Written Evidence submitted by Professor John Garry and James Pow for the Northern Ireland Affairs Committee’s inquiry into Devolution and democracy in Northern Ireland – dealing with the deficit (DDD0022)

Preface & Acknowledgements

John Garry is Professor of Political Behaviour in the School of History, Anthropology, Philosophy and Politics at Queen’s University Belfast. James Pow is a Research Fellow in the same School. The authors make this submission in their capacity as researchers at Queen’s University Belfast.

Different elements of this research have been presented at the annual meeting of the International Society of Political Psychology in Edinburgh in 2017, at the annual conference of the Political Studies Association of Ireland in Dublin in 2017, at the weekly seminar series of the Centre for Deliberative Democracy and Global Governance at the University of Canberra, and at the Northern Ireland Assembly’s Knowledge Exchange Seminar Series.

The authors would like to gratefully acknowledge funding for this work from the Economic and Social Research Council (ESRC), regarding the following two projects: “Randomly Selected "Politicians": Transforming Democracy in the Post-Conflict Context” (ES/M000257/1) and “Power-Sharing and Voting: Conflict, Accountability and Electoral Behaviour at the 2015 Northern Ireland Assembly Election” (ES/L005808/1).

1.0              Executive Summary

• A democratic deficit arguably existed before the collapse of Northern Ireland’s devolved institutions in January 2017.

• At the input stage of the democratic system, the nature of the electoral process results in the amplification of ethno-national politics over other ideological dimensions.

• Decision-making has been characterised by bargaining and negotiation over group interests, as opposed to deliberation over common interests.

• A low veto threshold has contributed to routine gridlock, or at least its routine possibility, reducing the perceived effectiveness of the devolved institutions.

• This democratic deficit in terms of representation, deliberation and responsiveness, is likely to grow in the event of the continued absence of devolved decision-making by elected representatives.

• We suggest that deliberative democracy can help to reduce this deficit, by involving citizens directly in deliberation and decision-making.

• More specifically, a citizens’ assembly could play a role in the decision-making process when locally elected politicians are unable to deal with certain issues themselves.

• Citizens’ assemblies, comprising a representative sample of ordinary citizens, have been used in Canada and Ireland to assist elected politicians with decision-making on controversial and constitutional issues.

• Survey evidence from Northern Ireland suggests that the public is broadly supportive of the idea of a citizens’ assembly playing a role in decision-making on contentious political issues.

• MLAs are more sceptical, but are more open to the idea provided a citizens’ assembly operates in an advisory capacity.

• Experimental evidence suggests that the public’s support for the idea is robust: people are just as likely to accept a (personally unfavourable) decision taken a citizens’ assembly as a (personally unfavourable) decision taken by elected politicians in inter-party talks.

• In any event – whether Northern Ireland’s devolved institutions are restored, or whether they are suspended indefinitely – the democratic quality of the decision-making process may be enhanced by the direct involvement of ordinary citizens in political deliberation.

2.0           Contextualising the Problem

2.1             Northern Ireland is currently without devolved legislative and executive branches of government. As per the Northern Ireland (St Andrews Agreement) Act (2006), the largest unionist party and largest nationalist party in the Northern Ireland Assembly must share power in a cross-community Executive, reflecting the terms negotiated in the Belfast Agreement (1998). When Sinn Féin, as the largest nationalist party, withdrew its ministers from the Northern Ireland Executive in January 2017, devolution has been in crisis. Efforts to restore the Executive have been unsuccessful. An Assembly election in March 2017 and subsequent inter-party talks chaired by the Secretary of State for Northern Ireland have, so far, been insufficient to bring about the restoration of a power-sharing Executive. If the Democratic Unionist Party (DUP), as the largest unionist party, and Sinn Féin, as the largest nationalist party, cannot reach an agreement to govern together, Northern Ireland will be left without devolved government for an indefinite period. This exposes a serious democratic deficit.

2.2             If we think of democracy as a decision-making process – with inputs, throughputs and outputs, it could be argued that recent developments are simply manifestations of existing deficiencies in the performance of Northern Ireland’s devolved democratic system. Indeed, an audit of the quality of democracy in Northern Ireland argues that the collapse of the Assembly in 2017 “reinforces the sense that Northern Ireland does not ‘do democracy’ very well” (Wilson and Nolan, 2017: 4).

