Over the course of the referendum I wrote several long pieces analysing opinion polls as they appeared in the newspapers to glean useful information about how the campaign for a Yes vote was doing, where we were stong, what weakenesses existed and what tactics this information suggested. I’ve gathered all these pieces together below in reverse order, the first has a lot of details on methodology and context that is not repeated in the later pieces.
The polls appear to have been very accurate in particular those carried out by Red C and also Behaviours and Attitude. Both companies showed likely results very close to the final outcome (66.4% Yes) if you presume the Yes campaign won most of the Don’t Knows over the 3 months between the start of serious campaigning around March 8th and the referendum in June.
Over the course of the referendum I wrote several long pieces analysing opinion polls as they appeared in the newspapers to glean useful information about how the campaign for a Yes vote was doing, where we were stong, what weakenesses existed and what tactics this information suggested. I’ve gathered all these pieces together below in reverse order, the first has a lot of details on methodology and context that is not repeated in the later pieces.
The polls appear to have been very accurate in particular those carried out by Red C and also Behaviours and Attitude. Both companies showed likely results very close to the final outcome (66.4% Yes) if you presume the Yes campaign won most of the Don’t Knows over the 3 months between the start of serious campaigning around March 8th and the referendum in June.
1. – Repealing the 8th – what the opinion polls are telling us – 04 April 2018
The Sunday Times with Behaviour & Attitudes have run two very useful polls that give a strong sense of how the campaign to Repeal the 8th Referendum is going. The overall story the poll results tell is bad for the Vote No campaign and promising for the Vote Yes campaign. If the referendum had been held at the time of the March poll then Repeal would have been carried by 64% to 36%, almost 2:1. The polling data also shows No has a soft vote that is very much larger proportion than the equivalent soft Yes vote. This means if anything between now and referendum day the polls are likely to drift towards repeal.
None of this is a reason for complacency, what the No side lacks in terms of numbers and support they make up for in terms of funding. Before the campaign had even started they were spending hundreds of thousands on online advertising, billboards, leaflets and free buses to what had to be their disappointingly small March 10th national march. Together for Yes may have far more support and more people out canvassing but will it have enough to defeat all that paid advertising?
As this is a long read we are also making an audio version available
That is a question that will be answered on referendum day and the answer will be determined by the work put in between now and then, both in terms of formal canvassing teams but also – and at least as important – in dozens of conversations at home, at work and out with friends. But here we want to look into the detail of the polls, in part because it helps guide those conversations. As well as looking at the results of each of these two polls we can also look at what changed between them, and presuming there are similar polls in future check back in from time to time.
The first poll was carried out over the first two weeks of February, the second in the second week of March. In terms of the overall question of which way people will vote in the Repeal referendum the two polls showed little change, both showed 49% of people were voting yes. There was bad news for the vote No campaign as not only were only 30% intending to vote No in the February poll but this actually fell by the time of the March poll to 27%. In the March poll some 20% of people Didn’t Know which way they would vote and 4% said they had decided not to vote. It’s when we exclude these last two groups that we get the 64% Yes to 36% No mentioned above.
This discussion will draw numbers from the four polls linked below but it mostly works off a Behaviour & Attitudes poll reported on by the Sunday Times in March so we recommend you download that and read this text with it open as we will be referring to specific pages of that report.
Margins of error
It’s useful to understand some technical details of polling so we don’t try and read more out of them than we should. The error on a poll of this size (900 voters) is a little over 3%. This is because a random sample of 900 people from the Irish population will not be exactly representative of the total population. There is a 95% chance that the percentages in a poll of this size will be within 3.3 points of the actual percentage if people voted in the referendum that day. In other words we can’t really say much about apparent difference that are less than 3%.
This error margin increases as the poll size reduces, a poll of 96 people will have an error margin of +/- 10%. This is important for the discussion that follows because the most useful aspect of these polls is when they break the results down to subgroups, for example Fianna Fail voters or Farmers. Because these are a subgroup of the total poll of 900 they will be smaller, many are about 300 people.
In some cases these subgroups are too small for any meaning to be derived, for instance in the first poll 80% of Green Party voters were voting Yes and by the second this had risen to 100%. The problem though is that there are not many Green Party voters and these percentages come from sub sample sizes that are tiny, just 9 people in the second poll. The error margin is too vast to draw any conclusion at all from those numbers.
Likewise the figures for farmers are also initially fascinating until you see there are only 36 farmers in the 2nd poll, again meaning such a huge error margin that attempts at interpretation are futile. So later on when we look at the number for various sub populations we will only be interested in those that show major gaps and major changes, in the order of 5% (a poll of 384 would have a 5% accuracy) and where the subpopulation contains hundreds of people.
RED C polls
In this piece we are focusing on the Behaviour & Attitudes / Sunday Times polls. Red C also carried out two polls in the same period, the second of which attracted attention as it appears to show a large swing from Yes to No in comparison with their first poll. It appears to us that something went wrong with that first poll as the B&A poll taken less than a week later had a startlingly different result which would require a 21 point swing against Yes in the period before any significant campaigning had taken place.
Note – we uploaded the wrong comparison in the first version of this article, thats corrected above.
Excluding RedC 1 the three remaining polls don’t have such wild swings although the different methodology and question phrasing of the Red C polls would complicate direct comparisons. For what it’s worth the No vote across the three polls goes from 30 (B&A 1) to 27 (B&A 2) to 26 (RedC 2). RedC 1 with a No of 20 would not fit well at the start of that sequence and if it actually measured anything as opposed to being the product of an error it was measuring something other than the impact of campaigning.
- Red C January – https://www.redcresearch.ie/wp-content/uploads/2018/01/SBP-Jan-2018-Poll-Report-GE16.pdf
- B&A February – http://banda.ie/wp-content/uploads/Sunday-Times-Report.pdf
- B&A March – http://banda.ie/wp-content/uploads/J.8878-Sunday-Times-March-2018-Report.pdf
- Red C March – https://www.redcresearch.ie/wp-content/uploads/2018/03/SBP-March-2018-Poll-Report-GE16.pdf
Soft and Hard votes will decide
In terms of strategy both sides of the referendum will pay most attention to 3 blocks of voters and mostly ignore two other blocks. The mostly ignored ones are called the ‘hard yes’ vote and the ‘hard no’ vote. These are people who are very unlikely to switch votes and canvassers will be trained to not waste time talking to them as the time is better spent with the three other groups .
We do of course want to encourage the ones voting hard Yes to definitely get out and vote – this could be key to winning Repeal – but arguing with a hard No would in most cases be a complete waste of time. In particular as the hard No’s are close in size to the percentage who opposed the very very limited reforms of the Protection of Life During Pregnancy act which was designed to deal only with cases where the women would die if denied a termination. The hards Nos are pretty much that ‘let women die’ bloc.
These are three groups whose opinions both sides will try and spend time shifting
1. The ‘soft Yes’ voters who are voting yes but unsure about aspects of a women’s right to choose. Typically they only want women to be able to access abortions for some reasons and not others..
2. The Don’t Knows that have not yet decided how to vote but who do intend to vote.
3. The ‘soft No’ who are against a women’s right to choose but perhaps think there are some circumstances eg Fatal Foetal Abnormality (FFA) where access to abortion should be granted.
