Introduction.

The Irish Constitution (Bunreacht na hÉireann) is an old constitution. Drafted and approved by the Irish people in July 1937, it came into effect in December 1937. It is the longest serving republican constitution in Europe.1 When adopted in the first half of the previous century, it asserted Ireland’s status as an independent and sovereign nation. Novel provisions changed the nature of the state including the fact this constitution can only be amended by the people. Ireland is thus part of this small group of countries, such as Australia, Austria and Switzerland, for instance,2 where a referendum is required for constitutional change.

Even though referendums represent “a very difficult and controversial democratic process”,3 many have been held to amend the Irish Constitution throughout its nearly 90-year of existence. Up to March 2026, 43 constitutional referendums have been held—which is almost one referendum every two years since 1937—of which 31 resulted in amendments. This represents a 72 % of “success” rate. In the period from 1972 to 2025 inclusive, there were more votes cast at referendums (approx. 55 million votes cast in 43 referendums) than at general elections (approx. 27.6 million votes cast in 15 general elections). For those positive referendums, these have resulted in constitutional amendments as regards many topics. Referendums on, mostly European, treaties and agreements, and referendums on the domestic political process top the numbers.4 Such variety and frequency of constitutional changes may suggest two contradictory ideas: firstly, that the Irish Constitution is no longer fit for purpose and that it should, perhaps, be completely redrafted; secondly, and on the contrary, that the Irish Constitution can keep being modified (i.e., some provisions revised or redrafted) to overcome, without a complete overhaul, the challenges of changing times under current processes. The second position might be the more plausible one especially since the Irish experience of constitution-making has been dynamised, in the last few years, by recourse to deliberative democracy processes.5 Half a dozen of citizens’ assemblies has been held in the last decade or so, from the 2012-2014 Convention on the Constitution6 to the 2023 Citizens’ Assembly on Drug Use. These experiences have resulted in major constitutional changes such as the legalisation of same-sex marriage following the 2012-2014 Convention on the Constitution, and the legalisation of abortion following the 2016-2018 Citizens’ Assembly.7 Citizen participation, achieved through the coupling of deliberative democracy (deliberative mini publics) and direct democracy (referendums), has indeed emerged as a defining feature of Ireland’s approach to constitutional drafting.

However, the Irish experience is less straightforward than it seems. It cannot be reduced to successful and steadily paced referendums informed by the deliberative input from citizens’ assemblies. Besides, there are other ways to amend the Irish Constitution, such as through advocacy and strategic litigation. While certainly paradigmatic, the Irish approach of combining referendums with citizens’ assemblies has also exposed notable limitations. The decisive rejection of the double referendum held on 8 March 2024 illustrates this point.8 The Irish voters not only opposed proposed constitutional changes—relating to the definition of the family and references to the woman’s role in the home—but also effectively dismissed the recommendations of the 2019-2021 Citizens’ Assembly on Gender Equality, which had informed the proposals put to the electorate, perhaps signalling the end of a successful coupling between the two processes.

The aim of this article is to revisit key aspects of the Irish experience by highlighting the limitations of referendums and citizen’s assemblies. From this emerges a more nuanced account of popular participation in constitution-making. The article is divided into four parts. The first provides an overview of alternative processes for amending the Irish Constitution beyond referendums and citizens’ assemblies. The second and third parts examine, respectively, referendums and deliberative mini-publics, focusing on their limitations. The final part ends on a more positive note by outlining a possible scenario for constitutional amendment under the current process, before concluding.

