Jeremy's Den

12/12/2019

An Dubhlachd Election - E-Day Ambient Update

Christmas Shopping Spree Election Day
56th anniversary of Kenya's Independence and 4th anniversary of the Paris Agreement on climate change
Also Bill Nighy's 70th birthday and Will Carling's 54th

Big Day 🔊
© Brian Eno, Phil Manzanera, 1975

So today is the big day, at least for the 3,322 candidates and their dogs and also my last attempt at making sense of this general election. The idea is to use only the on-the-day polls, which are in fact conducted on the last two or three days of the campaign and surface on social media on Election Eve or Election Day itself around 6AM. So this year we have six on-the-day polls in store. There were seven in 2017 and just one in 2015. But once again the big news is a brand new real Scottish poll, conducted by Survation on Tuesday and Wednesday. Before sharing its results, I have to remind you that Survation also published the very last full Scottish poll before the 2017 election. Back then it was the worst of twelve post-Dissolution Scottish polls for the SNP, but also the one closest to the actual result though it still overestimated the SNP and underestimated the Conservatives. So all things considered I'm inclined to trust this one anyway and it does paint a better picture for SNP than what we had yesterday, as you can see from the voting intentions and the possible distributions of seats. Fun part is that the Little Green Men are doing awfully poorly, 0.8% of the vote split between 22 candidates means 22 lost deposits. But they asked for it, didn't they?

I have also selected a sample of Scottish seats to watch tonight, which is purely arbitrary though you will find some here that have been mentioned in the media too. 22 seats are on my list, which might look a bit much, but that's what I think allows you to cover all the angles or the what-went-wrong and the what-went-right, and does not mean that what happens in the other 37 seats is irrelevant. There might be some upsets in these too, you never know. The vote shares in the table are my own projections, not even cross-checked with YouGov's MRP data. Text in bold dark red shows seats projected to change hands. Bold numbers on a solid background indicate the projected winner, while light backgrounds indicate the predicted runner-up.

I guess my list covers enough seats, and not just the main ones predicted to change hands, to assess who won the one-on-one fight between the Conservatives and the SNP, plus some juicy side stories. Will Glasgow definitely turn yellow or will a couple of red spots remain? Will the Tory tactical vote fail Ian Murray and make him vulnerable? Will John Nicolson stage a comeback in East Dunbartonshire... oops, sorry, meant Ochil and South Perthshire... just a slip of the tongue, wasn't it? Can the SNP actually unseat Jamie Stone and Christine Jardine, and why not Jo Swinson too? Could Labour do so badly that Martin Whitfield not only loses his seat but also finishes third behind the Conservatives? And how the fuck will the Kirkcaldy and Cowdenbeath bùrach end? You have my best answers in the table, not sure all will come true. But had to try anyway.

Just Another Day 🔊
© Brian Eno, Peter Schwalm, 2005

This day won't be remarkable for the couple of millions who have not registered to vote or the 25-30% who have but won't move their asses to the fucking polling place, hopefully fewer in Scotland than in England, as they will whine about the results first thing tomorrow anyway. Don't get too excited at the reports of strong early turnout as it was to be expected when the sun sets before teatime, and even just after lunch in some parts of Scotland if some of our MPs are to be believed. This might be the most important election in a generation but don't expect a record turnout as the last time it was officially above 70% was in 1997. The day-to-day plot of voting intentions is not that good for Labour when you compare it with what we had in 2017. There was indeed a Labour surge during the campaign but they underperformed most of the time while the Tories caught up with their 2017 vote over the last two weeks. So a 2017ish upset at 10PM is probably out of the realm of plausibility now.

The last Poll'O'Polls for this election reflects this. It includes the final six polls, conducted over the last three days by six different pollsters (BMG, Kantar, Deltapoll, Opinium, Panelbase and Survation). YouGov are conspicuously absent here but they probably decided they had nothing meaningful to say after they released their final MRP prediction on Tuesday. The final super-sample is 14,867 with a theoretical 0.80% margin of error as pollsters again went for bigger samples in the probably vain hope they would get better results than in 2017.

So we end up with the Conservatives leading by 10% over Labour, when the final 2017 estimate was a 7% lead and the actual result a 2.4% lead. But don't read too much into this. There is obviously still room for surprises before Huw Edwards takes the stage and the national average can't reflect all local situations as we have already seen quite often in earlier elections.

Here Come The Warm Jets 🔊
© Brian Eno, 1973

On the Last Day polling average my model predicts a 40-seat Tory majority. Which is pretty distressing but an improvement on what we had yesterday. The bizarre quirks of FPTP mean that the Tories have not just lost a handful of Scottish seats but also a handful of English seats as the tiny changes in the relative strength of parties here and there switched some marginals.

The other prognosticators have also published a final update to their predictions. They mostly also show a slight drop in the number of predicted Tory seats over the last week. What remains to be seen is whether the two MRP predictions will be more accurate than the rest as YouGov's was in 2017. One exit poll to go, please...

So the First Minister of England might find himself in a more precarious position than expected after the election. The many unknowns mean that he is probably closer to a 1992ish outcome than a 1983ish one and it did not end really well for John Major back then. Johnson may have eradicated one form of rebellion with his Stalinist purge of the last true Europhile One-Nationers but you never know where the next blow might come from. Could be from the ERG cultists when he totally fucks up the final Brexit Deal and ends up with some monstrosity that makes even Steve Baker throw up his jam-first scones. We'll know soon enough.