In these polls and it would appear in the internal polling of all the campaigns the key issues that are being used to identify the soft voters are, for the Yes vote whether they are voting Yes but oppose abortion on request up to 12 weeks and on the No side whether they are voting No but support access to abortion in cases of Fatal Foetal Abnormality (FFA) and threats to the health of the women. Fairly obviously someone who intends to vote No but feels abortion should be available in the case of FFA or where there is a threat to the women’s health should be possible to convince that they should vote Yes so this can happen.
The No campaign will target soft Yes voters by trying to steer the conversation onto the 12 week access on request issue. Typically this is done by trying to create categories of women who deserve access to abortion and those who do not. So even though the No campaign are against abortion in all cases they will try and chip away at the soft Yes, for instance through the favourite myth of women using abortion as contraception because they are lazy.
Likewise the Yes campaign will target the soft No’s by explaining that if they vote No they will be stopping women with (FFA) from accessing abortion in Ireland – only Repeal will mean the legislation required to access such medical care in Ireland can be passed.
To a very large extent this means that the No campaign will want to centre discussions on the 12 week issue while the Yes campaign will want to centre it on FFA and threats to the women health. Both will have to be able to provide answers to its own ‘soft Yes’ voters to avoid losing them but as far as possible they don’t want the conversation centered there.
The Sunday Times polls directly asked those polled on their attitudes on both these question so they give a very good indication of the relative side of these five important blocs and over time the movement between them. From looking at all these responses (p13 to p17) we can say that the March poll would divide the population up as follows;
- 40% hard yes (‘its a women decision’)
- 8% soft yes (‘for health threats & FFA only’)
- 20% Don’t know
- 4% Not voting
- 14% soft no (‘but should be available for health & FFA’)
- 14% hard no ( ‘not even for health & FFA’ bloc).
It may be a little bit hard to get your head around where we have produced these numbers from so hopefully the accompanying graphic will help. It hacks together three of the Sunday Times / Behaviour and Attitude graphics around the key questions to allow us to visually focus in on the soft voters for each side.
Both the images and the figures above show the difficulty the No campaign is in, it has a comparatively small hard No vote (51% of their 27% of the vote) and a comparatively large soft No which grew 4 points over the month. Together for Yes not only starts off with what would be a 64% Yes but this is mostly Hard Yes (83% of their 49% of the vote) and the Soft Yes is small and shrunk 4 points over the month.
About 42% of voters in total are either undecided or soft but the large hard Yes percentage at 40% means that for Repeal to win we only need to win 1 in 4 of those soft and undecided votes. The referendum can certainly be still lost with these numbers – in referendums Don’t Knows often become No.
If you look at the four different polls the weekend before the 2015 Marriage Equality vote you see huge differences in the polls according to which company carried them out but also that with then MRBI and Millward Brown polls that still showed high numbers of Don’t Know’s much more became No than Yes on the day. http://www.thejournal.ie/marriage-referendum-poll-analysis-2109857-May2015/
No is losing rather than gaining ground
There is worse news for the anti-choice campaigns when we look at the movement between the two polls. As already stated the March poll shows them losing 3% of the vote who become undecided. To be clear this is within the 95% probability but would be at a less stringent 90% probability. To explain at 95% probability there is one chance to 20 you are wrong, at 90% this falls to 1 in 10. A shift between the two polls can also be seen in the health & FFA question (p17) where the percentage of No voters who are actually in favour of access to abortion in these circumstances rises from 33% in February to 37% in march.
The February poll was carried out as the wording of a probable referendum became clear but before any campaigns had really been launched. The March poll was carried out before the Together for Yes campaign had been launched but right in the thick of the anti-choice campaign launches and in particular of their huge advertising spend that saw anti-choice billboards erected all over the country. This was also the context of their failed attempt to mobilise large numbers for a national demonstration on 10th March, slap bang in the middle of the week polling was carried out.
In other words Vote No were dominating news reporting and they had billboards, truck ads and leaflets all over the country while very little was visible of Vote Yes. To lose almost half of your soft No’s under such conditions is extraordinary. We shall see though that it’s a little more complex and there are some warnings for the Yes campaign as we dig deeper into the breakdown of figures.
Again at this headline level the No campaign got off to a terrible start, despite being far more visible and spending a huge amount they not only failed to dent the soft Yes they actually lost some of their soft No’s.
The No reaction and the Marriage Equality comparison
No spokespeople have looked at these polls and have access to their own internal polling which will be showing similar results. We get some clue as to their alarm from what they are tweeting about the polls and what they are tweeting about canvasses. On the polls they are actively trying to mislead their own supporters by comparing the eventual result of the Marriage equality poll with what they describe as a 76% Yes poll near the start of the campaign and then contrasting that with the 49% Yes above. http://www.thejournal.ie/opinion-poll-2015-same-sex-marriage-1882262-Jan2015/
What’s misleading is that the January 2015 ME poll excluded Don’t Knows to get that 74% Yes, if you do the same and exclude Don’t Knows from The B&A poll the Yes rises from 49% to 64%.
It’s true the Marriage Equality referendum eventually passed by 62%, a drop of 14% from that first poll. However… that first ME poll was also a Red C poll so in this case the real equivalent would be equally out of place first Red C poll back in January this year. Exclude Don’t Knows from that January poll and Yes is at 75 so a similar 14% drop from the first poll would see Yes winning at 61%. Red C polls consistently had too high a Yes vote in advance of Marriage Equality, something they explained by talking about a ‘shy vote’ that was unwilling to tell them how they really intended to vote for fear of taking an unpopular stance.
Observing the No spokespeople on Twitter we’ve seen a lot of effort going into trying to shore up canvasser moral by posting positive claims about canvassing which just happen to from the areas of the country that voted against Marriage Equality or only passed it by narrow margins. But we’ve also seen the replies from their canvassers in Dublin saying they are finding it very tough going. A lot of the online No paid ads have a very strong emphasis on trying to recruit canvassers, this from a campaign whose spokespeople have been boasting they have been organised for months. It’s hard not to conclude that they are already facing major demoralisation problems.
Who will vote?
Are both Yes and No voters equally likely to get out and actually vote on referendum day? The tables on page 24 where Yes & No voters are asked how likely it is that they will actually vote in a general election are our best approximation in these polls. Not too much can be drawn directly from this as it’s unlikely a general election will take place on the day of the referendum – but we might assume that the intentions to vote would be similar so large differences here might matter. As it turns out there are few surprise’s here to upset the calculation above, Yes and No voters are jointly a little more sure they will be voting than the Don’t Knows but there is no significant difference between the intentions of Yes & No voters which might tilt the result.
The table on p26 which combines likelihood to vote with attitudes to 12 weeks without restriction however shows those against unrestricted access to 12 weeks are quite a bit less likely to vote than those for unrestricted access. This is good news for Together for Yes as it suggest the potential soft Yes vote here is less likely to stay at home, but this is within the margin of error.
The table on p29 combines likelihood to vote with attitudes to abortion access for threats to health and FFA. It is the No voters that are less likely to vote here, if by a narrower margin.
