We, the People

Electoral Reform Society,

Posted on the 26th November 2015

2. Introduction
  1. Introduction
3. Lessons on designing a citizen's assembly
  1. Lessons
4. Overview
  1. Overview
5. Method of analysis
  1. Method
6. A Short History of Scottish Constitution Debate
  1. A short history
7. 1. British Columbia and Ontario: The Early Adopters
  1. Political Context
  2. Process
  3. Outcome
  4. Impact
8. Iceland: A [Not So] Radical Approach
  1. Political Context
  2. Process
  3. Outcome
  4. Impact


About this work

This publication was written by Jordan Kroll as part of her Capstone project placement whilst she was a student on the Masters in Public Policy Programme at the University of Edinburgh under the supervision of Juliet Swann, Campaigns and Research Officer at the Electoral Reform Society Scotland and Daniel Kenealy at the University of Edinburgh.

The subject matter was chosen by ERS Scotland, the research, methodology and drafting was undertaken by Jordan Kroll in discussion with Juliet Swann and Daniel Kenealy and the publication was reviewed and edited for ERS Scotland by Juliet Swann, Jessica Garland and Katie Ghose.



In July 2012, ERS Scotland began our ‘Democracy Max’ inquiry into what makes a good Scottish democracy with a deliberative event – the People’s Gathering. The People’s Gathering invited people from across Scotland to spend a day discussing and deliberating with their fellow citizens about aspects of Scottish democracy. Self-selecting, but chosen to be as representative as possible of the Scottish people, these citizens set the parameters for our inquiry. They were asked to consider what a good Scottish democracy might look like in 2030. Their ideas were wide-ranging and innovative and included a desire for increased citizen participation in decision making, not just in elections, but in day-to-day democracy.

Since that day, ERS Scotland has become increasingly interested in citizen-led decision making as a means both of making ‘better’ decisions but also of reinvigorating people’s engagement with our democracy and democratic institutions. We have worked with government, academia, campaigners and the third sector to demonstrate and encourage participative and deliberative decision making.

Recently, proposals to include these approaches involving citizens in discussion of constitutional issues have become more widely discussed. Indeed the Scottish Government’s White Paper ‘Scotland’s Future’, promised to take such an approach to writing a constitution for an independent Scotland. Many civic society groups expressed a hope that this kind of participative engagement be used to take the conversation about Scotland’s constitutional future forward, regardless of the result of the referendum.

But why should this be limited to a Scottish conversation? Scots have been engaged in a vivid discussion about democracy and democratic institutions. We believe that it’s time the whole country joined in with that discussion. There has been a quiet rise of participative democracy in different parts of UK, including ‘flat-pack democracy’ in Frome, community land buy-outs in Eigg and Assynt, employee owned companies sharing in decision making, and the use of more participative forums in many third sector gatherings. The challenge now is how to grow these, ensure they are led by citizens, and are fully inclusive.

To help shape how we might go about having these debates, ERS Scotland asked Jordan Kroll, a student on the Edinburgh University Masters in Public Policy to collate information about four notable constitutional conventions, and to tease out what we in the UK might learn from them.

This paper draws four fundamental conclusions.

Lessons on designing a citizen's assembly


In designing a citizen-led process, these key lessons should be considered:

1. Budget: An ample budget should be provided to avoid the problems that occurred in Ireland –these included limited public awareness, low levels of staff resource and a reliance on volunteers.

2. Politicians: Including politicians is essential, after all, the process seeks to reform their role so including them and achieving their ‘buy in’ to the proceedings and the outcome should be of paramount concern. To what extent is up for debate, however the Irish approach sets new standards and offers a useful example.

3. Ensuring full engagement and maximum buy-in from the public: Perhaps easier said than done, great attention should be paid to the engagement strategy to avoid the pitfalls that have been seen in the examples studied. How can the process generate higher levels of public awareness and involvement? Furthermore, engagement should not be approached as one size fits all. Methods used in Iceland, for example, will not necessarily translate well to other communities and this should be considered.

4. Process Design: It is crucial to ensure that the process is properly designed or else it may fall victim to a bad outcome. This was the case in Iceland; too much of what the Council did was ‘on a whim’, with decisions taken without consultation or full appreciation of the impact. A degree of flexibility should be allowed but not too much. The design should also take into account existing institutions, to ensure they can aid rather than hinder progress. Engaging institutions and holding them accountable for their role is essential. Processes like those in Ireland and Iceland arguably did not have strong enough mechanisms to ensure that the government acted on their proposals. Despite the recent referendums in Ireland; voting on marriage equality and lowering the qualifying age to stand for President, the other suggested changes from the Irish Constitutional Convention have yet to be acted on. This gives further force to this final recommendation.

Whatever the process looks like, it is vital that what Government will do to enact the recommendations of the constitutional convention are clear and enforceable from the outset. A constitutional convention should not be seen as a talking shop hidden in the long grass – rather, at its best, it is a living, breathing tool to renew our democracy and to re-engage our citizens in decision making.

We hope this publication encourages the continued development of innovative thinking about citizen engagement in British democracy. There is already much to celebrate about new models of decision making being tested and adopted across the country. A nationwide conversation about where power lies and how democracy works for people could embrace these innovations and develop them still further. We look forward to working with new and existing partners to play our role in that conversation.



The debate around the Scottish independence referendum thrust the issue of a written constitution or constitutional protections for Scotland into mainstream discourse. Following the no vote, many suggested the next step should be a general United Kingdom constitutional review as more powers are devolved to Scotland.

The Smith Commission process which recommended which further powers should be devolved to Scotland was largely between politicians. This was despite the Commission receiving 380 submissions from civic society organisations and a further 14,000 from individuals.

During this same period Wales has seen a Wales Act, and in March 2015, a cross-party St David’s Day Announcement heralding the next steps for devolution in the upcoming Westminster session. However, all of these developments are the outcomes of a traditional cross-party Commission process (the Commission on Further Devolution in Wales, or the ‘Silk Commission’). The Silk Commission included proposals for further devolution, of which many (not all) were adopted in Parliament. However, this devolution process allowed for little by way of real citizen engagement. Many have proposed the idea of a constitutional convention to fully involve citizens from all the nations of the UK in discussing how our democracy functions, from politicians to academics and from newspaper columnists to campaigners. How this process would be designed is still to be determined, but the inclusion of citizens is integral in ensuring the constitutional future of Scotland and the United Kingdom is shaped in a way that best represents the interests of the people. In deciding how this will be constructed, a compelling case is to be made for the use of a citizen-led deliberative process. These processes are lauded for their ability to potentially bring about increased levels of citizen involvement, further accountability for government actors, and overall added legitimacy to the political process. In the following report, four examples of citizen-led deliberative processes will be analyzed for key lessons and their overall pertinence to the constitutional context in Scotland and the UK.

Method of analysis


This analysis will focus on four case studies in British Columbia, Ontario, Iceland, and Ireland. The main areas of research will focus on each case study’s process, design and function, outcome, and overall impact. Process design and function will be evaluated using four different categories of analysis taken from Graham Smith’s seminal work ‘Democratic Innovations’, which are: inclusiveness, popular control, considered judgment, and transparency. Outcomes will be analyzed in examining whether the process achieved its intended goal and if there were any immediate changes as a result. Overall impact will focus on more long term impacts in both the area of the case study and beyond. Using this research, key trends will be identified and applied to the current context.

A Short History of Scottish Constitution Debate

A short history

In 1707, Scotland formally joined with England with the Act of the Union, a decision not considered by all to be in the country’s best interests at the time. Many perceived the terms of the union to be overwhelming skewed in favor of English interests and saw the move to be at the expense of their country’s national identity and sovereignty, a sentiment that still continues to resonate with many to this day (Macinnes, 2014). This conviction has manifested itself in everyday Scottish political thought with many seeing Westminster as unfit to best govern Scotland. Indeed, in a poll for the Guardian in the weeks before the independence referendum which tested reasons why people in Scotland were planning to vote Yes or No it was found that for 51% of Yes voters, their motivation was “Your feelings about Westminster and the types of politicians there”. This principle has underpinned an active political movement geared towards placing the powers that were lost in the Act of the Union back in the hands of Scotland through the reinstatement of a Scottish Parliament, the devolution of powers and beyond to the most recent example, the referendum on independence (Scotland.gov.uk, n.d.).