Taking Tiger Mountain 🔊
© Brian Eno, 1974

So this last update says we have 73 seats changing hands and 78 marginals. 73 changes is the same as what sealed Margaret Thatcher's victory in 1979 and her 43-seat majority back then also almost duplicates Johnson's predicted 40-seat majority. Quite an omen. Now let's have a last look at the full cartography of gains and losses and the alternate outcomes after redistribution of the marginals.

There's little change here from yesterday. A hung Parliament still remains a plausible result, but so does a 1983ish Conservative landslide. But the slightly reduced number of predicted Conservative seats in a hung Parliament opens the door to a new interesting option. A Labour-led Rainbow Min coalition (Greens-SNP-PC-SDLP-Lab) would bag 301 seats to Johnson's 306 so he could be ousted if the DUP chose to support Corbyn as they have hinted they might. Quite an interesting prospect even if it would be more likely to boil down to which brand of ministerial cars PM Jo likes best. As Always.

The Great Pretender 🔊
© Brian Eno, 1974

Less than three hours to go now until Huw Edwards reads out the exit poll and jaws drop to the floor. Or not. Only the punditariat's won't as they told us so, whatever the outcome, because this election was so unpredictable and we did tell you so, didn't we? Then we'll have the First Minister of England waffle-piffling his way through even his victory speech and promising he will uh.... get.... uh... Brexit... scoff... done, because that's the will of the seven people he met during the campaign. And Not-PM Jo will explain us why the LibDems did really well though they end up with fewer seats than on Dissolution Day and aye, she congratulates Amy Callaghan on her 2-vote victory and naw, she won't challenge the result in court and aye, she deeply regrets having massacred even more unsuspecting vowels than innocent squirrels during the campaign. Just before Keir Starmer comes out of hiding to read Jeremy Corbyn's resignation letter and farewell address.

But of course the worst part is that the First Minister of England will be in a position to go on pretending he is a leader. And we will be left wondering how the fuck he could get away with more lies than all other candidates since 1832 combined. The worst will be still to come obviously when the pretendy PM totally botches everything he touches and our only hope becomes that his corporate handlers will finally decide enough is enough and end up frankensteining a less buffoonish replacement. Might take years though and in the meantime he would have broken up the Precious Yoonion. Johnson as the best hope for Scottish Independence and Irish Reunification. Stranger things have happened.

But right now just brace yourselves for the Great Disturbance In The Farce at 10PM sharp when all prognosticators go FUUUUCK, those fucking pollsters fucked up AGAIN. Because, ye ken, energy fools the magician.

Marshall your forces, O great generals
The cost of victory exceeds the cost of defeat
(Andromeda, episode So Burn The Untamed Lands, 2004)

11/12/2019

An Dubhlachd Election - E-1 Flash Update

The Day Before Doomsday
83th anniversary of Uncle Teddy's abdication and John Kerry's 76th birthday

I'm Eighteen © Alice Cooper, Glen Buxton, Michael Bruce, Dennis Dunaway, Neal Smith, 1971

Black Juju © Dennis Dunaway, 1971

Welcome To My Nightmare 🔊
© Alice Cooper, Dick Wagner, 1975

GB-wide polls are again popping up at a sluggish rate so the main event is again a new full Scottish poll, this time the last one Panelbase conducted between 3 and 6 December. It has received only low profile coverage in The National, probably because it's again quite disappointing for the SNP. Only caveat here is that Panelbase is usually the least 'SNP-friendly' of all pollsters who field real Scottish polls. But even them overestimated the SNP vote in 2017 so there is every reason to carefully consider what they have to say right now, which is mostly the Conservatives closing the gap with the SNP and the array of seat projections delivering only a minor success for the SNP. What was once seen as the worst case scenario for the SNP (45 seats) has now become the best case scenario. The silver lining is that only one SNP incumbent is really endangered (aye, Stephen Gethins in North East Fife) but the downside is that the Conservatives are doing better than expected and could easily spin a minor setback, like losing 'only' Stirling and Gordon, as clear support for their arch-Unionist stance. And of course would be even louder if the SNP failed to clear the 40-seat hurdle, which can no longer be brushed aside as a totally implausible outcome. The key seat here is Lanark and Hamilton East as losing it in addition to my worst-case scenario would take the SNP's headcount down to 39 and signal a very bad Election Night indeed.

Now there may be some explanations for this in the SNP's campaign itself. Though not as disjointed as it was in 2017, it lacks real direction. I thought we had a deal that the SNP would stop trying to save England from Brexit against their will, focus on Independence and hammer the Tories about their mishandling of reserved matters. And now we have Nicola Sturgeon publicly admitting a second EU referendum could happen before IndyRef2 if Labour are the next government, and also that a vote for the SNP 'need not' be a vote for Independence. Which is, whichever way you look at it, totally shambolic and at odds with the earlier campaign strategy, especially as it also allows the other parties plenty of wiggle room to hit at the SNP over their handling of devolved matters. Was a recipe for disaster in 2017 and still is today. Then we also have the 'situation' in Kirkcaldy and Cowdenbeath that could easily turn into Nightmare On Bennochy Road: a split pro-Independence vote and Labour underperforming so the Tories sneak in through the back door and bag the seat on barely 30% of the vote, in what had been Labour territory since 1935. Back to the big picture, the updated trends of GB-wide polling remain surprisingly good for the Conservatives.