One area of major concern for Pro-choice campaigners would be the table on page 31 which breaks likelihood to vote down by age. Here only 60% of under 34s feel they would definitely vote but 80% of over 55s – the one bloc likely to vote No – feel they would definitely vote. If that gap played out in the referendum it would close the gap between Yes and No by a couple of points and in a close referendum could result in a victory for the No side. Both the US Trump election and the UK Brexit vote had the results they did because of the different turns outs of the similar age groups there in what were very close votes. Will this happen here, it would be a rather grim irony if the over 55s turned out in numbers to defeat the access to the healthcare that no longer has relevance to them but which the 18-35s are most likely to need and least likely to bother to go to the polling station for!
We can now move on to the other tables that divide into sub-populations based on gender, age, class and party support. Each of these tells an important tale providing we keep sample size in mind – the error margin increases to 4.5% and more for these subsamples.
Gender
For the most part there is little to learn here on of Yes voters as the differences between men & women are within the error margins, only 1% off the overall voting average intention for the Yes vote. On the No vote though it does appear that women are somewhat more likely to be No voters, a 5 pt gap and in comparison between the two polls this gap opened up during the first month of the campaign because the percentage of No voting men fell by 3 points. Why was there a 3% drop in men voting No? This may simply be because men are less likely than women to have already given serious thought to all the negative consequences of the 8th amendment, for example the way it reduces women ability to make medical decisions during a healthy pregnancy.
There is no significance difference otherwise in the Yes votes all the way down this table but opposition from women to Repeal, 12 weeks and even health and FFA is just about significantly higher. The higher No vote among women seems counter-intuitive but there is a simple explanation for this. The highest No voter population by far is the over 55s and in that sub-population women are over represented because men have started to die earlier. So the apparent gender difference may be a result of the comparatively large proportion of women in the over 55 sub-population – we will see this effect again below.
Age
It’s no surprise to see people from 18 to 34 are overwhelming Yes voters, 59% yes to only 18% no. And even that 18% is a substantial fall from the 25% No of the previous month. This is the second biggest shift anywhere in the tables, its a drop of 1/3rd and presumably reflects a rapid education process amongst younger people through discussion and online research. Likewise it’s no surprise that the over 55s are the one group where the No vote has a narrow lead, 35% Yes to 37% No. What is perhaps slightly more surprising is that Yes is also massively carrying the 35-54 age group, 54% Yes to 25% No.
This also very much reflects the composition of those pro and anti-choice campaign mobilisations for marches, canvassing and other street events. All the anti-choice mobilisations are overwhelmingly dominated by people aged 50 and over. The lack of young supporters isn’t just a branding problem for anti-choice organisations that can be fake fixed by putting all the young people on the front banner. It’s looks like its also going to really hit them in the actual vote. This incidentally is why getting a May rather than a June poll could be vital to getting the referendum passed and why if you are heading to a festival the weekend of 25 May you want to be sure to get down the polls first thing in the morning, wellies and all if need be.
It’s also worth pointing out again that this means the two age groups where abortion access may be something that will directly come into their lives are 3:1 voting for Repeal. The 55+ are group intending to vote against is very unlikely to face such a circumstance in their future. This is also an indication that a narrow defeat in the referendum is likely to lead to a second referendum within a few years.
The Red C poll has the age groups divided into 6 groups rather than the 3 groups discussed above. From their results the first two groups covering 18-34 are overwhelming Yes at 68%, the next 3 covering 35-64 are majority Yes and it’s only the over 65s that are majority No at 58%
Social Class
Here we have to be careful as the tables use a modified version of the UK NRS system which only has a tangential relationship to the way socialists talk about class. Rather than dividing class into workers and bosses the NRS subdivisions are more about what sort of work is done (manual or brain labour) and whether people are working or on welfare. For more on the problems of NRS see the section Class and Leave at https://www.wsm.ie/c/making-sense-brexit-tide-reaction-racist-vote
With that in mind we see little impact on class on the No vote (in the 25-28 range) but on the Yes vote ABC1s (55%) are more likely to be Yes voters than C2DEs (45%). However as E include almost all pensioners the difference here may be largely down to the weight of the over 55 No vote. There is also a F for farmer column but there were only 36 in the sample so the error margin is too large to say anything meaningful about the numbers here.
Regional breakdown
Over the page is the regional breakdown of voting intentions. This is important data although perhaps also frustrating as there is not that much we can do as individuals about the divisions that show up – they do have a lot to say about the deployment of campaign resources.
The towns and cities are voting Yes by 2:1, a 26 point lead. In rural areas Yes only has a modest 12 point lead. One figure that leaps out though is the huge drop in the rural No vote in comparison with the February poll, its down 10 points, 2/3rds of which became Don’t Knows. The drop is twice the error margin for this sub population size so its a real effect and we’d wonder what the cause of it might be – perhaps simply an impact of a previous lack of conversation about the need for abortion access being overcome by media coverage.
If the lead in urban areas is big in Dublin it’s huge, 3:1, a 40 point lead at 63% to 23%. The Yes vote is weakest in Connacht/Ulster, where Yes with 33% has a tiny 3 point lead. Yes leads by 20 points in the rest of Leinster and by 16 points in Munster. The Don’t Knows are highest in the regions where Yes is weakest, they are only 11% in Dublin but 30% in Connacht/Ulster. This may indicate that the ‘shy vote’ presenting as Don’t Know may shift more heavily to Yes than No because of the long term reliance of anti-choice movements on public shaming tactics.
Our impression is also that the Together for Yes campaign is weakest at the start of the campaign in Connacht/Ulster whereas Donegal has always been a strong point of the anti-choice movement. So this certainly suggests that canvassing and other outreach methods there may have a far bigger proportional impact than they would elsewhere. If you live in Dublin but come from Connacht/Ulster you’d do a lot better to canvass ‘back home’ and chase the 30% Don’t Know than the elusive 11% in Dublin.
This may also be a good point to insert a reminder that there is also a parallel pro-choice struggle along broadly similar lines in the 6 North-East counties of Ulster under British rule. The ‘carnival of reaction’ that followed partition created theocratic states north and south of the border but the continued impact of colonialism in the north has served to preserve clerical influence three decades after its disintegration started in the south. When Repeal is carried south of the border we may become the best available option for pregnant people in the north needing to access abortion. They will of be disproportionately disadvantaged by the requirement for two doctors visits 3 days apart to access abortion pills. This requirement is not simply a pointless inconvenience but something that if passed will deny abortion access to vulnerable segments of the population including those trapped in abusive relationships.
These regional sub populations are all around 250 people meaning the margin of error is around 6% so although there are fascinating fluctuations between the March and February polls almost all are around 3% and so meaningless within the error margin. A few are not, these are;
- In Leinster No has dropped 7 points, all it appears switching to Don’t Know.
- In Ulster Yes has dropped 9 points, however the Conn/Ulster population size is small that the others (161) so that 9% drop isn’t far off the margin of error at 8%. The drop split evenly between Don’t know and Won’t vote which indicates it may be a ‘shy vote’ in conditions where there is a lot of aggressive public shaming from the anti-choice campaigns.
12 weeks by region
The reason the No campaign want to centre the 12 week issue is clear when we look at regional variation on the question of unrestricted access to 12 weeks. If this was the question rather than Repeal Yes would carry Dublin and Munster but lose Connacht / Ulster and (by a tiny margin) the rest of Leinster. The concentration of population in Dublin and Cork would mean the referendum would still pass – even if 12 weeks on request was the question – but it would be very tight. This is why the No campaign is trying to centre the conversation on the 12 week question, they want voters to be thinking of that rather than FFA or health at the moment they mark their ballot box.
