Introduction

“A Constitution is not an Act of Government, but of a people constituting a government.”2 It is well over two centuries since this declaration of public ownership over constitution making by Thomas Paine, and we can now say that public participation in constitutional processes has become a norm. Even beyond democratic contexts where processes have been widely studied for their engagement with citizens, such as the three successive processes in Chile in the last ten years, even the government of Kazakhstan and the Palestinian Authority, neither known for their open, participatory form of law-making, embarked in recent months on constitution making processes which they claimed were based on broad inputs from the citizenry.3

Whether or not these processes are actually open and participatory, and whether or not participation made a significant impact on the text, that all governments now feel the need to claim popular ownership over constitution making reflects that it has indeed become a global norm and expectation. Given this, “the key question for decision-makers is not whether to involve the public, but how and when to involve them and in what ways”.4

This paper examines the how question, looking at recent experiences of constitution making—both those that succeeded and those that did not—and provides some general reflections on different methods of citizen engagement, the challenges encountered and considerations for improving these practices in the future. Most of what follows is drawn from my own personal reflections as a constitutional advisor in over a dozen countries, situating my own experiences and observations within concepts and frameworks from the ever-expanding literature on participatory constitution making. While this experience is by no means a comprehensive account of constitution making processes, the reflections that follow are based on issues which have recurred commonly across these processes, taking place over four different continents, and in diverse contexts including fragile and post-conflict states (e.g. Haiti, South Sudan, Somalia), post-authoritarian democratic transitions (e.g. Tunisia, the Gambia, Ukraine), popular revolutions in democracies (e.g. Chile) and stable democratic contexts in the absence of crisis (e.g. Barbados, Belize, Ghana).

But before discussing the “how” it is important to ask whether citizens should be making constitutions, and if so what do we expect from them?

I. Why Public Participation?

The reasons given for public participation are well-known, and have been the focus of many studies of constitution making processes.5 These include that it provides a forum for the negotiations of solutions to divisive or contested issues thus contributing to conflict resolution,6 can lead to better democratic outcomes,7 can result in citizens being better informed and more politically sophisticated in how their democratic system works8 and can strengthen compliance with certain aspects of the constitution.9

At the same time, a number of studies have been skeptical of these claims, finding them to be generally overstated, highly dependent on context and secondary to the role of political elites.10 Several studies have also made arguments as to why public participation can actually be harmful to the overall goals of democracy, human rights and constitutional legitimacy that it is purported to promote.11

Beyond these differences in the teleological reasons for public participation in constitution making, there are also questions about what we are doing when we engage citizens, and who are the citizens? When we design participatory constitution making processes, we must answer the boundary problem raised by Sir Ivor Jennings. “On the surface it seemed reasonable: let the people decide. It was in fact ridiculous because the people cannot decide until somebody decides who are the people.”12

When we say we have consulted the public, who are the public? Inevitably citizens will disagree on any constitutional design choice, so whose opinion counts? If one of the primary purposes of constitutions is to protect certain values and rights against the democratic power of the majority… how does this apply to participatory constitution making? If a majority of the citizenry favour a certain policy (for example the death penalty, or the exclusive use of a certain language) should that be seen as an imperative mandate for those sitting in the drafters’ chairs? Clearly, in this case, public participation is not meant to be a mere poll of opinions, but something else.

Beyond the tyranny of the majority, there are also concerns of a tyranny of the minority. Even the processes most lauded for their broad participation, have involved a tiny fraction of the population in their consultations. For example, according to the OECD, Colombia in 1991 achieved 0.06 % citizen participation: Iceland in 2010 only 0.3 % and Tunisia in 2011 a similar 0.06 % of the population. Chile’s case was extolled as being unprecedented because of its high rate of participation… which amount to only 1.13 % of citizens.13

And who makes up these tiny fractions of the population? The answer, very often, is not a representative sample of the citizenry. As discussed in more detail below, certain methods of public consultation result in strong self-selection by special interest groups with particular interests in specific provisions of the Constitution, but perhaps little knowledge, interest or even curiosity about the rest of the text.

These dilemmas are common to all participatory processes of constitution making. But as stated in the introduction, given that participation has become an expectation, they are dilemmas all processes must address. This requires thinking carefully about what we are asking the public to do when we ask them to participate.

