Consultation on an Interim Constitution for Scotland Written Submission, Dr Alan Renwick, University of Reading This response relates to Question 5 of the consultation, on the proposals for the development of a permanent written constitution by a Constitutional Convention. I am a political scientist with expertise in processes of constitutional and political reform. My recent pamphlet, After the Referendum: Options for a Constitutional Convention1 examines in detail the various ways in which a constitution-making body might be formed in light of wide experience in the UK and around the world. This submission draws on the analysis in that pamphlet. Summary There are seven basic forms of constitution-making body. The proposals set out by the Scottish Government has already rightly exclude several of these – such as a royal commission or a committee of politicians – as not adequately engaging the people of Scotland. I strongly endorse the Scottish Government’s intention to make the process more inclusive than this. Four more inclusive models exist o a civil society convention, on the model of the Scottish Constitutional Convention of 1989–95; o a directly elected assembly, on the model of Iceland’s Constitutional Council of 2011; o a citizens’ assembly, comprising randomly selected citizens, on the model of the electoral system assemblies in British Columbia, Ontario, and the Netherlands between 2005 and 2007; o a mixed assembly, the best model for which is the Irish Constitutional Convention of 2012–14. The first three of these are far from ideal. In terms of composition, something akin to the Irish model offers the best way forward for Scotland. This should comprise a majority of ordinary citizens chosen at random from the electoral register (though with the option to decline the invitation) and a minority of politicians chosen by parties. Inclusion of representatives from organized civil society would not be the best option, but could be defended. In terms of operation, the model of the three citizens’ assemblies is preferable. That is, the convention should have time and resources for extended learning, deliberation, and decision phases, backed up by consultation with experts, activists, and ordinary citizens. The convention’s recommendations should automatically be put to the people in a referendum. There should be no opportunity for political intercession at this stage. What the Design of a Constitutional Convention Needs to Achieve There are five main criteria for judging the options: 1. Will the process foster a debate that is based on reason rather than interest or passion? 1 Alan Renwick, After the Referendum: Options for a Constitutional Convention (London: Constitution Society, April 2014). 1 2. Will the reasoning be of good quality? Will the participants understand the options and their strengths and weaknesses? 3. Does the process allow for equitable inclusion of all parts of society? 4. Will the process foster public legitimacy? Will the public have confidence in the process and the recommendations that it generates? 5. Will the process foster political legitimacy? Will politicians feel an obligation to take the recommendations forward to fruition? The first four of these criteria are implicit in the consultation document. The fifth is not, but it is important: several excellent political reform bodies have produced well reasoned recommendations that have led nowhere because of lack of political buy-in. The danger of this is lower in Scotland than in the other cases, but it does exist: I note that the draft bill does not commit to putting the convention’s proposals to a referendum. This criterion therefore needs to be taken seriously. How Do the Options Measure Up against These Criteria? Civil Society Convention: This option may seem attractive. There is a precedent for it in Scotland, in the form of the Scottish Constitutional Convention of 1989–95. It offers a way to include voices beyond the political elite. The consultation document talks of the inclusion of “civil society groups”. But it is problematic. How is it decided who will be represented? In what proportions are they represented? What of those groups – such as the unemployed – that lack a strong voice in civil society? What of individuals who do not feel the spokespeople of organized civil society adequately express their perspectives? It is simply not possible to achieve equitable representation through the mediation of civil society: this option fails on the third criterion. For that reason, those who designed the Scottish Constitutional Convention in the 1980s themselves viewed it as only second best. They preferred a directly elected assembly, but recognized this was impossible without government support, which would not be forthcoming while the Conservatives were in power. Directly Elected Assembly A elected assembly is, if an appropriate electoral system is used, in theory representative of the whole of society. Furthermore, the Icelandic Constitutional Council, which is the only example of a fully directly elected constitutional assembly in any existing democracy, worked very well: its members were all independent of political parties; this meant that they could deliberate without the controlling influence of vested interests in the existing system. But it would be much harder to make this work in Scotland, whose population is more than sixteen times larger than Iceland’s. The candidates who can gain sufficient support in a polity of Scotland’s size are likely to be people who are prominent already – in politics, the media, and other spheres. Again, therefore, this option would not facilitate the representation of ordinary Scots. Furthermore, if, as seems likely, an elected assembly were dominated by political parties, then the parties’ interests, rather than impartial reasoning, would tend to dominate the debates and decisions Conversely, if, as in Iceland, party politicians were expressly excluded, there would be a danger of weak political legitimacy. In Iceland, despite the quality of the Council’s 2 deliberations and despite a positive referendum vote for the implementation of its draft constitution, politicians who felt excluded from the process have blocked its enactment. An elected assembly is thus likely to fail on the third criterion and also, depending on its composition, on either the first or the fifth. Citizens’ Assembly Citizens’ assemblies are composed of ordinary citizens who are randomly selected from the electoral register. Those invited to participate do have