In 1979, the first referendum on the creation of a Scottish Assembly was held but the positive result did not meet the minimum 40 percent threshold of electorate support required for passage (Scotland.gov.uk, n.d.). This did not signal the end of devolution for Scotland however. Supporters rallied and held the Constitutional Convention of 1989 to develop a detailed framework for future devolution, including a proposal for Scottish Parliament. The Labour Party used these recommendations as groundwork for their proposals for referendums on the creation of a Scottish Parliament and some limited taxation powers, which they chose to include in their election manifesto. When Labour won an overwhelming majority in the general election of May 1997, they followed through on their referendum promise. Held on the eleventh September 1997, a majority of voters chose to support the creation of a Scottish Parliament and its right to limited tax varying powers. The Scottish Parliament election was held in May 1999 followed by the first meeting of the new Parliament with responsibility for devolved powers in various areas of domestic concern such as healthcare and local government (Royal.gov.uk, n.d.). Unlike the first past the post electoral system used for Westminster elections, the Scottish Parliament is elected using an alternative member system (AMS). A form of proportional representation, this system was utilized to ensure the Parliament was more representative of the Scottish people while also making it extremely difficult for one party to secure a majority of the seats (Scottish.parliament.uk, n.d.). However, in 2011, the Scottish National Party (SNP) did secure a majority at Holyrood by winning 67 out of the 129 total seats available. Alex Salmond, as the Parliament’s elected First Minister, and the SNP Government declared this majority as a mandate from the people to further their party’s ambitions for independence by way of holding a referendum in September 2014.

86.4 percent of Scots turned out to vote to decide the future of their nation on 18 September 2014 with 55 percent of voters (around 2,001,926) choosing to vote against independence compared to 45 percent (1,617,989) who voted Yes. Prior to the vote, the leaders of the three main Westminster political parties (Conservatives, Labour and Liberal Democrats) signed a pledge promising more widespread devolution of powers to Scotland in the case of a No vote. This pledge was acted upon immediately after the referendum with the appointment of Lord Smith of Kelvin to chair a Commission with the remit: “To convene cross-party talks and facilitate an inclusive engagement process across Scotland to produce, by 30 November 2014, Heads of Agreement with recommendations for further devolution of powers to the Scottish Parliament.” (Carrell,2011).

This debate on the political future of Scotland started at a time when citizens were becoming more disenchanted with the political process. Data taken from the 2013 Scottish Social Attitudes Survey indicated that the level of trust in the Scottish Government fell from a peak of 72 percent in 2011 to 59 percent in 2013. Furthermore, only 38 percent of Scots trusted their government to make fair decisions for all (‘decisions that are fair to different groups of people in UK/Scotland.’) and only 46 percent felt the Scottish Government was ‘very or quite good’ at listening to the people’s’ views prior to making decisions (Scottish Government Social Research Group, 2014). Recognizing these trends, the Scottish Government committed itself to increasing engagement with the public beyond just elections to all stages of the policy process (The Scottish Government, 2011). The Community Empowerment Act, introduced to the Parliament in June 2014, and passed in June 2015 highlights this commitment by empowering communities to make participation requests in the areas that matter the most to them.

The debate around independence served to reinvigorate public interest and involvement in the democratic process in Scotland, and this appetite for engagement seems to be being maintained post referendum, with turnout in Scotland in the 2015 General Election several percentage points higher than in England and Wales. Grassroots groups established during the referendum campaign, including Common Weal and Women for Independence, remain active and in some cases stronger, post-referendum. Perhaps more surprising is the rise in political party membership. This has occurred across the political spectrum but is unarguably most obvious in the exceptional rise in membership of the SNP, from 25,642 on the day of the referendum to well over 110,000. It is up to political leaders, amongst others, to listen to these newly engaged citizens and capitalize on this momentum to continue to build upon this increased public engagement.

Across Scotland debates took place in pubs, on social media, buses, and more. This was not the politics of old, restricted to parliamentary debates and political elites; instead citizens genuinely felt part of the political process again (Renton, 2014). Perhaps the most telling sign was the 97 percent of the electorate that registered to vote, with 86.4 percent showing up to have their say on September 18th. Although 55 percent of voters chose to vote against independence, this arguably did not represent a preference for the continuation of the status quo (bbc.co.uk, 2014). Not least because, in the week prior to the vote, the leaders of the three main political parties (Conservatives, Labour, and Liberal Democrats) in Westminster signed a pledge promising devolution of more powers to Scotland in the case of a No vote. Following the vote, the Smith Commission was created to oversee the devolution process. The report from the Smith Commission was used to produce the Scotland Bill which is being debated in the Houses of Parliament in the summer of 2015.

Meanwhile, the newly elected Conservative Government, as well as introducing the Scotland Bill, have introduced the Cities & Local Government Devolution Bill to establish directly elected Mayors for combined local authorities in England and Wales and enable powers to be transferred to those regions; the Secretary of State for Wales has promised further devolution and a change to the Scottish model of reserved and devolved powers for Wales; we are to hold a referendum on membership of the European Union; and House of Lords reform remains a contentious issue.

If there was ever a time to consider where power lies in the UK, it is now. To exclude citizens from that process of consideration risks imposing unpopular policies and causing further disillusion with politicians and disengagement with politics.

1. British Columbia and Ontario: The Early Adopters

Political Context

British Columbia’s Citizens’ Assembly, the first of its kind in the modern age, set a precedent that has far-reaching impacts on other citizen-led cases in Ontario and beyond. Electoral system reform was the political issue at hand in both Canadian examples. Both provinces feature a first past the post system. Voters select the candidate of their choice and the candidate who receives the most votes is elected to the Parliament. The party that has the most representatives elected to Parliament is then asked to form the government.

First past the post does not necessarily produce election results representative of the electorates’ opinions, which was the case for a series of elections in British Columbia. In 1996 after winning a majority of the popular votes across the province, the Liberals managed to secure only a minority of the seats in Parliament. This disparity in opinion of the public and the makeup of Parliament prompted Gordon Campbell, leader of the disadvantaged Liberal Party, to include in the party’s election manifesto the promise of electoral system reform by way of creating a citizens’ assembly to review the state of British Columbia’s electoral system.

In 2001, the Liberals were able to win the election with 57 percent of the popular vote and won all but two of the seats in Parliament. The fact that the Liberals won 57 percent of the popular vote but were able to secure approximately 97 percent of Parliament seats again highlighted the issues with the first past the post system. Despite now being advantaged by the system, the Liberals followed through on their election promise, and oversaw the creation of a citizens’ assembly with the unanimous support of the entire Parliament (Fournier et al., 2011).

Ontario did not experience elections with irregularities in outcomes to the extent that British Columbia had, but had instead seen a series of elections that produced governments controlled by parties at different ends of the political spectrum. As a result of this extreme swapping of government control, there had been significant ‘staggers’ in policy. In 2001, Dalton McGuinty and the Liberal party put forward the suggestion of pursuing general democratic reform aimed at restoring the faith of disenchanted citizens in the political process. In 2005, after seeing the innovative democratic process that had taken place in British Columbia, McGuinty recommended that Ontario hold its own citizens’ assembly to review the electoral system (Fournier et al., 2011). The creation of a citizens’ assembly in Ontario was not met with unanimous support of Parliament like it had been in British Columbia. Not all believed, even within McGuinty’s own party, that there was a clear mandate for the Assembly like there had been in British Columbia and some saw it as a waste of government money (LeDuc, Bastedo and Baquero, 2008). Nevertheless the Assembly was approved by a majority of Parliamentarians.


As mentioned, the British Columbia example was the first of its kind in the modern age, and set the precedent for design and function for all examples that have followed it, including Ontario. Due to so many strong similarities, the following information applies to both Ontario and British Columbia unless noted otherwise.

