costs of political representation. The problem is that citizens may feel that the parties do not represent them correctly. They may also think that the political class should be renewed and that the constitution-making process may provide an effective avenue for political renewal.[9]

This posits a problem for constitution-making processes in democratic settings because the political crisis that triggered the demand for a constitutional replacement may be aggravated if the party system is ignored or harmed, but citizens may reject the process if they see that the parties have too much control. In other words, we may use the benefits that the party structure provides—e.g., lowering information costs to improve representation while having experienced politicians trying to achieve bipartisan agreements—but we need to find a way for the citizens not to reject the process. The puzzle is not easy to solve.

The push for political renewal has another risk that we should consider: it can harm the possibility that the constitution-making will work. The exciting Icelandic constitution-making process (2010-2013) failed when the Parliament lost interest in it despite a highly participatory—and arguably legitimate—process.[10]  The reasons why most of the recommendations of the Irish Convention (2012) and the Irish Citizens’ Assembly (2016) have not yet been implemented are plausibly connected to party interests.[11] The Chilean President Bachelet’s constitutional proposal (2017) did not work because the parties were not part of the relevant drafting stages.[12]

The Socially Inclusive and Politically Unbalanced Chilean Constitutional Convention

Even though 61.86% of Chilean voters (the turnout was 85.82% in the context of compulsory voting) rejected the constitutional proposal, it is implausible to argue that Chileans did not want a new constitution—the Constitutional Convention started to function after a 78.27% of voters approved the idea of enacting a new constitution in 2020. What happened? 

The Chilean political crisis was made visible after massive demonstrations took over the streets in October of 2019 in the context of high dissatisfaction with the ruling elites and the representative institutions,[13] and a multiparty agreement signed by most—but not all—political parties offered a constitution-making process as a way out.[14] It could be argued that the parties had tried to control the process and entrench their “institutional interests” by reproducing the electoral rules that are used to elect the members of the lower chamber of Congress.[15] Those electoral rules had benefited the parties in the past by establishing a proportional formula that maximizes large electoral pacts and list-formation while forbidding independent candidates to form non-partisan electoral associations. If an independent wanted to run for office, he or she was required to either get on the ballot space of a political party or run individually and without any partner-candidates that could help him or her to add more votes as a group.

However, the pressure to amend the procedure to elect the Convention members grew stronger. The modifications that Congress made to the electoral rules included 17 reserved seats for indigenous peoples and allowed independents to create parallel electoral pacts to have a competitive space on the ballot. Some independents ran in alliance with the party lists, but many formed non-partisan alliances. Many of those alliances were built on anti-party platforms and were successful.[16] The most successful of those alliances—the Lista del Pueblo—campaigned on the idea of former President Piñera dying. It obtained 27 members of the Convention. The Convention had 105 independents (out of a total of 155 members).

Many independents campaigned on single-issue platforms, including demands connected to environmental, feminist, regionalist, healthcare, and indigenous agendas. The electoral pacts of independent candidates did not form new political parties. They did not succeed in offering a holistic approach to public policies, nor to what the content of the new constitution should be, nor were they supported by a precise and easily recognizable political doctrine or ideology. We knew, for example, that the Lista del Pueblo was not conservative and was probably from the left. Still, we did not know its members’ positions on issues crucial for constitution-making. Many independent candidates had a hard time releasing platforms that voters could know. But voters still voted massively for independent candidates. They probably did so as a punishment against political parties they did not trust. Even though the political conditions that promoted that behavior was understandable, they also harmed the possibilities of the Convention to adopt positions closer to the median voter and damaged its representativeness. The only mechanism that could make the Convention accountable was the exit referendum. Even though this is sufficient reason to question those who argue that the exit plebiscite was a bad idea,[17] the exit referendum only played a modest role in moderating the positions of the independents.

