Ireland’s Citizens’ Assembly on Abortion as a Model for Democratic Change?: Reflections on Hope, Hype and the Practical Challenges of Sortition


Eoin Carolan

University College Dublin

Ireland’s recent citizens’ assembly and resulting referendum on abortion rights highlights the history-changing impact citizens’ assemblies can have…. Not only did the Assembly spark a long-awaited public discussion on abortion rights, it helped set the stage for last month’s national referendum on May 25th. 

Ireland’s Citizens Assembly is a progressive piece of democratic apparatus and its success should inspire democracies across the globe to adopt it, or something similar…. The success of the Irish case should be seen as a model for both the UK and the rest of the democratic world.

… Ireland is leading the way in democratic innovation—empowering its citizens to advise the parliament on future direction.


Ireland seems to be becoming the poster-child for progressive constitutional politics. The liberal-leaning results of successive referenda on marriage equality, abortion, and blasphemy ran counter to widely-held international conceptions of Irish society as “until recently …. deeply conservative”. The fact that these votes were preceded by deliberative assemblies, and (loosely) matched their recommendations has attracted particular interest, verging – in some accounts –on excitement. At a time of real concerns about polarisation, partisanship and deteriorating public discourse, has Ireland discovered the secret to a reasoned and deliberative process of constitutional change?

Not really.

Tempting as the deliberation-changes-conservative-constitutional-culture narrative may be, it presents a very simplified account of the Irish experience, especially with regard to the operation and impact of the two official deliberative assemblies held since 2012: the Constitutional Convention and, more recently, Citizens’ Assembly. The academic enthusiasm for these assemblies is not unusual: Canadian commentary after the British Columbia experiments was similarly noted for its celebratory, at times euphoric, tone. The problem, as I’ve argued elsewhere, is that this enthusiasm leads to the experience being oversold – and that this can, in turn, blind advocates to some of the trade-offs, challenges and risks that a ‘Citizens’ Assembly’ involves.

Every experiment in civic deliberation is, of course, different. In keeping with the Symposium theme, the aim here is only to point to certain aspects of the Assembly that may merit further reflection as counterpoints to more optimistic portrayals of the process.

Reviewing the international commentary on the Assembly process, three aspects seem to be singled out for the greatest praise: its representative ‘grass-roots’ composition; its deliberative procedures; and its effect on public opinion. In each case, the reality may be more complex than some of these accounts allow.

Looking first at composition, one of the more prominent rationales for mini-publics is that they comprise a randomly-selected but demographically representative sample of ordinary citizens. This representative character is an important element of their claim to legitimacy:

Suppose a microcosm is a good representative sample and that it engages in a substantive and balanced deliberation, what is accomplished? A representation of what the people would think under good conditions, a representation embodying both political equality and deliberation. (Fishkin, When the People Speak)

The Citizens Assembly, however, encountered significant controversy when it emerged that an employee of the polling company charged with recruiting members had done so by contacting persons suggested to the recruiter as potential candidates. An internal audit suggested that this was an isolated incident (although similar concerns had been articulated about the Constitutional Convention selection process after a married couple and next-door neighbours were amongst those ‘randomly’ recruited).

The larger significance of these incidents may be what they suggest about the representative nature of the eventual body: namely, that there is likely to be substantial over-representation of people with atypical levels of interest in political or constitutional issues.  The audit conducted into the Citizens Assembly is clear that the polling company found it extremely difficult to identify people willing to participate in the Assembly – so much so that they went back to the Assembly on a number of occasions with a request to introduce recruitment incentives that “would potentially allow us to recruit a somewhat more representative sample of people, rather than just those who are interested in political debate and discourse”. This problem may have been particularly pronounced in Ireland given the absence of any financial payments to Assembly members. Nonetheless, the potential for such bodies to be dominated by the politically enthusiastic does pose challenges to their presumptively representative character – especially when it is considered that those with strong pre-existing interest in political issues could be more likely to have strong pre-formed views.

This suggests a need for some caution about presuming representative legitimacy from demographic metrics alone. At the very least, it is important that consideration be given to how to ensure that members are genuinely representative of the ‘ordinary citizens’ on whose ideal so much of the justification for deliberative assembles is based.

Turning second to the deliberative process itself, the Assembly served yet again as a reminder that the design, structure and operation of a process – no matter how deliberatively-oriented – is not neutral or value free. Specific issues were raised about the clarity of agenda and degree of understanding of members and are discussed elsewhere in the Symposium by David Kenny. At a more general level, decisions about how to frame an issue; in what order it should be approached; from what disciplinary perspectives; using what kinds of ‘evidence’ and ‘experts’; and subject to what limitations; shape how the process develops. These decisions inevitably privilege particular kinds of input and require difficult trade-offs between abstract deliberative goods like expertise, independence, experience, impartiality, participation, openness and evidential scrutiny.

