But it is now complete and all the previews are here in one place in this post. The 18 Northern Ireland ones, that I haven’t posted before are at the end.

If you want it in spreadsheet form then subscriber Jon Skeet has built this neat site where you can see all my predictions against the seat, alongside other predictions and the time the seat will be announed. It will be updated live as results come in.

I’ve also gone back through some of the predictions I made early on when things looked marginally better for the Tories. And adjusted a few:

I’ve shifted three seats from Lean Conservative to Lean Labour: Faversham and Mid Kent, South West Devon, and Bridgwater.

And I’ve shifted five from Lean Conservative to Lean LD: Bicester and Woodstock, Witney, Surrey Heath, Mid Sussex, and Frome and East Somerset.

Needless to say these are all very tight races, by definition.

This leaves me with the following final prediction:

Labour: 445 (+247)

Conservative: 98 (-276)

Liberal Democrat: 62 (+54)

SNP 16 (-32)

Sinn Fein 7 (=)

DUP 6 (-2)

Green 3 (+2)

Reform 3 (+3)

Plaid Cymru 3 (+1)

Alliance 2 (+1)

SDLP 2 (=)

UUP 1 (+1)

Independent 1 (+1)

Speaker 1 (=)

I didn’t set out with a particular end number in mind. I’ve tried to go seat by seat. But it has ended up close to the average of the MRPs, albeit not with the same predictions as any of them.

There are, of course, big caveats. While every poll, and every other indicator we have, tells us there will be a big Labour majority, the scale of it, and what that means for the seat totals, is up in the air. I would not be shocked at any number of Tory seats between 30 and 200. There are so many close seats that we just won’t know till the exit poll.

Hopefully this has been a useful exercise (I’ve certainly found it useful to do).

Later on I’ll be sending out a “what to watch” post for tomorrow night.

Read more

Labour will certainly make big gains at the expense of the SNP but there are a lot of close seats. The Tories are looking to defend their marginals, a more realistic prospect than in England and Wales. The Lib Dems are looking to recapture some of their former strongholds.

Later this afternoon I will send out a post with all 650 seats in a single place, this will include the 18 Northern Ireland seats. It will also be updated to give a final set of predictions.

Then this evening I’ll send a post to the full subscriber list with a guide to what to look out for tomorrow night.

Read more

As with the North East and North West, there is a major question mark over whether the Tories can retain any of the 12 seats they notionally hold at the moment.

There will be several posts tomorrow. Covering Scottish and Northern Irish seats, but also “a what to watch guide” and final predictions. We’re nearly there…

Read more

The Tories notionally hold 33 of of the 73 at the moment, but many are marginals. There is a real question as to whether they’ll have a single seat in the north of England on Friday morning.

As well as the election I’m also starting to do a bit of publicity for my book “Failed State: Why Nothing Works and How We Fix It” - which is coming out in ten days. You can read an edited extract that the Sunday Times published yesterday. And there’s also this podcast I did with James Butler for the London Review of Books. I’ll write a bit more about here it once the election is over.

Read more

But before we get to that I wanted to address a question I’m getting asked more than any other at the moment. What happens if the Tories and Lib Dems get very similar numbers of seats? Who decides who the opposition is?

Well there is legislation - the Ministerial and Salaries Act 1975 - that states the Leader of the Opposition (LOTO) is that of the party that has “the greatest numerical strength”. But it adds that:

“If any doubt arises as to which is or was at any material time the party in opposition to Her Majesty’s Government having the greatest numerical strength in the House of Commons, or as to who is or was at any material time the leader in that House of such a party, the question shall be decided for the purposes of this Act by the Speaker of the House of Commons, and his decision, certified in writing under his hand, shall be final and conclusive.”

So if one party is clearly ahead then their leader is LOTO. But if it’s more complex - say, the Tories won 75 seats and the Lib Dems 73, and then two Alliance MPs from Northern Ireland said they’d sit with the Lib Dems - well the Speaker will have to get involved and make a call.

Moreover, this legislation only refers to the pay that goes with the titles. Other perks of being LOTO, like getting to decide what Parliament debates on 17 of the 20 days a session alloted to opposition parties, are just convention under the Standings Orders of the House. These can be changed by a vote. Or, like getting to ask six questions at PMQs, are purely at the discretion of the Speaker. He could propose that if two parties have very similar numbers of MPs that the perks should be split differently. In practice this would need to be agreed in a negotiation with the relevant parties before and changes or any vote in the Commons to amend standing orders.

Some of the more extreme potential scenarios indicated by the polls would require a more thorough re-think of how aspects of the Commons work.

Read more

Yorkshire, in particular, is the region where the scale of the coming Tory defeat becomes most obvious. Most of their majorities here are smaller and they are set to lose the vast majority of them.