The Tory vote has admittedly reached a plateau for some time now, but Labour is failing to actually catch up even with more potential LibDem voters switching. As always there are massive contradictions in recent polls. Once undecideds removed 59% are dissatisfied and 41% satisfied with the way Johnson handles his job as PM, while 68% are dissatisfied and 32% satisfied with the way Corbyn handles his job ad Leader of the Opposition. Yet the guy with a -18% rating on performance has a +16% rating on the PMability scale and roughly a 70% probability of winning the election with an outright majority. And the same people who now rate the NHS as a more important campaign issue than Brexit are also willing to grant Johnson a five-year lease on Number Ten when he's the one oven-ready to sell off the procurement of pharmaceuticals, the keyest component of the whole system, to Trump's Big Pharma donors who will raise prices threefold and even eightfold on some rare drugs. Go figure...

The Nightmare Returns 🔊
© Alice Cooper, Bob Ezrin, 2011

There was some bad news in this week's polling for Labour when YouGov released their last pre-election full Welsh poll. The previous one, fielded two weeks ago, showed Labour in better shape than before and still projected to hold a majority of Welsh seats. Now the Conservatives have almost closed the gap in voting intentions and Labour are predicted to lose eight of their 2017 seats. There is a symbolic dimension in this too as it shows the Far Western Pillar of the Red Wall in North-East Wales totally crumbling down. Five of the expected Labour losses (Aylin and Deeside, Clwyd South, Delyn, Vale of Clwyd, Wrexham) are within the boundaries of post-industrial Clwyd who would become a completely blue blob on the map for the first time in.... well.... forever.... as it never gave the Tories a full slate even in the olden days when it was Denbighshire and Flintshire, and not even during the Thatcher era.

Meanwhile, my updated Poll'O'Polls is still pretty worrying for Labour. It includes the last six ones conducted between 5 and 10 December. Super-sample size is 14,378 as pollsters went for bigger samples in the last mile, with a theoretical 0.83% margin of error. I deliberately did not include the massive surveys published by Focaldata and YouGov yesterday as both were conducted to feed their respective MRP models over a longer period and don't represent a snapshot of current public opinion, if any poll ever does. The massive samples do not mean they are less right or more wrong, or whatever, than others or that their tiny margins of error actually mean anything. Being right might very well again prove to be just lucky guess. Anyway here's what we have in the Poll'O'Polls....

As you might expect, pollsters contradict each other about the predicted Conservatives' lead over Labour. ICM and Deltapoll find it has gone down since their previous survey, while BMG, Survation and YouGov find it has gone up, and Savanta-Comres say it's either up or down depending on which one of their older polls you pick, as they conducted two simultaneously last week that delivered different results. All this is not really helpful overall. Especially as the rolling average has moved only very slightly, by roughly 1% in Labour's direction, and is now predicting the Conservatives leading by 8.5% overall and 12% in England outside London. Guess a repeat of the 2017 upset was always too much to ask anyway....

Elected 🔊
© Alice Cooper, Glen Buxton, Michael Bruce, Dennis Dunaway, Neal Smith, 1973

On such voting intentions my model predicts a 58-seat Conservative majority. This is slightly better for the Conservatives than three days ago because the seats they lost in England are offset by better results in Scotland and Wales. Never underestimate the colonies. Other prognosticators have updated their own projections and all that were published over the last week point to a Conservative majority, though they differ on its magnitude and mostly see it under 50 seats. The massive Tory majorities predicted by Forecast UK and Flavible definitely look like outliers here, especially when you remember Flavible's past record of wildly outlandish predictions.

Of course the most anticipated and most commented update is YouGov's new and last MRP estimate. Everybody in the commentariat remembers that YouGov were the only ones predicting a hung Parliament in 2017 though their seat count slightly underestimated the Tories and overestimated Labour and the SNP. So you might want to conclude that the Tory majority will be more solid than the 28 seats predicted by YouGov, and this is obviously possible. But you also might want to believe that a hung Parliament is still possible, and YouGov's published results also open the door to this possibility as what they actually say, after factoring in the margin of error, is that the Conservatives would bag 311 to 367 seats, and Labour 206 to 256. Focaldata, using their own version of MRP and a different data set, reach pretty much the same result. So expect the actual result to be weighted against the punditariat's best educated guess of a 30ish-seat Tory majority. That's real life media for you, lads.

Lay Down And Die, Goodbye 🔊

Under current polling we would have 78 seats changing hands, close to what happened in 1964 and 1900. One that brought a new government to power and one that was more a game of musical chairs barely changing the status quo. And we would also have 69 marginals. But of course the complete cartography of gains and losses and the possible alternate scenarios are pretty similar to what we had three days ago. And I again agree with YouGov's MRP: we could have a hung Parliament though it would certainly take one big fuck of a miracle to get it.