1. Preliminary considerations: Alternative constitutional amendment means

Other means of law reform exist in Ireland, primarily relating to legislative changes, though they may also involve amendments to the Constitution. Firstly, exploratory policy documents such as green and white papers,9 as well as reports and consultation papers from the Law Reform Commission allow for public consultation in different forms and at various stages.10 For instance, the 2013 Green Paper and 2015 White Paper on Defence initiated and shaped, respectively, policy in a major area where no policy review had been made in a long time.11 The former started a broad consultative process that, in turn, informed the latter which was also shaped by “an extensive consultative process” with input from the public in the form of written submissions, follow-on meetings with selected authors of those submissions and the setting up of various expert groups to refine the advice given to the Department of Defence. The 2015 White Paper set out a long-term framework for defence policy to around 2025 and continues to underpin subsequent reviews and updates.12 Law Reform Commissions are typically found in common law countries. The Irish Law Reform Commission13 aims to keep the law under review, that is to change, codify or consolidate it, as appropriate. Public consultation, allowed at various points in the reform process, from the first stage in the discussion to after the publication of the report by the Commission, is indeed “central to the work of the [Commission] (…) being neither an example of direct democracy nor deliberative democracy but [aiming to] capture public sentiment on the issues that the Law Reform Commission considers”.14 Despite allowing for some level of democratic participation in law reform, these mechanisms do not primary concern constitutional change.

Secondly, court mechanisms, especially through judicial review claims, represent another mechanism of law reform. These directly interest constitutional changes when challenging the validity of legislation, acts of the executive, or administrative decisions that applicants believe infringe upon their constitutional rights. In that (limited) sense, constitutional review can operate as a vital mechanism of democratic empowerment, albeit one which is outside of a constitutional reform process per se. Most constitutional challenges have arisen within ordinary litigation under Article 34 of the Constitution,15 resulting in hundreds of notable cases that have had an impact on constitutional provisions and have shaped individual rights, including cases related to privacy, equality, and freedom of expression.16 On example is the Norris litigation when homosexuality was decriminalised in Ireland. The Norris case contributed to a change in public opinion regarding LGBT people and, years later, assisted in another case—the Zappone case where the High Court refused to recognise a same-sex marriage that was contracted in Canada—to pave the way for the overwhelming support in the referendum on marriage equality in May 2015.17

These formal means would not work properly if it was not for a strong culture of advocacy and activism in Ireland. Such activism and advocacy take different forms including “signing petitions, attending public meetings, letter writing, protect and formal lobbying of politicians in an attempt to change the law [including constitutional law]”.18 Within this, strategic litigation plays a significant part. Citizens can use judicial claims to bring carefully selected cases to court to provoke a change by raising awareness, in the media and in the population, most of the time in the legislation but also possibly a change in the interpretation of the Constitution.

These alternative, or complementary, means of law reform, as reflections of a particular political culture, should not be overlooked when considering citizen’s participation in the context of constitutional reform process in referendums and within citizens assemblies.

2. The referendum: The indispensable half in the couple

Citizens participate directly in constitutional changes through referendums that give them, as a method of direct democracy, a voice by “aggregating individual expressions of opinion”.19 The singularity of the Irish case is that it makes such citizens’ involvement indispensable since, under Article 46 of the Constitution, any constitutional amendment requires the approval of the people in a referendum.20 This sets Ireland apart from other jurisdictions where the referendum is an option when it comes to amending the constitution (e.g., France under Article 89 of the Constitution, Chile under Articles 128 and 129 of the Constitution)21 or where no referendum is required to amend the constitution (e.g., Iceland under Article 79 of the Constitution).22 This section explores the central role of referendums in Ireland’s constitutional system, outlining key facts and figures, their legal framework and limitations as regards citizens’ involvement.

2.1. Facts and figures about Irish referendums

Not going as far as saying that Ireland is “governed by referendum”, it is right to say that such instrument has acquired a major role “as a means of making political decisions”.23 As mentioned in the introduction, this is not only their frequency but also the range of areas in which referendums have been proposed and passed over the years that give the extent of the practice. Since the accession of Ireland to the then European Communities, a total of ten referendums were approved concerning European Union treaties, albeit not without difficulties for the 2001 Treaty of Nice (October 2002, Twenty-Sixth Amendment) and the 2007 Treaty of Lisbon (October 2009, Twenty-Eighth Amendment).24

