Irish policy making and planning for a possible future united Ireland

Professor Jennifer Todd, Geary Institute, UCD

This paper is available to download here.

Introduction: Democracy, disengagement and disjunctures

Policy planning for a future united Ireland is foundational, difficult and dangerous. At basis, it involves the Irish government providing a model of a united Ireland to be put to all the people of the island in concurrent referendums in each jurisdiction. If planners get it wrong – if for example there is a yes vote for a model of united Ireland that no one really wants and which generates widespread disillusion – the consequences will be felt for decades.

It is even more dangerous because of global public disengagement from conventional forms of democracy, from expertise and from ‘change leaders’. The results of recent Irish referendums (see Bray, 2024) show starkly the need for insight into ordinary people’s views, their priorities, what influences them, how they are reasoning, and how far they will move in their views. And interlinking political and public views is particularly difficult in this case. Modelling a future  united Ireland is a foundational task that potentially touches on all aspects of life, and that has deep, often unspoken, emotional resonances which differ in each jurisdiction and there is already a disconnect between constitutional and everyday language.

In Northern Ireland, such disjunctures have had a major impact: for example the loyalist flags protest of 2012 (see Nolan et al., 2014) – which led to a polarising political dynamic and made the impact of Brexit much worse – occurred just at a time when politicians and commentators believed that power-sharing was at last stabilised.

In this paper, based on my recent research (2020-2024) with Prof Joanne McEvoy of University of Aberdeen, I outline a pathway to tackle these problems of public disengagement and public/political disjunctures.

Foundational policy problems

The Irish government has the central role in putting forward a model of a united Ireland in a future referendum in each jurisdiction. This is a particularly unusual kind of foundational planning and policy task. Irish policy makers will have to outline the form of a new state (a future united Ireland), whose remit will cover the whole island and which must be acceptable not just to the citizens in the Irish state but also to those in Northern Ireland. They must have this model available at a time chosen by another state – the initiation of a referendum in Northern Ireland is the defined task of the Secretary of State for Northern Ireland (see Good Friday Agreement, Constitutional Issues, Annex A), but the same question and model must be available for the referendum in Northern Ireland and the concurrent referendum in the Republic of Ireland.

There is some discretion on just how detailed the model of a united Ireland put forward by the Irish government must be. There is a strong argument that a ‘process’ model is preferable whereby many decisions are made in a transition period after a positive vote for unity in each jurisdiction. But most people in each jurisdiction want clarity on the type of united Ireland on offer before the vote.  

The task for the Irish state – and in particular for the politicians and civil servants within it – is particularly difficult for at least four reasons:

First, the Irish government has responsibility without power. The Irish state is the only body that can credibly and legally design a model to go forward in a referendum and it will have to implement the decision. But some of the key decision-making citizens are in another state. Information has to be gained from both jurisdictions, even though the Irish government has authority over only one. Even the triggering of the referendum is in the British, not the Irish, government’s power.

Second, the issue is exceptionally and existentially difficult. It touches on identity and loyalty as well as interest. It speaks to foundational assumptions about the political and social world. And it does so for different populations who have radically opposed perspectives and different paradigms of understanding of politics and society. Moreover, these differing perspectives were at the centre of group conflict in Northern Ireland. To open these opposing paradigms to discussion is necessary, because they frame judgements and even emotions. But this also threatens to retrigger group solidarity and conflict.

Third, the language of constitutional debate has provoked widespread disengagement, and silence, rather than reflection. Around half of the population on the island wants to distance from traditional nationalist/unionist debates[1].

According to the Northern Ireland Life and Times Survey (NILT), over half the Northern Ireland population distances from either nationalism vs unionism, or British vs Irish, or Protestant vs Catholic. Recent research shows that a large segment of the Southern population – notwithstanding their definite views about a United Ireland – shares very similar political views and values to those of Northern ‘others’.

The issues are not only emotive but those emotions have been silenced and thus not dealt with. Some try to avoid thinking about a possible future united Ireland because the issue seems to ensnare them in perspectives they do not much like but are unable to find a way out of. Their views are thus likely to swing widely, seemingly unpredictably.  

