Much as I appreciate your number crunching, hard work & expert judgment am I the only Labour supporter who finds it deeply depressing, disappointing & disillusioning that Starmer & Co seem terminally incapable of providing the bold left of centre policies & eloquent narrative to match that would actually inspire us to vote FOR THEM & not AGAINST OTHERS? 🤔🤬
What a terribly sad indictment of our ‘democratic’ process? Contemplating between 65-75% of voters feeling they should vote for a candidate they don’t really support, to thwart election of one who they would be aghast to ‘represent’ them. Surely FPTP needs to be buried, and very soon. So that a system that encourages everyone to express a positive view can be agreed, endorsed by the electorate and put in place before 2029.
Thanks for the analysis. I’d add some other caveats - leaders of both conservative and Labourcwill be different by the GE but the effect of that new leader on tactical voting will be paramount . I might tactically vote for a one nation Tory but never for Jenrick type candidate; similarly Andy Burnham is a social democrat who talks about PR and coalitions whereas Starmer would never get my vote . There was no analysis of the Green Party in your analysis and they are getting more and more popular .
Also a striking demonstration of how AV would have worked to keep out extremists. It really is time for Starmer to consider the fire-breaks needed to avoid replicating in the UK the destruction wrought by Trump in the USA.
It'd be nice to have something like STV and not worry about tactical voting, but of course Con, Lab, and Reform (support for PR now seems to have been quietly dropped) hope to win big under FPTP.
Farage will struggle to run a disciplined campaign in 2029 with candidates he barely knows, who don’t know each other, have not been vetted properly and have not formed into internal party blocs. Reform may be seen as radical right but there are lots of left wing or statist elements in it.
Reform will lack coherence and act without coherence and the other parties will pounce on this. Even if they run with an infusion of turncoat Tories, they’ll be seen as turncoats.
Farage will be looking over his shoulder every day for some unknown schmuck in his company / party making inflammatory statements.
Attila’s Huns may have been a rampaging horde, but they were a coherent and well established horde.
In a previous post you expressed support for the Alternative Vote. Does your current analysis of tactical voting cast any light on the impact AV might have on an election outcome.
You seem to be focussed on securing a progressive majority in the Commons. The surest way, surely, is through PR. If there is a natural majority of progressive voters then PR will straightforwardly translate into a progressive parliamentary majority. And it’s honest! It doesn’t rely on complex and unreliable strategising by often Ill-informed voters.
Really interesting Peter - thank you! Sounds like online tools to inform tactical voting going to be very important (like this one from GE2024: https://tactical.vote/). Is there a reason Greens aren't included in the analysis in quite the same way as other parties? Is there less data for them?
Tactical voting strategies operate inside a system that’s designed to produce exactly these dynamics, regardless of how voters distribute their preferences.
Information, capital and attention are coupled now.
When systems reward shock, speed and escalation, more information doesn’t create clarity.
It creates volatility.
Hype sustains capital.
Capital sustains infrastructure. Infrastructure sustains narrative. Institutions can’t regulate what they can’t scrutinize - and arousal moves faster than scrutiny.
Transparency doesn’t fix instability. These systems regulate nervous states, not markets or opinions.
The question isn’t how to tactically outmaneuver Reform. It’s why our media and platform infrastructure keeps producing Reform-shaped outcomes - and why that pattern will repeat regardless of voting strategy until we govern these systems as nervous-system regulators, not engagement engines.
Thank you for this excellent piece, Peter. I'm currently building a tactical voting model, using the British Electoral Study data to provide a basis for (reported) tactical voting in elections, and this more detailed YouGov data for the forward-looking basis. To really tie the bow on it, I'd need equivalent YouGov data from before a general election, so I can compare the proportions who say they intend to tactically vote against the proportions who claim to have done so. I've found equivalent YouGov data from Feb 2025, but the earlier data I can see isn't as detailed. You're not aware of any that I'm missing by any chance are you?
A little odd that it didn't include tactical voting in Labour-Green contests, though. There could be over a hundred of these if the tripling of the Green vote share is...
1. concentrated in constituencies with specific demographics.
2. Still an underestimate of their true support.
Voting intention polls tell us who people intend to vote for, not who they actually support, so if Green supporters still think a vote for them is a wasted vote in most of the country, they might still be giving a tactical voting intention, even now.
The YouGov ranked favourability poll at the start of October had the Greens at 19% of first preferences when their voting intention was 11%.
Green voting intention is now at 17%.