2.3             At the input stage of the process, elections are used to select decision-makers. A proportional electoral system (PR-STV) promotes fairness of representation and reduces the number of wasted votes. However, in practice, the electoral process places disproportionate weight on the ethno-national dimension of competition (Coakley, 2008). This is understandable given the nature of the historical structure of Northern Ireland’s party system, but the sheer dominance of this ethno-national dimension may only weakly represent the views of citizens for whom ethno-national politics is less salient. Since 2006, the annual Northern Ireland Life and Times Survey has consistently found that at least 40 percent of citizens identify as neither nationalist nor unionist (Ark, 2017), approximately four times the percentage of elected Assembly members who designate as such. Thus, the electoral process tends to amplify ethno-national division, rather than to purely reflect it.

2.4                                                                                                                                                                                                                                            The throughput-stage of decision-making is characterised by bargaining and negotiation at the elite-level. Debates typically reflect fixed partisan positions, as opposed to a more deliberative style of decision-making that necessarily involves a willingness to change initial preferences in pursuit of a common solution. This pattern is similar to decision-making in most other legislatures (Ferejohn, 2008). However, this is particularly problematic in Northern Ireland given the power-sharing context. The failure of co-governing parties to set aside narrow interests in favour of common interests can result in dysfunctional government.

2.5             Finally, any decisions taken constitute the outputs of the system. However, a low veto threshold makes gridlock a routine possibility. By attaching a petition of concern to a bill or motion, 30 members of the Assembly have the ability to require its passage by cross-community consent; that is, a majority of unionist MLAs and a majority of nationalist MLAs (Schwartz, 2014). This gives one community, or even a single political party, an effective veto. Thus, any policy enacted may reflect lowest-common denominator preferences. Indeed, Birrell and Gormley-Heenan (2015) argue that most legislation passed by the Assembly has been relatively inconsequential.

2.6             In sum, Northern Ireland’s devolved system of government may be criticised for an over-representation of the ethno-national dimension, a lack of meaningful deliberation in decision-making, and a low volume of substantive policy output. In the event of direct rule from London, these existing weaknesses over democratic representation, deliberation and responsiveness would likely grow in perceived magnitude.

2.7             It remains a significant achievement that the democratic institutions of Northern Ireland operated uninterrupted for the decade between 2007 and 2017. In a deeply divided polity, the challenge of effective institutional design – and successful operation thereafter – is profound. However, at the present juncture it is appropriate to consider the weaknesses of Northern Ireland’s devolved system and how they may be ameliorated.

3.0           Supplementing the System

3.1             If locally elected representatives are unable to form a devolved power-sharing government, alternative forms of decision-making may be considered. One such option would be to involve citizens directly the decision-making process through greater deliberative democracy (Dryzek, 2000). This could be achieved in a systematic way by establishing a citizens’ assembly (Smith, 2009).

3.2             There are three main features of a citizens’ assembly. First, its members, usually at least one hundred, would be randomly selected from the broader population – similar to the principles behind jury selection. The random sampling method, combined with a reasonably large sample size, is used to generate a membership that is representative of the broader population. In other words, to a high degree of probability, the members of the citizens’ assembly would mirror the Northern Ireland population in terms of gender, age, community background, political ideology, and so on. Second, the members of the citizens’ assembly would be tasked with deliberating on an issue or set of issues. That is, they would be provided with information, evidence and arguments on different sides of the issue(s). They would have the opportunity to ask questions of those presenting evidence and arguments, to individually reflect on the issue(s) at stake, and to discuss the evidence and arguments in small groups. Third, they would be asked with reaching a collective decision, usually by secret ballot. The decision may provide a purely consultative role, or else elected decision-makers may choose in advance to implement the citizens’ assembly decision, regardless of the decision itself.

3.3             Citizens’ assemblies offer a radical way of supplementing representative democracy, but the basic concept is far from original. The earliest forms of democracy involved randomly-selected citizens deliberating on political issues and reaching collective decisions, notably in Ancient Greece. More recently, citizens’ assemblies (and variants) have been used to assist elected governments on controversial and/or constitutional issues.

3.4             In the Canadian province of British Columbia, the provincial government established a citizens’ assembly on electoral reform in 2004. 160 randomly-selected citizens spent nearly a year (over a series of weekend sessions) learning about different electoral systems, listening to arguments on the positive and negative features of these systems, and carefully deliberating on the best way forward. The citizens’ assembly ultimately recommended replacing the province’s First Past The Post electoral system with the Single Transferable Vote. The recommendation was put to a referendum, in which it was endorsed by 57.7 percent of British Columbian voters. This fell short of a required threshold of 60 percent, and so the government did not implement the recommendation.

