The Week in Polls

Do 58% of people think Sherlock Holmes was real?

Sun, 01 Nov 2026 18:02:35 GMT

Welcome to the 237th edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

This time, I am fact-checking an apparent opinion poll finding used as part of a comedy quiz question.

This is followed by summaries of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

Does support for nationalising water companies stand up to scrutiny?
A poll carried out for a Reform councillor hits controversy; and
What teenagers think of being banned from using AI or social media.

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

Before we get to all that, this week my eyebrows were raised by Nate Silver’s claim on X that “I’m not sure what it would take to de-entrench the presumption that polls systematically underrate conservative parties or candidates. It’s not true historically, but it’s a cheapie take.”

I looked at detailed, up-to-date research on this in TWIP #225. The research I discussed says:

Drawing on over 38,500 election-poll-party observations across 32 countries and 372 elections between 1945 and 2025, we use a Bayesian hierarchical modeling framework to estimate latent election-party polling error.
We find that, on average, support for parties on the left is overestimated, while support for parties on the right is underestimated.

I’d call that a pretty good take, Nate.

And with that, on with the show.

Do 58% of people think Sherlock Holmes was real?

Is this a museum about a real person? Photo by huan yu on Unsplash.

A few days ago I was catching up on episode 2 of the BBC Radio 4 comedy quiz show Your Number’s Up. The format for one of the rounds sees the teams given a sequence of pairings of a factual claim and a mathematical sum, both of which have the same answer. An example might be “How many Liberal Democrat MPs were elected in 2024 or what is nine times eight?”

The format of those questions strongly implies that the claimed facts are correct - and hence people have a chance of knowing them, rather than having to do the sum.

Which is why I was intrigued by one question featuring the claim that “58% of British teenagers think that Sherlock Holmes is a real historical figure.” Moreover, one of the competitors expressed incredulity at this but the presenter responded with “Yes, 58%”.

Now, of course, this is just a comedy quiz show. But plenty of such shows pride themselves on their fact-checking and I am the proud and permanent owner of one bonus point on The Unbelievable Truth for having once pointed out a factual slip that show made. And that claim sounded like it came from an opinion poll (unless something very strange had happened with the census one time).

So… does that polling result stand up?

It certainly isn’t hard to find stories online claiming that it is true. Digging into them, and using the Internet Archive to check web pages which are no longer online, I found that it all seems to come back to a UKTV Gold press release from 2008:

Shame-faced Brits are increasingly confusing fact and fiction when it comes to historical knowledge – that’s the verdict of a compelling new study which found that most people believe that fictional figures such as King Arthur, Sherlock Holmes and Eleanor Rigby really existed.
The study, specially commissioned by UKTV Gold, tested the nation on its historical knowledge by asking 3,000 people a series of questions relating to famous factual and fictional characters.

A compelling and special study, perhaps, but alas one about which the press release gave no substantive information. Was this a proper poll, and if so, by whom? Or was it right at the other end of the spectrum, a voodoo online poll which one bored Australian pensioner could have spammed with multiple responses? Or something somewhere in between?

Less-than-robust findings have a history of popping up in PR puff pieces, so anyone reading that UKTV Gold press release should have wanted to know more detail before reporting on it, but if they ever did secure such detail, it didn’t make it into any of the stories I have found.

One mystery that the 2008 UKTV Gold piece adds, however, is that it is written up as being about all adults, yet the press coverage it generated, in at least some cases, translated this into being a poll of teenagers rather than adults (e.g. see the Daily Telegraph’s 2008 story which, knowing its audience, leads on the number of people saying Winston Churchill was fictional).

The story certainly spread and even the Special Adviser to the International Criminal Court’s Prosecutor on War Crimes from 2021-2026 took it at face value.

Yet other than introducing the confusion over whether the figures are about adults or teens, I can’t find any of these stories adding substance to flesh out the source of the data. We don’t know the exact question wording, for example, which matters as Sherlock Holmes, although fictional, was in part inspired by a real person.