In both cases each body was given the power to review and deliberate the issue of electoral reform. This was to be done throughout three stages: learning, consultation, and deliberation. Due to the complex nature of electoral systems, the learning stage consisted of six weekends of in-depth sessions to educate participants on the issue. After this stage participants then took part in a series of meetings (50 in British Columbia, 41 in Ontario) to consult the general public on their opinion on electoral reform in their province. Following this, participants began the deliberation stage to determine what the assembly’s policy recommendation would be. Once their recommendation was put forward, it was then translated to a referendum for the general population to vote on. For the referendum to pass and be put into law by way of a bill from Parliament a double requirement of 60 percent of electorate support and 60 percent within each district was set (Fournier et al., 2011).

Participants in the Assembly were drawn from the voter registration list using stratified random sampling. First they were placed into groups by age, followed by gender, and then district. This was done to create a sample that best reflected the composition of the general population. There was also a requirement to include representatives of the aboriginal population, one in Ontario and two in British Columbia. If a citizen was selected, they were sent an invitation letter asking if they would be interested in participating and if they were available to attend a meeting where the participants would be selected. If a citizen chose to attend this meeting, their names were then put into a draw to participate (Fournier et al., 2011). Citizens were able to choose if they wanted to participate in the event, which did create a natural self-selection bias. Those individuals who were considered to be more interested and active in the political sphere prior to the Assembly were more likely to want to participate. Individuals who were more educated and also those who had more free time were also more likely to choose to participate. In the end this created an assembly not necessarily as representative of the entire population as planned (Warren, 2008). Altogether, the British Columbia Citizens’ assembly was made up of 158 members (plus two alternatives) and Ontario’s had 104 (Fournier et al., 2011).

Steps were taken in the process design to ensure all participants had equal opportunity to voice their opinion in the learning stage and beyond. A commitment was made throughout the process to create an environment that promoted teamwork with an emphasis on building relationships and standards in communication among those involved.

Participants were placed in small, deliberative groups, along with a facilitator to moderate in each weekend session during the education stage. They then deliberated with each other on the topic at hand. Membership of these groups changed with each weekend session to further familiarize all participants with each other (Fung, Warren and Gabriel, 2011). Each of these would serve to be of benefit for the process throughout by creating an environment that was hospitable to different views and genuine deliberation. It is important to note that although steps were taken to ensure equal opportunity of participation, it did not necessarily transmit to equal participation by all. As Pearse notes in her section in Designing Deliberative Democracy (Pearse, 2008), some participants admitted to talking minimally during these small groups due to a variety of reasons such as lack of confidence in their understanding of the topic or simply being more shy at the beginning.

Throughout the entirety of the processes, mechanisms were designed to inform and further enable citizens to make their final decision using considered judgment by reflecting upon the views of others. Beginning with the first stage, participants were educated on the technical side of electoral reform through intensive learning sessions to familiarize them with the complexities of electoral systems in their province and beyond using a variety of education methods like textbooks, fact sheets, and academic lectures. In Ontario, interactive learning sessions replicating voting using different electoral systems were also used. Small groups were established to familiarize each of the participants with the thoughts and opinions of the others involved in the Assembly. Furthermore, in Ontario there was the inclusion of four extra working groups (political parties, government stability, women and underrepresented groups, and geographic representation) and four advisory committees (consultation submissions, evaluation, final report, and deliberation planning and monitoring) (LeDuc, Bastedo and Baquero, 2008). LeDuc argues that members of the Ontario Assembly may have been more self-reflective as a result of the choice to include these groups.

Following the learning stage came the public consultation stage, which was designed to involve the general public in the Assembly’s proceedings to garner their opinions and recommendations on the issue of electoral reform. A series of public meetings were held across each of the provinces where citizens were given the opportunity to present their ideas and also question the members of the assembly on its proceedings and the issue of electoral reform. Assembly participants often put in a significant effort to publicize these events. In British Columbia, members of the Assembly handed out fliers, wrote newspaper editorials, and more. A website was also created to run concurrently to the meetings to enable citizens to learn more about the assembly and also to submit their own proposals on electoral reform. If a citizen chose to present at a meeting, they would be required to also submit their recommendation via the Assembly website, email, or by post. Over 1600 submissions were received via the website in British Columbia (Fournier et al., 2011).

The objective of the consultation stage was to allow the participants to understand general public opinion. Although meetings and websites were used to do this, it is not certain the level of which the public bought into these processes as a whole. In British Columbia, where 50 meetings were held, only 2851 citizens attended out of the 4.1 million total population (Ward, 2008). At less than one percent of the population, it is not certain that the views of attendees were necessarily reflective of all in the province. Ontario chose to hold four additional outreach groups targeted at engaging citizens in socioeconomic groups that may have been left out of the process if ignored (Fournier et al., 2011).

Of the citizens who did choose to attend the public events in British Columbia, it was clear that many did not fully understand the issue due to its complex nature. This led to some presentations during the public meeting stage that misrepresented information as factual that was not. Also, presentations often went beyond the issue of electoral reform, as citizens would use the platform to express their general dissatisfaction with varying parts of the political process. On the website many submissions lacked breadth, while others were part of online lobbying campaigns and not necessarily completely reflective of the proposer’s understanding or opinion on the issue (Ward, 2008).

In the final stage, the Assembly was to deliberate and determine what electoral option they felt was best for their province. For British Columbia, deliberation lasted from September through to November 2004 and in Ontario, from February through to April 2007. Both were given six weekends to deliberate and prepare a report on their recommendation. Ultimately the British Columbia Assembly chose to recommend a single transferable vote (STV) electoral system, while Ontario’s chose mixed-member proportional (MMP) voting (Fournier et al., 2011).

Interestingly enough, British Columbia’s Assembly chose a recommendation opposite of what the public seemed to prefer. General sentiment at the public meetings tended to be in favor of a system that was more proportional and not the status quo, however most placed their support behind MMP and not STV (Fung, 2009). This decision could have been an indirect result of the process, which, although designed to educate and inform participants to make a decision most representative of the people, actually enabled participants to become quasi-experts on the issue (Thompson, 2008). Assembly members were no longer necessarily truly representative of the general public.


On May 15th 2005, the recommendation that British Columbia’s electoral system be changed from first past the post to a single transferable vote system  was put to the public vote via referendum. 57.4 percent of citizens in the province voted in favor of the change, with all but two electoral districts reporting a majority voting for STV. The proposal did not pass however, as it did not meet the required threshold of 60 percent set by the government. In general the outcome was surprising to many for a series of reasons. Only one political party, the Green Party, had openly endorsed the outcome of the Assembly. Gordon Campbell and the Liberals, the party that had been behind the creation of the Assembly, did not offer any support or critique of the recommendation. Leading up to the vote, polls indicated that the general public had little awareness of the Assembly or knowledge in the area of the referendum. Lack of political support, combined with little media awareness of the Assembly and lack of substantive education efforts, ultimately can be seen as potential reasons the referendum did not pass (Fung, Warren and Gabriel, 2011).

In Ontario the Assembly’s recommendation for a MMP voting system was put to a vote on October 10, 2007. Voters rejected the proposal with 63 percent voting against the Assembly’s recommendation. Much like British Columbia, the referendum in Ontario was adversely affected by low levels of public awareness as a result of inadequate education campaigns and media coverage throughout the entirety of the process. When media did cover the referendum, the coverage often failed to adequately explain the issue in whole and also tended to be negative. According to LeDuc’s media analysis, 45 percent of articles were negative and only 19 percent could be considered to be positive. Dalton McGuinty and the Liberals, the party that championed the process, followed Gibson’s lead from British Columbia and did not take a stance on the Assembly’s recommendation (Fournier et al., 2011).


Although the referendum did not pass in British Columbia, many took the result to be a clear indicator of the public’s dissatisfaction with the electoral system. The Liberal Government committed to hold another referendum in 2009 in which they sponsored the campaigns of those in favour and those against STV in an attempt to further educate citizens on the issue. Despite this, the referendum did not pass the second time around either.Research tends to suggest that this was due to the fact that the election was too far removed from when the Assembly occurred. Those who knew of the Assembly in the initial vote, tended to trust its decisions more than the campaigns informing them for the second referendum (Fung, Warren and Gabriel, 2011).

Despite the fact that neither province was able to secure electoral reform, much can be said of their experiments with citizens’ assemblies. Although not a perfect process, these cases provided a foundation for others to build from and improve upon.