Elsewhere I have shown how the Convention proposed a Constitution that was far away from the preferences of Chilean citizens.[18] Examples abound: Chileans may agree on having limited reproductive rights, but an arguably unlimited abortion is out of the question for most Chileans. Chileans may agree on recognizing indigenous peoples and the multicultural nature of the national state, but they disagree on establishing a plurinational state.[19] Chileans also differed on having a “regional state,”[20] and a majority favored keeping the traditionally bicameral Congress, including the Senate.[21] They preferred clear compensation rules in case of expropriation,[22] and most wanted to maintain a unified Judiciary—which the Convention proposed to modify in radical ways.[23] Social rights were probably the items Chileans cared about the most. However, the Convention did not do well in those cases anyway. The Government agreed on a pact to clarify them so that the voters could not be scared, but some of the public opinion’s preferences were unclear or unrecognized. For example, 74% of the Chileans seemed to prefer their health insurance, and 75% chose to select the institution that was going to administer and invest their pension funds.[24] However, the constitutional draft was unclear about whether those things could be done, which opened the door for legal interpretations far from public opinion’s preferences. The Convention could have been clearer. It just decided not to.

Instead of offering a negotiated document that large parts of the society could perceive as accepting their competing preferences, the Convention offered a unilateral platform that left out part of the center-left and the rightwing coalitions. True, those coalitions did not have sufficient representation in the Convention—the rightwing Convention members did not have the required one-third to oppose proposed norms and did not even achieve the necessary minority to present procedural legal challenges to the Supreme Court. Though historically, the center-left and rightwing groups perform better in elections. In 2009 and 2017, Chileans elected a rightwing president, and the rightwing controlled half of the Senate after the 2021 parliamentary elections.

The under-representativeness of the rightwing and center-left coalitions made it unnecessary for the left to find compromises with the right or the center-left. The result was a politically unbalanced Convention that, even though relatively socially inclusive in an identarian sense (indigenous peoples had reserved seats, half of the Convention was composed of women, and independents and members from less privileged classes could make it to the Convention),[25] the Convention had a severe problem of political representation. As the constitution makers were not able to seek reelection (which is natural in this type of organ), they did not have incentives to appeal to the median voter. Also, the exit referendum arguably proved to be a relatively poor way of aligning the preferences of the constitution makers with the preferences of the citizens.

Independents did not represent the majority of the voters. Indeed, most were from the radical left. The reserved seats probably did seem to perform better than other independents. A few votes elected them, and a poll shows that the Convention’s proposed text differed from the priorities of self-identified indigenous citizens.[26] Notably, the Rejection vote was particularly high in territories with high indigenous populations. The problem was partly explained by the design of the electoral rules that allowed independent candidates to bring together electoral pacts without establishing or joining real parties, the way the reserved seats for indigenous peoples were elected, and partly due to how Chileans distrust the party system. Instead of offering the voters predictable candidates working to build coalitions that could appeal to large sectors of the society, Chileans were offered a set of anti-party candidates in the context of a party crisis. Instead of using the opportunity to attempt to fix the party system, designers decided to appeal to the unstable and hostile majorities that claimed more radical changes. Political representation did not work correctly, as the election of the Convention involved Chileans voting against the elites and not in favor of a specific set of platforms.


Constitutions are supposed to express fundamental pacts accepted by all relevant sectors of society. A divided constitution, such as the Pinochet Constitution and the Constitution proposed by the Chilean Convention, cannot help legitimize political institutions in the eyes of the public. Had the Approval vote won, it would likely have imposed a Constitution that large parts of the society would have resisted. More participation from the parties could have solved that problem by promoting compromises that could have led to a more widely accepted outcome. Not the ideal outcome for a specific group, but a more optimal outcome for all the groups. There are normative-instrumental reasons to defend the existence of multilateral constitution-making processes that are carried out by consensual mechanisms: judges and prosecutors tend to be designed in more independent ways,[27] constitutional courts tend to be designed in stronger forms,[28] and liberal institutions are more likely in these contexts.[29] Unfortunately, the Chilean process could not be defended in that way. The result was a tragic option for Chileans. Let’s hope that the new process the parties are negotiating will lead to more compromises and fewer impositions.