II. Forms Of Participation:

In a previous paper,14 Erin Houlihan and I identified four principal roles for citizens in constitution making processes, which I update here to five based on recent experiences. In rough chronological order as to when they might feature in the process, citizens can act as:

– Initiators/Agenda Setters. Here the role of the citizen comes in before the formal constitutional reform process starts. From the Arab Spring to the Maidan Revolution, and more recently in Chile and Bangladesh, mass protest movements have demanded not only a change of government, but also a change of system—including calls for a new constitution (as in Chile), and sometimes explicit, substantive demands (as in Ukraine). In some cases, citizens have made formal decisions at referendum to call for a constituent assembly (e.g. Colombia 1990, Chile 2021).

– Electors. Perhaps most commonly, the role of citizens is indirect—limited to electing representatives to serve on a constitution making body. This may be in the form of a regular legislature which assumes a constitution-making role (as in Kenya in 2007), or electing members to a constituent legislature or assembly with an explicit mandate for constitution making (as in South Africa in 1994 or Chile in 2021).

– Contributors/Advisors. Public consultations, increasingly common though not always significantly substantive, put the citizen in the role of a contributor or advisor. They provide opinions, but political elites decide what goes into the constitution. An innovative form of this was the citizen-initiatives in the Chile Constitutional Convention which is discussed further below.

– Authors. In certain, limited recent cases the citizens themselves can be said to have been authors of constitutional change. This would include Ireland, where randomly-selected citizens assemblies put forward proposals for constitutional reforms which were decided at referendum.

– Approvers/Rejectors. At the end of a constitution making process, citizens may be asked at referendum to give a thumbs up/thumbs down decision on the final text. Referendums are not ubiquitous, but where they are used they are thought to provide greater popular legitimacy, and are common in particular in Francophone countries where a new constitution signifies a new republic.

While there is much to say on each of these roles and how they are manifested in different constitution making processes, the major focus—and I would posit greatest area for improvement—in most processes has been around the area of public consultations, and that is the sole subject of Part III. However, before turning to public consultations, in this section I provide some brief reflections on the other roles listed above based on some experiences in recent processes.

Firstly, at a time when trust in political institutions globally continues to decline,15 it should not be surprising to see mass protests demanding not only a change in personnel in political institutions, but also a change in the institutions themselves. However, in most cases where protestors are successful in removing the government and initiating a constitutional reform process, once the formal process begins it is commonly the case that it moves forward in seeming complete ignorance of the original reasons for its initiation, and revolves instead around the political divides inside the constitution-making body. A case in point is the Tunisia constitution making process of 2011-2014, where the revolution centered around social justice—captured in the slogan “Bread and Dignity”16—the debates in the Constituent Assembly, and resulting constitutional text, did little to seriously grapple with these demands, instead focusing on issues of religion and the state, gender and presidential powers—areas which were critical for the political parties, but perhaps not for the voters who had sent them there.

Recent, creative efforts in Bangladesh to address this challenge, and to safeguard the demands of the revolution through constitutional change also look to be in jeopardy. Following mass protests which forced the resignation of Prime Minister Sheikh Hasina in July 2024, a series of reform bodies were established during the transitional period, one of which was tasked with finding political consensus on a reform agenda. The resulting document—titled the “July Charter”—was released on October 25, 2025, and signed by 25 political parties, but with significant dissents and reservations to several key points. With reforms scheduled to be implemented by the parliament to be elected in February 2026, the transitional government sought to bind the parliament’s hands by calling for a referendum on the “July Charter”, whereby the people overwhelmingly approved a question asking them whether the parliament should implement the “July Charter” reforms. Nevertheless, once the Bangladesh Nationalist Party (BNP) won a sweeping landslide in elections, it wasted no time in asserting that the “July Charter” had no legally binding status, and it would be up to parliament (i.e. the BNP) to decide which reforms, if any, would be taken up.

This present a serious challenge in the light of the aforementioned distrust and disaffection of citizens in politics and political institutions—if citizens do not trust the public institutions meant to serve them, but also cannot change the institutions, what avenues for public frustration are left?

Secondly, a reflection on referendums. In my experience, I encounter two common misperceptions of constitutional referendums. The first is an assumption that a new constitution must be approved by referendum if it is to achieve enduring popular support and legitimacy. This is clearly not the case—a short list of prominent and long-lasting constitutions which were not approved by referendum but have achieved strong anchoring in public legitimacy includes Norway 1814, Netherlands 1815, the postwar constitutions of Germany, Japan, India and Italy, Brazil 1988, Colombia 1991, South Africa 1996 and Nepal 2015.