3.5             In 2012 the Irish government established a Constitutional Convention to consider a range of possible constitutional reforms. The membership of this body was a combination of (66) randomly-selected citizens and (33) elected representatives from across the political spectrum. Two of the Convention’s recommendations, to legalise same-sex marriage and reduce the age of eligibility for the presidency to 21, were put to a referendum in 2015. Irish voters endorsed the former recommendation and rejected the latter.

3.6             The Irish government subsequently established a Citizens’ Assembly in 2016, composed exclusively of randomly-selected citizens.[1] As with the earlier Constitutional Convention, the Citizens’ Assembly has been tasked with examining questions of constitutional and political significance, including fixed term parliaments, climate change, and abortion. On the latter, the Citizens’ Assembly submitted its official report to the Oireachtas (Irish Parliament). An Oireachtas Joint Committee has voted in support of a referendum on the matter; at the time of writing, the Irish government has yet to issue its formal response.

3.7             In the context of Northern Ireland, a citizens’ assembly may help to reduce the democratic deficit. The random selection of members may help to represent citizens who do not otherwise feel represented via the conventional electoral process. The intrinsic emphasis on deliberation and the weighing of arguments may help to reduce the partisan divisions associated with certain debates. Finally, the reaching of a collective decision by a citizens’ assembly could help inform any broader decision-making process.

4.0           Is the Support for a Citizens’ Assembly?

4.1             A citizens’ assembly can be justified as a highly legitimate decision-making body with reference to democratic theory. This does not, of course, mean that people would necessarily perceive it to be a legitimate way of making decisions. Therefore, in order to test the level of support for the idea, two surveys (with identical questions) were conducted. We asked the both the general public and MLAs what they thought of the idea of citizens making decisions on contentious issues such as flag display. By way of comparison, respondents are presented with a range of possible ways of making decisions, as reported in Table 1.[2]

Table 1: Public and MLA support for a deliberative democracy response to resolving contentious issues, compared to other types of responses [3]

4.2             The results in Table 1 suggest a high degree of support for the idea of a citizens’ assembly from the general public in Northern Ireland. 65 percent of respondents in a representative sample say a citizens’ assembly would be a good or very good idea as a way of deliberating contentious political issues. In contrast, however, just 17 percent of MLAs supported this kind of decision-making. Of the full range of options presented, a citizens’ assembly is the second most preferred option for the general public but the third-least preferred option for MLAs. Table 2 presents further differences in the attitudes between the public and MLAs towards citizen-based decision-making.

Table 2: Public and MLA support for aspects of deliberative democracy based decision-making

4.3             The survey evidence suggests that if a citizens’ assembly were established in Northern Ireland, MLAs would not support it making final decisions, but would be broadly supportive of it having the power to make recommendations. 61 percent of both MLAs and the public agree on this. To different extents, both MLAs and members of the public tend to agree that ordinary people would capable of deliberating on political issues and reaching a collective decision: just 17 percent of each sample think that ordinary people would be bad at making decisions if they were selected to serve on a citizens’ assembly. When it comes to the expected motivations of those selected to serve, less than a fifth of MLAs and the public think that members of a citizens’ assembly would just look after their own interests; a plurality in each sample think that members would try to come to a decision that is good for everyone in Northern Ireland.

4.4             However, there is a very large difference between the public and MLAs when they are asked a general question about the merits of elections, referendums and citizens’ assemblies as ways of making democratic decisions. Citizens’ assemblies attracted the greatest level of support from the general public, with support for referendums and elections not far behind. In contrast, there was a high degree of variation in responses from the MLA sample: 98 percent favoured elections and only 17 percent favoured citizens’ assemblies.

4.5             Additional qualitative open-ended questions were asked of MLAs in order to understand their views. A key theme emerging from the analysis of the responses is that while there is little support for a citizens’ assembly taking the final decision on a particular issue (an approach perceived as undemocratic), there is greater openness to the idea of a citizens’ assembly as a systematic way of hearing the reflective views of the people and generating potentially useful advice, or recommendations, to be considered by elected politicians.