Scholars and policymakers interested in constitution-making should note the lessons of the two Chilean processes compared to the experiences of Ireland and Iceland. If a democratic system exists, constitution-making processes should include the parties. If the parties are in crisis, then discussing with them the way out of that crisis is necessary. Appealing to external actors such as the Irish conventions elected by lottery may be a good idea. Still, parties should remain involved to secure the existence of a plausible agenda that can be carried out by experienced politicians that are actually accountable. Without them, the process can fail or, even wors, produce an outcome far from what the public wants, feeding social and political division. 

Suggested citation: Sergio Verdugo, The Paradox of Constitution-Making in Democratic Settings. A Tradeoff between Party Renewal and Political Representation?, Int’l J. Const. L. Blog, Sept. 24, 2022, at:

[1] Jon Elster, “Forces and Mechanisms in the Constitution-Making Process,” Duke Law Journal 45 (1995): 394–95.

[2] See a contemporary example in Joel I. Colón-Ríos, Weak Constitutionalism. Democratic Legitimacy and the Question of Constituent Power (New York: Routledge, 2012).

[3] See, e.g, David Landau, “Constitution-Making Gone Wrong,” Alabama Law Review 64 (2013): 924–80; Sergio Verdugo, “Is It Time to Abandon the Theory of Constituent Power?,” International Journal of Constitutional Law, forthcoming. See:

[4] On the connection between constituent power and populism, see, e.g., Luigi Corrias, “Populism in a Constitutional Key: Constituent Power, Popular Sovereignty and Constitutional Identity,” European Constitutional Law Review 12, no. 1 (2016): 6–26; Oran Doyle, “Populist Constitutionalism and Constituent Power,” German Law Journal 20, no. 2 (2019): 161–80.

[5] Verdugo, Sergio: The Chilean Political Crisis and Constitutions as Magic Bullets: How to Replace the Chilean Constitution?, VerfBlog, 2019/11/04,, DOI: 10.17176/20191104-162816-0.

[6] See Rosalind Dixon, “Constitutional Rights as Bribes,” Connecticut Law Review 50, no. 3 (2018).

[7] I have made the argument elsewhere. See supra note 3.

[8] See a good summary of these ideas, common in the literature on political parties, in the recent work of Tarunabh Khaitan, “Political Parties in Constitutional Theory,” Current Legal Problems 00 (2020): 1–37,; Tarunabh Khaitan, “Balancing Accountability and Effectiveness: A Case for Moderated Parliamentarism,” Canadian Journal of Comparative and Contemporary Law 7 (2021): 81–153.

[9] The argument in favor of creating an extraordinary organ for constitutional change—such as a constituent assembly—instead of sitting legislators can be connected to the need for political renewal and avoiding the problems of self-interested politicians. The tradeoff consists of the constituent assembly members’ unknown motivations. See a review of the arguments in Gabriel L Negretto, “Democratic Constitution-Making Bodies: The Perils of a Partisan Convention,” International Journal of Constitutional Law 16, no. 1 (2018): 254–79.

[10] Also, see Katrín Oddsdóttir, “Iceland: The Birth of the World’s First Crowd-Sourced Constitution?,” Cambridge Journal of International and Comparative Law 3, no. 4 (2014): 1207–20; Hélène Landemore, “The Inclusive Constitution-Making: The Icelandic Experiment,” The Journal of Political Philosophy 23, no. 2 (2015): 166–91.Tom Ginsburg, Iceland: End of the Constitutional Saga, I-Connect Blog, April 5, 2013, available at

[11] The literature typically focuses on the successful abortion and same-sex marriage referendums, but many recommendations are still pending. It is plausible to suggest that the problem connects to the interests of sitting legislators. See a discussion of the causes in Oran Doyle and Rachael Walsh, “Deliberation in Constitutional Amendment: Reappraising Ireland’s Deliberative Mini-Publics,” European Constitutional Law Review 16, no. 3 (September 2020): 440–65,