In fairness to the Citizens’ Assembly, these trade-offs appear – by contrast to the previous Convention – at least to have been consciously made. The process applied was one which tended to favour formal legal techniques and perspectives, reflecting presumably the familiarity with and expertise of the Chair (a former Supreme Court judge) in court procedures. That also meant, however, that the process was open to the criticism that the agenda and material provided were heavily legalistic in emphasis and tone (as discussed in more detail elsewhere in the Symposium by Mairead Enright).

Again, this highlights the need for some circumspection about the ‘deliberative’ quality of sortition. Attractive as the deliberative ideal may be in the abstract, practical implementation will likely produce imbalances in access, power and influence.

The third and final point of reflection concerns the ostensibly positive effects of the Assembly. One suggestion in the literature is that the process of having a deliberative mini-public provides an opportunity to inform the wider public on a specific issue. Another is that the recommendations of a deliberative mini-public will enjoy presumptive popular legitimacy as the reasoned outcome of a process that is fair, deliberative and representative.

Again, there is reason to question whether the Irish experience fully supports such claims.

First of all, it is unclear to what extent the wider public paid particular attention to the nuances of the discussions. The desultory viewing figures for the material presented at the Assembly call into question talk about public education from the deliberative proceedings themselves.

More pertinently, cause-and-effect assumptions – let alone hyperbolic claims that “[a] referendum was scarcely imaginable when the idea of an assembly was first mooted in 2016” –  are doubtful given that opinion polls in the years and months prior to the Assembly showed comfortable majorities for both a repeal of the 8th Amendment of the Constitution (which had restricted the possibility of making abortion legally available) and increased access to abortion (albeit with varying views expressed on the circumstances in which access should be permitted in its aftermath). This kind of analysis also risks underplaying the active work of repeal campaigners over many years prior to the Assembly and significant wider changes in Irish society over the past 30 years.

There is a sense that the ‘dramatic’ results of recent referenda are more to do with the delayed acknowledgment by an innately risk-averse political class of changes in majority social attitudes that happened – or began to happen – 20 or 25 years ago.

Indeed, the most significant effect of both the Convention and Assembly seems to have been – as argued elsewhere in the Symposium by Fiona de Londras – on the political class rather than the wider public. The process and results seems in some cases to have persuaded politicians to alter their views as to what was appropriate (or politically saleable). But such a view of the Assembly as a high-profile focus group is quite far from the more idealised visions apparent in some commentary.

It may also be important in this regard to distinguish between the effect on the decision to advocate or offer constitutional change; and the effect on the details of the change advocated. With both marriage equality and abortion, there were other political factors linked to the composition of coalition governments and parliament that seem likely to have had more impact on the decision to hold a referendum – something supported by the fact that the vast majority of recommendations from the Convention did not lead to the recommended referendums.

Finally, the Assembly experience also complicates any suggestion that the characteristics of the process confer presumptive validity on the outcome. There is an analogy here with the familiar legal assumption that the fairness of a process encourages disappointed parties to accept the outcome. The Asssembly’s recommendation for a liberalisation of Ireland’s laws on abortion was rejected by pro-retention groups who largely dismissed the process as unrepresentative, undemocratic and illegitimate.

There were, however, echoes of these complaints amongst some on the other side of the campaign when it (briefly) appeared that the Assembly might recommend against reform. In the second of its sequential votes, the majority voted to replace rather than repeal the 8th Amendment. ‘Repeal’ had been the theme of a long-running campaign against the 8th Amendment. While it is always questionable to read too much into Twitter comments, it was striking that complaints about the process quickly emerged amongst users (and some Assembly members) disappointed with the vote to ‘replace rather than repeal’. Concerns were expressed about alleged lobbying by members; about the clarity of the process; about the expertise, understanding or representativeness of members; about the involvement of the wrong experts; and about the perceived ‘undue influence’ of a paper from a senior counsel about the legal implications of various options.

Perhaps the telling aspect of these criticisms was that they were directed to aspects of the process that (in deliberative theory) would undermineits legitimacy as a whole. As such, they raise a question mark about the extent to which the perceived legitimacy or success of the Assembly is based not on the process but on its outcome.

Eoin Carolan is a Professor at University College Dublin (UCD) where he is Director of the UCD Centre for Constitutional Studies.

Suggested Citation: Eoin Carolan, ‘Ireland’s Citizens’ Assembly on Abortion as a Model for Democratic Change?: Reflections on Hope, Hype and the Practical Challenges of Sortition’ IACL-AIDC Blog (28 November 2018)