Scottish Momentum and The Guardian have it all wrong

First an update of my Scottish Westminster projection

Why an update when there hasn't been any new Scottish poll out for three weeks? Just because of that actually.

Polls have one thing in common with living organisms: they age. And unlike quality French wines they don't age well. This is a complete non-issue for UK-wide polling as we have two or three polls a week and my six-point rolling average automatically refreshes the data accordingly. But unfortunately we have far fewer full Scottish polls so the super-sample needs some pruning every now and then to stay as close as possible to the state of mind of Scottish public opinion.

We have had eight full Scottish polls in 2018, showing a definite SNP surge from 36% (YouGov in January) to 42% (Survation in July). But six of these are more than six weeks old which is like an eternity in a volatile political context. So I have decided to factor in only the last two full Scottish polls until a new one is published. Updated super-sample is 2,020 (2.1% MOE) and delivers these voting intentions:

The SNP on 40% is not a massive change from previous projections but the distribution of votes between the other parties is an important factor too. What we have now is Conservatives and Labour both down 3% from 2017 while Liberal Democrats do just a wee smitch better. The seat projection on SLLM ratings shows the impact is much more catastrophic for Labour than for the Conservatives. Summary: SNP 46, Conservatives 8, Liberal Democrats 4, Ian Murray 1.

At face value current polling projects 11 seats switching back to the SNP: all six Labour gains from 2017 and five of the 12 Conservative gains. LibDems would hold their four seats as they look immune from the overall Unionist debacle. So far.

But of course the higher MOE results in a higher number of marginal seats. Right now we have ten where the swing needed to switch the seat is within MOE.

Below is the statistically possible range of seats when you move the swingometer from the lower to the upper edge of MOE. Good news for the SNP is that they would still bag a net gain of four seats even if they missed all of the seven SNP-Unionist marginals. But of course this does not mean the SNP should be complacent as voting intentions might change significantly once the snap GE of 2018 is called. So please forget the 'we don't take anything for granted…. except your votes' attitude that worked so well last time aboot. Juist sayin.

News from the Scottish Branch Office of the British Labour Party

Scottish Momentum held their first Summer Camp (oops sorry, Training Event) last Saturday in Glasgow. It didn't register on anyone's radar except the ever-friendly Guardian's with Libby Brooks devoting a full-size article to it. First thing to remember: there are 1,300 Momenters (Momentumers just doesn't have the right ring to it) in Scotland. You read it right: that's ONE of them for every NINETY SNP members. Vanguard of the Corbynista Reconquista really?

I found the whole article quite confusing. To sum it up Scottish Momentum seem to have a problem with facts and electoral statistics. Which is quite embarrassing when you're hosting a 'training event'. Sometimes they get it almost right, sometimes totally wrong and sometimes the phrasing is ambiguous enough to be misleading.

What looks like the Guardian's own view

First of all of course we have the old tired talking point that 'Scottish marginal seats will be the key to winning the next election'. Which must be Libby Brooks' own view of the situation as it is not presented as a quote from Momentum. And of course is just as highly misleading as it ever was. Remember that Labour currently hold 262 seats counting suspended MPs. That is 60 seats short of a majority as Sinn Féin sitting out means House of Commons de facto has only 643 seats.

So for a few competitive Scottish seats (more on this below) to be the key and make this talking point a truth, Labour would first need to hold all their Welsh seats, which right now is surprisingly not a done deal. And then they would need to gain about 55 seats in England, which amounts to a clean sweep of all Tory-Lab marginals Doon Sooth plus some 20 seats beyond that to be on the safe side.

But the ever-friendly Guardian would never allow facts to get in the way of the narrative. Or would they? Guess we will get the same obfuscation from Owen Jones when he visits Scotland some time during the campaign, though the Jones Talking Point Of The Day is that Blackpool North is the key to the election. You just can't make that sort of cooshite up, can you? For the record Electoral Calculus has Blackpool North (if gained) as the 288th likeliest Labour seat so not quite the key.

To sum it up once and for all:

• at the 2017 GE Labour won 262 seats including 7 in Scotland; the best they could have done then was 16 Scottish seats for 271 overall and still far short of a majority
• on current 2022 GE projections Labour's best case scenario is 304 seats including ONE in Scotland; the best they could hope is 2 more Scottish seats and they would still be stuck with seeking a coalition with the Liberal Democrats
• pushing it to the limit a clean Labour majority would require 18 more seats than their current best case scenario or else it's back to some LibDem backroom deal; maybe Momentum could send us a list of which 18 Scottish seats they see switching, bearing in mind Scottish Labour fell far short from achieving this last year on a larger share of the vote

In any case 'Scotland is the key to Jeremy in Downing Street' is a deliberate lie compounded with delusion and should be treated and debunked as such.

Fun part later in the article is when Brooks describes 'Scottish Labour revival' as 'far more muted' than elsewhere in the UK. Then goes on to quote a YouGov poll that has Labour down 5% from 2017 in Scotland. Can't make that up either, can we? Sloppy journalism but pales in comparison to Momentum's sloppy analysis of electoral data.