FFA & health by region
The other significant shifts are around the FFA & Health access issue – here we see why the Together for Yes campaign will want to centre health & FFA access. The regional differences are huge here, Dublin has an 8:1 lead when you ignore Don’t Knows. Elsewhere the Yes lead on that question is a little over 2:1 except Leinster which is 3:1. If access to abortion for women whose health was under threat and for FFA was the question the referendum would be massively carried everywhere. The problem for the No campaign is that essentially this is the question for anyone who thinks abortion should be available in such circumstances even when they are unsure of the 12 weeks on request issue. All the more so as the 12 week limit is determined by that being the last point abortion pills can be used and the state has already demonstrated it cannot stop hundreds of women using these every year.
There were a couple of significant shifts between February and March on this question
- In Leinster support for FFA and health access rose 9 points to 61% and opposition fell 11 points to 19%
- But in urban areas support for FFA and health access fell 15 points to 63% while opposition rose 9 points to 18%.
These are real effects several times the error margin. The 15% urban drop is the largest change across both polls. A clue to the cause may be that while the drop is in urban areas it is not reflected in the largest urban area of Dublin.
Intentions by Party
Over the page we have the tables that provides a breakdown of voter intention by which party was voted for in the previous election. This is a particularly important subset as those most likely to vote in the referendum are those who voted in the last election.
It has to be said at the start that the size of the sub populations here for Labour (42) and the Green Party (9) are too small to draw any conclusions from the data. Otherwise we could waste time considering the plus and minus 20% changes since the last poll but that’s pretty much the error margin and less.
It’s probably not surprising to see that Fine Gael who are in government are over 2:1 Yes. Fianna Fail voters only give the Yes side a 10 point lead over No compared to the 22 point lead for the total population. It’s likely the Fianna Fail Yes is largely their urban vote and that their rural vote is No, one of the major difficulties Fianna Fail faces is the disintegration of their once powerful urban base because of their role in the crash and the pressure of Sinn Fein.
Realistically we can expect rural Fianna Fail to be sitting on their hands and quietly campaigning against the referendum. We’ve certainly see a couple of obvious sock puppet accounts on Twitter whose real owners appear to be rural Fianna Fail activists. As issue in rural areas will be Fianna Fail trying to use the referendum to damage rival candidates, another reason why understanding these polls is important as they are far more likely to try this if they think it will be a close vote. Likewise Fine Gael and Labour politicians in rural areas are less likely to campaign vigorously if they think the vote will be close. This is part of the reason the No spokespeople were very keen to talk up the second RedC poll even though they also must have spotted that the trend across all four polls suggests the apparent shift was an artifact rather than a real finding.
The most significant population bloc in terms of campaigning is found in this table and that is the Sinn Fein voters. In the February poll 26% were Don’t Knows or Won’t Vote but by March both these have fallen drastically resulting in a 7% rise in SF No vote to 30%. We need a little caution here though as the SF population size is only 133 so the 7% growth in the No vote is slightly under the error margin. Sinn Fein is the one political party where over the period the leadership have adopted a strikingly stronger position, in particular over the last few days (ie after the March poll had come out). It could well be this will have a major impact on the figures in the next poll, otherwise republicans will have to switch ‘No Freedom till Freedom for Women’ to ‘your on your own girls’.
To summarise this poll data means that it is likely but not inevitable that the Yes vote will win. The lead is significant and the trend is favourable but there are a lot of Don’t Knows and if they all become No’s and the over 55s turn out in droves while the under 35s fail to show than the result could be very close.
In addition we can expect a massive increase in the volume of dirty stealth ads to appear in the last week of the campaign. These will attempt to create moral panics and play on prejudices when there will not be enough time to address these. The conversations we have with workmates, friends and relative between now and then can be used to inoculate people against this tactic, the Cambridge Analytica investigation breaking when it did is useful in that regard to encourage people to question the funding and truthfulness of the dirty stealth ads as they appear.
We conclude by saying that a ‘women’s right to choose’ is no more subject to the popular whim than slavery. The logic of the passing of the 8th was the attempts to ban information and travel that quickly followed. A 2:1 majority did not stop our fight against those. If by some disaster Repeal is defeated then the day after pregnant people will still be illegally importing pills for use in Ireland and travelling to obtain abortions elsewhere. A no vote won’t stop abortion, it will just extend the period of illegality and force thousands to continue to travel or break the law here and risk a 14 year prison sentence.
Words: Andrew Flood (follow Andrew on Twitter)
20th April – Three additional opinion polls this April have shown that the anti-choice campaign has failed to reduce the number of people intending to repeal the 8th referendum at all, despite 3 months of frantic campaigning that has involved an enormous spend on misleading billboards, posters, online ads and glossy colour leaflets. With the Together for Yes campaign only gearing up last week this means the No may may well decrease between now and the referendum. However as our reports and analysis of these poles show the pro-choice Yes campaign cannot be complacent.
Two of these polls appeared on 28th and 29th April, the Daily Mail one had a very loaded to the No side question, the Red C seems to continue to over estimate the Yes relative to other polls.
The Red C poll in Sunday Business Post 29th April confirms little change against their previous poll since campaigns started in February.
- No vote static at 26%,
- Yes vote at 53% down 3% since last their poll but 3% is margin of error.
- Don’t Know up 3 to 19%
- Won’t vote static at 2%
The image above shows all 6 polls and demonstrate that despite Save the 8th & Love Both No campaigns spend of millions the No vote has not increase at all.
With Don’t Know excluded that Sunday Business Post / Red C poll has
- 68% Yes
- 32% No
Note this is a higher Yes than that which other companies are finding but the Yes to No ratio has remained close to 2:1 in all polls, all variation appears to be down to the company polling with Red C polling yes the highest.
Sunday 22nd April saw the appearance of yet another opinion poll of the referendum to repeal the anti-choice amendment added to the constitution in 1983. As with all 5 opinion polls that have appeared during the campaign it showed little change at the national level, the strong lead for Yes remains. But when the full data was published Monday we had a look and discovered some interesting trends within it.
The Sunday Times teased everyone Saturday night with tweet pointing out the Dublin Yes had fallen 8%. They probably a sold a good few papers the next day off of it but in doing so they buried the lead that otherwise no significant change in voting intentions had occurred in comparison with their previous polls.
You can view this Behaviour & Attitudes poll reported in the Sunday Times. In this poll 928 people were questioned meaning the margin of error is 3.3% error. Once that is taken into account we see the tiny changes in voting intentions are not significant despite what the Save 8th and Love Both spokes people initially tried to claim.
People were also asked if they supported unrestricted access to abortion in the first 12 weeks of preganancy, the period in which the abortion pill can be used. Despite the No campaigns huge huge spend on misleading posters, billboards, online ads & leaflets there was no impact on voter attitudes on this question. There is certainly an argument that the Save the 8th campaign being caught multiple times with its dirty tricks campaign at the start of the campaign rightly did permanent credibility damage to the No side. There was also no change in attitudes to allowing abortion to protect womens health & where there is fatal foetal abnormality detected.
