Chile’s constitutional processes (2020 to 2023)

Chile’s constitutional processes were a sequence of national votes on replacing the 1980 constitution: on October 25, 2020, 78.28% of valid votes approved writing a new constitution and 79.00% chose a fully elected Constitutional Convention; on September 4, 2022, 61.86% rejected the Convention’s draft; and on December 17, 2023, 55.76% voted against a second draft written through a different, more party-controlled process. The 2022 vote was compulsory and drew about 13 million voters, against about 7.6 million in the voluntary 2020 vote. The existing constitution, as amended, continues in force.

After large protests in 2019, Chile agreed to write a new constitution. In 2020, most voters said yes. A group of 155 elected delegates, with equal numbers of women and men and seats set aside for Indigenous peoples, wrote a draft. In 2022, with voting required by law, most voters rejected it. A second process, run more closely by political parties and experts, wrote another draft. In 2023 most voters rejected that one too. Chile still uses its 1980 constitution, which has been amended.

Chile’s case is contested on every side. This page attributes each interpretation to those who hold it. For other constitutional votes, see the Scottish independence referendum and the Iceland crowdsourced constitution.

Defined terms

Chile’s constitutional processes involve several specific terms.

• Plebiscito de entrada (entry plebiscite): the October 2020 vote asking whether to write a new constitution and what kind of body should write it.⁠^1,4
• Convención Constitucional (Constitutional Convention): the fully elected body of 155 members that drafted the 2022 text.⁠⁴
• Convención Mixta (Mixed Constitutional Convention): the alternative offered in 2020, combining elected citizens and sitting legislators; voters rejected it.⁠^1,4
• Plebiscito de salida (exit plebiscite): a vote approving or rejecting a finished draft.⁠^2,4
• Two-thirds quorum: the rule that the Convention approve its norms and voting rules by two thirds of members in office.⁠⁴
• Gender parity: rules requiring balanced representation of women and men among elected members.⁠^5,6
• Bases institucionales (institutional bases): twelve principles agreed by parties in Congress that bounded the 2023 draft and were enforced by a Technical Admissibility Committee.⁠^8,9

What happened

Chile’s constitutional processes unfolded over four years.

• 2019. Large-scale protests broke out, and protesters “demanded a new constitution because policy changes were seen as unable to fully address their concerns,” as Al Jazeera summarized the background.⁠⁷
• October 25, 2020. Entry plebiscite. “Do you want a New Constitution?”: Approve 5,892,832 (78.28%), Reject 1,635,164 (21.72%). “What kind of body should write the New Constitution?”: Constitutional Convention 79.00%, Mixed Constitutional Convention 21.00%.⁠¹
• May 2021. Election of the Convention: 155 members, gender parity, 17 seats reserved for Indigenous representatives chosen only by Indigenous voters (seven for the Mapuche), and at least 5% of each list reserved for people with disabilities, from 1,373 candidates.⁠⁵
• 2021 to 2022. The Convention drafted under a two-thirds rule and a nine-month term extendable once by three months.⁠⁴
• September 4, 2022. Exit plebiscite with compulsory voting: Reject 7,882,958 (61.86%), Approve 4,860,093 (38.14%).⁠²
• January 11, 2023. The Chamber of Deputies approved, 109 to 37 with two abstentions, a bill launching a second process, the Senate having already approved it.⁠⁸
• March 6, 2023. A 24-member Expert Commission, half elected by each chamber, and a 14-member Technical Admissibility Committee began work.⁠^8,9
• May 7, 2023. Election of a Constitutional Council under gender parity, with Indigenous seats added according to their share of the vote.⁠⁸
• December 17, 2023. Second plebiscite: In favor 5,470,025 (44.24%), Against 6,894,287 (55.76%).⁠³

How it worked

Chile’s two processes used sharply different designs, which is why they are often compared.

The first process: an elected convention. The constitutional procedure in force asked two entry questions, set the Convention at 155 citizens “elected especially for these purposes,” and barred members from running for elected office while serving.⁠⁴ It required the Convention to approve “the rules and its voting regulations by a quorum of two-thirds of its members in office” and prohibited it from altering that quorum.⁠⁴ The draft had to respect “the character of the Republic of the State of Chile, its democratic regime,” final judicial rulings, and ratified international treaties in force.⁠⁴ The exit plebiscite asked, “Do you approve of the text of the New Constitution proposed by the Constitutional Convention?”; “Suffrage in this plebiscite will be mandatory” for those with an electoral domicile in Chile, with fines for not voting, and if the text were rejected “the present Constitution will continue in force”.⁠⁴

Representation rules. Gender parity, 17 reserved Indigenous seats and space for lists of independents were designed to widen who sat in the Convention.⁠^5,6 Pre-election coverage noted that many candidates “do not respond to any political party, but are not necessarily independent either”.⁠⁵

The second process: experts first, then an elected council. The 2023 law created an Expert Commission of 24 members elected by Congress to prepare a preliminary draft in three months, an elected Constitutional Council to write the final text, a mixed commission to resolve conflicts between them, and a Technical Admissibility Committee of 14 professionals to enforce twelve institutional bases from a cross-party “Agreement for Chile”.⁠^8,9 ConstitutionNet reported the Council’s elected size as 55 in January 2023 and 50 in March 2023;⁠^8,9 the exact count, including added Indigenous seats, was not verified for this page. A 2024 study notes that the second process kept gender parity but dropped reserved Indigenous seats and mechanisms for independent candidates.⁠⁶

Evidence: measured results

Chile’s three plebiscites, as compiled by IFES from official returns, produced these results:⁠^1,2,3

The participation ratios are calculated here from the IFES figures.⁠^1,2,3 The 2022 plebiscite had compulsory voting.⁠^2,4 In 2023, invalid ballots (blank and null) numbered 650,651, more than double the 279,012 in 2022.⁠^2,3

Composition. The Convention elected in 2021 had parity built into its design.⁠⁵ Catherine Reyes-Housholder, Julieta Suárez-Cao and Javiera Arce-Riffo argue in PS: Political Science & Politics (2024) that legislators kept parity in the second process as a low-cost way to make an elite-controlled process appear legitimate, applying it within party lists so it did not threaten party leaders, and warn of “gender-washing” in weak parity norms in the expert draft.⁠⁶

Failures and criticisms

Chile’s processes have been criticized from every direction, and the criticisms are attributed here.