Some additional resources for those of you that have asked:

Seat declaration times are here (I’ll do a “key moments to watch” list next week)

All the MRPs in easily searchable and comparable form here, thanks to the brilliant work of subscriber Peter Inglesby

Also, as per many requests, I will try and put all these previews in a single searchable post when they’re done and finalised, and plans are afoot for a live updating website on the night itself.

Read more

Today we have the bulk of the West Midlands, minus Birmingham and Coventry, which are coming tomorrow. A huge variety of seats in here, from big rural constituencies in Shropshire and Worcestershire, to red wall ones in the Black Country. Plenty of classic swing seats too like Worcester and Redditch.

Current MPs in this batch include Jonathan Gullis, Daniel Kawczynski, and Marco Longhi. As you will see it’s not an area that’s been blessed with many high quality representatives in recent years.

Here’s how the predictions are looking up to yesterday, i.e. covering the South East, South West, London, East of England and East Midlands. Given the further deterioration of the Tory position in the polls since I started I may need to review some of those “lean Conservative” seats before finalising my predictions.

Read more

While this may seem like a fairly technical issue, the gap between the different approaches is wide enough to be the difference between the Tories continuing to exist as a major political party, and not. So it matters who’s right.

Then we have 29 previews covering some fascinating seats - including the Lincolnshire coastline constituencies, which have the biggest Tory majorities anywhere, but are vulnerable to Reform.

Plus loads of effective marginals in Nottinghamshire and Derbyshire, including the particularly mad contest in Lee Anderson’s Ashfield, as well as Robert Jenrick’s defence of Newark.

Read more

This batch covers the rest of the East of England, including a lot of very tight races and some high profile ones too e.g. Liz Truss and Nick Timothy. And then we get started on the East Midlands - including some seats where Reform will hope to do well and the extremely chaotic world of Leicester East.

But before that, Dylan has done a great map putting together all the MRPs from the past week to show where they differ and where they agree.

I’ve heard a lot of frustration from people that they have such a wide range - from More in Common predicting a Labour majority of 162 and 155 Tory seats, to Savanta predicting a 382 majority and 53 Tory seats.

The five mid-campaign ones, though, (More in Common, Ipsos, YouGov, Survation and Savanta) do, agree on 66% of all British seats - 420 of them. There are another 121 where four of the five agree. It’s just that the seats they agree on are mostly ones that will go to Labour. There are 368 Labour seats where there is complete agreement, but just 21 Tory ones. That’s less than the 23 they all agree will go Lib Dem.

Read more

First up Dylan looks at the story being told by the endless stream of polls we’re getting. By looking under the surface at crossbreaks for age, region, previous voting behaviour, and so on, he sets out, more clearly than I’ve seen anywhere else, exactly what’s changed and why.

Then we have 26 seat previews covering Cambridgeshire and Essex. The latter is home to three of the Tory leadership hopefuls: Kemi Badenoch, Priti Patel, and James Cleverly. It also contains the seat Nigel Farage is trying to win, and plenty of others where Reform would hope to do well. 25 of these 26 seats are currently Tory so it’s a crucial battleground.

A month of polls

After fairly little change in the first few weeks now, a month after the election was called, the polls are telling a different story. The much foretold narrowing of Labour’s poll leads has not materialised, with the Conservatives’ ratings falling sharply and Labour’s lead steady. The similarly expected squeeze of smaller parties has also failed to appear; support for the Lib Dems and Reform is increasing, while the Greens’ support is holding up better than expected. At the start of this campaign, the polls pointed to a worst-ever result for the Conservatives; today, they imply an even lower seat tally.

This multi-party polling picture is fast-moving and has many stories happening at once. This can create a tendency to infer things from cursory glances at the polls – for instance, Labour’s numbers taking a dip around the time Reform began surging has led to suggestions that Reform are now eating into Labour’s vote too. But is that really happening or is this something of an ecological fallacy? To really understand what is going on and what is moving the polls, we need to dig deep into our weekly polling tracker and beyond.

Read more

Then we head into the Anglia region and the commuter belt that sits above London with Bedfordshire and Hertfordshire, a traditionally very Tory area that has been shifting away from them demographically at some speed. Oliver Dowden and Grant Shapps’ seats are both in Herts.

You may have noticed there have been a lot of MRPs out over the last few days, and there’ll be another from Focaldata on Monday. Dylan will do some aggregate results from them all next week but, while they’re showing a fairly wide range of possible results, the broad picture is consistent with everything else we’re seeing. The only real question is exactly how bad it gets for them, with a 1997-style result now representing the upper end of the spectrum.