And of course the hung Parliament would soon deliver the New Coalition. Now it has finally dawned on Arlene Foster that Johnson is genetically engineered to lie even if you just ask for the time. So we can safely rule out any new Con-DUP trade deal even if the Irish Custom-Booth-In-The-Sea Frontstop that so angered Foster has fuck all chance of being implemented even in the unforeseeable future. So as always expect the LibDems to do the right thing: put personal interests over general interest and lend a hand in delivering Brexit. And Johnson's best stunt would be to buy them off with the promise of e second EU referendum and then actively campaign to defeat them, just as Cameron did with the Alternative Vote referendum. And of course never forget to lie through your teeth at every TV appearance, it worked so well already there's no reason to change strategy. So we're all fucked..... and some would actually love it.

A Runaway Train 🔊
© Alice Cooper, Bob Ezrin, Dennis Dunaway, 2011

So now it definitely looks like the UK is gonna serve a five year sentence without the possibility of parole. The last day of the campaign was a fitting end to all this clown show. Johnson's PR team telling ITV to fuck off while the First Minister of England ran for a fridge to hide was as hilarious as it was appalling. And now the blond buffoon will actually be elected and get a mandate to do what he does best: lying and waffle-piffling without any serious challenge from the Tory-cuddling state media and daily fish-wrappers. And so all will be fine for the billionaire corporate oligarchs as the dishevelled clown attracts all attention and they can freely puppetmaster him from the shadows. You know who they are: the garbage press owners and tax-evading entitled expats who brainchilded austerity and Universal Credit. The kind of lads who watch I, Daniel Blake weekly over their Sunday afternoon pink gins because it proves how successful they were: the masterplan was to make peoples' lives miserable and it worked beyond expectations.

And remember there is no such thing as the light at then end of the tunnel and there never was. All there is are the headlights of the runaway train coming at you at flank speed. Just don't whine, you asked for it.

Some more almost live coverage tomorrow. Just stay tuned.

You know the way to battle a lie?
With the truth? No, with a bigger lie!
(The Good Wife, episode Death Of A Client, 2013)

08/12/2019

An Dubhlachd Election - E-4 Flash Update

Four Days Until Doomsday
39th anniversary of John Lennon's murder and 28th anniversary of the dissolution of the USSR
Would also have been Jim Morrison's 76th birthday

People Are Strange 🔊
© Jim Morrison, 1967

The main recent event in the polling universe is of course YouGov's new Scottish poll, conducted between 29 November and 3 December and polling both Westminster and Holyrood voting intentions. Let's just say for now their Holyrood polling would return a pro-Independence majority and I'll come back to it after the General Election. YouGov's polls of Scotland have in the past steadily been the most 'SNP friendly' of all, be it their full Scottish polls or the Scottish subsamples of their GB-wide polls. This is not the case here as their findings are within margin of error of what Ipsos MORI found ten days before. So their data project onto a 10-seat gain for the SNP as I still rate Kirkcaldy and Cowdenbeath as a Labour hold despite the overall collapse of the Labour vote. But the Conservatives bagging almost the same vote share as in 2017 makes the SNP-Con battleground quite competitive and potentially uncertain. Here we have four predicted SNP gains (Angus, Ayr Carrick and Cumnock, Gordon, Ochil and South Perthshire) and four predicted Tory holds (Aberdeen South, Banff and Buchan, East Renfrewshire, Moray) decided by less than 4%. Where these seats go will tell whether we have a minor SNP success or a major one, and as always turnout will be the key, though I don't expect anything as bad as my earlier DoomSNP scenarios. Just to be on the safe side though, remember every vote counts until 10PM on Thursday.

The influx of new GB-wide election polls has slowed down over the last few days, which is quite surprising so close to Election Day. Only the now mandatory batch of Sunday polls delivered some fresh perspective. You have to wonder if pollsters think that it's already in the bag for Boris Johnson and that further polling is irrelevant. The day-to-day plot of voting intentions tends to support such views as it again shows Labour definitely underperforming when compared to the 2017 polls. This is something we have seen before this year and it only confirms two of the main points routinely raised by pollsters: the Conservatives have successfully managed to squirrel the Brexit Party vote away from their competitors but have probably reached their peak when current voting intentions match their actual 2017 vote share, while Labour still haven't gained back enough Remain voters from the LibDems and now need a more massive surge in four days than they achieved in 2017 over two weeks. Good luck with that.

In 2017 the Conservatives led by 21% on Dissolution Day and by 1% on E-5 (within margin of error of the actual 2.4% Tory lead on Election Day), Labour had closed the gap by 20% in 45 days. This year the Conservatives led by 11% on Dissolution Day and 11% on E-5, massively more than in 2017 and more worryingly Labour have failed to close the gap in 30 days. And the longer gap between dissolution and the election is certainly not the one factor that explains why Labour are doing so blatantly worse this year. Now there is renewed speculation that pollsters might fuck up as massively as they did in 2017, part of it bordering on conspiracy theories. Which conveniently overlooks basic facts: the very last batch of 2017 polls (those conducted on E-2 and E-1, 6 and 7 June 2017) got both the Conservative and LibDem vote shares right within margin of error. Only grossly underestimating the Labour vote made them look awful when a predicted 50ish-seat Tory majority turned into a hung Parliament. But if pollsters are wrong this year by the same amount they were in 2017, that would take the current Labour vote only to about 38%, still 3% short of their actual 2017 vote. And I definitely think this would possibly deliver a 2015ish result for Labour and not a 2017ish one. Better than current projections but still not enough to totally rule out a Tory majority.

