Analysis of the abortion (8th amendment) referendum, May 2018

The long-anticipated referendum on repealing the 8th amendment, known by proponents and opponents alike at the time of its initial adoption as the ‘pro-life amendment’, has delivered a result widely and understandably described as ‘seismic’, ‘historic’ and ‘momentous’. In a high turnout (64%, second among standalone referendums only to the 1972 vote on EC entry and similar to that at several recent general elections) over 66% of voters, amounting to 42% of the electorate, voted to remove the 8th amendment. Barring unforeseen developments, this sixth referendum on abortion (and of course the issue has been an undercurrent in some other referendums, particularly those on EU integration, as well) is likely to be the last on the subject, for which many will be thankful.


The Yes campaign was sometimes criticised during the campaign for being slow to get off the ground, but the outcome, a Yes vote larger than anyone had predicted, entitles it to claim retrospective vindication of its strategy. The government’s decision to give details of the legislation that it would bring to parliament if the 8th were repealed could have been high-risk, as it was plausible to imagine that some voters who were open to a degree of liberalisation might well baulk at the prospect of abortion without restriction in the first 12 weeks of pregnancy. With the benefit of hindsight, though, this was the correct strategy, since in the absence of explicit proposals the repeal side generally would have been accused of concealing sinister intentions and a Yes vote would not have constituted a mandate for any specific legislative proposal (rather like the endless debate in the UK since June 2016 as to exactly what the voters had in mind when they voted Leave).


The No side, in contrast, seemed to be fighting a campaign addressed to a different era – an era in which abortion was illegal and inaccessible everywhere and in which the issue was, therefore, whether Irish women should acquire the possibility of access to the procedure. The main pragmatic argument put forward by the government – that most or all of the abortions that would become legal if the amendment were passed would happen anyway given that Irish women are having abortions in England or in their own homes by administering pills without medical support – was simply not even addressed by the pro-life campaign. Issue ownership models of campaigning suggest that sometimes it makes sense to emphasise one’s own strongest issues and to do your best to ignore those of the other side, so that the campaign is fought on the terrain of your choosing, but that certainly did not work for the retain side on this occasion. Its attempt to frame the issue as, in effect, a referendum on the rights of the unborn child simply did not succeed.


One of the most telling posters was one from Love Both reading ‘Repeal Means Unrestricted Abortion up to 12 Weeks – Vote NO’. The statement (without the last two words) was a simple statement of fact, and the Yes side might well have used the same poster, simply changing the last word on the poster to ‘YES’. The fact that what many Yes voters saw as a positive feature of the repeal proposal was perceived by the No side as being so egregious that it self-evidently implied the need to retain the 8th illustrates the seemingly irreconcilable clash of values between the two sides.



Almost needless to say, the vote on 25 May ran very much along established liberal–conservative lines. The RTE Behaviour and Attitudes poll exit displayed what has become a very familiar pattern. Support for repeal was strongly related to age (88% of the youngest cohort supported it but only 41% of the oldest), related to class (highest among middle-class respondents, lowest among farmers), highest in Dublin and lowest in Connacht–Ulster. Religiosity was a strong determinant, with 61% of No voters (but only 16% of Yes voters) attending a service at least once a week, and 69% of Yes voters (but only 27% of No voters) attending a few times a year or less.


One relationship that has changed over the years is that between gender and liberalism. In 1983, women were less liberal than men on the issue (75% of women compared with 62% of men voted to insert the ‘pro-life’ amendment). That remained the situation in the 1992 abortion referendums, when men were more in favour of asserting the constitutional right to travel than women were and, remarkably, female No voters on what was called ‘the substantive issue’ were more likely to be motivated by a concern that the proposed amendment would not rule out abortion while male No voters were more concerned that it would not protect the rights of the mother. (This is discussed in the chapter on the 1992 referendums by Brendan Kennelly and Eilís Ward of NUIG in the book How Ireland Voted 1992 (Folens and PSAI Press, 1992).) At the time of the 2002 referendum, 31% of women, compared with 26% of men, declared themselves to be against abortion in all circumstances. By the current decade, though, the gender gap was in the other direction, with women more liberal than men. In the same-sex marriage referendum, 68% of women, compared with 57% of men, declared their intention to vote Yes (Behaviour and Attitudes poll, 1–15 May 2015). And in the RTE exit poll after the 25 May 2018 referendum, 72% of women reported voting for repeal compared with 66% of men. Over the last 35 years, men have become more liberal (the vote against the 8th rising from 38% to 66%) and women have become a lot more liberal (the vote against it rising from 25% to 72%).