Iceland: A [Not So] Radical Approach

Political Context

It is not often that a country receives a shock so colossal in magnitude that it is forced to reevaluate the very foundations on which its government has been formed, but such was the case in Iceland following the global financial crisis of 2008. The crash sent shocks through all segments of Icelandic society: All four national banks collapsed, workers were laid off, and businesses forced to close. Individuals who had considered themselves financially secure prior to the crash were now forced to consider issues like home foreclosure, unemployment, and more. Frustrated, citizens came out in force to protest a government they saw as corrupt and culpable for the crash their country had just experienced. These protests ultimately culminated in the Pots and Pans Revolution at the end of 2008 and into the start of 2009. Among the primary demands of the citizens protesting were for the current government to step down, the general director of the Central Bank to resign, and also, for a new constitution (Thorleifsdottir, 2013).

Iceland’s constitution was adopted in 1944 when the country formally gained independence from Denmark. At the time the constitution was intended to be only provisional as it was largely a copy of Denmark’s constitution that had been written in 1849. Politicians had pledged to update the document quickly after it was adopted, but their repeated promises had failed to amount to anything and the temporary constitution remained in place, despite Denmark updating its own version in 1953.

When the financial sector collapsed in 2008, the nation was left to reflect upon what had caused something so catastrophic. Light-touch regulation, corrupt officials, and an overreaching executive branch of Government were all identified as causes, but there was one underlying issue that could not be ignored: Iceland’s constitution had few powers to stop these things from happening. In particular it lacked an effective mechanism to check the powers of the executive branch, which it was considered had made a series of questionable but unchallenged legislative decisions that led to the financial crash (Gylfason, 2013b). The crash of 2008 provided Iceland with the constitutional moment it had long needed. Bowing to the pressures of the public, the government resigned and called for a new election that Spring. A new, left wing party coalition took control with Jóhanna Sigurðardóttir as Prime Minister (Iceland’s first female Prime Minister and the world’s first openly lesbian head of government) and in 2010 it proposed that the country undergo a formal constitutional review process (Burgess and Keating, 2013).



The review process, was to consist of a series of (arguably overly complex) stages with the intended goal creation of a new Icelandic constitution. Parliament first established the Constitutional Committee composed of experts to organize the Assembly, prepare a report based on their expertise and analysis of proposed ideas for a new constitution, and to organize the election of the Constitutional Assembly. The purpose of the National Gathering was to deliberate the central values of the country that would serve to form the guiding principles from which the constitution would be based. Drafting a new constitution would be the responsibility of the Constitutional Assembly, a body elected by the general public. They were expected to consider the recommendations put forth at the previous two stages (the Constitutional Committee and the National Gathering) when drafting the document (Gylfason, 2013bb).

The initial selection process for participants in the National Gathering was very similar to the processes in Ontario and British Columbia and also based on a similar project conducted by the Anthill Group in 20094. Participants were selected at random from the national registry and stratified by age, gender and area of residence. In total 950 participants were chosen and tasked with identifying the views and central themes of the public that would provide the basis of the constitution. There was extensive media coverage of this stage due to its relevance to the current political climate (Kok, 2011).

Due to the limited timeframe of the process (one day), it was intricately designed and participants followed a strict structure in determining these values. First, they were asked to envision what values they felt were central to their nation. Following this, they were broken into smaller groups of 8 led by facilitators to further deliberate with other participants on these values. Next they were then asked to vote on these themes and again placed in new groups. Later they were again asked to vote on these themes by identifying them by their importance. From this a series of recommendations were put forth and voted on. These were then grouped by theme and published with the relevant data. The themes identified as central to build the constitution upon were: country and nation, morality, human rights, democracy, preservation of Iceland’s nature, strong divisions of power and transparency within the government, international cooperation and peace, and values relating to well-being, justice, and equality (Kok, 2011).

Steps were taken in this process to ensure that the views of Iceland’s population would be well represented and play a central role in the building of the new Constitution. In general, surveys done of participants after the process tended to be very positive. Of those who participated, 95 percent felt the forum was a success and furthermore 93 percent felt the body’s recommendations would be used in the building of the new constitution (Kok, 2011). However, the recommendations suggested by the National Gathering were not a binding provision but were just meant to guide the Constitutional Assembly in their drafting of the constitution. The Assembly did not technically have to factor these themes into consideration when drafting a new constitution. Beyond this, any new constitution would be left subject to Parliament; essentially meaning the opinion of the people could be used at the discretion of the Assembly (Landemore, 2014).

Following the National Gathering, an election was held in November 2010 to select 25 individuals for a position in the Constitutional Assembly. The election used the Single Transferable Vote to select from 522 candidates. In total, 37 percent of the electorate participated in this election. Possible causes of this low turnout could be for a series of reasons such as the confusion stemming from the multiplicity of candidates, lack of interest, or understanding of the election. Gylfason is critical of the lack of media coverage leading up to the election. It is important to note, however, that nature of election did not necessarily facilitate easy coverage by the media. The vast number of candidates alone would have been difficult to sufficiently cover, along with the fact that those who did run often chose to not campaign extensively and instead simply ‘put their names in the ring’. This meant that little was known about the candidates beyond those who were already well- known prior to the election. Along with the lack of campaigning, the election also did not feature the traditional party politics that are often given the most attention in media. Both the Independence and the Progressive Parties chose not to field candidates, due to the fact they had received much of the blame for the financial crisis. Additionally many politicians were opposed to the perceived radical nature of the constitutional reform process (Gylfason, 2013b).

For the most part the lead up to the election of the Constitutional Assembly was fairly civil, but the same cannot be said for the aftermath. Some critics of the process claimed the election had only produced winners who were already well known and were also disproportionately from the left-leaning elite (Gylfason, 2013b). Furthermore, after the election, three individuals filed technical complaints, all claiming that the voting had not been done in secret. Siding with the plaintiffs, the Supreme Court ruled to invalidate the results of the election, throwing the entire constitutional process into disarray. Gylfason claims that the individuals who challenged the outcome had a vested interest in disrupting the process. Each of the three individuals who filed complaints had ties to the Independence Party, the party that from the outset was actively opposed to the process. He also claims that the judge who decided the case was too closely involved with the Independence Party and may not have made his ruling impartially.

Regardless of the reason why the Supreme Court chose to rule in favor of the plaintiffs, the decision could have easily served as a fatal blow to the entire process. With the decision by the Supreme Court calling into question the mandate of the constitutional review process and public trust in government institutions at an all-time low, it was up to Parliament to decide the fate of the process. In the end, they chose to sidestep the ruling of the Supreme Court by directly appointing the 25 individuals who had won the election to a Constitutional Council – basically created in lieu of the elected but invalidated Constitutional Assembly. Whilst this was a well-meaning attempt to ensure the process continued, it would also serve as a blow to the overall legitimacy of the process (Landemore, 2014).

When the work of the Council finally began, the 25 individuals agreed to work on three main tasks. These were: updating the human rights chapter of the existing constitution to include nature protection and further social and civil rights, establishing clearer divisions of the three branches of the government, and increasing mechanisms for direct democracy (Bergmann, 2014). The next move of the Council was perhaps the most groundbreaking. With a mandate that was not necessarily clear and an angry public, the Council made the choice to further open the process to the public through outlets like Twitter, Facebook, and YouTube. Each of their meetings was put online, along with all documents used during proceedings, in order that the public could actively follow their daily affairs. Individuals were encouraged to post recommendations for the Council to take into consideration when drafting the constitution. Opening the process drew the attention of media in Iceland and abroad, which quickly proclaimed the innovative process to be the first of its kind in creating a truly ‘crowd-sourced’ constitution (Valtysson, 2013).

The move of the Council to open the drafting process as much as possible to the general public was interesting, and one that the Council chose to do against the advice of many political elites who believed that the writing of a constitution was something best left to a small group of capable individuals. This decision was an attempt to open the process to regain the legitimacy that had been lost when Parliament chose to sidestep the ruling of the Supreme Court. Opening up the process more to the public undoubtedly increased levels of inclusiveness, popular control, considered judgment, and transparency; however, the true extent of this impact deserves further analysis. Beyond this, Ólafsson believes the Council also actively made it their goal to distance itself as much as possible from Parliament – an institution that was facing high levels of public distrust at the time. This choice to distance itself from Parliament and other government institutions may have had unintended repercussions on the process.

