The Day After
Also 78th Anniversary of the sinking of the Bismarck and David Mundell's 57th birthday
© Traditional, some time before 1841
What could possibly go wrong but did not
I must confess that, during the last few days before the election, I was worried that Scottish polls might prove as unreliable as before the 2014 EU election when they overestimated the SNP vote by 7% and underestimated the Conservative vote by 4%. I did not voice it though, as I had warned about polls possibly getting it all wrong already some time ago. And it certainly did not feel appropriate to repeat such a warning in the last mile before the vote, when the priority was to energize all SNP activists and all SNP voters and deliver at least three SNP MEPs. First a reminder of what the last batch of polls predicted, only recalculated on the actual 2019 turnout rather than the 2014 turnout I used before.
And now what actually happened. First important thing is that valid votes increased by 227.763 from the 2014 election and the SNP vote increased by 205.050, meaning the SNP bagged a mammoth 90% of all new votes. It's quite an understatement then to say that #ActiveSNP worked to perfection and beyond and all who took part in the campaign can be proud of what was achieved. Obviously quite a few, including myself, will feel a bit miffed that the SNP did not bag a fourth seat and that Margaret Ferrier will not sit in the European Parliament. But it was the result of a combination of factors, some of them unpredictable, and the SNP missed the fourth seat by 33.838 votes. Missing it by 500 would have been devastating, the actual result is just a wee smitch disappointing and takes nothing off the superb work done by all SNP activists and sympathizers over the last two weeks.
Then some others have many more reasons to feel devastated. First of all the Scottish Branch Office of the English Labour Party. Which looks more and more like a Twig Office by now. 9% is not just Scottish Labour's worst ever performance at an European Parliament election. It is also their worst overall since 1910 when they first stood in Scotland as The Labour Party. Same goes for the Ruth Davidson Team who scored their worst ever result in all recorded Scottish electoral history, that is since 1832. Some Tories would gladly take the whole UK back to Dickensian times and voters did just that to them. Well done lads and lasses.
What could possibly go wrong and went even worser
Just a casual look at the map makes my point. A vast ocean of turquoisish Brexit blue Doon Sooth with just a few traffic-light splotches every here and there. Then a yellow sea North Of The Wall with just two weeish orange islands to the Far North. So yes everything that could go wrong did and then some and beyond. Then I guess the Conservatives and Labour can only blame themselves for the multiple trainwrecks. So massively underperforming on already lousy polls is quite a feat in and of itself, and all the more remarkable when the two supposedly 'main' parties manage to achieve it simultaneously. Because it already happened like…. never.
Why Labour deliberately torpedoed themselves with shitloads of inconsistency is hard to fully grasp. Had it not been for Corbyn channelling his Inner Eurosceptic and missing every opportunity to actually listen to the people, including his own voters, events might have taken a decisively different turn. With the Conservatives already in tatters and Farage having nothing to offer but offensive populist demagoguery, they could have won this if their main concern had not been short-sighted internal parochialism. But they entangled themselves in a web of obfuscation and contradictions. Their loss. Guess the healing will take some time. It's quite different for the Conservatives. They expected the clobbering and were possibly quite relieved it wasn't even worse, like in zero MEPs. Only in a few weeks will they realize they just signed their own death warrant.
Now on to something completely different and a wee smitch of food for thought. Lets just wander through the mirror to a dreamworld where the LibDems, Change UK and the Greens somehow managed to stand on joint lists, however impossible, implausible or unbelievable it might look. It's my dreamworld scenario after all so I can do what I want, or can't I? This Coalition Of The Wishful would of course be an England-only thing because you can't include those Pesky Nats from Scotland and Wales in such a Grand Scheme, or else they might gain some more seats they would use as a mandate to destroy Our Precious Union. Anyway here's what the results would have looked like:
There's some irony here as United Remain would barely have dented the Brexit Party contingent but instead made both Labour and Conservatives look even more destitute than they are in the real world, and bagged just six more seats than LibDems and Greens combined at the real election. So, even if this was fleetingly an option under the guise of tactical voting, it would not have delivered such a dramatic change and the Brexit Party would still have emerged as the first party after the United Remain seats had been re-allocated to the various member parties. But it looked like a good idea a the time, or didn't it?
Did pollsters nail it?