This 3rd B&A/ST poll was done from 15-17th April . This was right at the end of the period when only Vote No posters were up, and in considerable numbers. Since then the Together for Yes & other Repeal the 8th posters are appearing in increasing numbers. The failure of the anti-choice organisations to erode ‘soft Yes’ votes when they were the most prominent voice by far is a serious setback for them. Now the question is whether the ‘soft No’ vote will be eroded by arguments for compassion & protection of health as the referendum approaches in 5 weeks.
‘Soft No’ votes can be understood as No voters who also want to protect womens health & allow abortion for fatal foetal abnormalities. Doing either however requires Repeal. The No campaign is in trouble here, this B&A poll on page 17 shows 33% of No voters want abortion in such cases & 11% don’t know. In this context 44% of the No could be won to Yes by the Together for Yes campaign as it unfolds.
Pages 16+17 of this B&A poll shows that the ‘Soft Yes’ is proportionately small, less than half as a proportion than the ‘Soft No’ so even apart from their failure to date the anti-choice campaigns are in a weak position. To illustrate
For voters who intend to vote Yes to repeal some 12% are against unrestricted access to 12 weeks and 8% don’t know, so you could say 20% of the Yes vote was soft. On the allowing access to abortion in cases of fatal foetal abnormality and where the persons health is threatened some 9% of Yes voters are against and 5% don’t know. This is a 13% ‘soft’ vote
Here its worth noting that although the 3 months of (mostly No) campaign have not effected voting intentions on repeal pages 16 & 17 does show both Yes and No soft votes have reduced between February & April polling, presumably as people educate themselves on these topics.
Soft Yes i.e. against 12 weeks fell from 21% of Yes to 14%
Soft No i.e. for health & FFA fell from 49% of No to 44%
So as an exercise if we take soft votes into account the unrealistic worst cases, assuming all Don’t Knows go against them as
Yes falls to 40% if all ‘soft yes’ lost
No falls to 16% if all ‘soft no’ lost
This sort of complete loss is very unlikely but this exercise demonstrates what a bad situation the anti-choice campaigns are in.
When you add the 2 other polls in the 3 month campaign period in (see image in 1st comment) and exclude Don’t Knows we see some differences probably due to different MRBI and Red C methodology but confirmation across all these polls that there have been no major changes in voter intention at the national level. Including Don’t Knows and Won’t vote makes little difference, again there is very little change over the three months of campaigning.
But to warn against Yes complacency. The 1995 Divorce referendum was almost lost despite a strong showing at start when the Yes vote fell sharply in the last days of the campaign. That said the fall was between the polls before the campaign started with a big drop during the campaign itself. The three B&A polls above are all taken with the campaigns underway. There does seems to have been a small drop before the campaign started as recorded in the MRBI and Red C January polls. Divorce data from the Irish Political Ephemera page
We are going to move on to looking at the B&A/ST data on regional voting intentions & how much can be said about the data on voting intention by age, gender, ‘social class’ and political party support. This is complex so you may want to have a look at what we wrote comparing their previous two polls where we discuss the issues around doing this, in particular then increase in the margin of error.
The change in voting intentions by region from the 3 polls carried out by B&A/Sunday Times, February to April show the urban V rural divide collapsing. In other words the Yes falls in Dublin & other urban areas but this fall is balanced by a No fall in rural areas so that there is no overall change of significance.
Yes has shifted from being well ahead in Dublin & other urban areas to being well ahead everywhere
Even in Connacht/Ulster Yes is now 13% ahead
The reduced Dublin Yes lead is still 27% ahead
Overall the Rural Yes is now 14% ahead, it was 1% behind back in February
With all the sub group discussions its important to keep in mind that as group size decreases the margin of error increases substantially.
When to look at this poll by region, gender & social class we see strong variations here as with the previous two B&A polls but no significant movement between these polls. Repeal the 8th remains the choice of every group except the over 55s which are Save the 8th by 4%, really with the margin of error included this makes them 50:50.
Another reminder that because these sub populations are small when looking at the raw B&A tables its important to keep an eye on sample size as what looks like a change may not be when margin of error for small sample is considered. For instance sequentially across the 3 polls it looks like the under 35s No vote goes from 25 -> 18 -> 23 but a lot of this may be because the margin of error for that sub sample size would be plus/minus 6%. So the first month drop is real but the subsequent rise may not be. On the other hand the under 35 Yes at 59% and No at 23% is a difference of 36%, many times the margin of error, so its real.
Keeping the margin of error in mind
-Yes does better than average with under 35s & ABC1s
– No does better than average with over 55s & C2DEs but pensioners are probably tilting that C2DE average towards No because of the over 55 effect. C2DEs are still voting Yes by a margin of 6%
A word of warning for Together 4 for Yes – when asked if they would vote in a general election (p31) 79% of the over 55s will definitely vote as against 60% of the under 34s. If the Yes vote falls and the referendum was close that higher over 55 turnout could defeat repeal the 8th. But of course intention to vote in an election isn’t the same as intention to vote in a referendum – anarchists vote in referendum only rarely if at all in elections, see https://www.wsm.ie/elections for why
Finally we look at the ST/BA poll and the data on how the vote breaks down by political party. And here we see the reason for Fianna Fails self destructive opportunism of trying to ride both horses at once, their voters are split 50:50, 41:42 to be precise. Rural Fianna Fail will want to be visibly No but that could spell the doom for urban Fianna Fail who already face decimation from the growing urban Sinn Fein vote that is hovering up what used to be the Fianna Fail youth vote/
All the other parties have far more Yes voters, the lead of Yes over No for each of the major parties is
Fine Gael 24%
Labour 30%
Sinn Fein 30%
April polls show No campaign not making ground – April 22nd
The Irish Times with MRBI published an opinion poll on the Repeal the 8th referendum 20th April that once more showed a strong Yes lead and a static No. It is the 6th poll of the year so we thought it useful to generate a side by side comparison of the Yes & No votes for all 6. The polls were carried out by MRBI, Red C and Behaviour & Attitudes.
IT/MRBI poll has
- Yes 47%
- No 28%
- Don’t Know 20%
- Won’t vote 3%
With undecided excluded that come down to
63% Yes
to
37% No
With the chart above we eliminated undecided from all 6 polls.
Other significant finds from this mornings poll included;
A narrow majority of Fianna Fail votes favour repeal
39% Yes
37% No
This is very bad news for No campaigns as it suggests there will be no further opportunism from Fianna Fail TDs who might otherwise think campaigning for a No would damage their Fine Gael or Sinn Fein rivals. This is now only the case in some rural areas and the advantage would be small but the Fianna Fail party nationally would have to be conferenced that any further prominent opportunism from rural Tis would further damage their chances of an urban recovery rather than the permanent loss of seats to Sinn Fein. Fianna Fail did a press stunt yesterday where several party members appeared with a Together for Yes banner, cynically we suspect this was because they were forwarned of todays poll showing No had failed to gain.
Most importantly the Yes vote appears to be very very solid with 80% of Repeal voters saying they would never change their mind. Its probable the sheer toxicity and disinformation of the No campaign as well as the saturation coverage in terms of billboards and online ads has backfired and solidified the Yes vote.
Also of great significance undecideds are leaning 2:1 towards Repeal – this is unusual in a referendum where the assumption is that a majority of undecided would opt for the status quo and vote no.