We do know, though, as The Register spotted, that the UKTV Gold write-up wrongly claimed that Dick Turpin was not real. One blogger did also express scepticism at the time.

Overall, however, it’s a good example of how an eye-catching survey figure can spread and be reported repeatedly without any of the stories containing basic information that would let us know how well founded, or not, the claims are.

To add to the proliferation of questionable claims about how many people think Sherlock Holmes was real, there was a similar clump of stories in 2011, this time from the then internet search engine Ask Jeeves, and again lacking the sort of details that would give us confidence in the accuracy of the numbers.

The story got another lease of life in 2013 because of the then Conservative minister, Michael Gove. As the BBC reported:

Tristram Hunt, a Labour education spokesman and historian, has attacked Education Secretary Michael Gove over his use of evidence.
It follows a Freedom of Information request showing Mr Gove’s claim about children’s lack of historical knowledge had been based on a UKTV Gold survey…
Pressed for further evidence, the Department for Education pointed to surveys commissioned by Premier Inn, Lord Ashcroft, Professor Robert Tombs for think-tank Politeia, the Sea Cadets and London Mums Magazine.

If you really want to go down the rabbit hole of what Premier Inn had said, the freedom of information request can be read here and The Independent reported more on the story at the time.

Returning to the main point, perhaps not surprisingly, when I contacted the UKTV Gold press office to ask them about a press release of theirs from 18 years ago, they didn’t get back to me.

The BBC’s press office, however, did:

The use of surveys will vary according to the nature of the output, with the requirements for entertainment programmes differing from factual and news reporting. The statistic relating to Sherlock Holmes and cited in Your Number’s Up came from a widely-reported UK Gold survey of 3000 participants and, although dating from 2008, would be understood by listeners as being quoted simply for entertainment value as part of a mathematical question.

I suspect readers of The Week in Polls will differ in their views on whether it’s reasonable to expect the quiz question’s answers to have been fact-checked to a decent standard.

But the BBC’s reliance in part on the claim having been widely reported does highlight that, once again, a large part of the problem with polls is how polls get reported. A figure of unknown provenance and (depending on the adults versus teenagers point) possibly also containing a major piece of misreporting received significant national media coverage - and from there it has had a life lasting decades.

Does AI or Wikipedia do better?

I gave several of the most popular AI services a spin to test them out, in each case using their default free version, not logged in, to ask “Do 58% of people think Sherlock Holmes was real?”

ChatGPT started its answer with “Yes!”, though several sentences on it added that “the survey’s findings were widely reported, but its methodology and representativeness were questioned.”

Perplexity did both better and worse. It too started “Yes” and then immediately added “but the claim needs important context”. However, that context was only that the figure came from 2008 rather than mentioning the rest of the problematic context. It did, though, spot the adults versus teenagers inconsistency.

Gemini also started with “Yes” and failed to go on to give any caveats.

Meta’s AI in my WhatsApp account seemed to do much better, starting its answer with “Yes and no”. But it then went on to say “the 58% figure is real”, confidently associating that with teenagers and saying it was true in 2008.

Claude did rather better, starting its answer with “No”. However, it had an advantage, as I could only use it logged in and so it will likely have benefited from my previous instructions to it to improve the quality of its fact-checking. (As I need to know about AI for my day job, I often run tests in Claude and some other LLMs, such as asking them to fact-check copies of this newsletter.) It also spotted the adult versus teenager inconsistency as well as the later Ask Jeeves finding, adding that “Both polls were commissioned by TV and media companies, so I’d treat them cautiously. I couldn’t find the original UKTV Gold tables, and I didn’t find any recent polling that reproduces the 58% result.”

As for a different form of knowledge, Wikipedia also (at the time of writing) presents the 58% figure as fact, by citing the Daily Telegraph story. A reminder that overconfident assertions are not simply an LLM issue.