The second misperception is that there is something inherently unfeasible or even undemocratic about approving a constitution at referendum as it is a package of proposals for which the citizens only get one vote, and should they be in favour of certain parts of the constitution, but a rejection of others, they cannot express these preferences. For example, this was a common critique of the recent “July Charter” referendum in Bangladesh.17 While one might certainly argue, and argue convincingly, that in most constitutional referendums the voters do not read the text, and may not understand the details on which they are voting, the “single-purpose” criticism of constitutional referendums is mistaken in my view on two counts. Firstly, it is entirely expected that citizens may not be in favour of all the policies of a candidate for president or a political party at elections,18 but we expect them to be able to weigh the pros and cons and vote accordingly for what is the best option which meets their preferences. The question before them at a constitutional referendum is between the new constitution and the status quo, not the draft constitution and their ideal constitution. In deciding between these two “packages” we should trust them to be able to make the same kind of judgment they make at elections. Secondly, many parts of the constitution are tied together as a holistic bundle. Voting for one part while rejecting another may undo the coherence of institutional design, as well as undoing a potentially significant political bargain upon which constitution making often depends. Thus voting on individual provisions would create perhaps more problems than it solves.

Thirdly, with regard to elections, with Kimana Zulueta-Fuelscher I have commentated elsewhere on the challenges and different modes of (s)electing constitution making bodies in fragile and conflict-affected settings,19 and I have also written on the nature of elections as “upstream constraints” on constitution making in divided societies.20 Here, I would like to mention one specific issue which recently arose in Chile.

In the Chile Constitutional Convention, one important innovation aimed at ‘bringing the people in’ was to allow independent candidates to run on lists, on an equal footing with political parties. This led to a large number (42 %) of the delegates being elected as independents, the vast majority of these on the left of the political spectrum and many who had run on one special interest. In speaking with colleagues who were working with some of the independents at the Convention, it was clear that many felt they were sent there for one specific purpose (e.g. animal rights) and would vote for a constitution which protected this interest, and against a draft which did not. When asked about other critical issues—for example the powers of the president, decentralization or judicial independence—some explained that these issues were not their concern. This led to a somewhat unruly negotiation process, whereby “(m)ost independents were single-issue activists seeking to participate in the Convention as environmentalists, feminists, or traditionalists rather than as agents responsible for negotiating across multiple, complex dimensions”.21 In my view the experience of independent lists was perhaps a necessary product of the political context at the time in Chile, and one to learn from—but not repeat—in future processes.

Lastly, I touch only briefly here on the citizen “as author” as I have had no direct experience with such mechanisms as citizen assemblies, and the Ireland experience is covered elsewhere in this journal.22 However, to the Irish experience I would add an interesting innovation from Mongolia where, under a 2017 Law on Deliberative Polling, a randomly-selected assembly of citizens must consider proposals, and can make proposals, for constitutional amendment to be placed before the parliament. This mechanism has been used twice, and both times was understood to be effective in either blocking amendment proposals, or suggesting amendment proposals.23 While it may be too early to judge the effectiveness of this mechanism, it may be an important innovation in considerations of how to bring the citizens into politics with a view to preventing abusive constitutional amendments democratic backsliding, or “constitution unmaking”24.

III. Public Consultations

Generally, when citizens, politicians or international advisors think about public participation in constitution making, they do not think about the roles and practices laid out above, they think primarily about public consultations. These may most often take the form of open townhall meetings (online or in-person); telephone/online/postal submissions or sectoral consultations, where the constitution-making body consults with a selected group (e.g. women, youth, business leaders, civil society leaders).

In many processes I have visited, these exercises are often not thoughtfully planned, with little clear strategy of how to engage ordinary citizens in a way which would make for meaningful consultation. If not carefully managed, in the best-case scenario, this results in a symbolic, box-ticking exercise, where the authorities can claim there was space for people to be heard. In other cases, the result can be delegitimization of the process as the public perceives the consultations as being superficial and unserious, or worse can result in polarization and conflict.25 In this section I discuss some of the common weaknesses of these exercises, mainly because the means it is not matched to the ends, with some reflections on where how they can be improved, drawing on comparative good practice.