If the two findings above are accurate the No campaign has no chance of winning on polling day. This along with their failure to significantly increase their share of the vote since February suggests outside of some major Yes mishap it’s over for them.
In this MRBI poll voters also reported a high degree of knowledge of post referendum legislation – only 15% said they were unaware – and Repeal voters showed the highest level of knowledge. This means the No strategy of fake claims and misleading posters has not only failed but probably backfired and instead is motivating Yes voters and alienating undecided’s.
But while there are strong grounds for optimism it’s not over yet. Dangers include the attempts by No to make the campaign bitter and nasty, particularly in Dublin, to try and drive down turnout, particularly of younger voters. They probably hoped to provoke a response in kind from the Yes campaign but it has stayed focused on compassion and women’s health.
The Irish Times headlined this as a slippage in the Yes vote in comparison with their January poll, something that is present until undecided are excluded when the apparent shift is then smaller than the margin of error. RED C showed a similar slippage between their January and March polls which we discussed in depth at https://www.wsm.ie/c/repeal-8th-opinion-polls-analysis but at this point we’d acknowledge that its likely there was a loss of soft Yes votes back then before there was significant campaigning.
The weakest point for Yes remains the 12 week unrestricted access which is why No will continue to try and centre that discussion and avoid the discussions being centred on protecting womens health and fatal foetal abnormality. There the No vote is very soft indeed, half of No voters in the B&A poll actually wanted abortion access in those cases, meaning No could lose half their vote if protecting women health becomes the main issue under discussion. The other half are the core 15% ‘let women die’ – a figure that has remained constant for the last few years.
From their messaging its clear that the No campaign recognise that they are not likely to erode many Yes votes on the 12 week issue and the Irish Times poll confirms that. The percentage saying 12 weeks goes too far (41%) is almost identical with the percentage saying abortion is wrong and should not be more widely available. (40%). 56% said they had reservations on 12 weeks but it was a reasonable compromise. Presumably a recognition that there is no other way of providing abortion in the case of rape and that 12 weeks is the current reality in Ireland because its the end of the period where the abortion pill can be used. The abortion pill may be illegal right now but the reality is women are taking it every week and that usage is increasing. Unless the state starts prosecuting women for its use – and that would carry a 14 year jail sentence without repeal – women will continue to use it regardless of the outcome of the referendum. Most voters want these women to be able to access medical care with the risk of jail.
This mornings poll should have been the strongest by far figures for No campaign as they were organised to campaign earlier and are spending huge amounts on billboards & ads while #Together4Yes was still in the process of raising funds. Two weeks back No spokespeople on Twitter were crowing that the Yes campaign was nowhere to be seen – and this poll would have been collected in that period. Last week saw the enormous crowdfunding drive by Together for Yes with 550,000 being raised through over 10,000 small donations, many with names and indeed moving stories attached. Just who is funding the No campaign on the other hand is murky and unclear – its widely understood that huge quantities of dollars have flowed in over the last years because extreme US christian groups see Ireland as a key battleground in their ‘crusade’.
What has also been striking on Twitter is that the No canvass groups remain smaller and appear to have become less frequent, particularly in Dublin while the Together for Yes canvass groups have appeared everywhere, including rural areas that didn’t see Marriage Equality canvassing and some of the groups are enormous. We’ve seen photos of canvass groups in individual Dublin constituencies that have had 50-70 people on them. We suspect the early start of the No campaign and the enormous amount of money they are spending on achieving saturation advertising everywhere from billboards to children’s computer games has really motivated Repealers to donate and canvass. A massive rebellion against the hated 8th amendment is very much in full swing and the status quo looks like its going down to a major defeat, and not just in the cities.
These are the 4 polls carried out since Repeal 8th referendum was declared. They show the impact of campaigning, mostly of the No side as Together for Yes got underway later. So both anti-choice groups, Save the 8th and Love Both have had no real impact on voters despite the enormous spend on posters, billboards & online ads.
Incidentally we have seek the claim that todays MRBI poll was the first time the Yes vote when don’t know are included fell below 50%. This claim is false, the yes vote was at 49% for both the B&A polls here, ie for the period of the campaign it has remained just under 50% for all but one poll.
There was also a poll in the Daily Mail but because of the loaded way the question was phrased (opposite) to include unrestricted access to 12 weeksthis means its not comparable with these other polls.
That Daily Mail question is not standard. In fact its what reputable polling companies have been using to see how soft the Yes vote is.
Behaviour & Attitudes for instance ask how people intend to vote on the 8th as one question and what their attitude to 12 weeks as a follow on – in their April poll 47% are for Repeal but only 43% in favour of 12 weeks, the gap showing how important the question asked is.
To combine the two isn’t just a loaded question, its most of the basis of the No campaign. It’s why reputable polling companies also have a soft No question about protecting health & cases where foetus will not survive (FFA). So if you are paying for a poll and the question you pay for is set up as the Daily Mail poll was then in effect you are paying for a result. If NO canvassing teams are demoralised it might fool them to have some hope to keep going
In that context the Daily Mail reported
- Yes 46%
- No 31%
- Don’t Know 16%
- Not Say 8%
Even in that context though when you account for margin of error and compare with their previous poll only this is the 7th poll to show the No campaign has had no significant impact
The polls to date
Pre-campaign
- MRBI January – https://www.irishtimes.com/news/politics/clear-majority-backs-abortion-on-request-up-to-12-weeks-poll-shows-1.3368816
- Red C January – https://www.redcresearch.ie/wp-content/uploads/2018/01/SBP-Jan-2018-Poll-Report-GE16.pdf
Polls during campaign
- B&A February – http://banda.ie/wp-content/uploads/Sunday-Times-Report.pdf
- B&A March – http://banda.ie/wp-content/uploads/J.8878-Sunday-Times-March-2018-Report.pdf
- Red C March – http://www.redcresearch.ie/wp-content/uploads/2018/03/SBP-March-2018-Poll-Report-GE16.pdf
- MRBI April – http://www.irishtimes.com/news/social-affairs/irish-times-poll-public-favour-repeal-of-eighth-despite-slip-in-support-1.3467503 with a nice viewer of the data at https://www.irishtimes.com/news/politics/poll
- B&A April – http://banda.ie/wp-content/uploads/J.9001-Sunday-Times-April-2018-Report.pdf
- Red C April – http://www.redcresearch.ie/wp-content/uploads/2018/04/SBP-April-2018-Poll-Report-GE16.pdf
Repeal the 8th isn’t in the bag yet – a warning from Marriage equality poll comparisons – April 30th story
Today we are warning that Repeal the eighth vote can’t be assumed to be already won even through No have failed to increase their vote. If the same last week shift comes into play as for Marriage Equality & Divorce we are looking at a result to close to call. We are going to explain why we are saying this in detail using polling figures from Marriage Equality and this campaign. Its going to get a little scary as we go through these but there is hope at the end.
The 1995 Divorce referendum looked to be in the bag from advance polls with a 2:1 lead but on the day was only narrowly carried 50.3% Yes to 49.7% No. A result so close that its said the difference was only one vote per polling box in the country. It appears with Divorce that either the polling companies had got it wrong or more likely a major drop in the Yes vote happened in the last 10 days. This was too long ago to draw detailed parallels between polling but we can do just this with the Marriage Equality polls.