Aside from EU treaties, a significant referendum was held in May 1998 on the British-Irish Agreement (the Good Friday Agreement). Concerned with removing the Republic’s territorial claim over Northern Ireland, it was approved by an overwhelming majority of the Irish electorate.25 Referendums about the confidentiality of government meetings (October 1997, Seventeenth Amendment), the abolition of the Seanad Éireann (October 2013, Thirty-Second Amendment, rejected), the age of the presidential candidate (May 2015, Thirty-Fifth Amendment, rejected), as well as several referendums about voting, especially about the proportional representation by the single transferable vote (or PR-STV) system, constitute the bulk of the more than ten referendums concerning the Irish political process. Another string of referendums has concerned court-related matters such as the referendum on the establishment of the Court of Appeal (October 2013, Thirty-Third Amendment), and the referendum allowing the ratification of the Rome Statute establishing the International Criminal Court (June 2001, Twenty-Third Amendment). Referendums on moral and societal issues have certainly been the most emblematic ones, including the six referendums on abortion which was a divisive issue until its legalisation in the May 2018 referendum. The most common societal issue has been, however, about families, “their form, rights and duties within them, and their dissolution”.26 Other miscellaneous issues such as the removal of the special position of the Catholic Church (May 1972, Fifth Amendment), the abolition of the death penalty (June 2001, Twenty-First Amendment) and the removal of the offence of blasphemy (October 2018, Twenty-Seventh Amendment), complete the list.

2.2. The legal framework of the referendum process

Article 46 (the amending provision) and Article 47 (the referendum provision stricto sensu) of the Irish Constitution govern the referendum process. This is complemented by law, namely a series of statutes passed between 1994 and 2001 (Referendum Acts 1994–2001), and judicial decisions, that regulate how constitutional referendums are conducted. This body of legislation and cases set out rules in various matters, such as campaign funding, advertising, broadcasting, and the role of the state during referendum campaigns. Additionally, a series of recommendations about the organisation of referendums were made by the 2016-2018 Citizens’ Assembly. The Assembly proposed, by a significant majority (98 %), that government should regulate spending and establish spending limits in campaigns, as well as should ban anonymous donations (72 %) and fund both sides of a referendum campaign to an equal extent (68 %).27

A constitutional referendum must be initiated, approved, voted upon, and eventually acted upon—hence several stages in the process. Firstly, every proposal for amending the Irish Constitution is initiated as a bill setting out the wording of the proposed amendment introduced in the Dáil Eireann (the lower house of Parliament). Although the bill may originate from the government, or from a member of the Oireachtas as a private member’s bill, the executive’s control of the legislature, means that the government effectively initiates the referendum proposal. Secondly, both the Dáil Éireann and the Seanad Éireann (the upper house of Parliament) must pass the bill and agree to put the specific proposal to the people. The Houses of the Oireachtas (the Irish Parliament) do not vote on the substance of the proposal but only decide whether to put the matter to the referendum. Once approved by the Oireachtas, the government makes an order specifying the polling day for the referendum. Once the referendum campaign begins, the Electoral Commission (An Coimisiún Toghcháin)28 provides impartial and unbiased information on the referendum proposal. This takes the form of statements on the main issues involved, statements that are subsequently disseminated to the public, through television, radio and other forms of media, with a view to encourage people to vote. Thirdly, the bill—formally titled “An Act to amend the Constitution”29—must be put to the Irish people within 30 days and 90 days following its vote by Parliament. All Irish citizens of voting age, who are resident and registered to vote, are entitled to vote in the national referendum. If the outcome is a “No” vote, the proposal fails and the process ends.30 If a simple majority votes “Yes”, the constitutional amendment is approved. In practical terms, a provisional certificate is published in the State Gazette (Iris Oifigiúl). If no petition challenging the result is lodged with the High Court within seven days of such publication, the certificate becomes final. The President of Ireland may then sign the amendment into law “… upon his or her being satisfied that the provisions of [Article 46] have been complied with.” The Constitution is thereby duly amended.

2.3. Citizen participation and the limits of referendums

Referendums are widely recognised as democratic instruments that carry significant political risk.31 They often oversimplify complex constitutional issues into binary choices and tend to polarise debates. This comes from flaws observed at both ends of those involved in the process, from citizens to political actors. Voters often lack detailed knowledge of the issues at stake due to persistent informational deficits despite the presence of institutional supports such as the Referendum Commission; at the other end, such deficit allows political elites to shape referendum outcomes through strategic framing, wording, and timing. Referendums also carry the inherent risk of majoritarian dominance as they allow a simple majority to determine outcomes that may affect minority rights.

