John Coakley [a]
In presenting what was billed as a major government initiative in Dublin Castle on 22 October 2020, Taoiseach Micheál Martin appeared to signal a fundamental policy shift on Northern Ireland on the part of the coalition government and of his own Fianna Fáil party. Rejecting Sinn Féin demands for a “border poll” to determine Northern Ireland’s future constitutional status, he indicated that this was off the agenda for at least the next five years. He opted instead for a “shared island” formula: promoting North-South cooperation rather than debating Irish unity.
This side-lining of decision making by referendum raises important questions about the meaning of the Good Friday agreement of 1998, but also about the interplay between public policy and public opinion. A central plank in the 1998 agreement was its elaboration of a pathway to Irish unity that left the ultimate decision to the people of Northern Ireland and the Republic. But what do we know about their opinions, about the likelihood that they would indeed support Irish unity, and about the manner in which these attitudes might have changed over time? Recent data from the biennial European Social Survey (ESS; fieldwork mainly early 2019) and from the annual Northern Ireland Life and Times survey (NILT; fieldwork mainly late 2019) shed valuable light on this question. [b]
For 2019, the headline findings on the “unity” question were that 54% in the Republic supported Irish unity as a long-term goal, as did 25% in Northern Ireland (unless otherwise specified, “don’t knows” and those refusing to answer are excluded in this note). For what the statistic is worth, weighting the data by the population of the two jurisdictions indicates that in 2019 only 46% of adults on the island of Ireland supported unity.
While these figures suggest that Irish unity is unlikely to be endorsed by referendum in the two jurisdictions in the near future, they raise important questions about the stability of public opinion in this area, about the character of support for and opposition to Irish unity, and about the extent to which the people’s views are likely to matter in any case.
Although it is difficult to track attitudes to Irish unity in the Republic over the years due to the shifting salience of the issue and wide variation in question wording, some comparison of the position over time is possible. Figure 1 reports responses to the question on long-term attitudes towards unity, comparing the contemporary position (2019) with that in the first ESS survey in 2003, and including corresponding data for Northern Ireland at the same two points in time.
An immediate conclusion to be drawn from figure 1 is that support for unity in the Republic appears to have dropped significantly, from 70% in 2003—a level close to which it was placed at the time in other surveys—to bare majority support. Opinion in Northern Ireland remained stable, with big majorities in favour of the Union.
The bottom part of figure 1 breaks down the Northern Ireland data by religious community background and helps to explain the pattern of support for constitutional options there: it is made up of two sharply contrasting components, a Protestant community profoundly committed to the Union and a Catholic community more attracted to Irish unity.
There is a striking absence of symmetry between the two communities: Protestant support for the Union is much more robust than Catholic support for Irish unity. This arises in part for technical reasons. In the more recent data, the “Union-unity” dichotomy has been altered significantly: the “Union” side has been further divided between those preferring this to be combined with direct rule from London and those preferring it to be combined with devolved government in Northern Ireland, allowing a broader sweep of Union supporters and tending to reduce levels of support for Irish unity. In fact, the relatively strong level support for Irish unity among Catholics in Northern Ireland (58% in 2003, 54% in 2019) does not represent the position in the early years of the twenty-first century: reported Catholic support for unity in NILT surveys dropped sharply in 2007 following the switch from two to three response options, and remained low (30-40%) from then until the “Brexit bounce” in 2019.
The relationship between attitudes towards unity and selected social characteristics is reported in figure 2, where the bars represent the proportions supporting unity. As the top part of the figure shows, there are some aspects of social behaviour that might be expected to have an impact on traditional nationalist attitudes that appear to have no bearing on preferences regarding Irish unity. Religious commitment, as measured by frequency of church attendance, is an example: regular churchgoers are not significantly more likely to support Irish unity than the religiously indifferent. Differences between men and women in this respect are similarly insignificant.
Variables related to social status, however, have a big impact: those linked to the wealthier half of the population as measured by income are much less likely than the rest to support Irish unity (47% of the former but 57% of the latter support it). Educational attainment has a similar impact: those with a secondary school leaving certificate or higher qualification are a good deal less likely to support Irish unity than those with lower formal qualifications.