I’m also conscious I’ve not written about the Labour manifesto (it’s not all that exciting). But you can listen to the Power Test podcast we did on it with Nick Pearce, who wrote Labour’s 2010 manifesto and Claire Ainsley, Starmer’s former head of policy. You can also read this feature I did for Prospect magazine on the challenge facing Labour, which nothing in the manifesto has changed.

Read more

These are mostly safe Labour seats, but precisely because it’s such a safe area for them there’s a lot of interesting factional in-fighting that tells us about some of the fault lines there’ll be in the new parliamentary party.

Dad and I recorded a fascinating interview yesterday which we hope to share in the next few days, and we also have a very cool guest post coming up soon.

Lib Dems, Tactical Voting, and 1997

In the last two weeks the Labour's share of the vote in polling has declined slightly - from around 45% when the election was called, to about 42% now. As discussed in last week's piece on Reform, the rise of Farage’s party is not driving this reduction; nor has there been a change in the number of people who voted Conservative in 2019 switching support to Labour.

Instead, Labour's fall in the polls seems to be partly due to pollsters' changing their methodologies in such a way that depresses the Labour vote share; and partly due to net losses to the Lib Dems. My instinct is that this is a loss of votes for tactical reasons rather than due to disillusionment, because the fundamentals underpinning Labour's position in the polls (Starmer approval ratings; Starmer Best PM lead over Sunak; Labour brand favourability; economic competency, etc.) remain strong.

In fact, Starmer's approval rating is the highest it's been since late 2022 according to YouGov  - the same firm that has put Labour's vote share on an anaemic 37%, compared to ~42% in the polling average. Other firms have found his approval rating rising too. In addition, Dylan’s poll breakdown from last week shows that the Lib Dem bump in the polls appears to be concentrated in the South, which is also suggestive of tactical voting.

The Lib Dem surge

Read more

Then there are 22 seat previews covering west London all the way round from Harrow and Brent to Sutton and Croydon (which is, admittedly, quite an expansive definition of “west”). The rest of London and then Anglia coming this week.

Broken Records: why the election will be unprecedented

If there has been one dominant theme in election commentary in the last year it has been a faith in precedent, or at least folk memories of precedent. The normal response to Labour’s sustained and substantial poll leads, which have continuously pointed to a generational defeat for the Tories since the second half of 2022, has been an assumption that the leads would come down as the election approaches, because that is what ‘always happens’.

It didn’t, Labour’s poll lead actually increased a little. Then came the presumption that the Conservatives would recover in the campaign, because polls ‘always’ narrow once an election is called. With three weeks to go, we have seen more evidence of poll widening than narrowing. Now we are left with a common belief that the models and projections, many of which show the largest-ever Labour majority and worst-ever result for the Tories, must be wrong as results like that ‘don’t happen’.

But we are now at the stage where the only thing that stops an unprecedented result, like those seen in most projections, are things that are themselves unprecedented – i.e., the rules of precedent are firmly out. Of course, to some extent, every election is unique, each one in the last decade especially so. It is hard to imagine how you would even define a ‘normal’ election result. Regardless, this one is set to be more extraordinary than most. So let’s look at the ways this election is set to be at the very least highly exceptional, as well as why we should not really be that surprised at some of the ‘outlandish’ possibilities.

Read more

It’s written by Mark McGeoghegan, an author, columnist, and pollster writing on Scottish politics and secessionism. He is also doctoral researcher at the University of Glasgow, completing a thesis on the political tactics of self-determination groups.

It’s a superb overview of Scottish politics since 2019 and the key battlegrounds this time, including some lovely maps.

There’s so much to come this week. We have a piece by Dylan tomorrow on all the records that could be broken on 4th July, plus loads of London seat previews over the next two days. And some really exciting pieces later in the week too.

The Election in Scotland

Winning a substantial number of seats in Scotland makes it easier for Labour to win general elections, albeit it’s only made the difference between victory and defeat twice: in 1966 and February 1974.

That might not mean much in an election in which Labour can expect a record-breaking win, but re-establishing themselves as the pre-eminent party of Scotland at Westminster could prove crucial in future, much closer elections.

Over and above the importance of Scotland as a battleground for this election, it is important in its own right – not least for all to the people who live here and the parties who seek to govern here – and as an ongoing battleground for the future of the Union itself. The constitutional question may be on the back burner for now, but it has not disappeared.

If Scots feel a tangible improvement in their material living conditions by the end of the next Parliament, the many pro-independence Scots currently intending to vote for Labour may well conclude that that vote was well-placed – and that the Union may be worth preserving, after all. If not, constitutional politics may be back on the menu rather quickly.

So, what is happening in Scotland in this election, what might we expect to happen, and why? I’ll set up the election campaign by laying out what happened in 2019 and the voting intention trends and key events since then. I’ll then set out the new Scottish political landscape and the standings of the parties, and lastly, I’ll get into the nitty-gritty of Scotland’s electoral geography and battlegrounds.