Ship Of Fools 🔊
© Jim Morrison, Robbie Krieger, 1970

Today's Poll'O'Polls includes the last six ones, conducted between Austerlitz Day and Pearl Harbor Day by five different pollsters (SavantaComres twice for two different clients and delivering different results, Panelbase, Opinium, Deltapoll and the ever-present YouGov). Super-sample size is 11,287 with a theoretical 0.92% margin of error, and it barely shows any change from three days ago, with Conservatives leading GB-wide by 10% and by 13% in England. And it's not really a case of outliers cancelling each other out though they actually do. But rather every new poll shows barely any movement from the previous poll by the same pollster, whatever their supposed bias is; Even the Guardian has to admit it's too little and too late for Labour though some of their readers are still desperately looking for signs of life.

One factor definitely working against Labour is Jeremy Corbyn himself as Johnson still beats him roughly 3-2 on the PMability scale. Which is quite similar to voters admitting they like the schoolyard bully better than the headmaster, and that would never happen or would it? There's a lesson in this: it's more efficient to bumble your way through incoherent waffle-piffling than to give serious answers to serious questions. And the host would probably shout over you anyway if you tried to be the adult in the room, just ask Nicola and Jeremy (The Other One).

Now a lot will have to be said during the post-mortem of this election on how the media handled two hot issues: Labour antisemitism and Tory islamophobia. The Tory-cuddling media were so eager to smear Corbyn and whitewash the Conservatives that they did not care they were treading into very dangerous territory. First they made islamophobia look like a very minor offence and then they made an alleged 'Jewish vote' something of a reality as long as it hurt Labour, while the very concept of a 'Jewish vote' is one of the pillars of real antisemitism. They played with fire and will get away with it. Says a lot....

End Of The Night 🔊
© Jim Morrison, 1966

If the actual election plays out the way current polling predicts, the one real upset on Election Night will be the LibDems actually ending up with fewer seats than on Dissolution Day. Which proves one of the current common wisdom points about PM Jo: the more people see her, the less they like her and the ghosts of dead squirrels glance down at us from the hereafter with vengeful grins on their faces. Then we will have a 54-seat Conservative majority as my model projects, or possibly anything between a 40-seat and an 80-seat Conservative majority as other models predict on the same polling data. So Johnson would do better than Heath in 1970, Thatcher in 1979, Major in 1992 or Cameron in 2015. Quite a feat to write his children about, once he figures out how many letters he should send, for someone who reached his Peter Principle peak as an absentee editor for The Spectator.

Here we have 77 seats changing hands, which might sound like a small number compared to real tsunami elections. But not every election has to be The End Of Civilization As We Know It and a similar number of changes was enough to take Harold Wilson to Number Ten in 1964 or Margaret Thatcher in 1979, which was definitely not her landslide election, 1983 was. And again we have Labour here back to a 1983ish number of seats. The cartography of predicted gains and losses is again merciless for Labour with 21 incumbents going down North of the Red Wall and another 14 in the Midlands.

So don't expect any Portillo Moments in the wee hours of Friday The Thirteenth. And no Balls Moments either as most of Labour's frontbench are entrenched in deep sinkholes, and aye I still do love those mixed metaphors. Unless you count on some really upsetting upsets like Jo Swinson or Dominic Rabb being unseated, which would not be undeserved but is also not that likely. Then I guess this election will again prove rebellion and defection don't cut it with voters as none of the best known rebels who are still standing will be re-elected (Grieve, Gauke, Milton, Field) and neither will the most prominent defectors (Lee, Soubry, Leslie, Umunna, Gyimah). So those who chose to hit the road and head for greener pastures, be it retirement or the £100k-a-night have-dinner-with circuit, probably made the wisest choice. For fuck's sake even Bercow would have been endangered by an official Conservative challenge in true blue Buckingham.

Take It As It Comes 🔊
© Jim Morrison, 1966

Current polling would spawn 73 marginal seats and the alternate possible outcomes look pretty similar to what we had three days ago. And all pretty bad for Labour as they all mean some variation on Back To The Future. 1983ish on the median projection, 1923ish on the Con Max scenario and just 2015ish on the Lab Max scenario. Then the question would be: would Jeremy Corbyn stay as leader if the result was just as bad but no worse than what Ed Miliband achieved? Genuine question though I can't remember any case when a past Labour leader was granted a third chance at losing a general election.

Making the 'Con Min' scenario become reality would require slashing the Conservative lead to about 4-5%, slightly higher than the final 2017 margin because changes in voting patterns would work marginally in favour of the oppositions in a number of marginals. And once again we would have this loveable situation where PM Jo would face the existential dilemma between doing the right thing or getting the right car. Which she would have to solve in the time Huw Edwards would need to read out the exit poll, and we surely have no doubt about her final decision once she gets over the prospect of daily meetings with Voorhees Johnson as his Deputy First Minister of England.