Geographically the pattern was very familiar, with the usual liberal and conservative suspects in their accustomed places. Of the 14 constituencies most in favour of repeal, 11 were in Dublin and the other three were in its commuter belt (Wicklow and the two Kildare constituencies). The most conservative constituencies were rural, with Donegal taking over from Roscommon (in 2015) as the only constituency to record a No majority. And whereas Roscommon in 2015 was only a little more conservative than several other constituencies, this time Donegal was out on its own, with the Yes vote more than 7 percentage points below that in the next most conservative constituency. The only constituency that seems to deviate much from the expected pattern was Cork North-Central, a predominantly urban constituency (albeit now with a sizeable rural component, which may help explain the outcome), which recorded a lower Yes vote than, for example, Clare, Cork East, Cork South-West and Wexford. As usually happens in referendums, in complete contrast to the general election pattern, turnout was highest in Dublin and lowest in Connacht–Ulster.


Comparing the votes with those at previous ‘liberal agenda’ referendums shows a very high degree of correspondence. The RTE exit poll found that the Retain vote on 25 May contained a sizeable number of people who said they had voted in favour of same-sex marriage in 2015, but this finding should be treated with great scepticism. Of those respondents who could recall how they voted in the 2015 referendum, 81% claimed to have voted Yes and only 19% to have voted against the proposal, compared with the actual distribution of 62–38. In other words, about half of those who voted against same-sex marriage in the previous referendum told the 2018 exit poll that they had supported it, and were either misremembering or misreporting their past behaviour. Almost certainly, the great majority of those who voted No in 2018 had also voted No in 2015, despite what they told the exit poll.


If we compare the 2018 vote to repeal the 8th amendment with the 1983 vote to insert it, the two most striking features are the rise in turnout (only 53% voted in 1983 compared with 64% in 2018) and, of course, the huge increase in the vote against the amendment, from 33% to 66%. The raw number of votes in favour of the 8th fell only slightly, from 840,000 to 720,000, while the number of votes against it more than trebled, from 420,000 to over 1.4 million. The pro-choice vote in Dublin rose by slightly less than that elsewhere (by 27% in Dublin and by 37% in the rest of the country), so support for each position is more even across the provinces than it was in 1983 and the standard deviation of the percentage Yes vote is much less than in 1983 (7.5 compared with 12.4). There was an increase of over 40% in the liberal vote in Cork NW, Cork SW, Donegal, Galway E, Roscommon and Wexford, which are predominantly the constituencies that had most strongly supported the insertion of the amendment in 1983, while only in Dun Laoghaire, which had recorded the highest No vote in 1983, did it increase by less than 20%. A tide of liberalism has swept across the country; as the most liberal constituency (Dun Laoghaire) voted in 1983, ie 58% against the 8th amendment, so the fifth and sixth most conservative constituencies (Offaly and Limerick County) voted in 2018. In Chapter 5 of his book Public Opinion, Politics and Society in Contemporary Ireland (Irish Academic Press, 2008), Pat Lyons concludes that both turnover in the electorate (older more conservative voters being replaced by younger more liberal ones) and actual changes of opinion among continuing members of the electorate played a part in the evolution of attitudes regarding the liberal agenda during the time period he examined.


Correlating the voting patterns across the country confirms the high degree of continuity. (This exercise requires some amalgamation of constituencies given the frequent redrawing of constituency boundaries. There are 39 more or less comparable units for comparisons between 2015 and 2018, and 35 for comparisons between 1995 and 2018.) Comparing the percentage Yes votes in the May 2018 referendum with those cast in the same-sex marriage referendum three years earlier produces an exceptionally high Pearson correlation (r) of 0.95 (a figure of 1 would indicate a perfect relationship while 0 would indicate no relationship at all). The correlation with the divorce referendum of 1995 is also very strong (r = 0.91). Even though the substantive issues in those three referendums might appear to be completely different, it is evident that most voters saw each as related to a strong underlying moral issue liberal–conservative dimension.


When we relate the vote in May 2018 to voting at the 2016 election, we find quite a strong negative relationship (r = –0.57) between support for repeal and support for FF, and quite strong positive relationships with support for Labour and the Greens (in both cases r = 0.54). Support for repeal in 2018 and for Fine Gael and for Sinn Féin in 2016 is weak and negative (r = –0.11 and r = –0.03 respectively), while the correlation with independent support is stronger and negative (r = –0.45). The RTE exit poll showed that current supporters of all parties except FF voted for repeal, but the implications of this for the next election are not clear. At the 2016 election only 2 per cent of voters said they saw abortion as the most important issue in determining their vote, and while this does not mean that voters do not regard abortion as an important issue – the high turnout on 25 May shows that not to be the case – it does suggest that voters distinguish clearly between referendum issues and election issues.