It is widely acknowledged, and feted, that the Icelandic Constitutional Council created an innovative series of mechanisms to enable greater participation by the public, but how much did the public and the Council buy into these participatory tools? According to Bergmann, the Council received several thousand submissions through social media and 370 by way of traditional correspondence. While a substantial amount, it is important to note that due to the limited timeframe the Council was unable to read and respond to each submission (Landemore, 2014). This begs the question that although the Council was willing to engage the public, to what extent did they value their opinion? Creating a space for public discussion did not necessarily translate to deliberation by the Council. Was the public truly being empowered through this process as there is without any mechanism in place to ensure their opinion had any impact in the drafting of the constitution? In this sense, the Icelandic process, while seemingly more far-reaching in widespread public engagement than an example like British Columbia, was perhaps not as innovative as has been suggested as it still failed to engage the public beyond simply providing for submission suggestions and ideas that would not necessarily receive due consideration.

Ólafsson is critical of the decision-making of the Council, pointing out that often their decisions seemed to lack any clear methodology in design and were more on a whim than in line with a consistent procedural approach. While not necessarily poor ideas, they often appeared not fully thought out and were designed in a way that did not necessarily then facilitate what the Council had initially intended. Such was the case with the decision to open the process more to the public. Although in theory this move further empowered the public to become involved in the process, it was designed in a way that only superficially did so. There was also a tension in how the Council interpreted their role compared to what others believed it to be. The Constitutional Committee was created to advise the Council (then Assembly) on updating a constitution. In the document they created, it was clear that the assumption was that the constitution was only going to be revised and not completely rewritten. This was not what the Council interpreted their role to be and undertook the process with the intention of rewriting the entire document (although they did end up using sections from the original Constitution) – a lofty goal for a three-month process (Gylfason, 2013b). Perhaps the most telling example of this ill-thought through thinking was the Council’s choice to distance itself from Parliament and other political elites, with some members thinking that if they were able to reach consensus on the document this would alone be enough to compel Parliament to accept the draft (Landemore, 2014). In the event, this decision did the opposite of what they had intended as they inevitably alienated MPs from the process.


On July 29, 2011, after almost a four-month process (an option of a one-month extension was available if needed) the Council presented a draft constitution to Parliament. Parliament had not made a formal plan for what to do following the submission of the draft and was left to decide the next steps. The decision was made to hold a non-binding advisory referendum on October 20, 2012 that would feature six questions relating to key provisions of the constitutional proposals (Landemore, 2014). These questions were:

  1. Do you wish the Constitution Council’s proposals to form the basis of a new draft Constitution?
  2. In the new Constitution, do you want natural resources that are not privately owned to be declared national property?
  3. Would you like to see provisions in the new Constitution on an established national church in Iceland?
  4. Would you like to see a provision in the new Constitution authorising the election of particular individuals to the Althingi more than is the case at present?
  5. Would you like to see a provision in the new Constitution giving equal weight to votes cast in all parts of the country?
  6. Would you like to see a provision in the new Constitution stating that a certain proportion of the electorate is able to demand that issues be put to a referendum?

Although the results were non-binding, it was thought that this vote would reinforce that a

majority of the public was in support of the proposals and would further compel Parliament (mainly aimed at the critics of the process) to support the draft. When the referendum was held, every question received a majority of votes in favour of yes with a voter turnout rate of 49 percent (Landemore, 2014). While not a majority of the population, the result did broadly indicate that the electorate was in support of a new constitution and the ideas proposed by the Council. It would be up to Parliament to guarantee its passage.

Unfortunately, this process has not led to a new constitution for Iceland. Although many supported it initially, the bill ultimately stalled in Parliament for a variety of reasons. Those who were against the process from the outset, the Independence and Progressive Parties, intensified their criticisms once Parliament began to initiate the steps to put the proposal into a bill. They were strongly opposed to certain provisions like collective ownership of natural resources, with some labelling the process a pet project of former Prime Minister Jóhanna Sigurðardóttir and those on the left (Bergmann, 2014). It was not these two parties alone that were against the process though, as the coalition that initially had backed the process was not able to maintain the necessary level of support among all in their ranks leading up to Parliamentary action on the draft (Gylfason, 2013).


Although the Icelandic example did not produce a new constitution, it has undoubtedly had an impact both in Iceland and internationally. In Iceland, the project at the very least provided an outlet for the public to channel their frustrations into. It could be argued that this process has had more of an impact on the international community than in Iceland itself as many have hailed the process as groundbreaking and the first ‘crowd-sourced’ constitution suggesting this is something that could provide the framework for Constitution-making in the modern age. However, the fact remains that, Iceland still awaits an updated Constitution. The two political parties (Independence and Progressive) most associated with the crash were reelected into a coalition following the April 2013 election (Burgess and Keating, 2013). Time will tell if this signals a return of status quo or whether this process is representative of a change in the nature of Icelandic politics.

Ireland: A Hybrid Example

Political Context

Beginning in 2007, the nation of Ireland was in financial crisis. For most of the 1990s and 2000s, Ireland had enjoyed a period of rapid economic growth fuelled by business-friendly policies and an increasingly well-educated and productive workforce. The first phase of this growth (1995-2002) can largely be characterized as catch-up growth in which Ireland’s economy rapidly converged to the level of other economically successful nations. This changed in 2002 when Ireland’s growth became far more dependent on the expansion of credit from banks, particularly in the housing sector. Due to the lack of strong government regulations or oversight, banks were able to lend essentially at will and take on riskier investments than normal, actions which led to the creation of a speculative bubble that inevitably burst in 2007. Starting in this year, banks began reporting major deficits in their loan portfolios as a result of borrowers’ inability to pay their debts (Ec.europa.eu, 2012).

During this time, financial problems were not isolated to Ireland. Much of Europe and the rest of the world were undergoing similar crises that put a strain on banks everywhere. Constrained by this, Irish banks were unable to access funds from interbank lending that may have normally been an option for assistance. This put further pressure on the Irish Government to offer assistance in saving these banks from collapse. Already faced with a significant budget deficit as a result of a large decrease in tax revenue, the decision to back-up the banks’ liabilities pushed the government further into economic trouble. In an attempt to mitigate their financial problems, the government implemented a series of austerity measures that would end up being insufficient to the task. When it became clear that the country was unable to solely continue to fund itself, they were able to secure a €85 billion financial assistance package through the assistance of the International Monetary Fund and the European Union (Ec.europa.eu, 2012).

A financial crisis of this magnitude predictably permeated through all of segments of Irish society with high levels of unemployment, slashes in government spending, and more; people were angry and their trust in government nearly nonexistent. In 2011, the first general election since the onset of the crisis, much attention was focused on fixing what had gone wrong and preventing it from happening again. Political parties promised reform to what the public perceived as broken and highly untrustworthy institutions. In addition to institutional reform, each party also included in their respective manifestos the promise of increased citizen engagement (Fianna Fail, 2011; Finn Gael, 2011; The Irish Labour Party, 2011) Farrell points out that these cross-party commitments to political reform and citizen engagement had never happened before in Irish history. The detail of this cross party agreement was aided by the results of a pilot constitutional convention, ‘We the Citizens’. We the Citizens was a project initiated by a group of political scientists from various Irish universities, the Political Science Association of Ireland (PSAI) and the Irish Universities Association and was grant funded by The Atlantic Philanthropies. The findings from the pilot were submitted to the Irish Government and although the final Constitutional Convention was not modelled exactly on the pilot, it is fair to say it had considerable influence on the inclusion of proposals in the party manifestos in 2011.