Firstly, town-hall meetings, which I would define as any meeting open to the public where views are heard and recorded by members of the constitution making body. These usually take the form of open, physical meetings around the country, often taking place in schools, municipal halls or open spaces, and recently increasingly also taking place online—for example the Ghana Constitution Review Committee and Barbados Constitution Review Commission both held “X-spaces” to engage thousands of youth citizens.26

As means to gather information, I have found such meetings regularly disappointing. Firstly, there is a self-selection bias in the citizens who attend. This not only means that the participants cannot be said to be a representative sample of the population, but also because those that ‘self-select’ are often from organized special interest groups which can sometimes lead to the degeneration of these meetings into shouting matches between opposing groups, rather than deliberative gatherings. The initial meetings of the Barbados Constitution Review Commission were dominated by LGBTIQ+ activists and conservative Christian groups who gave long speeches focused on the same issue, namely whether the Constitution should declare Christian values or whether it should recognize same-sex marriage. Indeed, some of the same individuals traveled to different consultations, repeating the same arguments and dominating the CRC’s initial agenda. This experience is by no means limited to Barbados, and while this is certainly not to say that these groups did not have valid concerns and desires to contribute, that they should dominate the agenda was counterproductive to the goal of understanding the general public sentiment.

Secondly, with little prior information or civic education, the general citizenry is often ill-prepared to discuss the constitution and propose changes. This not only decreases the value of the submissions, but it can also contribute to the spread of misinformation. Further, most submissions focus on perceived grievances with government, and often these shortcomings are not necessarily constitutional in nature. A memorable example from the Gambia came from a local farmer who complained to the Constitution Review Commission that he was no longer allowed to shoot hippopotamuses who were damaging his crops. It falls to the constitution making body to interpret these sentiments and determine how they might connect with the constitution, but as inputs are often contradictory and the list of complaints soon becomes very long, most do not feature in the constitution making body’s recommendations.

Having said this, the advantage of town-hall meetings is that they are open to everyone, so all citizens know they have the opportunity to be heard. They also take the constitution making body around the country, rather than sitting in the capital city. This not only has a symbolic importance, it also helps raise the profile of the constitution making process in the consciousness of citizens, increasing the likelihood that they feel this is their process. Furthermore, from Fiji to the Gambia I have heard members of the constitution making body say that they learned a great deal about the state of their country by going to places they have never been to—and would never go to—in their lives.

Thus, town-hall consultations are by no means useless, but should be carefully planned to cater for their shortcomings, while maximizing their advantages. Barbados provides interesting insights here—after the first few meetings, the Constitution Review Commission changed its strategy in small but important ways: they produced an agenda for different meetings, so that one issue could not dominate all the discussions; they installed a large clock at the front of the room so that participants knew they had a limited time to speak in order to allow a larger plurality of contributions; and they brought independent experts to correct any glaring misinformation.

Another example of catering for the weakness of townhall meetings in terms of gathering information, while maximizing their communicative function, was in the Gambia. As village meetings would often be dominated by the voices of male heads of households, the Constitutional Review Commission (CRC), also sought to convene separate “focus group discussions” with women, youths, persons with disability in each locality.27 They also accompanied the consultations with a broad communication plan—including a song, a music concert and an advance team which went ahead to each venue for consultations to receive the permission of local traditional leaders, and explain to the public what the process was for, and why it was important. At the very least, this gave more marginalized communities the opportunity to be heard in their own right, but also from what I witnessed, there ended up being clear links between these focus group discussions and the draft constitution produced by the CRC.

If townhall meetings are sub-optimal means of obtaining data/information during public consultations, what alternatives might be sought? From discussions with diverse constitution making bodies, sectoral consultations often arise as providing a targeted and informed forum for discussions. The disadvantage here is that they suffer from the Jenning “boundary problem” referred to above, in that the constitution making body must decide which sectors and which groups it will consult with. But if this is carefully and consultatively planned, and combined with open “townhall” meetings, sectoral consultations can be an invaluable form of bringing citizens’ voices into the constitution making process.

On the issues that concern them, these groups are more likely to be informed to make prepared submissions, and the quality of information received by the constitution making body is likely to benefit as a consequence. Further, if the right sectoral leaders participate, this can greatly increase general public buy-in for the process, and the eventual constitution.

There are generally two different methods of sectoral consultations. One is to bring together a national convention, with delegates selected by different sectors. This has been common in Francophone Africa, as well as in examples such as the 2004 “Bomas Convention in Kenya”. Often consisting of thousands of participants, these can be unruly occasions, and of limited use in gathering reliable information. However, it is important to remind ourselves here that a democracy does not consist only of a multitude of individual voices. Democratic societies have always been made up of cross-cutting communities—from political parties to sports clubs, from church groups to trade unions. While these conventions may not be ideal for data gathering, having the leaders of these communities come together to discuss, debate and commit to a consensus deal can be a powerful means of donating societal ownership and legitimacy to the constitutional project.