A recent focus group illustrates this volatility (see Todd, McEvoy, and Doyle, 2023). The participants were chosen as representative of the relatively small proportion of the Southern population without a defined constitutional preference. And yet, once issues about a future possible united Ireland were raised, they responded intuitively, emotionally and forcefully – when asked about changing the flag and anthem, they answered ‘No, no, no, no’. When asked about joining the Commonwealth, they said it was like ‘spitting on your ancestors’ graves for everything that they fought for’. Once they heard their own conversation, they pulled backed: at the end of the 90 minute focus group, participants were saying ‘sure that’s never going to work’, ‘We have to be more openminded, ready for some change as well’.. It was clear that they held contradictory views, and swung between them. Was this simply because they didn’t have a defined constitutional preference? It is implausible that those with even more definite initial intuitions also have more coherent and thought-out views. But the data does not let us say for sure.

Fourth, this is a situation where people do not have accurate information about conditions in the other jurisdiction (see Todd, McEvoy, and Doyle, 2023). In part, the information has not been available until the recent period. In part, some information is still not available because factors are measured differently, or not at all, in each jurisdiction. In part, accurate information cannot yet be attained – for example, the cost of unity to the Irish state depends in part on negotiations with the UK, which have not yet happened. In this situation, misinformation and partial information is likely to be weaponised.

However, there is one positive factor. Although the planning task is very difficult, it is not yet immediate. Unlike the German government in 1989, the Irish government has a number of years – one might guess a decade – to accomplish it before a referendum happens.

Methodological issues

Reliable surveys are essential for policy (see RIA, 2024). Yet while surveys give a snapshot of opinion in time, they do not show respondents’ strength of opinion, priorities, willingness to change their minds, or how their beliefs and preferences are interconnected, or how far they are caught in perspectives they are unhappy with. They allow us to count preferences, not to assess unease with the questions.

Of course survey experiments can be developed to deal with these issues, but only some of them at any time, and we have to already know what people may be worried or uneasy about before we start to measure this in surveys. The problem here is that we don’t know.

Absent such analysis, we are left with a map of people’s overt preferences today. They are, as expected, divided. Unionists differ from nationalists, and Southern nationalists from Northern nationalists. Learning from this, analysts attempt to find the highest common factor in the radically opposed views, the options that would be least bad, and that would maximise losers’ consent in a future referendum. Thus, for example, it has been argued that a devolved Northern Ireland under Irish sovereignty is the best option because it is least disturbing to Northern unionists. It also would have the benefit of minimally disturbing institutions and practices in the Republic. But there is a potential problem: this is a model of a united Ireland that no one really wants, one that is uninspiring as well as costly. It is very likely to lead to later disillusion. 

Are there ways to anticipate where people’s views might move in the future? Ways to see how far they might re-evaluate their immediate intuitive responses and come to converge on the values and shape of a more inspiring society?

Deliberation gives a possibility of assessing this. But the problems outlined above – inchoate public unease, whose extent and patterns are not yet identified, over a large set of issues that are intrinsically interconnected, that tap into norms as well as interests, that are highly emotive, with deeply opposed paradigmatic perspectives on them – demands a form of deliberation that aims to find points of normative commonality and convergence on problems and process, rather than to arrive at an informed choice between pre-given options. It requires deliberation, centred on narrative as much as on argumentation and aimed at understanding as much as informed choice (Curato, 2019). Moreover the deliberative net has to be cast very widely, going beyond representative sampling to look at many diverse clusters of the population whose perspectives, unease, and potential for change may be distinctive, prioritising access rather than representativeness, and inclusion of different perspectives not just different social categories. And, since we do not yet know the issues at stake for participants, the deliberative method has to allow them – to the maximum degree possible – to set their own agenda of discussion, define their own priorities, and interrogate one another. 

This mode of deliberation is particularly necessary at an ‘upstream’ phase in the process when the questions and issues involved still need to be teased out. Moreover, the widening of the engaged public and the participatory strategy means that the educative effect of this sort of deliberation is maximised. 