On December 1st of 2012, the Irish Constitutional Convention held its first session. A July 2012 Resolution of the Houses of the Oireachtas (Ireland’s Parliament) gave the body twelve months to deliberate and prepare recommendations on a series of predetermined issues. The Convention would be made up of 100 individuals: 66 were to be Irish citizens selected at random (much like the other examples), 33 were to be elected politicians, with the appointed chair bringing the group total to 100. Rules governing the conduct of the body were to be determined by the group, and they were free to seek opinions from experts or whomever else they felt would be beneficial. Decisions would be made using a majority voting system and in the case of a tie, the chair would cast the deciding vote. In total, the group was given a budget of €300,000 for the entirety of the process (Harris, 2014). The issues the body was given to examine and make recommendations on were the following:

  1. Reduction of the Presidential term of office to five years and the alignment with local and European elections
  2. Reduction of the voting age to 17
  3. Review of the Dáil electoral system
  4. Irish citizen’s right to vote at Irish Embassies in Presidential Elections
  5. Provisions for same-sex marriage
  6. Amendment to clause on the role of women in the home and encouraging greater participation of women in public life
  7. Increasing the participation of women in politics
  8. Removal of the offense of blasphemy from the Constitution

Beyond these specific issues, Oireachtas also gave the Convention the power to make recommendations on additional issues providing there was enough time following the completion of review of the prescribed list. Their recommendations were to be submitted to the House of Oireachtas, which would then respond to each of the recommendations within four months of receiving them. If a recommendation put forth by the Convention was approved, Oireachtas would then set a date for the recommendation to be voted on by referendum.

The Ireland example followed the path of case studies like British Columbia in many ways and also the pilot assembly, We the Citizens, which had taken place in 2011 as referenced above. Citizens were stratified and selected at random to ensure that they were representative of the general population and as inclusive as possible. Once participants were selected, they met for initial plenary sessions where various experts with various positions on the issues would present evidence. They were then broken into small groups for deliberation. These deliberation sessions were led by a facilitator and aided by a note- taker. This ensured that participants were given equal opportunity to speak if they wanted to (Farrell, 2014b).

Following these deliberations, a series of public meetings were held to hear opinions from the general public, similarly to the outreach in BC and Ontario. Throughout the entirety of the process there was also a public website that detailed the purpose of the Convention, posted videos of its proceedings and other relevant documents. This was intended to enable greater transparency and public awareness. There was a space on the website for citizens to submit their own comments and proposals, much like each of the other cases had done (The Convention on the Constitution, 2014).

Although very similar in design, there are important differences between Ireland and the other case studies. Perhaps most notable was the decision to include 33 politicians directly in the Convention, a decision which came about largely from a political compromise. As mentioned above, leading up to the election in 2011, all parties had included a commitment in their respective manifestos to consider increasing citizen engagement. Finn Gael’s manifesto detailed an ambition to create a Citizens’ Assembly to deliberate issues of electoral reform. Labour’s manifesto proposed a plan for a Constitutional Convention that would be made-up of one-third citizens, one-third politicians, and one- third experts. When the two parties joined together in a governing coalition, their compromise was to have a body composed of 2/3 citizens and 1/3 politicians which would consider a wider range of issues including, but not limited to, electoral reform (Farrell, n.d.).

The choice to include politicians directly in the process was controversial. Many experts on citizens’ assemblies advised against it and argued that the body be made up solely of citizens (Farrell, 2014b). They argued that including politicians directly could run the risk of these individuals attempting to commandeer the process. While this could be a possibility, it was also suggested that politicians might interfere with the process from outside if they were not engaged enough. In the Iceland example, the decision made by the Constitutional Council to not engage with politicians during drafting perhaps led to politicians feeling alienated from the process. Arguably this, combined with other factors, led to the ultimate defeat of the Constitution. Ultimately, Oireachtas decision to include politicians has been judged to be for the better. Of the participants polled, many said they felt they had equal opportunities to speak up and participate at all times. Additionally, politicians were often more comfortable with asking questions and offering opinions, especially when it came to the more technical issues like electoral reform and Oireachtas operating standards, areas that citizen participants did not always feel confident in discussing, thus aiding the dissemination of information (The Convention on the Constitution, 2014).

Like the other case studies, a series of mechanisms were put in place to enable greater popular control, considered judgment, and general inclusiveness. These mechanisms, however, were not perfect and often constrained by factors like the low budget (Farrell, 2014a; Harris, 2014). For instance, the limited budget and staff meant they were unable to put together a substantial media campaign to increase public awareness of the process. Professor Farrell speculates that only around two to three percent of the general public were aware of the Convention. The inclusion of politicians did help with this publicity vacuum as they were able to use their platform to spread news to constituents and other members of Parliament. To try and increase public awareness and involvement, the Convention itself chose to hold a total of nine public meetings in various places around the country (The Convention on the Constitution, 2014). Farrell points out that often these meetings were not held in geographically ideal areas and general public awareness of their occurring was low. Issues that were already considered more salient, like same-sex marriage, tended to draw far more public attention. Harris suggests that to further increase inclusiveness on all issues steps could have been taken to further enable citizens to participate by providing practical services like childcare to generate greater possible involvement of women.

The website was also greatly limited by the lack of financial resources. Again, the already discussed issues received most attention: 2500 submissions were received, with 2326 submitted via the website. Over 1000 of these were in regard to same-sex marriage (The Convention on the Constitution, 2014).


In total, the Convention made 38 recommendations. These were both from the 8 areas they were delegated power to deliberate on plus an additional two (Daíl reform and economic, social, and cultural rights) which were added for consideration on the request of participants (The Convention on the Constitution, 2014). Of these recommendations, 18 would require a national referendum (Verma, 2014). As noted earlier, the Irish Government had committed itself to responding to each of the Convention’s recommendations within four months. To date, the Irish Government has scheduled dates for only two referendums – on marriage equality and reducing the required age for presidential candidates. A third on lowering the voting age was initially to be included with the same-sex marriage and age of presidential candidate polls but this was not proceeded with. A fourth on removing blasphemy from the Constitution has been agreed upon but has yet to be formally scheduled (Kelly, 2014). Despite the commitment to respond within four months, many recommendations have now seen the deadline for response pass with no action taken. There is still hope that additional referendums will be set on the remaining issues.


Although the extent of the impact of the Convention on Irish politics and beyond is still largely to be determined, the process design and function should be lauded. Given the limited budget and the wide- ranging issues that were considered, the Convention could be judged a success. Not only did it deliberate and make recommendations on more technical issues like Dáil reform, but also traditionally controversial issues like same-sex marriage. The choice to include politicians, which was met with skepticism from many experts initially, has provided a successful alternative to the traditional citizens’ assembly design. The extent to which it will influence future processes is still to be seen.

Key Trends


Although there is no clear right or wrong way to design a citizen-led process, key lessons from previous examples can be identified and serve as a guide for future Scottish or UK-wide processes. Whether this be a constitutional convention or something else, it is essential that the inclusion of citizens serve as the central tenet it is designed around. Beyond this, the following lessons that have been identified should be considered:

Budget and Timetable

An inadequate budget can severely limit the parameters and impact of a citizen-led process. In British Columbia, a process that was designed to examine solely the issue of electoral reform, had a budget was approximately 5.5 million Canadian dollars. This provided financial resources for an adequate staff and extensive public engagement (50 public meetings). This can be contrasted with Ireland, where the Convention was tasked with considering eight constitutional policy areas, but only had a budget of around €300,000. When interviewing Professors Farrell and Harris, each mentioned how the lack of a proper budget severely limited the Irish process in nearly all aspects from the evidence and education phase to public outreach. Due to this constraint, the Convention was unable to hire more than the bare minimum of staff. This led to the outsourcing of important tasks such as website creation to someone who had volunteered for free. As a result, the website was rarely updated and many complained about difficulty navigating the page due to poor design. In the final report published by the Convention, the need for sufficient resources to further increase engagement of the public was highlighted if another convention were to occur.

The process must also be allotted enough time to ensure that it can devote the proper amount of attention to each the issues it will be deliberating, engage the public, and write reports. Both British Columbia and Ontario devoted a year to their Assemblies, and a similar amount of time was dedicated to the Irish process. In Iceland, the Council was given just four months to revise the Constitution and chose instead to draft a new Constitution in that timeframe, a mammoth task for the short time allotted.


There is a lot of information to consider when discussing the role of politicians. Should they be involved and further, to what extent? Are their roles simply to act as champions?