The other form of sectoral consultation is to identify a particular sector or issue, and invite qualified individuals or groups to participate or make submissions. This was the approach taken recently in places such as Ghana28 and Haiti29, where—in the former in particular—the sectoral consultations were combined effectively with public “townhall” meetings, and consultations with state organs to generate a broad composite picture of the demands for change.

A third form of brining the people in through sectors is by having specific sectoral representation included in a constitutional reform commission itself, rather than it consisting solely of a group of experts. Thus, I have seen religious community leaders, trade unionists, business leaders, youth leaders, civil society representatives and many others included directly on relatively small constitution making bodies. Many of these had no prior experience or knowledge of constitutions, or even public law. Where organized well, and combined with the requisite expertise, however, these bodies can be effective ways to bring in the most important societal voices to the process.

Lastly, a major challenge with public consultations—in particular when conducted without careful planning—is they can be more detrimental than helpful in achieving objectives such as popular support and legitimacy for the constitutional project.30 This occurs in particular when citizens feel they have been used to provide a veneer of inclusivity, but their opinions have not been listened to. There are three different approaches I have seen to thoughtfully avoiding this trap, none of them remotely complicated, but all of them often ignored in many processes.

Firstly, many constitution-making bodies set out on public consultations without any system to gather, organize and analyze the data. The result is reams of paper, or terabytes of computer files, which are never considered in any systematic way. The benefit of having a team of data analysts on board from the start—as was the case in Kenya (both 2000-2004 and 2009-2010), Zimbabwe31 (2013) and the Gambia (2019-2020) should not be underestimated.

Secondly, in its final report the constitution-making body should explain clearly how public submissions were considered, what the most common opinions were, and if they were not translated into recommendations, why this was. While consultation does not infer a duty to take all opinions on board, it should include a duty to listen and carefully consider, and demonstrating how this was done in a final report should be a regular practice. A good example is the final report of the Constitution of Kenya Review Commission of 2004, which not only included a specific chapter on “Owning the Process”, explaining how the public were consulted, but also in each of its chapters containing recommendations, including a subsection with “What the People Said”, before explaining how these views were incorporated or not in the final recommendations.32

Thirdly, the Constitutional Convention in Chile worked into its procedures a process for citizen initiatives whereby any petition—which had to include a concrete proposal, written in a particular format—achieved a certain number of signatures, it would be included and discussed in the agenda of the Convention. Again, there was no commitment that the proposal would be adopted, but citizens knew ahead of time how the procedure through which their submissions would be treated, and guaranteed to be considered and discussed in the constitution making body if it garnered sufficient public support.

None of these recommendations is particularly groundbreaking or innovative. Indeed, they all flow from the basic but fundamental principle of considering the purpose for public participation and adapting the most suitable means. Nevertheless, it is remarkable and somewhat disheartening to see how many processes do not take this path.

IV. Conclusion

As stated at the outset, nearly all processes of constitution making in modern times utilize, or claim to utilize, some form of public participation. Thus, the practical consequences of the question of whether a process should be participatory seems to be redundant. As further detailed here, this participation can take up several different roles including the citizen as initiator, elector, advisor, author and approver. These roles can take different forms and often are combined in the same process.

One particularly visible and often-considered role is through public consultations. However, these are too often conducted without careful planning, and with little consideration of matching means and ends. I have sought to capture some practical considerations, drawn from recent comparative practice, as to where the relevant challenges were more considerately addressed.

I would end here with two general notes. Some critics of public participation have posited that, in the end, constitution making is actually dependent on an elite bargaining, and there is no constitution which is actually derived from the general will of the citizens. This may be true, to an extent, and thus the title of this paper (taken from the original talk upon which it is based), may be a utopian dream, as there is no possibility to actually let the citizens draft the constitution. However, I do believe that constitutional politics—like all politics—can be shaped and constrained by public pressure, and well-organized public participation can therefore shape, if not override, any eventual elite pact.

Secondly, given the number of recent constitution making processes which have been launched as a consequence of public pressure, but which have eventually stalled or failed,33 more consideration needs to be given to alternative pathways for the citizen as ‘initiator’ role. If wholesale constitutional change is increasingly unlikely, what other means and processes might we conceive to meet the demands of the increasing number of citizens who want to reimagine their political society?