My real regret this week is that we will never see the First Minister of England interviewed by Andrew Neil as he had Stanley send a note saying he can't come today, nor any other day for that matter though he found time for some entertainment shows, so Paisley Andy is preparing to pitbull an empty chair on Tuesday. Which I doubt will be enough for a 30-minute show but might be fun to watch for a shortish moment. Then all that will prove is that Johnson can be trusted on just one thing: avoiding any questions about his untrustworthiness. There is every reason to believe a fucking lot of people will have a rude awakening on Friday. But buyer's remorse and regretting not having voted won't help, now's the time to choose: get the fuck out to vote and kick out the Tories. Or else....

You weren't there for the zombie invasion in Finchmere
It died down after they were barred from the local pub
(Midsomer Murders, episode Send In The Clowns, 2019)

05/12/2019

Westminster Projection - E-7 Update

Snap Winter Election E-7 and also Snap Autumn Election E-7
15th Anniversary of the Civil Partnership Act coming into effect and Sajid Javid's 50th birthday

One week left before the Election Of Two Seasons, the one set in both the Astrological Autumn and the Meteorological Winter, whichever suits the punditariat best to explain low turnout or high turnout, predictability or unpredictability. Only certainty so far is that 76 MPs elected in 2017 will not be re-elected because.... aye, you guessed right... they decided not to stand this time. Which is not a record even if only 31 stood down in 2017, but 90 did in 2015 and a massive 148 in 2010. Among recent campaign developments we have Labour deciding to focus on Leave voters in the North and Midlands. Which sounds like a good idea as Labour are definitely not in good shape north of the infamous Mersey-To-Humber Red Wall. The Eastern end of the Wall looks shaky with Great Grimsby, a Labour seat since 1945, predicted by a dedicated constituency poll to switch on a 10% swing from Labour to the Conservatives and neighbouring Scunthorpe, a Labour seat since 1987 which they held on an outright majority in 2017, now projected as an ultra-marginal that could go either way by 600ish votes. The Western end in Merseyside looks more solid but only thanks to a number of deep Labour sinkholes there. Quite revealingly Southport, the only Tory seat in the area in 2017, has switched from a close Con-Lab battleground to a Con-Lib marginal.

The voting intentions in the five regions, as predicted by a recent Survation poll, confirm a massive Labour slump compared to the 2017 results. Interestingly this generally benefits the LibDems more than the Conservatives, with also a strong residual Brexit Party vote in the North. Which is the most dangerous combination for Labour: core Remainers switching to the LibDems while the New Model Blackshirts still bag a sizeable part of the Leave voters. Adding insult to injury, Survation's sample of 3,082 also rate Johnson higher than Corbyn on the PMability scale by 46% to 30%. In a three-way race Johnson gets 44%, Corbyn 25% and Swinson 13%. YouGov have delivered much better news for Labour, this time from London with a new full-scale poll of the Imperial Capital that has Labour on 47%, Conservatives on 30% and LibDems on 15%. This amounts to a 6% swing from LibDems and Greens to Labour since the last similar poll a month ago, with Labour now projected to hold 48 London seats, only one down on the 2017 result when earlier polling had them seven seats down. Not coincidentally, most of the polling here happened after the London Bridge terrorist murders and part of it after Boris Johnson's appearance on the Andrew Marr Show, so I can't help seeing the results also as a reflection on how Sadiq Khan, Jeremy Corbyn and Boris Johnson handled the aftermath of the attack. Meanwhile the overall trend of polls GB-wide remains fairly bad for Labour.

The Conservative vote may have reached a plateau probably because they have now exhausted the supply of Brexit Party votes who could switch back to their previous party of choice. The Brexit Party vote has now gone down almost to a similar level to UKIP in 2017 and their downfall after their peak vote in the wake of the European Parliament election has been noticeably faster than UKIP's after the 2016 referendum. Of course their recent slump is a direct fallout of them not standing in Tory constituencies as some polls predict their vote share could still be as high as 6% in constituencies where they stand. These seats being overwhelmingly held by Labour, it's easy to see how the New Model Blackshirts can seriously hurt Labour in a number of marginals and do Johnson's bidding there. As always there are some paradoxes in current polling. Those who say Labour is closing the gap are right, but those who say the Conservatives are consolidating their lead are right too. Just pick the pollster and poll that suit you, as over the last week we have anything from Conservatives leading by 7% and down to Conservatives leading by 15% and up. Polls will be polls.

There has been a lot of speculation lately about a number of constituencies delivering upsets as they vote against the national trends, which I think is just another stunt from the punditariat to cover their asses whatever the actual election results are. Bear in mind that a number of constituencies always defy the national trends at every election or else we would never have Portillo Moments, or would we? Then what little constituency polling we have so far does not hint at a massive number of upsets. The only notable one would in fact be a non-upset: Labour holding Portsmouth South, one of their surprise gains in 2017, with an increased majority and some LibDem tactical voting to defeat the Tories there. This despite LibDems throwing the kitchen sink and the electric battle bus at the seat, and one of PM Jo's infamous bar charts claiming LibDems would come rushing at flank speed from third place and gain it. Then we have had The Guardian headlining that Dominic Raab could be unseated by the LibDems, which he certainly deserves given his global incompetence and the Foreign Office's dismal handling of the Harry Dunn case, but the actual poll has him leading by 5% so there's a bit of wishful thinking here, like in similar claims that Iain Duncan Smith could be unseated in Chingford and Woodford Green. Otherwise Tory Rebels David Gauke and Dominic Grieve would both lose their seats to official Conservative candidates, which comes as no surprise as party dropouts standing as independents very rarely win. So don't hold your breath, there is every chance the results will actually be quite close to the overall trends, like what we have in this week's Poll'O'Polls.