These processes are often born out of some form of public disenchantment with the political process (see all cases previously discussed). Given this, the instinct might be for politicians to remove themselves from the process as much as possible or lead to an approach that actively removes politicians from the process like in Iceland.

In British Columbia, while Gordon Gibson and the Liberal Party instigated the process, they made the choice to abstain from supporting the outcome when the referendum was called and instead only supported the process as a whole. Little definitive evidence exists as to whether this decision proved a hindrance to the overall success of the referendum but still must be mentioned and considered when designing a process. Should politicians only support the process or do they have a greater role?

The cases of Iceland and Ireland are particularly interesting due to their similarities (both created to consider wide constitutional issues and born out of the financial crisis) and their differences (largely the role of politicians). In Iceland, following the decision by the Supreme Court to invalidate the results of the Constitutional Assembly, Parliament chose to appoint each of the individuals that had been elected to the Constitutional Council. Following their

appointment, the Council chose to distance themselves as much from Parliament in attempts to regain the legitimacy lost from their decision to sidestep the ruling. When their draft was put to Parliament for consideration, many politicians stated they felt alienated from the process, including those who had initially supported the process.

In Ireland, the decision to include politicians was met with skepticism initially. Many experts warned their inclusion could lead to their possible commandeering of the process, but this has not proven to be the case. According to Farrell, participants overwhelmingly indicated they felt politicians’ participation was positive to the process. Politicians were more comfortable speaking, especially on the more technical issues that citizen participants did not always feel comfortable querying. Furthermore, politicians often acted as champions of the process outside of the Convention when speaking to constituents and fellow politicians. This was useful in helping to mitigate some of the effects of a limited budget, particularly in raising awareness amongst the public and other decision makers.

Outside Engagement

British Columbia, Ontario, and Ireland had a similar approach to public engagement. A website for each was created to inform the public of the Convention and to also serve as a space for citizens to submit their policy proposals. British Columbia received slightly over 1600 submissions, 2152 in Ontario, and just over 2300 in Ireland. In addition to the website, each assembly held meetings for the general public to attend. Ireland held by far the least with only nine, but this was arguably due to budget constraints and also the fact that the process design did not require they hold these meetings (they were suggested by the participants themselves). Figures are not available for each case regarding the number of citizens that attended these meetings but consider British Columbia; of the 4.1 million population at the time, only 2851 citizens attended. While a substantial amount on its own, this is less than one percent of the population as a whole. Does this qualify as a substantial amount of the population that has been engaged? A lack of public awareness and failure to generate wider engagement has affected each of these processes. In Ireland, Farrell speculates that less than 2 percent of the general population was aware of the Convention. How can this problem be improved upon in future processes?

Iceland’s approach to engagement has led to worldwide acclaim and the label ‘first crowd-sourced constitution’. Unlike the other examples, it solicited proposals beyond the traditional website. Citizens were able to post their recommendations via sites like Twitter and more. However, as noted in the earlier analysis, this did not necessarily lead to any more democratic a process. The Council was not required to look at all of the recommendations, and this led to many being ignored. Furthermore, while perhaps a hallmark of a new, technological approach to engagement in a citizen-led process, it should be acknowledged that this method of engagement would not necessarily transition well into all future examples. Internet penetration is Iceland is around 95 percent, whereas it is much lower in other countries – for example under 80 percent in Scotland.

Process Design

British Columbia and Ontario followed identical structures with three stages: learning, public consultation, and deliberation. The design of the structure was largely determined prior to the start of the actual process. Participants were given the power to examine one issue, electoral reform. They were to make a recommendation that would then be put to referendum for the general public to vote on. In Ireland, the Convention was given eight issue areas to deliberate and was permitted to consider additional areas if there was time. Compared to British Columbia and Ontario, the Irish Constitutional Convention’s powers were quite broad. On the other hand, they were limited largely to an advisory role. Their recommendations were not promised to go to referendum but instead to be reviewed by Parliament within four months. Essentially, Parliament was not bound by law but only by promise to review these recommendations. This is arguably the biggest criticism of the process as many of the recommendations have passed the four-month window and have yet to be reviewed by Oireachtas.

Iceland had significantly more flexibility of process design than the other case studies. Following the Supreme Court ruling, the Constitutional Council made the decision to further open the process in an attempt to regain the legitimacy that had been lost. Additionally, they made the decision to actively disengage from politicians as a way to distance themselves from an institution that suffered from high levels of public distrust. As mentioned earlier, not all commentators feel this flexibility was a positive. Ólafsson points out that this enabled the Council to make decisions seemingly on impulse without any clear methodology. This was arguably detrimental to the process, especially in regards to the decision to actively avoid interacting with politicians. Instead of acknowledging Parliament as an existing institution which they desired to reform, the Constitutional Council took the decision that reaching a consensus on a draft Constitution would be enough to compel Parliament to act. This did not prove to be the case.


Lessons on designing a citizens' assembly

In designing a citizen-led process, these key lessons should be considered:

  1. Budget: An ample budget should be provided to avoid the problems that occurred in Ireland – these included limited public awareness, low levels of staff resource and a reliance on volunteers
  2. Politicians: Including politicians is essential, after all, the process seeks to reform their role so including them and achieving their ‘buy in’ to the proceedings and the outcome should be paramount. To what extent is up for debate, however the Irish approach sets new standards and offers a useful example.
  3. Ensuring full engagement and maximum buy-in by the public: Perhaps easier said than done, great attention should be paid to the engagement strategy to avoid the pitfalls that have been seen in the examples studied. How can the process generate higher levels of public awareness and involvement? Furthermore, engagement should not be approached as one size fits all. Methods used in Iceland will not necessarily translate well to Scotland or elsewhere, and this should be considered.
  4. Process Design: It is crucial to ensure that the process is properly designed or else it may fall victim to a bad outcome. This was the case in Iceland; too much of what the Council did was ‘on a whim’, with decisions taken without consultation or full appreciation of the impact. A degree of flexibility should be allowed but not too much. The design should also take into account existing institutions, to ensure they can aid rather than hinder progress. Engaging institutions and holding them accountable for their role is essential. Processes like those in Ireland and Iceland arguably did not have strong enough mechanisms to ensure that the government acted on their proposals.
    Despite the recent referendums in Ireland, voting on marriage equality and lowering the qualifying age to stand for President, the other suggested changes from the Irish Constitutional Convention have yet to be acted on. This gives further force to this final recommendation.

Whatever the process looks like, it is vital that what Government will do to enact the recommendations of the constitutional convention are clear and enforceable from the outset. A constitutional convention should not be seen as a talking shop hidden in the long grass–rather, at its best, it is a living, breathing tool to renew our democracy and to re-engage our citizens in decision making.


Bbc.co.uk, (2014). Scottish independence referendum – Results – BBC News. [online] Available at: http://www.bbc.co.uk/news/events/scotland-decides/results [Accessed 9 Nov. 2014].

Bergmann, E. (2014). Iceland and the international financial crisis. Palgrave Macmillan, pp.169-183.

Burgess, S. and Keating, C. (2013). Occupy the Social Contract! Participatory Democracy and Iceland’s Crowd-Sourced Constitution. New Political Science, 35(3), pp.417-431.

Carrell, S. (2011). Salmond hails ‘historic’ victory as SNP secures Holyrood’s first ever majority. The Guardian. [online] Available at: http://www.theguardian.com/politics/2011/may/06/scottish-elections-salmond-historic-victory-snp [Accessed 1 Sep. 2014].

Community Empowerment (Scotland) Bill 2014. Available at: http://www.scottish.parliament.uk/S4_Bills/Community%20Empowerment%20%28Scotland%29%20Bill/b52s4-introd.pdf [Accessed 1 Sep. 2014]

Ec.europa.eu, (2012). Ireland’s economic crisis: how did it happen and what is being done about it?. [online] Available at: http://ec.europa.eu/ireland/key-eu-policy-areas/economy/irelands-economic-crisis/index_en.htm [Accessed 21 Sep. 2014].

Farrell, D. (n.d.). The 2013 Irish Constitutional Convention: A bold step or a damp squib?. Unpublished Manuscript, In: J. O’Dowd and G. Ferrari, ed., 75 Years of the Constitution of Ireland: An Irish-Italian Dialogue, 1st ed. [online] Available at: http://www.openpolitics.ro/wp-content/uploads/UCD_Mar13_final1.pdf [Accessed 25 Sep. 2014].