Marriage Equality also saw a sharp drop in last week. There were 4 polls published a weekend before by the same range of companies polling this time around – the top left table shows each poll and then actual result first for raw data, then with Don’t Knows excluded, as you can see all 4 polls overestimated the share of Yes by 7 to 9%.
This table below compares each companies poll with the actual result & then calculates how much of Don’t Know became No and in the two cases where that was over 100% how much of the Yes vote became No.
Using these percentages in the bottom table we recalculate all the polls by applying the error ratios for each company from the Marriage Equality referendum to each of their polls on the Repeal the 8th vote. IF (ands its a big IF) it was same error this time around the referendum would only narrowly be carried in the B&A poll and narrowly defeated according to both MRBI and Red C polls.
Adjusted figures if the polls are as far out on Repeal as they were for Marriage Equality
To be very clear this is not a prediction, simply a warning that we can’t be complacent (who is!). The campaigns are very different in intensity & length and there is some evidence in B&A polling that the Don’t Knows are splitting to Yes.
One major difference is that the No campaign have ran a very intense campaign over months with a huge spend on misleading ads since February. Despite this the polls show a failure to increase the No vote at all. All the variation in the polls of the last 3 months have between polling companies. For each company their polls across the months show no change when you account for the 3% margin of error.
But the No campaign is not focused on convincing additional voters to No but in trying to make the referendum so nasty and confusing that undecided & Soft Yes voters will stay at home or, our of fear, opt for the status quo. This would explain the often bizarre nature of the No campaign, their failure to discourage aggression, lying and indeed the active encouragement and even participation of their spokes people in such tactics. Disengagement and fear helps them, in 1983 only 54% of the population voted in the 8th amendment referendum which saw similar tactic and worse.
But this may not work this time around. Much has changed in Ireland, in particular rural areas, and polls indicate that the Yes voters are quite determined and sure of their understanding of issue. 80% of Repeal voters say they would never change their mind in the last MRBI poll
Together for Yes is already canvassing areas that never had organised canvass teams during the Marriage equality referendum. Its expected that as May 25th approaches the Yes canvass will substantially grow in size & scope, as happened with Marriage Equality. Canvassing is something you can do that will make a difference , you can sign up for. Together for Yes have already distributed three times as many Yes posters as were distributed during entire Marriage equality campaign – in general you can say Together for Yes learned from and then built on the Marriage Equality ref experience. Something that is of course also true of the No side, in that case its the same organisations and individuals running this campaign as who campaigned against Marriage equality (and indeed sex education, contraception and divorce, if they have been around long enough).
All the polls so far were from the period before Marriage Equality had even really launched but the No campaign has been in full swing since February. So it may be we will see the Yes vote increase and No vote decrease in the next set of polls.
So again this scary comparison with the Marriage Equality polls in NOT a prediction but simply a warning that the strong and consistent Yes lead does not mean Repeal is already in the bag, its very much up to YOU to be part of a Together for Yes win by talking to friends, fellow workers and families and by getting formally involved in canvassing and other work. As the referendum vote approaches people will pay more and more attention to whats said and written, particularly in final days. So its not too late to get involved, indeed we expect that the early starters will really appreciate new faces and new energy in helping Yes win on the 25th.
Repeal vote still too close to call when likely Don’t Know to No taken into account 5 May 2018
The referendum to remove the clause in the constitution that limits what medical care, including abortion, women in Ireland can access approaches at the end of this month. With another poll appearing this morning we have updated our graph of how that May 25th vote would look IF the polling companies had a similar margin as they had for the Marriage Equality referendum a couple of years back. As you can see they suggest if nothing changes the result of the May 25th, Repeal vote will be too close to call until the count on the 26th.
We are going to post what the polls actually say in the first comment to this story, all the actual polls show Repeal passing. This graphic isn’t a scientific prediction, simply a warning that in reality the result may be too closed to call and so every effort made in the remaining 3 weeks will matter. In 1995 Divorce passed by a tiny margin of votes, about one per ballot box – something similar is possible this time too.
This latest poll is from the Sunday Independent and carried out by KMB. It shows an increase in the No vote. Their raw numbers were:
Yes 45%
No 34%
Don’t Know 18%
No response 4%
In the 2015 Marriage Equality referendum there was a sharp drop in the last week with many Don’t Knows switching to No. There were 4 polls published the weekend before the vote by same range of companies polling this time around and we calculated the error made in each companies poll then and applied to the poll data for the same company this time around. KMB were pretty close, so their adjustment is small. This is explained in more detail in our first piece on this topic, below
This poll would appear to indicate that the No method of spending an enormous amount on online advertising, billboards and posters to spread misleading fear-based messaging is having an impact. We estimate that while Yes have 5 or more times as many volunteers No has 10 or more times the cash. Whereas the source of Yes funds is pretty transparent very little is known as to where 95% of the No spend is coming from, speculation is a lot is coming from outside the country in contravention of the rather toothless SIPO rules which are powerless to prevent online ads being placed from elsewhere.
The question is what will turn out to be more persuasive over the next 3 weeks, the Yes strength of personal contact through canvassing or the No strength of a deluge of online and offline paid advertising. One thing that can be said is Facebook’s failure to introduce meaningful ad transparency is in danger of deciding another vote.
We can also see the problem with referendum campaigns coming through – by their nature they are short term attempts to convince people to come over to a side rather than to change fundamental attitudes. The huge age discrepancy between Yes & No illustrates how central the tactic of playing on deep-seated prejudices against women are for the No campaign. Such prejudices have shifted substantially for most younger voters – there is almost an opposed pre & post 1960s feminism vote at that level, something that can also be observed in the Trump and Brexit votes.
BTW we’ve seen anti-choice spokespeople claim this result reflects how people felt about ‘who won’ the Late Late Repeal show. This isn’t possible as that broadcast was 27th and this polling was carried out from April 18 to 30th, i.e. 3/4 before that broadcast.
No fails to close gap as we enter final week of campaign – 17 May
Its just over a week before Ireland will have a referendum to remove the clause in the constitution the equates the life of a women with that of a foetus. The Irish Times/MRBI have published a poll showing Yes has a 16 point lead over No with one week to go to the referendum. The detail of the poll also reveals a surprisingly large soft No vote still exists. 22-35% of No voters should be voting Yes according to their opinions on wanting more abortion access for women.
But we also continue to warn that if a similar last minute drop in support fort Yes happened as occurred with the Marriage Equality referendum the strong lead in this poll would reduce so that the vote was too close to call. In other words there should be no room for complacency, its likely every vote will count as happened with the 1995 divorce referendum which pass with a tiny margin of 50.28% Yes, equivalent to a couple of votes per ballot box.
No still has a significant soft vote with;
A. 35% of No voters feel access to abortion up to 12 weeks on request is a reasonable compromise. Presumably because they know proving rape cannot be made conditional on accessing abortion – trial take weeks, not months and conviction rates are low. The recent Belfast trial is probably in many peoples minds. But 12 weeks on request can’t happen without a Yes vote to Repeal, its completely impossible with the 8th in place.
B. 22% of No voters say the law needs to change to recognise a women’s right to choose to have an abortion. This is a VERY soft No indeed as clearly Repeal the 8th has to happen to make this possible, there is no way this 22% should, be voting No while holding this opinion.