There is a certain irony in hearing the No side now invoking the principle of pluralism and minority rights as a reason why their concerns should not be dismissed out of hand, given that this claim was previously the preserve of Ireland’s liberals, who would no doubt say that respect for minority rights was not always apparent during the decades of their opponents’ ascendancy. The feasibility of respecting minority rights is obviously easier when that minority seeks simply the right to live its lives according to its own moral standards than when, as in the present case, the minority demand is that its moral code be enshrined in law and the constitution so that the whole of society is compelled to live by it. The case is particularly tenuous when that demand has just been emphatically rejected by the people in a referendum. This is, no doubt, an issue that will manifest itself in discussions over the kind of conscience clauses to be incorporated in the proposed legislation, and in the way these operate once the legislation comes into effect. The strong relationship between age and conservatism suggests strongly that the tide will continue to flow in a liberal direction, and battles over church control over so much of the education and hospital systems lie ahead unless the church disengages without demur.


There is another potential irony, this one concerning the prospect of a united Ireland. In the last days of the campaign, the Orange Order, not a body to which many voters in the south would instinctively think of looking to for a cue as to how to vote, expressed its hope that the outcome would be No. It would be quite an irony if, after decades of resisting a united Ireland on the ground that this would mean becoming part of a confessional Catholic state, some unionists – and not just unionists – now come to feel uncomfortable about the idea of a united Ireland because those southerners are just too liberal. More probably, though, were voters in Northern Ireland ever to get a chance to vote directly on abortion, it might turn out that their views on the issue are not so different from those of their southern counterparts.


Overall, the 8th amendment had an unhappy existence. For 35 years liberals chafed at its existence, and conservatives may be coming to realise the damage that it did to their cause. Its proponents in the early 1980s intended it to have the effect of preventing either parliament or the courts from legalising abortion under any circumstances, and it can be fairly be identified as a pre-emptive strike given that the issue was nowhere near the political agenda at the time. The amendment’s prohibitionist absolutism blocked the possibility of gradual incremental reform, except to the very limited extent permitted by the 1992 X case judgment (which was entirely unforeseen by the amendment’s proponents), and led to a build-up of a demand for more radical change. When the reform option is blocked off, the outcome is sometimes revolution.


The retain side seemed to realise, far too late, that reform might help lessen the pressure for more radical change, and in the final days of the campaign some of their spokespersons suggested that the 2013 Protection of Life during Pregnancy Act, suitably amended, could deal with many of the hard cases, even though they had vehemently opposed this legislation at the time because it conflicted with their absolutist position.


If anti-abortion pressure groups had not called for insertion of the 8th amendment in the early 1980s, and if the leaders of the two main parties had not agreed (or capitulated, as their critics saw it) to their demands, then there would never have been an 8th amendment. In that situation, the 1992 X case could not have arisen, and if some analogous case had somehow reached the courts, the Supreme Court would have had no basis for stating, as it did in its X case judgment, that abortion was legal in certain circumstances. The issue would most probably have remained dormant well into the current century, raised only by groups far from the political mainstream. Perhaps in the 2010s politicians would have moved to legalise abortion in the ‘hard cases’ such as fatal foetal abnormality, but those pressing for unrestricted abortion during the first 12 weeks of pregnancy would have made little headway. Most politicians would have resisted such a proposal, out of principle or out of a belief that such a liberal regime, easily labelled with what used to be the dread phrase ‘abortion on demand’, would be unacceptable to voters, or both. Because the issue would not have been constitutionalised if there had not been an 8th amendment, there would never have been a referendum on the subject and the level of public support for significant liberalisation would never have been discovered.


There is every reason to believe that the laws on abortion as of January 2019 will be much more liberal than if the 8th amendment had never been inserted into the constitution. The pro-life amendment of 1983 was an attempted pre-emptive strike that, as sometimes happens with pre-emptive strikes, eventually rebounded powerfully against those who launched it.

One thought on “Analysis of the abortion (8th amendment) referendum, May 2018

  1. A Mairtín,
    Subsequent to the results following referenda seeking to facilitate change to Bunreacht na hÉireann relating to the European treaties of Nice and Lisbon, on foot of the vote last week on removing the protection afforded the unborn in Ireland for myself I feel the real question ought be, when are we having our second referendum this decade on the matter?
    is misé

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s