Read more

In the polling average Reform are still 8 points behind the Conservatives. Moreover, YouGov had Labour on 37% when their average is 43%. Clearly, as we’ve been saying throughout the campaign, the Tories are in really serious danger, but we’re not quite at Canada 93 yet.

We are seeing meaningful shifts in the averages, though, as we can see in Dylan’s break down of polling sub-samples from the past week.

Some notable points: the Tories are dropping across most demographics, but particularly amongst younger voters. The Lib Dem bump seems most concerted in the south, which fits with it being related to tactical voting, as that is where most of their targets are. We can also see 2019 Lib Dem voters revert back to the party from Labour. Again consistent with tactical voting. Reform are taking Leave voters off the Tories, and have gone past Labour into second place with that group.

None of this is good news for the Conservatives, and if these trends continue I’ll need to shift some of the seats I’ve predicted they’ll narrowly hold to other parties in my final predictions.

Today’s previews cover the Bristol area which contains the seats belonging to Jacob Rees-Mogg and Liam Fox, as well as Bristol Central - one of the big Labour vs Green battles. I also start on the London seats, on my home turf, with Enfield, Barnet and Haringey.

Read more

Poll-wise we’ve seen both the Conservatives and Labour dip a bit this week. As per Josh’s post on Tuesday this is largely because the Tories are losing votes to Reform and Labour to other smaller parties. “Don’t know” numbers are also starting to fall, which reduces the Labour score a little as “don’t knows” are more evenly split between them and the Conservatives. Labour’s lead, though remains over 20 points, higher than it was at this point in 1997. Moreover, it was dropping then and it isn’t now.

Today we have previews of 22 seats across the rural south west - in Dorset, Wiltshire, and Gloucestershire. All these seats are currently held by the Conservatives and there are three cabinet ministers in this batch.

Below is the Sankey diagram of my predictions up to yesterday (by subscriber Graham Foreman), covering the south east plus Cornwall, Devon and Somerset.

Read more

Then we start our seat previews for the south west, beginning with Cornwall and Devon. A subscriber (@gj4man on twitter) has made this lovely sankey diagram of all my predictions for the south east. It’s really striking how many effective marginals there are. It’s going to be a fascinating election night

Read more

Yet those scenarios - the Tories below 100 seats, or even the Lib Dems as the opposition - still seem so bizarre that they’re not being discussed much in media coverage. Indeed we’re still getting headlines about “Starmer not sealing the deal”, as if it was a traditional campaign. Even at this point the difficulty of considering the unprecedented holds. Will it still if the position is the same, or even more extreme, in a week’s time?

We have loads more great posts coming this week. Today it’s a bumper load of seat previews - looking at the large number of seats in Hampshire and Kent. The latter, in particular, could see some of the biggest swings, and stories, of the night.

Read more

MRPs take large polling samples and identify the demographic and political characteristics affecting people’s voting intentions and then project them onto constituencies based on their demographic and political characteristics. While the exact variables vary from model to model, in an era of more complex political behaviour, they offer a better way of predicting individual seat races than traditional swing models that assume uniform behaviour in each seat.

In the last week, four polling companies have released the results of their early campaign MRP models. While offering quite a varied set of results, they really just reflect the range of recent polls and their implications. More in Common’s echoes the narrower end of the pack, a reasonable best case for the Conservatives and a Labour majority of a mere 114; Find Out Now (with Electoral Calculus) and Survation’s 300+ Labour majorities are more what you would expect if the polls with the largest Labour leads came true; while YouGov’s model and its headline of a nearly 200-seat Labour majority is slap-bang in the middle and how current polling averages would likely translate into seats. Each of the pollsters offers oodles of analysis of their own results in isolation, so what we are going to do is take a look at them collectively – what do they all tell us about the election and where do they agree and disagree.

It should be noted that all four MRPs were done before Nigel Farage became leader of Reform and before Sunak’s D-Day blunder, so should be seen in that light.

Read more

Redfield and Wilton 17 (+3)

BMG 16 (+5)

Survation 15 (+7)

Techne 15 (+3)

Focaldata 14 (+2)

In each case the highest ever Reform score for that pollster. Looking at the raw data we can see that nearly all these votes are coming from people who voted Conservative in 2019. A big question for the rest of the campaign will be whether this surge fades, sticks, or grows. We will have to wait and see, but there is very little loyalty to the Tories out there at the moment, and Sunak having to apologise for skipping D-Day commemorations won’t have helped.

Two of the polls saw a 4 points drop in the Labour vote as well (Redfield and Survation). The three others saw little movement. This looks to me like noise, there was no clear pattern in the polls with a Labour drop (YouGov had a 4 point drop yesterday too, on their new methodology). But it’s one to keep an eye on.

Read more