Still, referendums, as instruments of direct democracy, empower Irish citizens in constitutional reform processes. However, this is per se “a narrow moment” of democratic participation. The empowering factors can be found at multiple stages of the Irish referendum process. Firstly, the scope of the referendum is large since any constitutional matter can be amended. In a series of cases from the late 1980s,32 the Supreme Court affirmed the supremacy of popular sovereignty by referendum as a key principle of the Constitution. If there are, in theory, no limits to the people’s right to amend the Constitution by way of referendum, this means Ireland could legally abolish fundamental rights and judicial review, for example, although it would obviously be a no-go in political terms. In that sense, Ireland represents one of the most extreme examples of total constitutional amendability.33 The Irish case resembles the Icelandic case whose constitution does not contain any provisions that are explicitly unamendable. However, as opposed to Ireland, constitutional amendments are hard under Icelandic constitutional law because of a very demanding amendment procedure which is also the case for the constitutions of the Netherlands, Belgium and the Nordic Countries (i.e., Denmark, Sweden, Norway and Finland), for instance. The Irish case contrasts with other jurisdictions whose constitutions contain material limitations on constitutional amendments such as France (Article 89 concerning the republican form of government), Italy (Article 139 concerning the republican structure of government), Germany (Article 7 concerning the federal form of government, the inviolability of human dignity and the protection of human rights), and Portugal (Article 290 concerning the republican form of government, the separation of church and state, the protection rights and freedoms, and the independence of the courts). Further, any constitutional amendment, no matter how minor or technical, requires approval in a referendum under Irish law.

Secondly, the requirements for the referendum vote are quite open. No minimum turnout, or any other threshold, is required, and a simple majority vote (i.e., more than 50 % of the votes cast) is sufficient for the amendment to pass.34 Three referendums passed with a turnout lower than 35 %, namely the Sixth Amendment of the Constitution on orders made by the Adoption Board (July 1979, around 28.6 % turnout) and, on the same day, the Seventh Amendment on how certain university graduates elected members to Seanad Éireann (same), as well as the Twenty-First Amendment on the prohibition of the death penalty (June 2001, turnout at about 34 %).

Thirdly, although no judicial review of the contents of a bill to amend the Constitution is allowed,35 and an approved amendment approved by the voters is immune from challenge in the courts, there have been instances of such challenges at the initiative of citizens as constitutional litigants.36 These judicial challenges have essentially dealt with the issue of government funding and of information provided to the public during the referendum campaign. While the courts have understandably been reluctant to interfere with the democratic process,37 they have attempted to clarify what the government can and cannot do in the referendum context. Consequently, the government cannot be one-sided in a referendum campaign by publishing official information materials explaining the referendum in a non-neutral manner (i.e., advocating a “Yes” vote). This was held as unfair in a referendum.38 On that occasion, Chief Justice Hamilton put significant emphasis on the democratic process involved:

Once the Bill has been submitted for the decision of the People, the People were and are entitled to reach their decision in a free and democratic manner [our emphasis].

The use by the Government of public funds to fund a campaign designed to influence the voters in favour of a “Yes” vote is an interference with the democratic process and the constitutional process for the amendment of the Constitution and infringes the concept of equality which is fundamental to the democratic nature of the State.39

However, also recognising that neutrality is difficult to achieve in such a context, since no referendum can be put to the people without the political support of the government, the courts later acknowledged that members of government could support referendum proposals in their private capacity or as members of government or political parties provided it stopped short of direct public expenditure to support one side—in other words, it is ok to take side in the referendum campaign but not with public money.40