Research findings

Our research over the last 4 years has explored how the large numbers of people disengaged from constitutional discussion thought about that discussion. Our research methods included interviews and focus groups (in the COVID-19 years online) with about 100 people North and South, accessed through community organisations. We spoke primarily with women, migrants and youth, because there was already significant research that shows that they are disengaged from constitutional discussion. We also spoke to politicians North and South, from almost every political party.[2] The discussions were loosely semi-structured around the current constitutional discussions, respondents’ engagement with this and perception of it, and their own sense of priorities in the debate, their involvement and perception, and their interest and what they thought were priorities. In 2022, we held four in-person deliberative cafes – small informal local deliberation of about 3 hours in a local hotel – with over 60 participants. In all of this research, our participants were transversally defined – for example we accessed them as border women or as gender activists, not as unionists or nationalists although they might also be that. They were, for the most part, disengaged from constitutional discussion. They were very diverse, from all parts of the island, all religions and none, and with diverse political perspectives. Our method was to raise the broad issue of constitutional change and North-South relations and ask for their perceptions and priorities around the issues, so as to encourage their participation in defining the issues in debate.

Our findings surprised us. First, it is sometimes argued that greater democratic participation increases the likelihood of polarisation especially in situations of conflict (e.g. Horowitz, 2021). Instead, we found very substantive convergence amongst the diverse participants on meta-issues of process and values. They agreed that there should be conversations – not debates – about the issues but starting from experiential ‘organic’ basis, not from ideology. They agreed that as many people as possible should be involved. They agreed that bread-and-butter socio-economic and rights’ issues were priorities, much more than the precise issues of institutional design. They agreed, for the most part, that time should be taken to get the picture right. Participation of these transversally defined groups increased convergence, and provided confidence that participatory discussion, properly organised, could be constructive. Transversal access encourages such convergence (see McEvoy and Todd, 2023).       

Among the convergent views, women in particular argued for ‘conversation’ not debate, and, in focus groups, discussion often wound its way back to earlier interventions, trying to hear, understand and integrate different experiences and perspectives rather than win debates. The women saw this as beginning with ‘pre-conversations’ which helped understand other perspectives and develop shared values, even if they have different preferences. Participants saw it as something that could take place locally – in local ‘knit and natter’ groups – networked together.

Second, we were surprised that amongst the obstacles to participation, the constitutional discourse itself was most emphasised (Todd and McEvoy, 2024). It was seen as ideological, rather than based on experience, as provoking knee-jerk responses rather than engaging in thought, and as non-organic. The very terms ‘constitutional change’ and ‘Irish unity’ were criticised. This was a very immediate intuitive response by many – one gender activist said she felt that she had ‘rocked up to the wrong meeting’. On the other hand, the vast majority of our participants were concerned about the shape of a future society, interested in how others saw it and wanted to continue the discussion, but on terms that spoke to real experience.

Third, our participants emphasised issues of practical life – bread-and-butter issues and rights, not primarily identity and institutions (McEvoy, Todd, and Walsh, 2022). It wasn’t just the financial cost of unity, but the form of life that would result, and the changes to their lives, and those of their families. They wanted improvement – not just the end of barriers of religion in Northern Ireland, but the end of barriers of class in Dublin. In this sense, they wanted constitutional change that would be inspiring.

Arising out of these findings, we devised small scale deliberative cafés to see if, and how, open-ended deliberation that began with issues of shared North-South concern (for example, three cafés discussed cross-border healthcare) would scale up and out to political and constitutional issues (McEvoy and Todd, 2024). Each 3 hour café was divided into three 50-minute parts, interspersed with information sheets, and refreshments – there was an initial sharing of experience on the issue, a second session on policy implications, leading into a question/answer session with a zoomed-in expert, and a final shorter session on political and constitutional implications.

We found that the narrative discussions sharing experience led to clear and credible collective definitions of the shared problems – a step change from individual experience to convergent collective definition. This provides one criterion for judgement on policy initiatives and constitutional change. To put it crudely, any model of a united Ireland that did not alleviate present problems of cross-border healthcare would not be good enough.

On this basis, participants’ policy discussion was intense and exhaustive. The participants wanted to hear the experts’ views and pressed for more evidence and argument. They were disappointed when no clear policy recommendations emerged. Our participants would clearly have continued discussion longer if it were possible in this way to help create better policy instruments.

But our participants did not scale up to constitutional debate. Some said that constitutional change wouldn’t make any difference. They were frustrated that they had no answers to their policy questions. They were frustrated that their deliberation was unlikely to have policy relevance.