the opportunity to turn the invitation down, which tends to skew participation towards those who are more interested in politics. It is possible, however, to stratify the sample, as in an opinion poll, to ensure it is representative in terms of factors such as age, gender, and income. The citizens’ assemblies that have taken place have been genuine cross-sections of the whole of society. Those existing examples also show that, with appropriate support, ordinary citizens working in citizens’ assemblies can understand apparently abstruse topics and engage in high quality reasoning. Appropriate support includes the use of trained facilitators to help all members express their views and the development of a programme of ‘learning’ sessions with experts and activists. Citizens’ assemblies thus perform very well on the first three criteria. Evidence from past cases suggests they also perform well on the fourth: citizens who understand the work of these assemblies are more likely than others to support their recommendations. The problem is the fifth criterion: no citizens’ assembly has yet led to actual change. Part of the problem has been that the assemblies have been called when the political elite was far from unanimous in its desire for any change at all, which will clearly be less of an issue in Scotland. But part of the issue has also been that politicians have felt no buy-in to the process and no need to help the assemblies’ proposals towards enactment. Pure citizens’ assemblies are therefore at risk of failing the fifth criterion. Mixed Assembly There have been examples of various mixed assemblies. The best by far is that of the Irish Constitutional Convention, which operated between 2012 and early 2014. 66 of the Convention’s members were ordinary citizens chosen on the citizens’ assembly model; 33 were politicians chosen by the parties; and the government appointed an independent chair. Concerns were expressed at the Convention’s inception this mixture would not work: that the politicians would dominate and the ordinary citizens’ voice would not be heard. That is not what happened. Detailed research by Jane Suiter and colleagues2 shows that members of all types overwhelmingly felt able to express themselves. The Convention members deliberated well and produced clear, reasoned conclusions. While the Convention concluded its work only recently, the evidence so far is that politicians have taken it much more seriously than in the cases of the pure citizens’ assemblies: the politician members of the Convention have often acted as ambassadors for the proposals during Dáil debates; the government has so far promised three referendums stemming from the Convention’s recommendations, and more may follow. The evidence we have therefore suggests that a mixed assembly is most likely to satisfy all five criteria. 2 Jane Suiter, David M. Farrell, Eoin O’Malley, and Clodagh Harris, “It’s good to talk: Citizen-politician deliberations in Ireland’s Constitutional Convention of 2012-14.” Paper delivered at the General Conference of the European Consortium for Political Research, University of Glasgow, 3–6 September 2014. 3 What Should the Mix in a Mixed Assembly Be? The Irish model involves a majority of ordinary citizens and a minority of politicians. The Scottish Government’s consultation document expresses a desire to include representatives of civil society. So what would the best mix be? As noted above, civil society representation cannot meet the criterion of equal representation of the whole of society. Random selection of ordinary citizens and selection of politicians, if done appropriately, can meet this criterion. The optimal assembly therefore does not include civil society organizations. The case for including civil society organizations would be similar to that for including politicians: it may be through that their support is needed in order to broaden the legitimacy of the convention’s proposals. Given the past example of the earlier Scottish Constitutional Convention, it may be that in Scotland, more than elsewhere, civil society leaders may make life difficult for constitutional recommendations if they do not feel they have been included in the process that generated them. It may therefore be pragmatic to include civil society groups. Nevertheless, the majority of the convention members should be ordinary citizens. Otherwise, there is a danger they would feel marginalized. If, therefore, it is felt necessary to include civil society groups, then the appropriate balance may be either one sixth each politicians and civil society group representatives and two thirds ordinary citizens or one fifth each politicians and civil society group representatives and three fifths ordinary citizens. Procedural Matters Beyond these various points relating to composition, I emphasize three procedural points: A convention of the kind proposed here needs time. It will need time for a ‘learning phase’, when it hears a wide range of ideas about the sorts of issues a constitution should determine and what options are available. It will need time for a detailed ‘deliberation phase’, when it explores various options in detail and receives feedback on them from a wide variety of sources. And it will need time for a ‘decision phase’, when it gradually makes decisions on the many aspects of a constitution. All of this is likely to take at least a year, with the convention meeting for weekends every two or three weeks. Such a convention will also need support. Expert advice should be available at all stages. Politicians, civil society organizations and ordinary citizens should be able to express their views through written and oral submissions. Members of the convention should take part in meetings throughout Scotland. Twitter, Facebook, and other media should be used to gather ideas and encourage debate. Trained facilitators should be available during the convention’s sessions to help all its members participate fully. As indicated above, similar models have been used in Ireland and Canada with remarkable success. A clear procedure should be laid out for what happens after the convention submits its constitutional recommendations. The Scottish Government should commit to putting these recommendations to a referendum automatically and to respecting the outcome of that referendum. A constitution is a document that sets out (among other things) the rules of 4 the game that politicians must observe. It is, therefore, not for politicians to intercede in its content. 5
© Copyright 2026 Paperzz