Elsewhere from me…

I have written often about how slowly much of Whitehall moves. It’s a particular puzzle because this lethargy even encompasses things which you think might move quickly - such as a zero-cost measure to change one line of legislation which would make one of the most popular people in Britain happy. But it’s even more of a puzzle when money is involved, as I wrote about in my newsletter covering Parliament, A Lord’s Eye View. Take a read here:

If you spot a factual error…

Borrowing an idea from Stuart Ritchie and others, if you spot a factual error in an edition of this newsletter, I will give you a free 12-month subscription to the paid-for version (or an extra 12 months on your existing one). Factual errors qualify, but grammatical errors or disagreements over interpretation do not, and my decision is final. Messages about both of those are always welcome, though.

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters, and although it is too soon to see if the Budget moved the polls, bear in mind that Budgets don’t usually do much to the polls.

The table is also online here, along with a more detailed explanation of its content. It is updated regularly throughout the week as new polls are released.

Next are the seat projections from MRPs, also sorted by fieldwork dates, listing those which are based on dedicated fieldwork (rather than models built on averages of other polls).

This time there are new figures from the Good Growth Foundation (GGF) and from PF Global (previously Stonehaven).

Finally, here is a summary of leadership ratings, with a graph from The Times:

Detailed numbers from the latest polls are updated regularly throughout the week on my website here.

For historical figures on both voting intention and leadership ratings stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

If you are subscribed to other email lists of mine, please note that unsubscribing from this one will not automatically remove you from the others. If you wish to be removed from all my lists, simply reply to this email to let me know.

How strongly do people really support water nationalisation?, and other polling insights you shouldn’t miss

The following ten findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

It’s not uncommon for the apparently very strong support for a policy shown by a simple headline question to wilt away when follow-up questions dig into the arguments for and against it. That indeed was a big theme of my recent piece about polling on Europe.
New YouGov data shows something very similar for water nationalisation. Asked the headline question, “Do you think that water companies should be nationalised and run in the public sector, or privatised and run by private companies?” people back nationalisation by 74%-6%. However, when asked

Brexit, ten years on: what the polls say about Europe now

Sun, 25 Oct 2026 18:01:27 GMT

Welcome to the 236th edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

This time, I’m digging into a wide range of polling data to look at where public opinion now stands on Europe.

This is followed by summaries of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

What impact have the party conferences had on the polls?
People’s top priority for the budget; and
What people think of Andy Burnham appearing on Gogglebox.

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

Before we get to all that, another fake pollster update from the US:

Kudos to for catching this one and there is more here from him and his colleagues on this story.

And with that, on with the show.

The polls on Europe: clear views on Brexit, less clear on what to do

Sometimes your psephological dreams come true.

I’d been thinking for a while about returning to the topic of where the polling stands on Europe, especially as I’ve not done a longer piece on this since way back in edition number 94. But each time, I’d rapidly concluded that there just wasn’t enough data available to give a proper answer that went beyond the sugar rush of a headline number or two. But there’s been a burst of extra data recently, so let’s dig in and see where public opinion stands, a decade or so on from the biggest public policy decision made in Britain this century.

Some of that data - as regular readers will not be surprised to see - comes from , who recently wrote:

A few weeks ago, Andy Burnham delivered one of the more accomplished conference speeches of recent political times … At the end of it was a curious section on UK-EU relations. Burnham, someone who years prior had signalled his desire for the UK to rejoin the EU – but who had recently downplayed that view - now launched into a stark denunciation of Brexit and its impacts. The section built to what felt like was going to be an eye-popping crescendo, and then, well – sort of just stopped. All options are on the table, we’ll have a meeting about a meeting, but anyway: that’s all for now folks.

So what does public opinion think of the things on the table? In what follows, I’ve sometimes picked only the one poll to illustrate each point along the way, but other polls and other pollsters show the same picture too.

A negative verdict on Brexit

The headline figures, as illustrated by the YouGov tracker and substantiated by other pollsters, show that by a large margin people say that Brexit was the wrong decision:

The figures have been pretty flat for the past two years, which can be read two ways: that it shows the robustness of anti-Brexit feeling in that the figures have been sustained even after a change in government from Conservative to Labour (plenty of other political trendlines show a sharp movement at that point), and/or that it shows the limitations of the theory that time is relentlessly marching on to make the anti-Brexit position stronger.

