Categories
politicalbetting.com » Blog Archive » How steep is Starmer’s mountain?
politicalbetting.com » Blog Archive » How steep is Starmer’s mountain?
Since the last election resulted in a substantial Conservative majority, many have said that Labour has a mountain to climb to win the next one. The implication is that the result this time significantly influences the result next time. The results in 2024 are influenced by the results in 2019.
0
This discussion has been closed.
Comments
The chances of 7, 8, or 9 are: 17% + 14% + 11% = 42%
The chances of 6, 7 or 8 are: 14% + 17% + 14% = 45%
https://twitter.com/jamespomfret/status/1262219612176838656
I guess it’s part of the human tendency to see patterns even where there aren’t any,
Naive. Recessions do not harm the chances of governments in following elections in remotely like the way people seem to think they do.
I awoke at 6am. The recent cacophony of birdsong as been drowned out by the usual noise from traffic on the A24. Lockdown is over if anyone is in doubt.
However, I think politics and football are quite different. Footballers have to remain fit and motivated. It's hard to keep doing it season after season. First past the post means that only two parties can realistically aim to be the largest party - I'm sure the top two parties at UK elections are autocorrelated!
I think when analyzing the chances of the LOTO becoming PM, we should look at the circumstances at which that has happened in the past:
1945 - Attlee
1951 - Churchill
1964 - Wilson
1970 - Heath
1974 - Wilson
1979 - Thatcher
1997 - Blair
2010 - Cameron
It seems to me that Blair and Cameron took the approach of trying to avoid scaring voters. Thatcher put forward a more radical plan for change. I find it hard to see Starmer following the Blair/Cameron approach, so I think the question is - "will enough English voters want to vote for change in 2024?"
Thatcher was in the right place at the right time. Economics was a key part of her success even if a lot of people didn't like what she was doing. I don't think economics have been a key part of elections since her time. The dividing lines have changed. Perhaps this crisis will change that and an economic crisis will provide something of an opportunity for Starmer. But if that doesn't happen, I think Starmer is up against it.
Finally, Scotland is a huge headache for Starmer. Perhaps the fear of the SNP being in government at Westminster will fade over time. I suspect it won't, and the Tories will continue to benefit from SNP's success for as long as Scotland are returning MPs to Westminster.
The test has been applied to a % of the vote, whilst the question and conclusion are couched in terms of a binary result.
How does that linkage model the relationship between the two?
But may I point out one thing you write:
- “Labour has to overturn forbiddingly huge majorities in Tory seats“
Yes, that is true. But Starmer’s real problem is that Labour has to overturn forbiddingly huge majorities in Tory seats AND in SNP seats. That is a much harder task, because the two objectives require mutual contradictory strategy and tactics. As the Liberal Democrats discovered to their cost: in the age of the internet you cannot send vastly different messages out to the electorate in different geographical areas, because anybody anywhere can read, and redistribute, your two-facedness.
My thought is that if return is made voluntary (and I think compulsion would result in mass truancy) it is the most deprived children who are least likely to return.
Maybe as simple as poor people feeling that social advancement through education is not on the cards for their children. Interesting though.
I think that there are many in Quintile 1 (certainly not all) who simply don't value education. I'd be curious to see a breakdown of truancy or opinions on education in normal circumstances, wouldn't surprise me to see such a relationship.
To get a stable, healthy, Labour majority requires Tory and SNP seats. But SNP seats are very much secondary.
I know this has little to do with your article but the very fact of Coronavirus between 2019 and 2024 renders all previous correlations null and void anyway, with the sole possible exception of the Second World War. When the populist successful leader was booted out by a thumping Labour landslide.
Desperate news for Carlotta, it looks like the Scottish Government have hired at least 600 Contract Tracers since Carlotta swore late yesterday that ZERO had been hired. 600 hired on a Sunday night and starting work today seems pretty good going.
How do you police the 14 day quarantine? Simple. Random flights go into mandatory quarantine - so you don’t know if you’ll get home or not - and steep fines for those who break self quarantine.
There has not been, nor can there ever be, a randomised double blind trial on the preventative efficacy of face masks. So WHO will never confirm the science of it.
For the rest of the world with an ounce of common sense, it's a no brainer.
Whereas for Scotland, you have the Nationalists dominating since 2015, Labour before that for thirty years, then the Conservaties having majorities if you go further back. There's no single relationship you can apply the DW statistic to, unless you use very small subsamples of general elections (e.g. 2015, 2017 and 2019 for the Nat domniation), and then the results would be pretty worthless. Durbin and Watson didn't even calculate critical values of their statistic for fewer than six observations. I understand what you're saying, but I'm not sure I agree. Mathematically, because there are so many more English seats than Scottish ones, Labour can win a majority without winning any seats north of the border. But even if it won every seat in Scotland, it would still be far short of a majority in England, unless it picked up some English seats. A large number of Scottish seats would make Sir Keir's task much easier, obviously, but they are not essential.