These detailed figures are available here https://www.irishtimes.com/news/politics/poll
Both these large soft No blocs suggest that if people fully understood the referendum and the implications of a No vote as a continuation of cruel regime of the 8th then there should be at least a 10% switch from No to Yes with the referendum then passing 63% Yes to 39% No. This may explain why No has fallen back on distrust, fear and confusion messaging along with trying to make referendum campaign as toxic as possible. They need not to allow the 22% to 35% soft No voters to consider the implications of a No victory.
Overall the Irish Times / MRBI poll has responses as follows
44 Yes
32 No
17 Don’t Know
5 Won’t Vote
2 Refused to say
The poll was collected Monday & Tuesday of this week, the controversial Clare Burns Live shows which has a strong No bias in speaker section and time was broadcast Monday night so No might have expected a much stronger showing in this poll as they were very loudly declaring a victory after than broadcast. This poll however shows no significant change in comparison with the KMB poll taken a couple of weeks beforehand.
Dublin remains strongly Repeal with Ulster/Connacht which has seen large scale suppression (tearing down) of Yes posters being the weakest, close to 50:50. It’s also the smallest regional bloc of voters.
The detail of the poll also reveals that the Urban V Rural divide has close a lot during the campaign, according to this poll
Rural V Urban
Yes 39% V 46%
No 36% V 30%
Age remains the biggest dividing line, over 65s are the only group strongly intending to vote No
18 to 24 V 65+
Yes 52% V 30%
No 27% V 47%
Older voters won it for both Trump & Brexit, will the same happen here and the same sense of a betrayal of grand children’s futures by their own grand parents – in this case where ether consequences don’t even directly effect them? Talk kindly to your grandparents about this.
With this question its interesting that with the 65+ and 18-24 age groups the Don’t Knows lean towards No strongly but all the in between age groups lean towards Yes. With the 65+ we can say this is the influence of clerical ideology over a life time. With the 18-24 this is the group most likely to see the enormous No spend on misleading online adverts and which doesn’t necessarily have the life experience to measure this against. In terms of voter mobilisation it may be that No’s early contempt towards younger (student) voters may hurt them here.
The IT/MRBI last polled in April, this is the shift between their April poll and this one
April->This poll
Yes 47->44
No 28 ->32
DK 20->17
Won’t vote 3->5
Won’t say 1->2
It’s unusual for Won’t vote and Won’t say figures to rise in the course of a referendum. This very much reflects the hate and fear messaging of the No campaign along with the widespread attempts to sow confusion and doubt. It’s clearly scared some intending voters from voting and made others fearful of speaking. This also happened in 1983 leading to a tiny turnout of 53% on the day. No have been busy in the last few days setting the ground for declaring the referendum to be a fix, a low turnout would also help them in that respect so its useful to be aware that they are trying to engineer a low turnout. At times this has been blatant, as with releasing press releases saying colleges should not close on the 25th to allow students to vote.
A lot of people are very sure of how they will vote, this is obviously good for Yes and bad for No as it suggests the pool of voters who can be shifted is not large. Yes has very much larger canvass teams so they are much more likely to be able to sway this pool in the final week.
65% never change
22% extremely unlikely
9% unlikely but open
4% possibly
We presume there will be at least another 2 polls out in next few days if all major polling companies are polling. The MRBI ‘weekend before’ poll was closest to being correct for the Marriage Equality Referendum result, more on this important comparison for the previous polls and how we do this calculation
That said in bad news for No and going somewhat against our Marriage Equality ref comparison more Don’t Knows are leaning to Yes than No
Lean Yes 31%
Lean No 24%
Not leaning 44%
But if we assume all No leaning Don’t Knows become No votes this poll becomes one very close to what happened with MRBI and their Marriage Equality referendum polling, bringing us back towards a too close to call result. Again we say this is a warning and not as a prediction of actual outcome, there is far more for the No campaign to sweat over in this poll which is why they are preparing the ground for their defeat and claims of a rigged referendum.
Canvassing and leafletting for Yes is going to have a massive impact in the remaining week of campaigning. So too will talking to your friends, workmates and relatives. Divorce passed in 1995 by a tiny fraction and its looking like it could be very close for Repeal so do have those conversations. No want a low turnout, we want a high one so vote early on the 25th, and post to social media that you have done so to encourage others to do likewise.
Final – All Polls shows Yes lead for Repeal but it could still be close if previous patterns hold – 21st May, the weekend before polls
Two new opinions polls were out Sunday morning on next Fridays referendum to remove the hated 8th amendment to the constitution that equates the life of a woman with that of a foetus. Both this mornings polls are good news for the Yes to Repeal campaign with an increased Yes vote since the same companies’ previous polls.
The Red C poll shows a 3% increase for Yes, 1% increase for No since their last poll.
The B&A poll shows a 5% increase for Yes to 52% with 5% fall for No to 24%. With Don’t Knows excluded, this is 63% Yes 32% No.
The MRBI poll published by the Irish Times on Thursday wasn’t so good, it had a 5% drop for Yes in comparison with their previous polls, suggesting a very much tighter 58% Yes, 42% No vote.
What is important when looking at these polls if not the date they were published but rather the dates data was collected over. All 3 polls have collecting periods overlap (see graphic). The MRBI was collected over a short 2-day period 14/15 May and has the worst result for Together for Yes . The B&A poll was collected over a long 12 day period and appears to have the best result for Yes. The Red C poll was collected over 6 days and with don’t knows excluded has the same result at the B&A one.
As the polls overlap we have to consider that all three are correct even though the numbers differ. B&A and RedC are similar when Don’t Knows are excluded but MRBI is radically worse for Yes. As MRBI collected towards the end of the period it MIGHT be showing a sudden No swing although we consider this unlikely. Its more likely that the difference reflects long standing differences in polling methodology.
When we adjust each polling companies results by how far out that company was on the 2015 Marriage Equality referendum vote its the Red C poll that gives Yes a clear win, the other two would be too close to call. Why?
With their weekend before Marriage Equality poll B&A had
Yes 63
No 26
Don’t Know 11
BUT result was
Yes 62
No 38
So ALL B&A Don’t Know went to No in that Marriage Equality referendum and so did some yes votes. If the same happened on May 25th the Yes would only have 51%, way too close to call.
https://www.wsm.ie/c/repeal-8th-warning-marriage-equality-poll-comparison
This isn’t intended as a prediction, just an illustration that Don’t Knows might break very heavily for NO and in that case 2 of these 3 polls say its still too close to call while just one is a Yes win.
In other words while the polls are good news for Yes its all still to play for in the final few days. At this stage getting out the Yes vote is going to be central as in a close result who votes could decide it. eg Over 65s are most likely to vote and are heavily No, this age group was key to Brexit passing in the UK.
There remains a sharp urban rural divide so a sunny day in the west and heavy rain in Dublin could have a big impact. If your a Yes vote make sure you actually vote and spend the week talking to other Yes voters encouraging them out too. We can win this but its not yet won.
Here are our 8 reasons we are voting Yes to Repeal the hated 8th https://www.wsm.ie/c/8-reasons-vote-yes-to-repeal
Words: Andrew Flood (follow Andrew on Twitter)