Ireland has been widely viewed as a positive model of citizen involvement in referendums for approximately forty years, a trend that has been notably reinforced by the recent use of deliberative mini-publics in constitutional reform. On many occasions, Irish citizens have been, through their binding vote, the ultimate decision-makers on constitutional matters. However, the extent of their direct involvement stops here. Even though public discourse, lobbying and deliberative mini publics can influence political agendas, citizens do not initiate the referendum process. Other jurisdictions have explored extending the right of initiative of the referendum to involve citizens more directly. In France, for instance, if the option of a referendum at the initiative of the citizens (référendum d’initiative citoyenne), including for constitutional matters, has gone nowhere,41 the possibility of a referendum called by a mix of parliamentarians (one-fifth of Parliament members) and citizens (one-tenth of the registered electorate) has been inscribed in the constitution since 2008 under the referendum of shared initiative (référendum d’initiative partagée) provision, albeit never used.42 Such provision is somehow irrelevant in the Irish context given that the government has to consult the people on constitutional changes, thus conferring an automaticity to the constitutional referendum. Apart from having no part in initiating it, Irish citizens do not draft referendum proposals either even though, again, the recommendations of citizens’ assemblies may influence their wording. The drafting process of the amendment bill43 is done by Attorney’s General Office, the legal adviser to the government. During this highly technical part of the referendum process, the Office44 ensures that the wording is legally sound and consistent with the rest of the constitution, that it is unambiguous so that the courts can interpret it clearly in future; and limited to what is necessary to achieve the proposed change. In other words, the focus is on whether the phrasing of the amendment bill can withstand constitutional litigation.45

However, what has raised concern about the empowerment of citizens in the constitutional drafting process is the apparent erosion of a “referendum culture” in Ireland.46 This interesting perspective taken by Kenny presupposes to look at the referendum through a more socio (cultural)-legal, rather than purely constitutional, lens. Such “referendum culture” deems to exist thanks to the generally enticing and flexible environment surrounding the referendum process in Ireland, articulated around four major components: firstly, the no nonsensical approach to referendums adopted by Irish people who can display “a healthy scepticism of government proposals and vague referendum questions”; secondly, their equally sensible “ability to separate primary preferences about constitutional change from secondary preferences about likely subsequent legislative outcomes”; thirdly, a rather good “control of misinformation despite a lack of formal rules”; supported by, fourthly, “a robust role for civil society and citizen engagement to help limit elite dominance of the [referendum] process”,47 hence the importance of highlighting the culture of activism and advocacy from the outset.

According to Kenny, such culture cannot be taken for granted. Because of its functional nature which is so dependent on the frequency of referendums (i.e., less referendum experience, less cultural solidification), this referendum culture is prone to disruption by various contingencies. This is different from saying that the referendum process is not perfect per se as indeed many referendums have had deep problems.48 What is meant here is that “[t]here is no guarantee that such a [referendum] culture will not be radically disrupted or undermined at any point.”49 And this point in time may have happened with the failed referendum of 2024. The double referendum, held on 8th March 2024, was about the expansion of the constitutional definition of the Family to include non-married families, on the one hand, and on the replacement of a long-controversial provision on the role of women and mothers within the home with a provision recognising the importance of care, on the other hand. Whatever the narrative was—either that, in this precise case, the proposals were badly drafted, badly argued and badly rejected as a rebuke of elite dominance or, more broadly, that the erosion of “a healthy referendum culture” was already under way50—Irish politicians will certainly think twice before using the referendum again. This is not necessarily disastrous per se. From an outsider’s perspective, it will probably change the perception—often tainted with a mix of admirative and intrigued sentiment—that other jurisdictions hold toward the Irish case; the decreasing use of referendums will arguably set aside Ireland as a modèle du genre. Internally, it will signify that Ireland has reached a new stage in its political practice of direct democracy, coming to the realisation that things cannot perdure unless there is a strong need for constitutional change.

A related problematic question would be about the connection between the referendum process and the use of deliberative mini publics. In that regard, would the erosion of the former consequently reshape the role and relevance of the latter, thus weakening the dynamics of the couple referendums-citizens’ assemblies in constitutional change?

3. Citizens’ assemblies: The new dynamic in the couple

Deliberative mini publics have proven effective in recent years in empowering citizens in constitutional reform, thus making citizens’ assemblies as the other dynamic half of this coupling in democratic practice. As an instrument of deliberative democracy, citizens’ assemblies bring together randomly selected citizens to deliberate on issues of major national importance. Through the adoption of recommendations, they propose constitutional reforms and other policy responses. Here again, the Irish case is singular. Ireland not only has extensive experience with citizens’ assemblies—having convened around half a dozen, from the 2012-2014 Convention on the Constitution to the 2023 Citizens’ Assembly on Drug Use—but this experience has also resulted in major constitutional changes, including the legalisation of same-sex marriage and of abortion.

