Adrian Kavanagh, 29th May 2018

With a national turnout rate of 64.1% – just 1.0% lower than the national turnout level for the 2016 General Election – the Regulation of Termination of Pregnancy (Repeal of 8th Amendment) referendum was passed with 66.4% of the valid votes (almost two-thirds of all the valid votes cast).

Figure 1: Turnout levels at Irish referendum elections, 1937-2018

In actual voter number levels, the turnout at this referendum was 2,159,655, making this the referendum contest with the highest ever number of voters since the foundation of the state. In percentage terms (as shown by Figure 1), this marked the highest turnout level for a referendum contest, or set of referendum contests, since the three referendum contests that took place on the same day as the November 1992 General Election contest. This was also the highest turnout level for a “stand alone” referendum contest since the 1971 referendum on Membership of the EEC.

This contest and the 2015 Marriage Equality referendum contest both followed immediately on two low turnout referendum contests (the 2012 Children’s Referendum and the 2013 Seanad Referendum), which in turn were preceded by the higher turnout EU Fiscal Stability Treaty referendum contest. This suggests strongly that turnout levels at referendum contests are strongly reflective of the issues being voted on – with “moral issues” and European Union contests attracting the higher turnout levels – while turnouts at referendum contests may also be pushed up by having other electoral contests on the same day (as was especially evident with the 1992 referendum elections, but was also evident in the case of the referendum votes that took place on the same day as Local and European Election contests in 1999 and 2004).

Figure 2: Regulation of Termination of Pregnancy (Repeal of 8th Amendment) referendum: Margin (%) by which the Yes side won per Dáil constituency

As suggested by Figure 2, the highest percentage votes shares for the Yes side were associated with the more urban constituencies, as in the Marriage Equality referendum contest and other similar “Moral Issues” referendum contests. The large urban-rural divide evident at other “moral issue” contests, such as the 1995 Divorce Referendum, was not as evident in the May 2015 vote on Marriage Equality, but the extent of urban-rural polarisation increased somewhat, relative to the May 2015 contest, in the case of the Repeal of 8th Amendment referendum vote.

With the other cities and the Dublin commuter belt constituencies included, the share of the vote won by the Yes side in the more urban constituencies came to 72.0%, while the share of the vote won by the Yes side in the other (more rural) areas was 59.7%; a difference of 12.3%. While the urban-rural difference in terms of the vote share won by the Yes side came to just over twelve percent, the margin increased slightly relative to the 11.6% difference (Urban average: 67.4%, Rural average: 55.8%) between the Yes vote in these same urban and rural areas at the 2015 Marriage Equality referendum. By contrast, the share of the vote won by the No side for the 1992 Protection of human Life in Pregnancy referendum in the more urban constituencies came to 41.8%, while the share of the vote won by the No side in the other (more rural) areas was 58.7%; a difference of 16.9%. While the share of the vote won by the Yes side for the 1995 Divorce referendum in the more urban constituencies came to 58.0%, while the share of the vote won by the No side in the other (more rural) areas was 40.6%; a difference of 17.4%. So the overall trend does point towards a reduced level of urban-rural polarisation in voting trends at moral issue referendum elections relative to the 1980s and 1990s.

Figure 3: Percentage of valid votes cast in favour of the amendment

At the constituency level, the largest percentage share of the vote for the Yes side was the 78.5% share of the vote won in Dublin Bay South (the same constituency also returned the highest Yes vote (74.9%) for the 2015 Marriage Equality referendum), followed by Dún Laoghaire (77.1%), Dublin Fingal (77.0%), Dublin Central (76.5%) and Dublin Rathdown (76.1%). In all of these constituencies, at least three-quarters of the valid votes being case were in favour of the proposed amendment. The lowest vote share for the Yes side was recorded, by contrast, in Donegal, where only 48.1% of the valid votes cast were in favour of the amendment, while relatively low Yes vote shares were also associated with the constituencies of Cavan-Monaghan (55.5%), Mayo (57.1%), Roscommon-Galway (57.2%), Offaly (58.0%) and Limerick County (58.1%).

Reflecting the pattern observed at previous contests, turnout levels at the referenda tended to be higher in the more urban constituencies (65.2%, on average) and especially in Dublin (65.7%). Turnout levels in rural Ireland averaged out at 62.9%; 2.3% lower on average than in the more urban constituencies. This contrasts with the general trend in turnout patterns usually observed at general and local election contests, where turnouts tend to be highest in the more rural constituencies. The highest turnout level by constituency at these contests was found in the Wicklow constituency (74.5%), with high turnout levels also being found in Dublin Bay North (71.6%), Dublin Final (70.3%) and Dublin Rathdown (70.1%), while the rural constituency with the highest turnout level was Cork South-West (67.3%). Reflecting the trend evident at past contests, Donegal recorded one of the lowest turnout levels nationally for this election (57.1%), but on this occasions turnout levels were actually lower in Dublin Central (51.5%) and Dublin Bay South (54.9%).

Figure 4: Turnout Difference, by Dáil constituency, between the May 2018 Referendum contest and the 2016 General Election

When contrasting turnout levels with those for the 2016 General Election, it can be seen that the difference between the overall/average turnout levels for these two contests at the national level was relatively small, amounting to a difference of just 1.0%. The average turnout levels in Dublin were actually, on average, 3.0% higher than the general election turnout level for the city. By contrast, in the more rural constituencies the average turnout level at the referendum was 3.9% lower than the average turnout rate for the 2016 General Election in those constituencies. At the constituency level, the referendum turnout in Dublin Final was actually 5.3% higher than for General Election 2016, while the referendum turnout was also notably higher in the constituencies of Dublin Mid-West (4.5%), Dublin South-West (4.2%), Dún Laoghaire (3.9%) and Dublin Rathdown (3.8%), Dublin Bay North (3.6%) and Wicklow (3.6%).  By contrast, the referendum turnout was 8.3% lower than the general election turnout in the Kerry constituency, while the general election turnout level was also notably higher in Mayo (7.0%), Tipperary (5.9%), Roscommon-Galway (5.9%) and Donegal (5.8%). In all, the referendum turnout level was higher than the turnout level in the preceding general election in 19 of the 40 Dáil constituencies, while the general election turnout level was higher in the other 21 constituencies.