Of the 2022 draft, from critics. Eduardo Alemán and Patricio Navia, in the Journal of Democracy (2023), argue that “the convention’s proposal was radical and out of sync with most Chileans,” that inexperienced members made errors and did not correct course, and that the declining popularity of President Gabriel Boric, its leading backer, hurt it.⁠¹⁰ Hartmut Rank of the Konrad Adenauer Foundation argued that the draft read “more like a left-green-secular government programme than like a constitution,” criticized its “388 articles and 57 transitional clauses,” and objected to provisions on the Constitutional Court, the judiciary and Indigenous consent.⁠¹¹ Legal scholar Roberto Gargarella, as summarized by one commentator, argued that delegates ended up “acting and deciding too far from the society they wanted to speak to”.⁠¹²

Of the 2022 rejection, from supporters of the draft. Supporters described it as including strong environmental protections, recognition of Indigenous peoples’ rights, gender diversity in public institutions, a national health system and labor protections, and some called it among the most progressive constitutional texts written.⁠¹² Some argue that a single yes or no vote forced voters who objected to one provision to reject the whole text, a criticism of referendum format that Gargarella also makes.⁠¹² Others point out that the electorate changed between 2020 and 2022: about 7.6 million voted when voting was voluntary and about 13 million when it was compulsory,⁠^1,2 so the two votes did not measure the same population.

Of the 2023 draft. Al Jazeera reported that the second draft was written by a body dominated by the conservative Republican Party, that it would have reinforced property rights and free-market principles and set limits on immigration and abortion, and that critics said it curtailed women’s and LGBTQ rights; Indigenous representatives also rejected it.⁠⁷ After the vote, President Boric said the result showed the country had become polarized and “The country needs everyone”.⁠⁷ ConstitutionNet had reported in March 2023 that the second proposal was “expected to be more moderate than the first one”.⁠⁹

Of the second process’s design. Reyes-Housholder, Suárez-Cao and Arce-Riffo describe the second process as “elite-controlled” in its design, with party-list parity and no reserved Indigenous seats.⁠⁶ Others expected the second design to yield a more moderate text; ConstitutionNet reported that expectation in March 2023.⁠⁹

Competing interpretations

Analysts read Chile’s sequence in several incompatible ways.

The drafts were out of step with voters. On this reading, voters wanted a new constitution but not either of the specific texts: the first was seen as too far left and the second as too far right. Alemán and Navia’s argument about the first draft.⁠¹⁰ and Al Jazeera’s reporting on the second⁠⁷ are consistent with this view.

The electorate changed. On this reading, the 2020 vote reflected a smaller, more mobilized voluntary electorate, and compulsory voting in 2022 and 2023 brought in millions of voters with different views⁠^1,2,3 This explains part of the swing, but not why the 2023 draft also failed.

Process design shapes legitimacy. On this reading, the first process gave independents and newcomers control without the coalition-building experience of parties, and the second gave parties control without broad public trust. Reyes-Housholder and colleagues’ analysis of what was kept and dropped.⁠⁶ is relevant here.

Fatigue and polarization. On this reading, by 2023 many voters were tired of the question; analyst Jennifer Piscopo, quoted by Al Jazeera, said voter fatigue made a third attempt uncertain⁠⁷

What transferred and what did not

Chile’s processes are closely watched by constitution-makers elsewhere.

Transferred within Chile. Gender parity carried over from the first process to the second.⁠⁶

Dropped within Chile. Reserved Indigenous seats and mechanisms for independents did not carry over.⁠⁶

Transfer abroad. Chile’s parity design is discussed internationally, but whether other countries adopted specific features from it was not established by sources verified for this page.

Application to this framework

Chile’s processes inform several principles of My Vote Forward.

• Agreement on direction is not agreement on a text. A 78% vote to write a new constitution.⁠¹ did not predict support for either draft. The framework separates priorities from opinions on specific proposals for this reason.
• The population that participates can change the result. The shift from voluntary to compulsory voting⁠^1,2 is a denominator lesson; see measurement.
• All-or-nothing ballots hide partial agreement. Gargarella’s critique⁠¹² parallels the framework’s interest in recording support for components, not just packages; see collective priorities.
• Who drafts matters for legitimacy. Elected newcomers and party-selected experts each lost public trust in different ways. See governance and representation.
• Limits in advance. Both processes set boundaries before drafting: Article 135’s limits⁠⁴ and the twelve bases⁠⁸ The framework favors stating constraints before participation; see proposals.

Unresolved questions

• How much of the swing from 2020 to 2022 came from changed views and how much from a changed electorate under compulsory voting? Individual-level research was not verified for this page.
• Would a provision-by-provision ratification design have produced a different outcome, and would it have produced a coherent constitution?
• What was the exact composition of the 2023 Constitutional Council, including added Indigenous seats? Reports differed.
• Will Chile attempt a third process, and if so, what design would command broad trust?
• Did parity in the Convention change the content of the 2022 draft, and did it affect voters’ judgment of it?