A few points based on the '100% certain to vote' column for YouGov.
When they do a poll they discount all people who don't know how they will vote (unless its a GE - it which case for the final poll they try to work out what the don't know will do - and this is where the devil in the detail is). I don't believe that YouGov did this finer detail for any of their EU polling.
SO: If Remainers are more likely to not know how they will vote, then they may be more difficult to predict. (Note, I didn't know what I was going to do until in the booth, so in theory I wouldn't show up on this polling).
In earlier polls, you would expect a lower number for all groups for 'certain to vote'. You are possibly more likely to say you will definitely vote, if you KNOW, who you are going to vote for.
So here's a comparison from the YouGov 8/9th May survey and the 19th - 21st May surveys.
One of the interesting things is the % of 18-24 year olds saying they are definitely going to vote, doesn't shift at all. It stays solidly on 39%. It suggests no one is engaging at all with the 18 - 24% group in this election. All the other age groups shift. The purple circles are the 65+ age group which is the largest group in numbers - this goes up 5%. But actually the biggest % shift is in the 50 - 64 age group with an 11% shift. The group that reduced May's majority at the GE were the 25 - 49 age group though (it wasn't the real youth vote, but the 'middle aged vote' swinging at the last minute from Conservative to Labour that made the most significant difference).
What bothers me is really where momentum has been. In the 8th/9th survey 55% of Leavers said they were certain to vote. This was LESS than the 58% of Remainers who said they were certain to vote. BUT this switched by the 19th/21st. Leavers were more likely than remainer to vote with 63% to 61%.
The changes in the polling fortunes for the the Brexit Party and the LDs has come from 2017 Con and LD voters becoming far more likely to vote. 2017 Labour Voters haven't been motivated as much in either direction. 2017 LD voters are THE most likely group to vote. But theres loads less of them than 2017 Conservative voters. The apathy of 2017 Labour Voters is really showing here.
And thats why YouGov have put the Brexit party as high as they have in their polling.
BUT the big cavet is the 'don't know group' who were not sure how or if they would vote. And these would fall MORE into the remain column than the Brexit column precisely because remainers have a more difficult decision to make. And thats the only hope I can really see here tbh.
A low turnout means that this election is particularly volatile in terms of polling accuracy.
And thats what seems to be being said on twitter:
Adam Ramsay @AdamRamsay
SNP, Lib Dem and Green sources all agree massive turn out in heavily Remain and young areas of Scotland, but lower turnout elsewhere. #EuropeanElection2019
The issue with this is, that the Brexit Party have a broad base appeal geographically whilst the LDs /Greens in particular have traditionally been very concerntrated so you EXPECT high turnouts in Remainy areas.
Sam McBride @SJAMcBride
The turnout picture from figures I've seen & from political sources is of many polling stations across rural & urban areas of Northern Ireland where turnout is down & none where it is up on the recent council elections - that suggests less than half of the electorate has voted.
And it doesn't seem to be holding up at all in NI.
My suspicion is that Labour might under perform compared to polling, whilst LDs might do slightly better. The Brexit Party, I just don't know how it will go. I fear the worst tbh, because the Brexit Party do have that 'clear message'.
I find it interesting that the Netherlands has shown something of a centre left resurgance - I'm guessing the closest thing we have to this is a labour remainer grouping, that no one saw coming - is this like a Corbyn 2017 effect? Or is this group particularly difficult to predict atm? And does this cut across Europe as a whole?
The Netherlands is VERY different to us, but there have been broad trends going on over the last few years across borders - this deindustrialisation / provisional town pattern being one of the most significant. Significant polarisation is another.
Turnout is everything. I don't think we are as motivated as the Netherlands, which is why I don't think Sunday will be pretty. And whilst we don't have an exit poll here, we should have some idea of turnout tonight
The verification process, IS being done tonight in many places to check the number of votes cast.
The lower it is, the MORE unpredictable I think the result will be. And the worst it will be for Labour / Conservatives.
Cas Mudde @CasMudde
I sat at @deafspraaktv when exit polls for The Netherlands in #EP2019 came through and I almost fell from my chair. PvdA biggest party?!?! Social Democrats as big as (two) far right together (PVV & FvD)! Some quick thoughts. #Thread [PvdA = centre left]
1. Important: this is an exit poll and claims a margin of 1 seat. While exit polls have long been very reliable, this has been less so in recent elections, in part because differences and parties are smaller.
2. NONE of the polls predicted anything like this. In fact, several polls had PvdA at half of the 18% they got in exit poll. Only Kantar came close, but had VVD still bigger. [pollsters got it wrong]
3. All polls had FvD much higher. This could have been correct, in which case the Rutte-Baudet debate cost FvD dearly (I'm skeptical because debates tend not to have that much impact, although this one was allegedly well watched). [FvD = far right]
4. Alternatively, debate led to higher turnout among undecided PvdA voters, who were triggered by fact that Rutte limited #EP2019 to choice between right and far right. Turnout is estimated at 41.2%, which is almost 4% higher than in 2014.
5. PvdA scored more than THREE TIMES as high as in disastrous 2017 national elections! But also almost TWICE as high as provincial elections just few months ago. This seems to indicate Timmermans Effect. [Is it more comparible to a Corbyn effect? Or a more broad centre left revival?]
6. Frans Timmermans (PvdA) was Spitzenkandidat of S&D and #EP2019 was partly presented as struggle between Weber (EPP) and Timmermans (S&D). Perhaps Dutch voters rallied around their local boy?
7. It could also be part backlash to anti-Timmermans spot by desperate "radical left" SP, which created massive backlash to SP and, perhaps, both attention and sympathy for Timmermans.
8. Rutte-Baudet debate was, according to exit poll, debate between shared second biggest party and shared fourth biggest party. This should lead to reflection within media, including public media, about their role in cynical move of Rutte.
9. Even if this is just one country, and an exit poll, it will affect the broader framing of #EP2019 -- in this sense, the Dutch functioned as the first primary in the US. Journalists will look for "revival of Social Democrats" as well as "rise of populism". [the narrative will be interesting...]
10. All in all, this was NOT what I, or anyone else I know or follow, had expected. Several other countries will have higher turnout. Perhaps there will be more surprises. Game on! #EP2019 #TheEnd