Should anything be done?

I like the Ipsos long-term tracker for what issues people say are most important because it uses an open-ended question. People can say whatever they like in response, rather than being given a list to pick from. That means it avoids problems over what you do or don’t include in the list, and any difficulties over how to term an issue - such as whether it is Europe, Brexit, rejoining, remaining out, the EU, the European Community or the EEC. Ipsos’s open-ended question has sailed through all that and so here are merely the past 25 years of results:

There is no sign of an uptick in people wanting politicians to make the UK’s relations with Europe a top issue. The issue continues to rate in the low single digits, even when people are allowed to pick more than one issue.

(Though note that the uptick pre-referendum came late in the day. High politics secured a referendum despite the absence of a public appetite for the issue to dominate politics.)

If something is done, what should it be?

Just because people view something as having been wrong, that does not necessarily mean people think that the previous situation could or should be restored or that it is a priority to do so.

Some of the nuance in that comes out in the views of the small, but not non-existent, chunk of Leave voters who now think it was the wrong decision:

As Labour are in government now and the EU and civil servants would be involved in moves to undo the Brexit vote, those sorts of figures point to the possibility that currently unhappy Leave voters would not be won over by a plan to have Labour ministers work with their civil servants to negotiate something new with the EU.

That said, there is clear and consistent support for doing something. Again, the headline figures show a large lead for undoing Brexit when a simple two-option question is posed, as the table from What UK Thinks shows:

However, those figures are not as solid as may appear. For example, that 62%-38% lead with More in Common for joining comes from underlying data which has (only) 47% saying they would vote to join. It becomes 62% by excluding those who said they would not vote or don’t know. Those are standard, reasonable exclusions. However, the starting point being 47% shows that this is not as solid a lead as many a social media posting implies.

(This point is not specific to More in Common. The latest YouGov figures, for example, for similar reasons start from “only” 52% saying they would vote to rejoin.)

What’s more - and this is a crucial twist on those figures - when poll questions offer a pro-European option short of full rejoining, it often polls better than full rejoining:

Indeed, the level of support for rejoining the EU varies significantly depending on how it is described in a polling question:

Or as pollster Michael Turner of Freshwater Strategy argued in City AM:

On face value there is majority support for rejoining [in our poll] (50 per cent). And this rises further if rejoining means frictionless trade and the ability for Brits to live, work, study, and retire anywhere in the EU (62 per cent).
But the toplines are a mirage…
If rejoining means paying more than ever into the EU budget, just 30 per cent would support the move, while 37 per cent oppose. Forcing them to confront joining the Eurozone reduces support to just one-quarter (26 per cent) and a majority of voters oppose (50 per cent).
The EU has already made it clear that they will never grant Britain a bespoke, buffet-style deal. Any referendum on rejoining will not be a vote on abstract goodwill; it will be a vote on how many more billions UK taxpayers will have to pay, whether to adopt the Euro, how many more politicians to add to the payroll, whether to relinquish fiscal control, and whether to accept regulatory alignment without a veto.

The latter part of that quote is debatable, as others argue that much more flexibility will exist in possible terms with the EU. The point for a polling newsletter, however, is how much the numbers change as the proposed details change.

One other cause of softness in the ‘do something’ figures is people’s belief in who would benefit from doing something. To quote YouGov again:

A small majority of Britons (51-55%) believe that rejoining the EU would have a positive impact on the UK’s economy, businesses and level of international trade, while nearly half (46-47%) expect such a move would prove beneficial to the UK’s diplomatic influence in Europe and our national security…
The public, though, are less convinced that they themselves would benefit. Just 31% of Britons believe their household finances would be improved by the UK rejoining the EU, relative to 32% who expect it would make no difference and 17% who feel they would be worse off if the UK returned to the EU fold.
Likewise, just 30% of Britons expect that the NHS would be better off if the UK were inside the EU.

To summarise this section, the starting point may look like a large lead for voting to rejoin, but the more you dig into the figures, the softer they become.