My current super-sample includes the six most recent polls, conducted between 28 November and 3 December by six different pollsters (YouGov, Deltapoll, Survation, ICM, Kantar, SavantaComRes). Super-sample size is 9,458 with a theoretical 1.01% margin of error, and show both Conservatives and Labour slightly up. But the Conservative voting intentions have on average risen a wee smitch more than Labour voting intentions. You might want to always see the bright side of life and believe that it gets worse before it gets better. But there are many cases where it gets worse before it gets worser. I don't really know which one to anticipate right now because, ye ken, this election is sooooo unpredictable. Or so the punditariat say. What polls say is that the Conservatives' lead GB-wide has gone up from 9% a week ago to 10% on current polling, while their lead in England outside London has gone up even more significantly from 11.5% to 13%. The Guardian have slightly different numbers, but they weight each poll equally, which effectively means there is no weighting by individual sample size, and their polling averages also include older data than mine.

We have reached this point in the campaign where Labour should actually do better and close the gap with the Conservatives if they want to end up with a 2017ish upset on Election Night. But the day-to-day plot of voting intentions does not show such a trend. Ironically the most recent data-points here (which are two days old as we don't have polls in real time) show all three main English parties roughly where they were in 2017 at the same point in the campaign. But Labour have underperformed over the last week and we don't have a conclusive Labour surge as was seen two years ago. And I venture this will last as long as the First Minister of England manages to play the media like a fiddle and avoids serious scrutiny. The Conservatives would be wrong to stop bullying media that don't agree to be turned into CCHQ's subservient useful idiots. It has worked wonderfully so far with Johnson avoiding all embarrassing questions and getting away with a stash of lies that goes bigger by the day. I have a hunch Johnson will finally agree to be interviewed by Andrew Neil but will manage to appear in a slot of his choosing and very late in the campaign, late enough to make sure even a major trainwreck will have no impact at all on the postal votes and as little impact as possible on the on-the-day votes. And will peddle another bunch of lies anyway because he can't help it, it's his nature.

On current polling, my model predicts that the Conservatives would bag a 50-seat majority and Labour would again clear the 200-seat hurdle though still embarrassingly close to their 1983 result. Both parties would gain a few seats thanks to the LibDems entangled in a downward spiral. There is now a plausible scenario in which LibDems will bag fewer seats than they had on Dissolution Day and go through their own Groundhog 1983 Moment when the SDP-Lib Alliance lost half their incumbents, though conventional common wisdom still says they did well back then. Right now LibDems are predicted to lose all Red and Blue defectors' seats except South Cambridgeshire, Heidi Allen's old seat where she has now stood down, pretty much like 25 out of 30 SDP incumbents met an untimely demise in 1983. Jeremy Corbyn might remember that part of the story well as his own first election back then was that rare (and possibly unique) occurence of one new candidate defeating two incumbent MPs at the same time. See the Saga Of Islington North I told you in an earlier article some months ago.

Even in London, where the ABC1 Remainer vote once made seats like Bermondsey or Vauxhall look like sure gains, LibDems would gain only Richmond Park and would have Zac Goldsmith's abysmal record to thank for that. So perhaps PM Jo should now refrain from campaigning in rural areas that only the diesel battle-bus can reach and where squirrel huggers keep her at arm's length, and devote her time solely to urban areas within range of the electric battle-bus. If the LibDem vote falls further this could lead to a slight recovery for Labour, though I don't think they can do better now than a 2015ish result. This would work though only with the odd combination of potential LibDems voters defecting to support Remainer Labour MPs in Leave constituencies in the North and Midlands, and Remainer Labour MPs in Remain constituencies in the South. If I were a Labour spad I would advise Jeremy Corbyn to take a step back in such constituencies and send Keir Starmer to the frontline as he looks more able to convince Lab-To-Lib defectors to come back home on the pledge of a second EU referendum without simultaneously totally alienating Lab-Leave voters. Of course Vince Cable's strong support for the so called People's Vote, bundled with his merciless takedown of PM Jo's now brain-dead pledge to revoke Article 50, might also help Labour. Just sayin'

Eric Clapton with Bob Dylan and Robbie Robertson

On current polling 81 seats would change hands, again the same number as in 1900. But this is just the theoretical outcome and I have a hunch only 80 would actually switch, see more on this below. Did I mention already that 1900 was the first time Winston Churchill was elected, as MP for Oldham, the first of five seats he would represent? Then we all know Winston crossed the floor before the end of the term, served as a Liberal Minister in a position he had specifically requested, and then sailed on to become the most calamitous First Lord of the Admiralty of the whole 20th century. But I digress as my point here is that we definitely lack any kind of Churchillian rhetoric in the present campaign, even from Boris Johnson, though his talking points might well also be Powellian and Huxleyan at the same time, like in 'Rivers of blood, sweat and tears with my Brave New Deal'. Back to the electoral math, the balance of gains and losses and the cartography of switches are still quite bad for Labour, though marginally les than a week ago.