4.6             In order to probe the views of the public further, we conducted a survey experiment on a representative sample of 2,000 citizens.[4] We know from Table 1 that people generally like the idea of a citizens’ assembly, but how robust is this level of support? Would people actually accept a decision taken by a citizens’ assembly if it was a decision with which they personally disagreed? In the experiment, we ask all respondents for their position on a salient issue of contention: Irish Language policy.  Those who say they oppose an Irish Language Act are presented with a hypothetical decision that one is to be introduced; those who support an Irish Language Act are told a decision has been taken whereby an Act will not be introduced. In other words, all respondents are presented with a decision with which they personally disagree. Respondents are then randomly assigned to one of six treatments, corresponding with six different ways of reaching this decision, and asked whether or not they accept the decision. Results are reported in Table 3.

Table 3: Percentage of respondents who say they would accept a (personally unfavourable) decision on an Irish Language Act, in each experimental group

4.7             In each of the six groups, a large majority of respondents say they would accept the hypothetical decision on Irish Language policy, even though they personally disagree with it. When respondents were told that a citizens’ assembly took the decision, 75 percent said they would accept it. In terms of statistical significance, this is not significantly higher or lower than the levels of decision acceptance across the other treatments.[5] The only exception is the referendum treatment: 94 percent of respondents in this group say they would accept the personally unfavourable decision on Irish Language policy; this level of acceptance is significantly higher than that across each of the other groups. People may not necessarily support the idea of holding a referendum on a contentious issue to the same extent as other options, including establishing a citizens’ assembly. However, in the event that one is held, its decisions are more likely to receive broad acceptance.

5.0           Further Research & Conclusion

5.1             Further work will be carried out as part of an ESRC-funded research project on attitudes towards Brexit in Northern Ireland.[6] A citizens’ assembly will be held in February 2018, allowing Northern Ireland citizens the space and relevant information to consider the challenging issue of Brexit and the border. Once they have become informed about the relevant issues and arguments – and reflected upon them – we will ask the citizens to put forward their own views. The exercise will focus on two main questions. First, how difficult would it be for Catholics/nationalists to accept a policed North/South border and what level of protest would they condone? Second, how difficult would it be for Protestants/unionists to accept a policed East/West border and what level of protest would they condone? We seek to provide policy-makers with evidence-based answers to these questions. This evidence will prove to be especially useful when policy makers are considering the likely implications of the post-Brexit border for peace and stability in Northern Ireland.

5.2             In the absence of a functioning Assembly and Executive at Stormont, it is arguably all the more imperative to listen to the views of ordinary citizens themselves. The greater application of deliberative democracy, through the use of citizens’ assemblies, offers one systematic way of achieving this goal. Survey evidence shows that both the public and MLAs are supportive of a citizens’ assembly playing a role in decision-making (provided it operates in an advisory capacity, for the latter). Experimental evidence finds that people’s support for the general idea of a citizens’ assembly is robust, in that people are just as likely to accept an unfavourable decision taken by a citizens’ assembly as one taken by more conventional forms of decision-making, including by elected politicians in inter-party talks.

29 December 2017

References

Ark  (2015) Northern Ireland Life and Times Survey [online]. Available at: http://www.ark.ac.uk/nilt/2015/ [Accessed on 29 December 2017].

Birrell, D. and Gormley-Heenan, C. (2015) Multi-Level Governance and Northern Ireland. Basingstoke: Palgrave Macmillan.

Coakley, J. (2008) ‘Ethnic Competition and the Logic of Party System Transformation’, European Journal of Political Research 47 (6), pp 766-793.

Dryzek, J. S. (2000) Deliberative Democracy and Beyond: Liberals, Critics, Contestations. Oxford: Oxford University Press.

Ferejohn, J. (2008) ‘Conclusion: The Citizens’ Assembly Model’. In: M. E. Warren and H. Pearse (eds.) Designing Deliberative Democracy: The British Columbia Citizens’ Assembly. Cambridge: Cambridge University Press, pp 192-213.

Garry, J. (2016) ‘Deliberative Democracy in Northern Ireland’ [online]. Available at: http://www.niassembly.gov.uk/globalassets/documents/raise/knowledge_exchange/briefing_papers/series5/garry-briefing.pdf [Accessed on 29 December 2017].

Schwartz, A. (2014) ‘Petitions of Concern’ [online]. Available at: http://www.niassembly.gov.uk/globalassets/Documents/RaISe/knowledge_exchange/briefing_papers/series3/schwartz200314.pdf [Accessed on 21 May 2017].

Smith, G. (2009) Democratic Innovations: Designing Institutions for Citizen Participation. Cambridge: Cambridge University Press.

Wilson, R. and Nolan, P. (2016) Beyond Voting: Deliberative Democracy in Northern Ireland. Belfast: Building Change Trust.