What is really sad is the amount of "progressive" people saying they would vote for their 4th choice Conservatives over their 5th choice Reform UK. This will be distorting voting intention polls to give a false impression of right-wing vs progressive support, so causing Labour to think that drifting right is the right strategic ploy (obviously the under-estimation of Green support will be doing the same).
So in a Reform/Con seat 28% of 21% of Labour voters would vote Conservative? I.e 5.88%?
Wouldn't some of that already be captured in an MPR as it models take in the last GE, where there may have been some tactical voting, particularly for Lib Dems etc?
Why try to estimate all GB seats from polls where the sample sizes for Scotland & Wales are far too small to be meaningfully or statistically representative? Especially given the increased multi party combinations given the high voting levels for SNP and Plaid Cymru (and the differentiation of Green parties) this is a fool's errand - why not just do the 543 seats in England? You and the polling companies don't do the whole of the UK, for good reason. That reasoning should be extended to Scotand & Wales and only the representative England sub-sample is usable to project only England's seats. The devolution context also clearly marks Scotland & Wales as different polities, yet alone the broader constitutional aspirations. The census data used for MRP models also has a different background and collection/publication timetable.
I'm disappointed that an alleged expert like Peter Kellnet persists in this totally unscientific "GB" process despite acknowleding the deficiencies - I can only assume that he and the polling companies are so entrenched in a unionist philosophy and ideology that they cannot stomach the notion that what we need us separate representative polls for each of England. Scotland and Wales.
I’d be intrigued to see some polling results asking views on English independence - I just have a feeling there could be quite strong support in England for disconnecting from NI, Scotland …and even Wales? Hopefully I’m wrong, but no one ever seems to ask the question?
Much as I appreciate your number crunching, hard work & expert judgment am I the only Labour supporter who finds it deeply depressing, disappointing & disillusioning that Starmer & Co seem terminally incapable of providing the bold left of centre policies & eloquent narrative to match that would actually inspire us to vote FOR THEM & not AGAINST OTHERS? 🤔🤬
Driver Andy
At the moment I'd be happy with a display of basic political competence. I'm not holding my breath.
What a terribly sad indictment of our ‘democratic’ process? Contemplating between 65-75% of voters feeling they should vote for a candidate they don’t really support, to thwart election of one who they would be aghast to ‘represent’ them. Surely FPTP needs to be buried, and very soon. So that a system that encourages everyone to express a positive view can be agreed, endorsed by the electorate and put in place before 2029.
Thanks for the analysis. I’d add some other caveats - leaders of both conservative and Labourcwill be different by the GE but the effect of that new leader on tactical voting will be paramount . I might tactically vote for a one nation Tory but never for Jenrick type candidate; similarly Andy Burnham is a social democrat who talks about PR and coalitions whereas Starmer would never get my vote . There was no analysis of the Green Party in your analysis and they are getting more and more popular .
First past the post must be changed
Also a striking demonstration of how AV would have worked to keep out extremists. It really is time for Starmer to consider the fire-breaks needed to avoid replicating in the UK the destruction wrought by Trump in the USA.
It'd be nice to have something like STV and not worry about tactical voting, but of course Con, Lab, and Reform (support for PR now seems to have been quietly dropped) hope to win big under FPTP.
Where are the hundreds of numbsculls for reform coming from? This is all hypothetical there are no valid candidates
That is the biggest question in all of this.
Farage will struggle to run a disciplined campaign in 2029 with candidates he barely knows, who don’t know each other, have not been vetted properly and have not formed into internal party blocs. Reform may be seen as radical right but there are lots of left wing or statist elements in it.
Reform will lack coherence and act without coherence and the other parties will pounce on this. Even if they run with an infusion of turncoat Tories, they’ll be seen as turncoats.
Farage will be looking over his shoulder every day for some unknown schmuck in his company / party making inflammatory statements.
Attila’s Huns may have been a rampaging horde, but they were a coherent and well established horde.
Reform are not.
In a previous post you expressed support for the Alternative Vote. Does your current analysis of tactical voting cast any light on the impact AV might have on an election outcome.
You seem to be focussed on securing a progressive majority in the Commons. The surest way, surely, is through PR. If there is a natural majority of progressive voters then PR will straightforwardly translate into a progressive parliamentary majority. And it’s honest! It doesn’t rely on complex and unreliable strategising by often Ill-informed voters.
Could you do the same analysis for a LibDem vs Tory seat?
Really interesting Peter - thank you! Sounds like online tools to inform tactical voting going to be very important (like this one from GE2024: https://tactical.vote/). Is there a reason Greens aren't included in the analysis in quite the same way as other parties? Is there less data for them?
