Our cup runneth over….

Well. final stretch of the stability/fiscal/austerity treaty, and a rake of polls to get through, all of them showing the YES side edging it, but not by enough to be sure.

First out of the blocks was MRBI, released on Friday evening and published in yesterday’s Irish Times.  As you’ll remember, the final MRBI was closest for both of the last two General Elections than all other companies, and this was also the most recently taken of the weekend polls (23-25 May), so a lot of people would be watching this closely.

The poll shows a decrease in the number of undecideds, with them breaking more or less pro-rata, Yes up 9 to 39%, No up 7 to 30%,  with “No Opinion” down to 22%. After DKs are excluded this came to 57-43.  The movement was from the previous MRBI which had been taken 5 weeks ago, and suggests that, despite a number of gaffes on the YES side, the voters are inching towards passing the treaty, albeit unenthusiastically.

Second to reach us was the RedC poll for the Sunday Business Post.  As with most RedCs, this was collected between Monday and Wednesday (21-23 May), and so was a slightly earlier poll, measuring opinion about half a week earlier than the MRBI offering.  This showed the Yes side down 4% from their previous poll to 49%, the No side up a corresponding amount to 35%, and DKs unchanged at 16%.  When the DKs are taken out, this comes to 58-42.

Third out was the Sunday Times  B&A.  This found that the Yes side had gained 3 points to 45%, and the No side was also up 3% to 30%, with DKs down to 24%.  DKs excluded this amounts to a 60-40 split.  The fieldwork was earlier still at 18-23 May.

Finally, the Sindo commissioned a real poll this week, with MB/Lansdowne offering arriving late to the party.  The reporting is awful sh!te and is trying to pump up interest in a poll with its thunder stolen, but the method I’m sure was excellent.  The figures were Yes 42% (+5), No 28% (+4), and with undecideds excluded it came out as 60-40, unchanged from the previous Lansdown offering.  I can’t find a reference to when the fieldwork was taken, but would guess around the time of the RedC or slightly earlier based on previous examples.

So what to make of it?  Well, the number of DKs varies, but we’d expect that, both through the various polling methods (RedC for example only count those who self describe as 70%+ likely to vote) and of course nearly a week in the difference in fieldwork, when voters might be likely to be making up their minds.  And yet, despite that, the variance in Yes/No is within the margin of error, with the yes chronologically moving 60-60(assuming MB/L 2nd taken)-58-57.  It might look like a slight trend to the NO side over the week involved, but 3% max.

Given the differing levels of DKs, this suggests that they are breaking very much pro-rata with perhaps a slight boost to the No side, but the DK numbers are getting too low to be enough to defeat the referendum if breaking in those proportions, or even a bit higher.  And remember, most DKs are non-voters who prefer to put it in those terms.

If there was indeed a 3% fall in the YES support over the last week, and it was repeated in the final week of the campaign (OK, big if), you’d expect it to be about 54-46 on the day.  The No side would then be hoping that turnout would help them, but it would still appear too much of an ask to me.  Say, for example the turnout among yes ‘voters’ was 50%, the ‘no’ side would need about 59% of their people to turn out.  That would mean a No voter was 18% more likely to vote than a Yes voter, and this seems unlikely to me given their respective demongraphics, and so if the referendum is rejected, I believe it will be a result of actual movement from the Yes to the No column.

Might that happen?  Yes, it may.  Certainly the Sindo are talking up the influence of their columnist $hane Ro$$ who comes out against the treaty today.  And it’s true that the financial journalist who backed Sean Fitzpatrick and Michael Fingleton is not without his supporters among the wider electorate.  But a lot of that is already priced into these figures I think, I certainly reckoned he was voting No by now.  If anything, he may get a few more Yes voters out, as many referendum voters vote ‘against’ the politicians they despise rather than in favour of the ones they offer grudging support at election time.

So, all in all, still plenty to play for.  Advantage Yes, I think, but close enough for both sides to scrap for every vote…