Even so, people do say they want political parties to commit to doing something, as polling for the European Movement shows:

Even Conservative voters would back some manifesto commitments to getting the ball rolling on joining the EU.

The demographic trends

Despite that potential fragility in support for undoing Brexit set out above, perhaps there is good news for pro-Europeans in demographic trends?

has recently made this point with customary gusto:

Around 3.8 million Leave voters have died since the referendum ten years ago, compared with 2.1 million Remain voters. This means that 14 million survivors voted Remain, while 13.6 million voted Leave: a pro-EU majority of 400,000…
This matters. The referendum campaign by Brexiters to “take back control” promised that a vote to leave the European Union would both enrich Britain’s economy and enhance its democracy. It has long been clear that the economy has got worse, not better. Moreover, the recent story of immigration – the biggest specific concern of Leave voters – is that of losing control, not taking it back.
The one remaining argument for sticking with Brexit is that any attempt to reverse the referendum would be fundamentally undemocratic. The people spoke; their verdict stands.
That is true only if the votes of the dead still count.

Others have made similar calculations too. And yet… I have a strong prior against viewing demographic trends as a sure guide to future political choices. (This is, to be fair to Peter, a subtly different point from the one he makes.)

My bookshelves alone are full of works pointing to long-term demographic changes and extrapolating from them to future political doom or nirvana for one party or another in one country or another. Nearly all of them have aged embarrassingly badly.

Demographic change happens, but people also change.

As I’ve previously highlighted in The Week in Polls, although the Conservatives did very well with young voters at the 1987 general election, perhaps even being the most popular party among them, this did not set the scene for decades of political success. Rather, it was the prelude to one of their worst election runs ever: they only just held on in 1992, and lost three times in a row after that. After 1987, the Conservatives did not win another decent working majority until 2019 (their majorities in both 1992 and 2015 were under 30, and they did not win a majority at all in 2010 or 2017).

Thirty-two years from doing so well among young people to winning again with a good majority!

So while the demographic change argument certainly has some mileage in helping to support the legitimacy of change, it would be foolish to think that means some remorseless movement in a pro-European direction is going to keep on happening year after year. It may, it may not.

How might the debate play out?

Back to Steve Akehurst, and the fragility of those headline ‘let’s rejoin’ numbers:

Once stress tested, the pro-Rejoin majority is a lot tighter - more like 52-48% than the 64-36% you see in standard polling…
We conducted two stress testing experiments here, the full approach and wordings of which are in the full report. The first one used conjoint methodology to expose people – fairly neutrally – to the likely policy obligations and benefits of both a Rejoin and Closer Relationship (eg customs union) style deal, as well as some of the pro and anti arguments they’d likely hear from advocates in a referendum campaign. We did the same in the second experiment except this time we exposed people, via a mock social media feed, to some screeching media headlines they’d see on either side of the debate, including bad faith or hostile ones…
Out of the back of this, in our conjoint, average headline opinion turned out to be more like 52-48% for Rejoin, a reversal of the now somewhat cursed 2016 ratio…
Rejoin is largely pulled down by the Euro, budget payments and voter exhaustion – advocates will need to eliminate or neutralise these issues, while opponents can seek to exploit them

It’s worth noting that his work also finds that:

[A] closer relationship – a customs union style deal - would appear to come through more comfortably, although of course not a totally free lunch [and] several things still push people towards the EU and even Rejoin: especially travel, the economy and defence and security.

That’s notable for parties such as the Liberal Democrats, who are arguing for a staged approach to this issue and focusing on aspects of it like the trio he mentions.

What people expect

So what will happen? Here are people’s expectations:

And is a referendum necessary to achieve that? Here is what the public currently thinks:

The relevance of Prohibition

There are certainly many pieces of good news for pro-Europeans in this data. Brexit is not seen as a success and people overall aren’t desperate to hold on to the status quo. But support for change is brittle in several ways, and it would be easy for pro-Europeans to blow it by failing to respect the more rounded story the data tells us once you look beyond the simple headline numbers.