However, although citizens’ participation in deliberative democracy has largely been successful, the process has not been without flaws and shortcomings.

3.1. The success of deliberative democracy in Ireland

The first two citizens’ assemblies, namely the 2012-2014 Convention on the Constitution and the 2016-2018 Citizens’ Assembly, were undoubtedly the most influential ones. They played a pivotal role in shaping the debates that led to the Marriage Equality referendum legalising same-sex marriage in 2015, and the Repeal of the Eighth Amendment referendum legalising abortion in 2018. In this context, citizens’ empowerment operated on three levels: direct participation, policy influence, and enhanced democratic legitimacy. Through sustained engagement with sensitive issues, ordinary citizens contributed, through their recommendations, not only to shaping policy but also to the process of constitutional drafting. This participatory dynamic conferred a strong popular mandate on the resulting reforms, grounding them in informed public opinion.

Instead of rehashing what has been presented elsewhere about these two assemblies,51 a few pointers are given about the overall experience. Firstly, the experience of deliberative democracy in Ireland points towards the ambitious perspective that was taken from the start. Building on a limited previous informal experience that took place in a particularly difficult context (i.e., 2011 We the Citizens Assembly, in the aftermath of the economic crisis),52 deliberative democracy was envisaged at national—and not local or regional—level, and for a multiplicity of topics. The first two assemblies were given a fairly broad mandate. The 2012-2014 Convention on the Constitution was entrusted with no fewer than eight topics, all with constitutional implications. These issues included reducing the presidential term, lowering the voting age, reforming the electoral system of the Dáil Éireann, granting voting rights to Irish expatriates, removing the constitutional provision on blasphemy, revising constitutional provisions regarding women’s roles in the home and their participation in public life, and, of course, introducing same-sex marriage. The 2016 Citizens’ Assembly had participants debate five topics, not all however with constitutional implications. These concerned institutional aspects (such as the length of parliamentary terms and the referendum procedure), a number of challenges in specific areas (such as the challenges related to an aging population and those related to climate change), and, of course, the amendment of the Eighth Amendment to the Constitution on abortion.

Secondly, various formats and working rules were tested, choices and improvements were made, without, so it seemed, too much apprehension towards trial and error. For instance, while there was initially a sense of bias due to the presence of political members within the 2012-2014 Convention on the Constitution, it eventually seemed that the “mixed” composition did not allow politicians to exert undue influence on ordinary citizens. The mixed format was not repeated though, and the government subsequently opted for assemblies composed exclusively of ordinary citizens. In terms of improvements, the sortition process was refined, whether it was to change the company hired for that purpose in order to use a more elaborate process (starting from the 2019-2021 Citizens’ Assembly on Gender Equality), or to open the selection criteria to ensure better representation (e.g., the inclusion of any adult citizen residing in Ireland from the 2022-2023 Citizens’ Assembly on Biodiversity Loss). Similarly, governance structures and rules were fixed resulting in a now classic model of assembly, comprising a steering committee, an expert advisory group, and a secretariat assisting the chairperson. Apart from the ongoing consolidation of working rules,53 improvements have included, for example, from the 2019-2021 Citizens’ Assembly on Gender Equality, the involvement of independent researchers in charge of compiling and analysing the contributions received from the public and, generally, examining the effectiveness of the deliberative function. The consolidation of these rules and structures, together with a lengthy experience, have been undoubtedly an element of success, positioning Ireland as a specialist of coherently designed and efficient deliberative instruments.