Now back to Scottish Momentum

All 2017 GE data I will use are the official results as published by the Electoral Commission here. And to avoid any ambiguity when I use 'X-Y marginal' I mean X won the seat and Y finished second. Might sound obvious but precision never hurts.

The cornerstone of Momentum's demonstration is that 'a good 20' Scottish seats are marginals. Which is approximately true if you allow the concept. If we agree that a marginal seat is where the winner's margin is below 4% then 76 seats overall qualify (outside Northern Ireland). Including 20 in Scotland. Not 'a good 20' just exactly twenty.

Moving the threshold to 5% doesn't change the big picture. Then we get 91 marginals overall including 22 in Scotland. Which might pass for 'a good 20'. On a good day.

Here comes the really disingenuous part. Both Libby Brooks and Jessica Galloway (Scottish Momentum's organiser) carefully avoid to mention how many of these 20 seats are the actual SNP-Labour battleground. The casual reader may thus be misled to believe all or most of them are. Which is far from the truth.

Only eight qualify as SNP-Lab marginals. You can make that nine if you add Lanark and Hamilton East, a three way SNP-Con-Lab marginal. Remember this is based on the 2017 results and current polling has altered the picture dramatically. It is also quite misleading to include SNP-Con, SNP-LD and Con-SNP marginals in the count as Labour came third in all these seats, even those they had held before 2015, and don't stand the slightest chance to gain them back anytime soon. Ironically the only valid point Momentum have is to include the Lab-SNP marginals as battlegrounds since these are the truly endangered seats right now.

What's wrong with Momentum's approach?

The main fault is that their whole demonstration is based on a 'what if?' scenario. What if the next GE follows the same patterns and delivers the same results as the last one? Of course 'Groundhog Day elections' do happen. Think December 1910 vs February 1910 or 2001 vs 1997 or 2010 vs 2005 in Scotland. But they are the exception rather than the rule. To expect a repeat of 2017 at the next GE is wishful thinking at best, delusional at worst, with the quickly shifting political landscape we have this year.

Jessica Galloway never refers to changes that have occurred among Scottish public opinion in the last fourteen months. It's as if she deliberately disregards recent polling and what it implies, and is trying to convince her members (and herself) that the exact same seats that were the battleground last year would be the battleground at the next GE, snap or not. A close-up on the nine infamous '2017 battlegrounds' I identified earlier shows this if far from being the case. Here is how they voted last year and how they're projected to vote next time on current polling.

 

None of these seats would be marginals at next GE. All have switched to 'Lean SNP' and even to 'Likely SNP' in one case. Average swing here would be 3.5% from Labour to SNP while polls predict a 3.1% national swing from Labour to SNP. So the '2017 battlegrounds' would deliver even better results for the SNP than the national average.

Last but not least Jessica Galloway claims that 'the SNP are losing traction with younger and more active people in politics'. I don't know about 'more active' as pollsters don't crosstab this one. But I have a hunch about 'younger' as pollsters do crosstab that one. Here is what the latest Survation poll says:

Both the SNP and Labour do better than their national average among younger voters, but barely so for Labour among 25-34yrs olds. More interesting is that the SNP have a 18.5% lead over Labour nationally but 22% among 18-24s and 27% among 25-34s. So much for 'losing traction'. Young voters appear to be Momentum's core target but they misread them. Not a good start….

Another interesting result in the Survation poll (and one Scottish Momentum probably missed too) is that the SNP would hold 95% of their 2017 voters with 3% switching to Labour. While Labour would hold 88% with 5% each switching to the SNP and the Conservatives. Food for thought.

And now what next?

We have all witnessed how Labour's own bùrach has cost us what were probably the best opportunities to trigger a snap GE so far. But there will be others. Dinna fash.

Credible sources have said the 48 signatures (15% of sitting Conservative MPs) required for a vote of no-confidence in Theresa May have already been collected. All that is needed is for the 1922 Committee to give it a go, which can be done on the very first day Commons return. And then we have seven sitting days left before the Conference Recess, during which the non-confidence vote can happen and succeed.

As a confidence vote in another government will obviously not take place within the fourteen days required by the Fixed-term Parliaments Act, we have our snap GE which would then probably happen sometime between the November Recess and the Christmas Recess. 22 or 29 November being my best educated guess. All this would certainly make for some interesting moments during the Conferences.

The parts have been cast, the stage has been set, the players are ready.

Now entertain conjecture of a time

When creeping murmur and the poring dark

Fills the wide vessel of the universe.

From camp to camp, through the foul womb of night,

The hum of either army stilly sounds,

That the fixed sentinels almost receive

The secret whispers of each other's watch.

Fire answers fire, and through their paly flames

Each battle sees the other's umbered face.

Steed threatens steed in high and boastful neighs

Piercing the night's dull ear, and from the tents,

The armorers accomplishing the knights,

With busy hammers closing rivets up,

Give dreadful note of preparation.

Let the dance begin.

Cha togar m' fhearg gun dìoladh