How to make the best of those opportunities and risks? Well, as I concluded my 2024 deep dive on this same issue:

It took decades of campaigning for Prohibition campaigners to win in the US, getting alcohol banned in the early twentieth century. Once they’d won, their victory seemed set to last. Yet in less than a decade and a half their victory had been undone and the idea killed off politically. So can pro-Europeans take heart from drawing parallels between Prohibition and Brexit?
Find out in an edition of my podcast, Never Mind The Bar Charts, where I discussed just these parallels with Professor Ben Ansell, who delivered the Reith Lectures in 2023.

Elsewhere from me…

If you spot a factual error…

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters, with an increasing number of pollsters now having the Conservatives ahead of Reform UK.

The table is also online here, along with a more detailed explanation of its content. It is updated regularly throughout the week as new polls are released.

Next are the seat projections from MRPs, also sorted by fieldwork dates, listing those which are based on dedicated fieldwork (rather than models based on averages of other polls). Note that for both the Electoral Calculus / Find Out Now and the YouGov MRPs, the seat changes shown are since the previous comparable MRP. In both cases that previous one was not particularly high profile.

Finally, here is a summary of leadership ratings, with a graph from The Times:

Detailed numbers from the latest polls are updated regularly throughout the week on my website here.

For historical figures on both voting intention and leadership ratings stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

What was the impact of party conference season?, and other polling insights you shouldn’t miss

The following ten findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

I was fairly disparaging last week about people who overreact to polls during party conference season, rather than waiting to see where the numbers settle down.
So let’s turn to Opinium’s summary of where the numbers may now be settling:

The battle on the right: eight decades of polling data

Sun, 18 Oct 2026 17:01:05 GMT

Welcome to the 235th edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

This time, I’m diving into polling data stretching back decades to show the long-term picture for the battle on the right of British politics.

This is followed by summaries of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

New evidence on why inheritance tax is so widely disliked despite being paid by very few people;
News on the future of polling (self-)regulation; and
A new JL Partners poll.

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

Before we get to all that, last week I covered the insights from polling carried out for the Holborn & St Pancras by-election. Another poll is now out, this time from Survation. It paints a similar picture, though with rather less emphasis on the role of eventual winner Sagal Abdi-Wali’s own standing in the seat. You can read the details here.

I’ve also been digging into various obscure corners to unearth more political polling from the 1950s to the 1980s to fill out the latest edition of PollBase. Along the way, I found that opinion polling response rates really aren’t what they used to be:

That latest quarterly release of PollBase is now out. As well as another three months of data, changes this time include additional Gallup data from the 1951 Parliament, taken from the Bodleian Library’s set of Gallup/BIPO reports in the Conservative Central Office archives. Thank you to Marcus Collins for sharing his copies of this data with me.

Finally, my eyebrows were raised by Matt Goodwin writing that “Yesterday, for the first time since 2024, the Tories have moved ahead of Reform UK in a national poll … Today, another poll by More In Common found the same.” Why the eyebrow movement? Because as of the day he published that, three other polls in September and October had already shown this too.

And with that, on with the show.

The Conservatives and the right’s vote: what the long view shows

There are two things wrong with much of the coverage of political opinion polling and the reactions to it: it moves too quickly and it moves too slowly.

It moves too quickly, as in people get over-excited about one poll that shows a small movement within the margin of error. It also moves too slowly, as in a trend starts getting regularly recounted long after it started and even after it has ended. You may recall how this newsletter started reporting Reform’s slip in the polls months before it became common to mention in the media, for example.

We’ve had two choice examples of this behaviour recently. First, in some quarters there has been an overreaction to polls taken at the height of a possible post-conference bump for the Conservatives, before we’ve had a chance to see whether it subsides, as post-conference bumps for many parties often do.

Second, there has been recent coverage of the Lib Dems that often talks about the party being overtaken by the Greens in the polls… just as the polls have been showing the Lib Dems regularly ahead of the Greens again.