Thirdly, and most of all, the success is measured by output, that is by the follow-up of the recommendations made by the citizens assemblies.54 As said above, the 2012-2014 Convention on the Constitution and the 2016-2018 Citizens’ Assembly were emblematic in that regard. The former recommended amending the Constitution to allow marriage equality which 62.1 % of voters approved in the May 2015 referendum, making Ireland the first country to legalise same-sex marriage through a popular vote. The latter recommended the repeal of the Eighth Amendment of the Constitution—that recognised the equal right to life of the pregnant woman and the unborn—which was approved by 66.4 % of voters in the May 2018 referendum, which subsequently led to the enactment of legislation permitting abortion services in Ireland. Both assemblies have enshrined the deliberative practice in the Irish political landscape by consolidating a follow-up process that has been roughly the same since. Under this process, the assembly’s report, which contains its recommendations, is submitted to the Oireachtas. The Oireachtas may then convene a debate and, where appropriate, vote to establish a joint parliamentary committee to consider whether, and how, to act on those recommendations. The committee then submits its own report to the Oireachtas, which transmits it to the government. The government, in turn, determines the appropriate moment to follow suit on the committee’s recommendations. A key aspect is the way the deliberative mini public has allowed some original ways to play out in political practice. The 2016-2018 Citizens’ Assembly is emblematic in that regard having established a sort of democratic continuum between the assembly’s work and that of the national representation during the vote on abortion legislation. Such continuum was manifested by the recommendation proposing a 12-week period to allow abortion, proposal which eventually made its way to the Oireachtas, the Joint Committee on the Eighth Amendment to the Constitution and the government in the final draft bill. In this instance, the Citizens’ Assembly clearly had a decisive influence on subsequent reform by introducing the 12-week period—an issue that the Oireachtas had not previously anticipated—into the political agenda. Scholars have argued that, through deliberate institutional design (“designed coupling”),55 the 2016-2018 Assembly had a substantial impact on the outcome eventually adopted in Irish law. In other words, it was possible to identify a traceable connection between citizen input and legislative output in relation to the Eighth Amendment (abortion) referendum because the Citizens’ Assembly’s recommendations were taken up by the Joint Oireachtas Committee, which helped draft the proposals that were later put to the people.56

The solid foundation established by this initial experience paved the way for further experiments in deliberative democracy. Subsequent citizens’ assemblies were thematic rather than generalist, having had to address one issue, or one area, only. Such change in approach was probably due to the difficulty of mobilising a group of citizens over a long period of time and on a variety of topics. At the same time, constitutional reform has become less prominent, as none of these later assemblies have resulted in a successful referendum. This was particularly evident in the failure of the 2024 double referendum, despite being informed by the noteworthy recommendations of the 2019-2021 Citizens’ Assembly on Gender Equality, including the proposed removal of the reference to “housewives” from the Constitution. Nevertheless, these assemblies have continued to advance deliberative democratic practices and have contributed to significant legislative developments beyond constitutional change, particularly in areas such as gender equality and climate action. However, these achievements have not eliminated persistent criticisms of citizens’ assemblies.

3.2. The problem with citizens’ assemblies

Citizens’ assemblies face persistent criticism, particularly concerning their representativeness, effectiveness, and overall legitimacy. Various problems can be identified in that respect.

The first problem is that, although part and parcel of recent constitutional reform processes, the experience of deliberative democracy in Ireland has fallen short of a “legal systematisation”.57 Legal issues about the decision to establish citizens’ assemblies, about their nature as public bodies, and about the nature of their rules of governance and of the acts they adopt, have not been clarified.58 Citizens’ assemblies have emerged outside any formal institutional framework and, so far, none has been established on a statutory or legal basis.59 It is essentially a government decision, albeit with parliamentary endorsement. Each House of the Oireachtas is invited to adopt a resolution which approves the “calling”, or “establishment”, of such a body and those matters already decided upon by the government. What the government decision does is that it (only) sets out the responsibility of the citizens’ assembly to report to the Houses of the Oireachtas and the ensuing government’s obligation to respond to that report.