Towards an ecology of deliberation

In order to build on this research and improve it, our present research sequences and alternates café sessions. Schematically, we envisage three x three hour sessions, each of which has Northern and Southern participants:

  • An initial three hour narrative deliberative session which shared experiences and aimed collectively to define the problems to be resolved.
  • An interactive deliberative session with local politicians and policy makers along with participants, with the question of how to remedy the problems. This builds accountability – the absence of which so frustrated our participants in earlier research – into the process.
  • A subsequent ‘constitutional’ session where deliberation is structured around the construction of a constitutional model (either united Ireland or united Kingdom) that would help alleviate the problems.

Our research focusses on the local level and aims to provide a replicable model of cross-local North-South deliberation that can be rolled out across the island.

There are alternative pathways, and with greater resources, an alternation of smaller scale and larger scale deliberation would be possible, with translation of the results of one session to the next.

The concept of a deliberative ecology (see Mendonça et al., 2024) is particularly appropriate for this sort of North-South deliberation in a changing, weakly institutionalised, island-wide arena where linkages vary over time and are in process of being built, and where there is not one social system but a plurality of interlocking ones, open at the edges, and open to further transformation. A deliberative ecology would (like the deliberative system approach outlined by Suiter (2021) but on a North-South basis) combine and link small local deliberative arenas with larger citizens’ assemblies and meso-level sectoral meetings. It would focus on changing local, and cross-local priorities. And it would accept that the ‘public’ whose decisions were to be anticipated and informed by deliberation is itself plural, in an important sense still to be formed across the island(s).

This would provide a resource to help policy makers answer the extremely difficult questions they will face in the coming years as a referendum on constitutional change comes closer. 


Bray, J. (2024). Family and care referendums comprehensively defeated. The Irish Times, 9 March:

Curato, N. (2019). Democracy in a time of misery: From spectacular tragedies to deliberative action. Oxford University Press.

Horowitz, D.L. (2021). Constitutional Processes and Democratic Commitment. New Haven, CT: Yale University Press.

McEvoy, J., & Todd, J. (2023). Constitutional inclusion in divided societies: Conceptual choices, practical dilemmas and the contribution of the grassroots in Northern Ireland and the Republic of Ireland. Cooperation and Conflict58(3), 393-413.

McEvoy, J., & Todd, J. (2024). Scaling up from everyday concerns to territorial politics and constitutional debate: deliberation among women in the Irish border area. Territory, Politics, Governance, 1–19.

McEvoy, J., Todd, J., & Walsh, D. (2022). Participatory Constitutionalism and the Agenda for Change: Socio-economic Issues in Irish Constitutional Debates. Irish Studies in International Affairs 33(2), 140-171.

Mendonça, R.F., Veloso, L.H.N., Magalhães, B.D., Motta, F.M. (2024). Deliberative ecologies: a relational critique of deliberative systems. European Political Science Review. Published online 2024:1-18. doi:10.1017/S1755773923000358

Nolan P, Bryan D, Dwyer C, et al. (2014). The Flag Dispute: Anatomy of a Protest. Belfast: Queen’s University.

Royal Irish Academy (RIA). (2024). North and South surveys – data now published. Royal Irish Academy News, 13 February:

Suiter, J. (2021). A Modest Proposal: Building a Deliberative System in Northern Ireland. Irish Studies in International Affairs, 32(2), Analysing and Researching Ireland, North and South 2021, 247-270.

Todd, J., & McEvoy, J. (2024). Obstacles to constitutional participation: Lessons from diverse voices in post-Brexit Northern Ireland and the Republic of Ireland. The British Journal of Politics and International Relations26(1), 170-186.

Todd, J., McEvoy, J., & Doyle, J. (2023). Time for Deliberation, not Decision, on the Shape of a New United Ireland: Evidence from the ARINS Survey Focus Groups. Irish Studies in International Affairs 34(2), 122-148.

[1] According to the Northern Irish Life and Times (NILT) survey, over half the population are either neither nationalist nor unionist, or neither solely British nor solely Irish, or no religion: much fewer distance from the blocs on all three dimensions. This was analysed using the NILT variables: Community relations, NINATID; political attitudes, UNINATID; Background, RELIGION

[2] The DUP did not agree to talk with us.