So we’re suffering both from knee-jerk, precipitate coverage and from coverage that is slow to catch up and out of date.

Rather than indulge in either of those, therefore, this week I am digging into the history books.

Here is a graph of the total vote for the right in all the national opinion polls from 1945:

The usual caveats apply to this sort of data. For example, “all the polls” means “all the polls recorded in the most comprehensive data set available, though that data set isn’t actually complete, as there are yet more obscure records to hunt through, such as newspaper reports of opinion polls in the 1950s from the less famous polling firms”. However, reassuringly, in putting this graph together I did a fair amount of data cleaning and checking on the dates associated with polls. There are a fair few improvements to that data ready for the next version of PollBase but none of that work changed the graph in any way I could spot. There’s enough data in there to weather the problems with it. It is also a very simple average, without the complexity of proper adjustments for polling errors.

One thing that strikes me when looking at the graph is how little attention is paid to the long, sustained recovery in the right’s vote share from 1995 to early 2017, driven predominantly by the Conservative Party. When was the last time, for example, you heard a political commentator talk of the importance of Kemi Badenoch learning lessons from that multi-decade period of growing popularity?

But of course, “the right” is not just the one party, so let’s look at how the battle between parties on the right has played out. This graph starts in 2009, as before that it’s a pretty boring horizontal line hugging 100%.

I’ll let readers pick their own interpretation of the final part of the graph: does that show a collapse in the Conservative position as the dominant part of the right, a collapse that the recent recovery has barely dented, or does it show how Kemi Badenoch has turned around a catastrophic trend and put things back on the right course? Take your choice, dear reader.

Nutters, spinners and losers: The Thick of It with Sam Freedman

It was back to one of the all-time classic political TV comedies for the latest episode of my podcast with , Political Fictions. We had with us a very special guest, the co-author of Britain’s most popular political Substack newsletter, Sam Freedman:

Top author and newsletter writer Sam Freedman joins us this week to talk about The Thick of It special episodes ‘Rise of the Nutters’ and ‘Spinners and Losers’.
We talk about how prescient the episodes are about both the rise of Gordon Brown and the Conservative rebranding under David Cameron, top political advice about how you can ensure you get a good desk in an office, Cory gets massively distracted by the cricket, and we end by wondering whether the show has been misunderstood.
Sam’s newsletter is Comment Is Freed, and you can listen to Mark’s previous chats with Sam about his writing here and here.
If you are in the UK, you can watch these episodes on BBC iPlayer.
Cory’s email newsletter is Paperback Rioter and Mark has a family of email newsletters.
Subscribe to our newsletter for some additional bonus content and to get alerted whenever a new episode comes out.
Our theme tune is “Monkeys Spinning Monkeys” by Kevin MacLeod/incompetech.com and licensed under the Creative Commons: By Attribution 4.0 License.

You can also listen to our episode on Apple Podcasts, Spotify, the web or YouTube.

If you spot a factual error…

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters. A shout-out to Trajectory for providing details on where their panel data comes from, something I’ve previously highlighted as an area where polling transparency could be improved.

The table is also online here, along with a more detailed explanation of its content. It is updated regularly throughout the week as new polls are released.

Finally, here is a summary of leadership ratings, with a graph from The Times:

Detailed numbers from the latest polls are updated regularly throughout the week on my website here.

For historical figures on both voting intention and leadership ratings stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

Why is inheritance tax so unpopular?, and 9 other polling insights you shouldn’t miss

The following ten findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

Even among those unlikely ever to pay inheritance tax, it is a very unpopular tax. That unpopularity heavily skews debate over making changes to it even though it might otherwise look a tempting tax to change, given that only a small minority have to pay it. Why is it so unpopular?

Holborn and St Pancras: what the polling says about Labour’s win

Sun, 11 Oct 2026 17:00:50 GMT

Welcome to the 234th edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

This time, I’m taking a look at what a constituency poll tells us about the reasons for Labour’s win in this week’s Parliamentary by-election.