This is the wrong diagnosis.
Tactical voting strategies operate inside a system that’s designed to produce exactly these dynamics, regardless of how voters distribute their preferences.
Information, capital and attention are coupled now.
When systems reward shock, speed and escalation, more information doesn’t create clarity.
It creates volatility.
Hype sustains capital.
Capital sustains infrastructure. Infrastructure sustains narrative. Institutions can’t regulate what they can’t scrutinize - and arousal moves faster than scrutiny.
Transparency doesn’t fix instability. These systems regulate nervous states, not markets or opinions.
The question isn’t how to tactically outmaneuver Reform. It’s why our media and platform infrastructure keeps producing Reform-shaped outcomes - and why that pattern will repeat regardless of voting strategy until we govern these systems as nervous-system regulators, not engagement engines.
Thank you for this excellent piece, Peter. I'm currently building a tactical voting model, using the British Electoral Study data to provide a basis for (reported) tactical voting in elections, and this more detailed YouGov data for the forward-looking basis. To really tie the bow on it, I'd need equivalent YouGov data from before a general election, so I can compare the proportions who say they intend to tactically vote against the proportions who claim to have done so. I've found equivalent YouGov data from Feb 2025, but the earlier data I can see isn't as detailed. You're not aware of any that I'm missing by any chance are you?
Having retired from YouGov almost ten years ago, I'm afrfaid I don't have that information. Try asking Anthony Wells: anthony.wells@yougov.com
Thanks Peter. Much obliged. I'll drop him an email.
Really helpful number-crunching, Peter.
A little odd that it didn't include tactical voting in Labour-Green contests, though. There could be over a hundred of these if the tripling of the Green vote share is...
1. concentrated in constituencies with specific demographics.
2. Still an underestimate of their true support.
Voting intention polls tell us who people intend to vote for, not who they actually support, so if Green supporters still think a vote for them is a wasted vote in most of the country, they might still be giving a tactical voting intention, even now.
The YouGov ranked favourability poll at the start of October had the Greens at 19% of first preferences when their voting intention was 11%.
Green voting intention is now at 17%.
What is really sad is the amount of "progressive" people saying they would vote for their 4th choice Conservatives over their 5th choice Reform UK. This will be distorting voting intention polls to give a false impression of right-wing vs progressive support, so causing Labour to think that drifting right is the right strategic ploy (obviously the under-estimation of Green support will be doing the same).
We desperately need more ranked preference polls, and publicly available ones rather than the under-reported and not-on-the-website poll from last month. https://d3nkl3psvxxpe9.cloudfront.net/documents/GreenParty_Ranking_251002_W.pdf
Is this the MPR it's taken from?
https://yougov.co.uk/politics/articles/53059-yougov-mrp-shows-a-reform-uk-government-a-near-certainty-if-an-election-were-held-tomorrow
So
Reform UK 27
Labour 21
Lib Dems 15
Conservatives 17
SNP 3
Greens 11
Plaid Cymru 1
Others 5
So in a Reform/Con seat 28% of 21% of Labour voters would vote Conservative? I.e 5.88%?
Wouldn't some of that already be captured in an MPR as it models take in the last GE, where there may have been some tactical voting, particularly for Lib Dems etc?
Why try to estimate all GB seats from polls where the sample sizes for Scotland & Wales are far too small to be meaningfully or statistically representative? Especially given the increased multi party combinations given the high voting levels for SNP and Plaid Cymru (and the differentiation of Green parties) this is a fool's errand - why not just do the 543 seats in England? You and the polling companies don't do the whole of the UK, for good reason. That reasoning should be extended to Scotand & Wales and only the representative England sub-sample is usable to project only England's seats. The devolution context also clearly marks Scotland & Wales as different polities, yet alone the broader constitutional aspirations. The census data used for MRP models also has a different background and collection/publication timetable.
I'm disappointed that an alleged expert like Peter Kellnet persists in this totally unscientific "GB" process despite acknowleding the deficiencies - I can only assume that he and the polling companies are so entrenched in a unionist philosophy and ideology that they cannot stomach the notion that what we need us separate representative polls for each of England. Scotland and Wales.
Should this illustration manifest it would probably mean a Scottish independence referendum. What are the latest poll suggesting for that result.
I’d be intrigued to see some polling results asking views on English independence - I just have a feeling there could be quite strong support in England for disconnecting from NI, Scotland …and even Wales? Hopefully I’m wrong, but no one ever seems to ask the question?