I also think that sending vastly different messages to two groups of supporters can work, at least for a while, even in the age of social media, if it's what the two groups are desperate to believe. It's more difficult, of course, but possible. So you had Labour doing better than expected in 2017 by dog-whistling to Remainers in cities and on social media, but pretending it would implement the referendum result in its manifesto and in eurosceptic areas. It didn't work for them in 2019, of course, but I wonder if that's because the Conservatives were more competent this time, rather than Labour being dishonest?
https://www.dailyrecord.co.uk/news/scottish-news/coronavirus-scotland-despite-8500-people-22039844
So it must be an abject failure. :-)
The message "more of your friends and family are dead because the Government were incompetent" will play well both sides of the border. Maybe even better in Scotland.
Votes:
Conservatives - 35.7%
Labour - 35.5
Seats:
Labour - 54.1%
Conservatives - 36.7%
Contrast those results with 2019:
Votes:
Conservatives - 47.2%
Labour - 33.9%
Seats:
Conservatives - 64.7%
Labour - 33.6%
I think Labour were quite fortunate to do as well in 2005 as they did. Back then, FPTP really was working well for them. That isn't the case now.
To win most seats next time, I think Labour are going to need a serious revival from the Lib Dems. Whilst that's not impossible, the Lib Dems are going to continue to suffer from a lack of press coverage as they are not the third party.
The mechanism and risk for spreading there will be transfer from person A to person B via the priest's hands. The risk from the priest himself is small relatively because there is only one of him, and multiple people in the congregation.
The only way to stop transfer risk is either to break the chain where people receive communion - eg by going to mini wafers dispensed with no priestly contact and protestant style individual glasses properly sterilised, or the priest sterilising his gloves between each communicant.
Hmm. Displacement activity?
Plus I guess some risk for the priest if infectious people remove their masks at the rail.
Most of it seems to have come from Labour. If that is so, it cost them a great many seats - e.g. Blyth Valley, Durham North West, Delyn. So if it goes back, they should have a decent shout of regaining many of them.
However, another way of looking at it is that in 38 seats, the Brexit Party vote was larger than the Labour majority. So if Nigel Farage had not been a dimwitted egomaniac, the Tories might have picked up Doncaster North, Normanton, Alyn and Deeside, Torfaen, both seats in Newport and all seats in Coventry if Leave voters had coalesced around them. So if that Brexit vote shifted to the Tories, Starmer’s task is even harder.
Therefore, I am reluctant to make firm predictions about the next election. Starmer could win, or force a draw, but he needs the dice to fall correctly. We could see considerable churn in both votes and seats - I could see Labour gaining Cheltenham (repeat Cheltenham) and falling further in Wales, for example, under his leadership.
Full PPE for all staff
Because they use air drills they need 30 mins between patients to let wet particles settle
Plus a full wipe down after 30 mins
He reckons one patient per hour at best.
https://twitter.com/Peston/status/1262144064310390786?s=20
Angela Merkel having a good crisis as CDU recovers lost ground
With an election next year and no obvious successor will she run again ?
https://www.welt.de/debatte/kommentare/plus208029615/Coronakrise-Angela-Merkel-auf-dem-Hoehepunkt-ihrer-Kanzlerschaft.html
Two very minor points are, with n=27, this is on the lowish side to get good estimates of auto correlation, which is similar to 2 years & three months of monthly data, but I do realise we often have to put up with small sample sizes. The second is that the GDP example for auto-correlation is not great, because the first thing you do in this type of time series analysis is to subtract out the average. I think a better way to explain it is: if GDP increases this quarter, then there is a higher chance that is will increase next quarter as well, whereas if GDP decreases this quarter, then there is a higher chance that is will decrease next quarter as well.
As a related aside, this TfL bailout is not good for Sadiq Khan's chances of re-election. Yes fares revenue was down 90% and a decent number of drivers were sick/dying. But the reduction in service provision was branded him putting workers at risk and the massive congestion charge increase is branded as him putting workers at risk...
Rejoice, life is better under Labour.
(Ironically the fiasco of leaving the ERM is what propelled us to success)
I suspect it varies around the country.
In some places - Hartlepool and the Yorkshire mining constituencies I think it damaged the Conservatives, in other places it might have damaged Labour.
I will say that I've never seen such an expansive campaign as what TBP did in South Yorkshire - activists infesting town centres, masses of leaflets and even cars with speakers crawling through residential areas.