Farrell, D. (2014b). The Irish Constitutional Convention offers a potential route-map for renewing UK democracy. [Blog] Democratic Audit UK. Available at: http://www.democraticaudit.com/?p=8625 [Accessed 20 Oct. 2014].

Fianna Fail, (2011). Real Plan. Better Future. [online] pp.31-32. Available at: http://election.fiannafail.ie/pages/read-the-plan [Accessed 21 Sep. 2014].

Finn Gael, (2011). Let’s Get Ireland Working. [online] pp.16-19. Available at: http://electioncamp.eu/sites/default/files/5%20Point%20Plan%20A4%20Brochure%20WEB.pdf [Accessed 21 Sep. 2014].

Fournier, P., van der Kolk, H., Carty, R., Blais, A. and Rose, J. (2011). When citizens decide. Oxford: Oxford University Press.

Fung, A. (2009). British Columbia Citizens’ Assembly on Electoral Reform | Participedia. [online] Participedia.net. Available at: http://participedia.net/en/cases/british-columbia-citizens-assembly-electoral-reform [Accessed 14 Sep. 2014].

Fung, A., Warren, M. and Gabriel, O. (2011). 53 The British Columbia Citizens’ Assembly on Electoral Reform (BCCA). In: B. Stiftung, ed., Vitalizing Democracy through Participation: Reinhard Mohn Prize 2011, 1st ed. [online] Bertelsmann Foundation, pp.53-67. Available at: http://www.bertelsmann-stiftung.de/cps/rde/xbcr/SID-2F361BED-59CF835B/bst_engl/British%20Columbia_engl.pdf [Accessed 9 Aug. 2014].

Gylfason, T. (2013a). Democracy on ice: a post-mortem of the Icelandic constitution. [online] openDemocracy. Available at: https://www.opendemocracy.net/can-europe-make-it/thorvaldur-gylfason/democracy-on-ice-post-mortem-of-icelandic-constitution [Accessed 13 Oct. 2014].

Gylfason, T. (2013b). From Collapse to Constitution: The Case of Iceland. In: L. Paganetto, ed., Public Debt, Global Governance and Economic Dynamism, 1st ed. Springer, pp.379-417.

Harris, C. Interviewed by: Kroll, J. (14th August 2014).

Houses of Oireachtas Resolution, (2012). Found at: Convention on the Constitution, (2012). Resolution of the Houses of the Oireachtas of July, 2012. [online] Available at: https://www.constitution.ie/Documents/Terms_of_Reference.pdf [Accessed 21 Sep. 2014].

Institute for Government, (2011). Scottish Devolution (1997-9). Institute for Government Policy Reunion. [online] Available at: http://www.instituteforgovernment.org.uk/sites/default/files/scottish_devolution_0.pdf [Accessed 15 Sep. 2014].

Kelly, F. (2014). Referendum due on removing blasphemy from Constitution. The Irish Times. [online] Available at: http://www.irishtimes.com/news/politics/referendum-due-on-removing-blasphemy-from-constitution-1.1945835 [Accessed 15 Oct. 2014].

Kok, A. (2011). Icelandic National Forum 2010 | Participedia. [online] Participedia.net. Available at: http://participedia.net/en/cases/icelandic-national-forum-2010 [Accessed 14 Sep. 2014].

Landemore, H. (2014). Inclusive Constitution-Making: The Icelandic Experiment. Journal of Political Philosophy, p.n/a-n/a.

LeDuc, L., Bastedo, H. and Baquero, C. (2008). The Quiet Referendum: Why Electoral Reform Failed in Ontario. In: Annual Meeting of Canadian Political Science Association. [online] Available at: http://www.cpsa-acsp.ca/papers-2008/LeDuc.pdf [Accessed 12 Sep. 2014].

Macinnes, P. (2014). BBC – History – British History in depth: Acts of Union: The creation of the United Kingdom. [online] Bbc.co.uk. Available at: http://www.bbc.co.uk/history/british/empire_seapower/acts_of_union_01.shtml [Accessed 6 Nov. 2014].

Ólafsson, J. (2011). An Experiment in Iceland: Crowdsourcing a Constitution. In: Epistemic Democracy in Practice.

Pearse, H. (2008). Institutional design and citizen deliberation. In: M. Warren and H. Pearse, ed., Designing Deliberative Democracy The British Columbia Citizens’ Assembly, 1st ed. Cambridge University Press, pp.70-84.

Renton, A. (2014). Scottish Independence: How the Better Together Campaign Almost Threw It All Away. Newsweek. [online] Available at: http://www.newsweek.com/2014/09/26/scottish-independence-how-better-together-campaign-threw-it-all-away-271370.html [Accessed 9 Nov. 2014].

Royal.gov.uk, (n.d.). Queen and Scottish Parliament. [online] Available at: http://www.royal.gov.uk/MonarchUK/QueenandGovernment/QueenandScottishParliament.aspx [Accessed 1 Sep. 2014].

Scotland.gov.uk, (n.d.). History of devolution. [online] Available at: http://www.scotland.gov.uk/About/Factfile/18060/11550 [Accessed 6 Nov. 2014].

Scottish Government Social Research Group, (2014). Scottish Social Attitudes Survey 2013: core module – attitudes to government, the economy, health and social care services, and social capital in Scotland. pp.7-20.

Servini, N. (2014). Crabb’s wind of change. [online] BBC News. Available at: http://www.bbc.co.uk/news/uk-wales-28608716 [Accessed 6 Oct. 2014].

Smith, G. (2009). Democratic innovations. Cambridge, UK: Cambridge University Press, pp.8-28.

The Irish Labour Party, (2011). Labour Party Election Manifesto 2011: One Ireland – Jobs, Reform, Fairness. [online] pp.45-48. Available at: http://www.labour.ie/download/pdf/labour_election_manifesto_2011.pdf [Accessed 21 Sep. 2014].

The Convention on the Constitution, (2014). Ninth Report of the Convention on the Constitution: Conclusions and final recommendations. [online] Available at: http://static.rasset.ie/documents/news/conclusions-and-final-recommendations.pdf [Accessed 6 Mar. 2019].

The Scottish Government, (2011). Commission on the Future Delivery of Public Services.

Thjodaratkvaedi.is, (2012). The Referendum – October 20, 2012 – Iceland – Referendum. [online] Available at: http://www.thjodaratkvaedi.is/2012/en/referendum.html [Accessed 15 Sep. 2014].

Thompson, D. (2008). Who should govern who governs? The role of citizens in reforming the electoral system. In: M. Warren and H. Pearse, ed., Designing Deliberative Democracy The British Columbia Citizens’ Assembly, 1st ed. Cambridge University Press, pp.20-49.

Thjodaratkvaedi.is, (2012). The Referendum – October 20, 2012 – Iceland – Referendum. [online] Available at: http://www.thjodaratkvaedi.is/2012/en/referendum.html [Accessed 15 Sep. 2014].

Thorleifsdottir, T. (2013). From the people to the people, a new constitution. [Blog] openDemocracy. Available at: https://www.opendemocracy.net/thorhildur-thorleifsdottir/from-people-to-people-new-constitution [Accessed 12 Sep. 2014].

Ward, I. (2008). An Experiment in Political Communication: The British Columbia Citizens’ Assembly on Electoral Reform. Australian Journal of Political Science, 43(2), pp.301-315.

Valtysson, B. (2013). Democracy in disguise: the use of social media in reviewing the Icelandic Constitution. Media, Culture & Society, 36(1), pp.52-68.

Verma, N. (2014). Deliberative approaches to political reform: David Farrell on the Irish Constitutional Convention. [online] The Constitution Unit Blog. Available at: http://constitution-unit.com/2014/05/29/the-irish-constitution-convention/ [Accessed 20 Oct. 2014].

Warren, M. (2008). Citizen representatives. In: M. Warren and H. Pearse, ed., Designing Deliberative Democracy The British Columbia Citizens’ Assembly, 1st ed. Cambridge University Press, pp.50-69.

Read more publications...