Labour is projected to lose eight members of the current Shadow Cabinet, including the two Scottish ones, Paul Sweeney and Lesley Laird. But that's just the basic math as recent developments mean that the SNP will not gain Laird's Kirkcaldy and Cowdenbeath after all as they are no longer standing there. This is obviously a very embarrassing situation for the SNP and I guess nobody will wager even one penny on Neale Hanvey winning as an independent. And even if he did that would be yet another embarrassment for the SNP as it again points to improper vetting of the candidates. The worst part is probably that Hanvey was elected to Fife Council for the Dunfermline Central ward in 2012 then lost his seat in 2017 and lost a by-election in Inverkeithing in 2018, ironically one that was triggered by Lesley Laird resigning from the Council after one full year double-jobbing as Councillor and MP. Which leads to the sadly unavoidable conclusion that the SNP's vetting was at fault already twice since Hanvey's now infamous Soros post. All things considered I have now downgraded the SNP to 44 seats and upgraded Scottish Labour to 2 seats. Unless of course we have a major upset there and Green candidate Scott Rutherford comes from nowhere to defeat Laird. Though I certainly wouldn't bet one penny on that either.

Current polling would also deliver 69 marginal seats. Conservatives over-performing could take their majority up to 108 seats in their best case scenario. Then of course a further LibDem slump might take some of the 23 Lab-Con marginals out of the danger zone, but that would still deliver a Tory majority of possibly 70 to 80 seats. More than enough for Johnson to cruise to a No-Deal-Crashout by the end of next year, which remains the most likely outcome of the whole 'Get Brexit Done' march down Desolation Row.

As usual, scenarios where the Conservatives under-perform are more promising. A renewed Con-DUP pact would just scrape past the majority hurdle by a mere two seats, though this won't happen as long as the Custom-Booth-In-The-Middle-Of-The-Irish-Sea option remains on the table. So we're back to the usual Groundhog 2010 outcome. A revisited Con-Lib Coalition would bag a 48-seat majority here due to Sinn Féin bagging more seats here than on the direct projection. Though you again have to wonder what the real purpose of a Con-Lib government could be during the Brexit transition period. Unless of course Deputy PM Jo is delusional enough to think she could strongarm Johnson into holding a confirmatory referendum on the terms of the final Withdrawal Agreement, in the unlikely event one could be successfully negotiated before the end of June.

Revisited by Steve Howe and Jon Anderson

Predictor sites again flourish as they did in 2015 and 2017. Some old ones have vanished, some new ones have appeared and some pillars of the community have never interrupted service even outside election periods. Now you can even get on-the-fly seat projections from Forecast UK and Flavible, supposedly factoring in all the most recent polls, as well as regularly updated ones from Electoral Calculus. Or you can just read my blog, can't you? Anyway all prognosticators have agreed over the last ten days on a solid Conservative majority. The only question being: will it be massive like in 80ish seats, or more perilous like in 30ish seats? Just bear in mind though that a week before the 2017 election, projections were something like 360 Tory and 210 Labour seats, and even Britain Elects on Election Eve still predicted 353 Tories to 219 Labour. So there is still room for some surprising twists, though I still consider them far less likely this year. Time will tell and I would love to be proved wrong if that means Johnson fails to bag a majority or even enough seats for a minority government, however implausibly remote that last possibility might look.

You might want to look at today's projections through rosy glasses as Labour have after all virtually gained some seats since last week. Quite true but at this rate Labour would reach the credible threshold for a minority government in late January and bag a majority in mid-March. So their best hope remains to contain the Conservatives below 320 seats. Now many will argue a hung Parliament would be a better outcome as it would require Johnson to agree to some modicum of compromise after the election. But I don't see that happening either, quite the opposite in fact. I think it would make Johnson even more determined to fight on a hard line and again play the populist card. And then trigger another snap election within twelve months, now that the Fixed Term Parliaments Act has de facto been nullified already. It has happened in the past and there is no reason to believe it won't happen again. I have a definite feeling Tories already have that in mind as their Plan B for next year.

The saddest part is that Johnson will probably get his majority next week after all and that his disgusting electioneering of the London Bridge terrorist murders will probably help as millions are more than ready to take every bit of anti-Labour and more specifically anti-Corbyn smear as self-evident truth. Of course the BBC openly acting as CCHQ's very own Propagandastaffel is just another sign that something is deeply rotten in the Kingdom of England. As is the Prime Minister of England acting like a weird blend of Donald Trump, shamelessly lying without any remorse and bullying journalists who do not want to be his useful idiots, and Vladimir Putin, using the state media as channels for his own self-promotion and the protection of the entitled oligarchy's interests. What will happen after the election is stuff that will give your nightmares nightmares when the Lord of the Lies unleashes his wildest plans unfettered, with Labour left humptydumptyfied by their worst walloping in forty years. But of course we still have seven days to make that unhappen, or haven't we?

Cetero autem censeo… aye, I do repeat myself.... sometimes. Flash Update scheduled for Sunday E-4, three days and thirteen polls from now.

Let us not be driven by fear into an age of unreason
Remember we are not descended from fearful men
(Edward R. Murrow, American journalist, 1908-1965)