BTW there's no chance Labour will win Cheltenham - they lost their deposit there in 2019:
https://en.wikipedia.org/wiki/Cheltenham_(UK_Parliament_constituency)
On the basis of the ONS survey there will be around 170,000 who would test positive. 20,000-30,000 are actually testing positive per week. So doesn't that mean about 150,000 undiagnosed? How the precise definition of symptoms bears on that doesn't seem likely to be decisive.
n=27 is on the high side of the DW tables, which run from 6 to 30 - at least they do in my book. The subsamples are certainly very small, which is why I didn't put much emphasis on them. Above 30, the book recommends the use of the normal distribution.
What about the one for bringing in standing passengers on airplanes?
Personally, I live alone and have used a little over a litre of surgical spirit since March 15 ish, going out at most about 3 times a week plus exercise, and carrying a small bottle but mainly a ss-dampened cloth.
A colleague has used up his previous 20l supply as he has been the safety person for a village voluntary delivery network. He has now rigged up home electrolysis to manufacture a different suitable product.
- How many in application
- How many in vetting
- How many hired
- How many in training
- How many trained
That goes for all jurisdictions in the UK.
I would suspect that that the above numbers issue is to do with where people are in that flow.
I guess the question I am asking is whether there are many examples of datasets where in the full data you do see autocorrelation, but in a subset of the size looked at here, there isn't?
The other thing is that a few general elections ago there was a prediction model touted that combined positive autocorrelation on short timescales (ie for the next election to stick with the devil known) and anti-autocorrelation for longer timescales (ie of a few general elections the power of political folk tales is strong).
A more complicated model would require more data to find significance, and the combination of the two effects might muddy the waters for the test as you have applied it here.
As you've probably guessed, I've done more than my fair share of statistical analysis in my career, and most of the conclusions you get from real world data are nuanced.
It's also worth bearing in mind that there are other tests for serial autocorrelation which could give different results.
The thing is, it's exactly the sort of seat where a very large chunk of that Liberal Democrat vote would go with the soft left under Starmer (now Labour is not led by a racist nutter) rather than the soft right Davey is positioning the Yellows in. At the same time, much of the Tory vote will have no love for Johnson. I could see a Starmer-led Labour party picking up 10,000 remainers from the Tories who were panicked by Corbyn, and the same number from the LibDems who are unenthused by Davey.
So yes, Cheltenham is a seat where churn could be very interesting.
At the same time, I can easily see Labour losing both seats three stops down the railway line in Newport for exactly the same reasons.
We also see and recognise a first time incumbency bonus for MPs driven partly by this funding but also harder working MPs aware that they have only just gained the seat and need to consolidate. Many Tory MPs are in that position now with seats in previously alien territory. The incumbency bonus can be looked at another way. Since 1979 (41 years depressingly) there have been 2.5 changes in government (allowing for the Coalition). If Fishing was right surely we would see more changes of government than that?
Finally the point was made recently by TSE, IIRC, that 1970 was the only election in modern times where a clear majority for one party was replaced by a clear majority for another. Again, if Fishing was right surely this would be happening as often as not.
I am not qualified to challenge the statistical analysis but the results suggest to me that the Tories with a majority of 80 are pretty much nailed on to win most seats next time out. They may return to minority status, as in the May period, but I do not believe Labour will be the largest party.
The hope for Starmer is that the who? he has promoted into the Shadow Cabinet turn out to have some umph about them. Have literally no clue who the new Shadow Home Secretary is or what he represents, but by not being Diane Abbott he is already obviously a huge improvement. And should any of them fail Starmer has the big hitters currently deployed in Select Committees who could be brought back in.
Contrast a calm and measured shadow cabinet with clear ideas as to how the post Covid post Brexit world could be better vs the dregs of the Tory Party ideologically insisting the huge shower of shit raining down on people's heads is what they voted for/better than Corbyn could actually pull off a win no matter how many seats they need to win.
I have found the discussion around schooling and the inequalities exacerbated by Covid fascinating. So many Tories here and in the wider debate suddenly so concerned that the poor lack the resources to get a good education. Yet it is their policies that have made that happen. No room to study? The bedroom tax. No money for a laptop? The welfare cap and the third child tax. No broadband? Of course Labour's free broadband policy was a stupid pointless gimmick because everyone has broadband, right?
Our society lacks resilience to deal with the stresses of this kind of shock, but that has been a deliberate policy choice. Spare us your crocodile tears now, and don't force the rest of our kids into an unsafe rush back to school simply because schools have become the last remaining functioning part of our threadbare welfare state.
Like so many Tories suddenly becoming concerned that an interruption to economic growth may cost lives. I just think it's nice that they're concerned about lives, rather than the effect of their investments. That has to be a positive.
It is not impossible but a big ask and essentially requires a Labour landslide
Conversely, the Conservatives' loss of better off voters in big cities looks as if it permanent, too, but so far, the trade off has worked to the Conservatives' advantage.