[12] Sergio Verdugo and Jorge Contesse, “The Rise and Fall of a Constitutional Moment: Lessons from the Chilean Experiment and the Failure of Bachelet’s Project,” Int’l J. Const. L. Blog (blog), 2018,

[13] The Chileans distrust of the representative institutions is not new, of course. See, e.g., Juan Pablo Luna and David Altman, “Uprooted but Stable: Chilean Parties and the Concept of Party System Institutionalization,” Latin American Politics and Society 53, no. 2 (2011): 1–28; Juan Pablo Luna, “Chile’s Crisis of Representation,” Journal of Democracy 27, no. 3 (2016): 129–38.

[14] For the reasons each party had in opening the door for a constitution-making process, see María Cristina Escudero, “Making a Constituent Assembly in Chile: The Shifting Costs of Opposing Change,” Bulletin of Latin American Research, 2021.

[15]  See a version of this argument in Benjamin Alemparte, The Institutional Interest of Political Parties in Chile’s Constitution-Making Process, Int’l J. Const. L. Blog, Nov. 17, 2020, at:

[16] There were 79 electoral pacts, and only five of those included political parties. See Jorge Fábrega, “Ordenamiento Ideológico En La Convención Constitucional Chilena,” Revista de Ciencia Política 2021, no. ahead of print (2022),×2022005000106.

[17] See, e.g., Roberto Gargarella, «El “plebiscito de salida” como error constituyente», IACL-AIDC Blog (6 de Septiembre de 2022). Disponible en:

[18] Sergio Verdugo, “El Poder Constituyente Impopular,” Actualidad Jurídica 46 (2022). See

[19] A pollster showed that 74% of the Chileans prefer to recognize Chile as a multicultural state, and not as a plurinational state. See CADEM (2022b): La opinión pública frente a las propuestas constitucionales.

[20] They instead prefer to keep a decentralized but unitary state. Only 24% agreed on the regional state. See supra note 19.

[21] 53% wanted to keep the Senate and only 35% were in favor of replacing the Senate with the Territorial Chamber that the Convention finally ended up approving. See supra note 19.

[22] The Convention offered to change a clear rule on what is supposed to be compensated by a general and rather vague (for the Chilean legal doctrine) principle on “just compensation” that was supposed to be interpreted in the future by legislative majorities. However, 63% of Chileans seemed to favor the rules they knew and only 25% agreed on deferring the way of paying the compensation to future legislators. See supra note 19.

[23] Only 11% favored the idea of establishing parallel judicial systems as part of the legal pluralism advocated by the indigenous demands that were advocated by the Convention. See supra note 19.

[24]See supra note 19.

[25] The Convention’s composition looked more like the Chilean population than the Congress. See Carmen Le Foulon and Valeria Palanza, “Elecciones a La Convención Constituyente: Innovación y Renovación,” Puntos de Referencia – Centro de Estudios Públicos 580 (2021).

[26] See, e.g., the poll by Centro de Estudios Públicos 87, February-July of 2022:

[27] Julio Ríos-Figueroa and Andrea Pozas-Loyo, “Enacting Constitutionalism: The Origins of Independent Judicial Institutions in Latin America,” Comparative Politics 42, no. 3 (2010): 293–311.

[28] Tom Ginsburg, Judicial Review in New Democracies. Constitutional Courts in Asian Cases (Cambridge University Press, 2003).

[29] Gabriel Negretto, “Constitution-Making and Liberal Democracy: The Role of Citizens and Representative Elites,” International Journal of Constitutional Law 18, no. 1 (2020): 206–32; Gabriel Negretto and Mariano Sánchez-Talanquer, “Constitutional Origins and Liberal Democracy: A Global Analysis, 1900–2015,” American Political Science Review 115, no. 2 (2021).

Print Friendly, PDF & Email
Published on September 24, 2022
Author:          Filed under: Developments

Leave a Reply

Your email address will not be published. Required fields are marked *