The interplay with age is particularly intriguing: younger people (defined here as under-35s) are much more likely to support Irish unity than the middle-aged (35-54 years old), but support picks up a little again among the older age groups (55 or more). This may well reflect life-style effects during people’s politically formative years, with a young generation unfamiliar with Northern Ireland other than as a moderately peaceful place, an older generation who lived through the conflict but retained traditional aspirations, and an intermediate generation who experienced a conflict that tested traditional nationalist values.
It would also be reasonable to expect a strong association between attitudes towards Irish unity and party support, since the origins of the southern parties lay historically in bitter divisions over the relationship with Britain and partition, and the roots of Northern Ireland’s parties are even more obviously located in the conflict.
Figure 3 describes the constitutional preferences of party supporters separately for Northern Ireland and the Republic. It shows sizeable differences between those sympathetic to the traditionally dominant parties in the Republic. Fianna Fáil supporters are much more likely to favour unity (58%) than Fine Gael or Labour supporters (43% in each case). Not surprisingly, Sinn Féin supporters are most enthusiastic of all on unity (71%), though almost a quarter of Sinn Féin supporters would prefer to see Northern Ireland remain in the UK, in apparently direct conflict with party policy. Electoral research has, however, shown that voters support parties for a range of reasons, by no means always related to policy or ideology, so apparent anomalies of this kind are common.
The bottom part of figure 3 shows a very striking relationship between position on the constitutional question and supporters of the five main parties in Northern Ireland. Sinn Féin supporters are more strongly in favour of unity than in the Republic (81%), while SDLP supporters appear to follow the same pattern as southern respondents; in particular, their profile resembles that of Fianna Fáil. Supporters of the two main unionist parties remain rock-solid in support of the Union (97% in the case of the DUP, and 93% for the Ulster Unionists), with the Alliance Party, reflecting its cross-communal composition, spanning a wide variety of preferences.
There is little in these data, then, to suggest that Northern Ireland is on the verge of a pro-unity majority, and support for unity in the Republic is far from overwhelming. But we need to be careful about prediction: there are other polls, other questions and other forces of which account must be taken.
First, the NILT survey series relied on in this note is not the only dataset that addresses the constitutional question. A number of recent polls in Northern Ireland (notably those conducted by Lord Ashcroft and by Lucid Talk) have shown supporters of Irish unity vying with supporters of the union in a finely matched battle. Use of a different measurement instrument may, then, then, reveal a higher level of support for unity than that reported above.
Second, the thrust of the question used here is about preferences as regards the “long-term policy for Northern Ireland”—an enquiry that leaves the timing of any constitutional change to the relatively remote future. One might expect support for a radical constitutional change such as Irish unity to drop off a little as the issue becomes more concrete. However, when asked how they would vote if there was a referendum tomorrow on the future of Northern Ireland the proportions supporting unity actually increase to 61% and 29% respectively in the Republic and Northern Ireland (to recall, the proportions supporting unity as a long-term policy were 54% and 25%). Care is needed, then, in drawing inferences about probable electoral behaviour from patterns of long-term attitudes.
Third, the great unknown is the impact of new forces set in train by the Brexit process: the new quadrilateral relationship between the EU, the UK, the Republic of Ireland and Northern Ireland, the future constitutional structure of the UK itself (including prospective Scottish independence), and the detailed outworking of the Ireland-Northern Ireland Protocol. The new geopolitical realities that will result from this upheaval are likely to create a framework for the relationship between the Republic and Northern Ireland that supersedes both fundamental decision making by referendum and plans for a “shared island”. The prospects for unity of the island may well be determined less by Irish public opinion than by perceived British interests and, in particular, European geopolitical realities.
[a] Geary Institute for Public Policy, University College Dublin and School of History, Philosophy and Politics, Queens University Belfast.
[b] For background information and data, see www.europeansocialsurvey.org/ and www.ark.ac.uk/nilt/