In the run-up to the election we had a massive batch of polls, eight published over three days 20-22 May, with a grand total of 16,384 respondents. So it seems like a strong and valid enough basis for comparison with the actual results. Data are for Great Britain only as pollsters did not include Northern Ireland. So here is the comparison between what the last batch of polls said and the results:
I mentioned earlier that the 2014 polls closest to the election did not miss the actual result by much. This is not the case this year and sometimes by a significant margin. You must bear in mind that pollsters never publish their raw results but weighted results. The classic and basic weighting is based on demographics, region and recalled votes at previous elections. Then data are usually 'super-weighted' (or should I say 'over-weighted'?) by likelihood to vote at the next election. There are many weak links in the chain, with 'likelihood to vote' probably the weakest of all.
Anyway pollsters here missed the weight (pun fully intended) of the combined Europhile votes. They correctly saw a trend towards increased LibDem and Green votes but failed to grasp its magnitude. Missed by 7.5% total, which is way beyond the usual margin of error. And it has direct impact on the seat projections when these are allocated on PR. Incidentally I am always utterly flabbergasted by the way various media outlets perpetuate the urban legend that d'Hondt (or 'highest averages method', to give it its proper technical name) is 'complicated' or 'complex' while it is neither but rather basic maths. All it takes to make it work properly is to remember that 'average' here does not mean [ Votes / Seats ] but [ Votes / (Seats + 1 ) ]. Simples. Not exactly quantum physics, or is it?
Anyway here is the comparison between the seat distribution generated by the latest batch of polls (based on individual regional seat projections from the various crosstabs) and the actual result. For all 73 seats as we have now the official Northern Ireland results. Interestingly the third Northern Ireland seat switched from the UUP to the Alliance Party as I predicted, and the Alliance did even better on first preferences than the two available NI polls predicted.
At least pollsters got the SNP and Plaid Cymru right too, but these were the easiest to get. Then I am quite ready to consider that a number of voters had last minute second thoughts that polls could not properly capture, which might be admissible if the very last poll wasn't also the most off of all with Labour on 23%, Conservatives on 13%, LibDems on 12% and Greens on 7%. So all the more reasons to handle polls with care now more than ever. But of course the main concern here is Scottish polls, which have proved highly unreliable at almost every election for years and especially in 2014. So here is what we have:
Like in GB-wide polls there are multiple errors all over the place, the most noticeable being an overestimated Brexit Party and understimated LibDems. And I won't even mention getting the order all wrong between Conservatives, Labour and LibDems. But I just did, didn't I? Never mind because what matters here is the SNP vote and polls were remarkably close to the actual result. And as this was a PR election you can realistically expect the SNP to do better by a few points at an FPTP election where smaller parties (think Scottish Greens) would sit out most constituencies. This is important because it validates ex-post-facto all past polls that had the SNP in the low forties for the Westminster and Holyrood constituency votes, and in the mid-to-high thirties for the Holyrood list vote. Not to mention these polls that Unionists hate when they say the Yes vote is now on 48-49% and were uncoincidentally fielded by the same pollsters who also polled the EU election in Scotland-only polls.
Now on to something completely different, to end on a lighter note. I am totally chuffed that I don't have to roll back any of my comments about Change UK. The useful idiots were predicted to do poorly and they did even poorlier. Ending up trailing the SNP and barely ahead of the remnants of UKIP is certainly not what best befits Chuka Umunna's high self-esteem. And they achieved exactly what I predicted and beyond: channel the English Europhile vote to the real parties, LibDems and Greens. Guess that could be the final toll of the bell for a hardly-existant party with ectoplasmic ideology, and the hypothetical merger with the LibDems might well end up as a not-too-friendly takeover. Not that I will miss them anyway.
And now what?
Both Nicola Sturgeon and Ian Blackford nailed it in the aftermath of the election. And Paul Kavanagh's take was both right on target and entertaining as usual. The people of Scotland have spoken and sent a crystal clear message. Now even the most pig-headed Unionists can't acknowledge it. Oh wait…. of course they can and they will, just as they always ignored anything that does not fit the never ending #SNPBad narrative. But is it bound to last long? This election clearly makes the case that Ruth Davidson's obsessive anti-Independence rhetoric does not work. And also that the SNP were definitely right to link stopping Brexit, Independence and European identity in the campaign.
Now that we have conclusive evidence that the SNP can actually bag 50+ seats at the next GE and possibly oust all Scottish Conservative MPs, and also that a renewed pro-Independence majority is the most likely outcome of the next Holyrood election, the SNP have all the reasons they need to stay on course on the path to Independence and act swiftly and decisively on the triple-lock mandate. Introducing framework legislation for IndyRef2 next week is the much welcome and much anticipated first step. Then the closer we get to a Halloween no-deal Brexit, the better it will get and the more likely it is the 49% Yes will morph into 50+. Nothing's gonna stop us now 🔊.
Wha daur meddle wi' me
© Dougie MacLean, 1977
No comments:
Post a Comment