As for their legal status, citizens’ assemblies are regarded as “emanations of the executive”. It remains unclear, however, if the parliamentary resolution (approval) is there as merely an additional safeguard against the abuse of the power to establish a citizens’ assembly, or whether such assembly is a collaborative venture by the two branches of power, executive and legislative. One of the recommendations of the 2016-2018 Citizens’ Assembly supported introducing citizens’ initiatives that would allow citizens to bring questions either directly to the Oireachtas or to a referendum if a defined level of public support was achieved, thus strengthening direct democratic participation. This was not taken further. One can assume that this does not really matter as long the process is regularly used by the government and works well. Likewise, and as explained above, Irish citizens’ assemblies have managed their own internal governance by incrementally refining a number of operational rules. However, this has been achieved without any legal basis as such with, perhaps, the exception of the 2019-2021 Citizens’ Assembly on Gender Equality that published its rules of governance.60

Yet, the thorniest problem has been the effect attached to the recommendations. A total aggregation of more than 350 recommendations has been adopted by citizens’ assemblies over the years.61 Whether implementation has taken the form of constitutional amendment or the adoption of governmental policies, plans or other actions, approximately 150 to 160 of the 355 recommendations (around 42-45 %) have been implemented or have made significant progress to date. Only seven or eight recommendations have led to a formal constitutional change though, thus giving a realistic picture of the impact of citizens’ assemblies in the constitutional reform process.

Overall, the issues of their ad hoc form, of their loose procedural framework, and of the uncertainty attached to their output make citizens’ assemblies an uncertain instrument in terms of constitutional drafting. Should they gain further momentum as one of the chosen devices of deliberative democracy, there needs to be some degree of institutionalisation. This does not seem to have been the case despite the extensive experience. This is further compounded by another difficulty that relates to their articulation with the constitutional referendum process. As above explained, citizens’ assemblies are arguably regarded as “a trusted forum of considered deliberation that voters felt they [can] heed”,62 thus reverberating on the referendum process and in the information that people receive during the process. Yet, the executive has control over the whole process—from the establishment of a citizens’ assembly to the initiation of the referendum and the linkage between the two (i.e., which recommendations to take into account and in what form).63 The ensuing concern is whether the suspected fading of referendums in the aftermath of the 2024 failures will also mean a certain irrelevance of the deliberative process in the context of constitutional change. In conclusion, while citizen assemblies have proven effective in certain contexts, these experiences underscore the importance of clear communication, public engagement, and political will in translating deliberative processes into successful constitutional reforms.

Conclusion.

This article has shown that citizen empowerment in Irish constitutional reform does not reside in a single mechanism, but rather in a virtuous interaction between referendums and deliberative mini-publics, complemented by other means. Each avenue enables participation at a different stage and in a different form: referendums provide the people with ultimate decision-making authority, although with no initiative part while citizens’ assemblies offer spaces for informed deliberation and agenda-setting, although with a limited scope. Taken together, these mechanisms form a distinctive, if imperfect, system of constitutional participation.

Yet, the coupling of direct and deliberative democracy might not be so enduring. Both referendums and citizen’s assemblies reveal structural tensions. Referendums, while constitutionally central, are episodic and vulnerable to political miscalculation, public fatigue, and poorly communicated proposals, as illustrated by the 2024 double referendum. Citizens’ assemblies, although widely praised for their deliberative quality and past successes, remain politically contingent, legally under-institutionalised, and dependent on executive follow-through. Neither of these two mechanisms, taken alone, can guarantee meaningful and sustained citizen influence over constitutional change. The Irish experience therefore suggests that citizen empowerment in constitutional reform is best understood as a dynamic process. Its effectiveness depends not only on institutional design, but also on political culture, trust, clarity of communication, and the willingness of political actors to respect and integrate citizen input which arguably remains the main bone of contention. The recent setbacks in the referendum process do not negate Ireland’s achievements, but they do underscore the fragility of participatory practices. Strengthening the coherence between deliberation, decision, and enforcement—while preserving democratic legitimacy—will be essential if Ireland is to maintain its distinctive model of citizen involvement in constitutional change. This is all the more relevant that recent discussions around citizen empowerment continue to evolve, notably around extending voting rights in presidential elections for the Irish diaspora (i.e., Irish citizens residing abroad)64 or extending the judiciary’s role in scrutinising executive actions, reinforcing the principle that governmental decisions, especially those with significant public impact (in that case, an alleged secret defence pact between Ireland and the United Kingdom, established post-9/11), are subject to legal accountability and governmental transparency.65