This is followed by summaries of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

What a former Conservative Director of Strategy makes of the party’s response to its 2024 defeat;
Polling, polling and more polling about the state of the Conservative Party; and
Attitudes towards AI companies.

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

Before we get to all that, a quick follow-up to last time’s piece about the state of play in Kemi Badenoch’s own seat. The Daily Telegraph has since reported:

Reform UK will target Kemi Badenoch’s constituency with a US-style primary to find a candidate to fight the Essex seat at the next general election.

My usual reminder as well: please do hit reply if you’ve spotted a poll, focus group or new piece of poll-based research that you think should be in contention for coverage in a future edition. Thank you!

And with that, on with the show.

Polanski may have lost, but was he the reason Labour won?

The popularity of Andy Burnham and of Labour’s candidate, now MP, Sagal Abdi-Wali was key to the party’s win in the Holborn & St Pancras by-election. That’s the finding of a constituency poll in the seat carried out by Convergent Opinion for Persuasion UK during the campaign:

The data tables are here. Note that the sample size is reasonable, but small enough to warrant caution about some of the cross-breaks.

Note how low down the list Zack Polanski comes; though he failed to win, he wasn’t a huge drag on the Green vote either, something also illustrated by the net favourability scores for him and others:

Rather, what most cost the Greens were warmth towards the Labour government and a paucity of belief that the Greens could win:

One other finding from the poll that caught my eye concerned whether Andy Burnham’s focus on the north and Manchester has undermined his image as someone interested in the people of London:

Read more about the poll from Steve Akehurst on Bluesky and Morgan Jones in the New Statesman.

State of Play: the classic BBC drama, part 2

Photo credit: BBC.

The latest episode of my podcast sees Cory and me returning to the classic BBC political drama State of Play, as Cory finally sees the wisdom in my praise for the show:

Mark and Cory conclude their discussion of the classic TV series State of Play. Mark reports back on his correspondence with a cast member and also auditions for a role as a defence counsel, whilst they also talk about which elements of the plot remain realistic today and which are complete MacGuffins.
The Guardian article Cory refers to can be read here and the previous podcast episode, covering State of Play episode 1, is here.
Cory’s email newsletter is Paperback Rioter and Mark has a family of email newsletters.
Subscribe to our newsletter for some additional bonus content and to get alerted whenever a new episode comes out.
Our theme tune is “Monkeys Spinning Monkeys” by Kevin MacLeod/incompetech.com and licensed under the Creative Commons: By Attribution 4.0 License.

You can also listen to our episode on Apple Podcasts, Spotify, the web or YouTube.

Elsewhere from me…

If you spot a factual error…

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters. It’s too early to see where things settle down after the conference season, but Labour’s improved poll performance is being sustained.

The table is also online here, along with an explanation of its content. It is updated regularly throughout the week as new polls are released.

Finally, here is a summary of leadership ratings, with a graph from The Times:

Detailed numbers from the latest polls are updated regularly through the week and are online here.

For historical figures on both voting intention and leader approval stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

Do the Conservatives know why they lost in 2024?, and 9 other polling insights you shouldn’t miss

The following ten findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

Lord Andrew Cooper, pollster and former Conservative Director of Strategy, has written for ThinkLabour about his doubts over the Conservative Party’s reaction to its 2024 defeat. He writes:

Is Kemi Badenoch in trouble in her own constituency?

Sun, 04 Oct 2026 17:02:14 GMT

Welcome to the 233rd edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

This time, I’m taking a look at what the recent MRPs tell us about the state of play in the constituency of Conservative Party leader Kemi Badenoch.

This is followed by summaries of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

Polling on Andy Burnham’s conference speech;
Big news for the British Election Study (BES) that underpins so much of what we know about political opinion in the UK; and
How Rupert Lowe’s polling is emulating Zack Polanski’s.

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

Before we get to all that, my eyebrows were raised this week by The Guardian’s story “Failure to protect free speech could cost Labour votes at next election, says Liberty”. Why? Because the story contains all the basic errors to avoid when reporting on such policy polling - as it is based on a poll that asks just about one policy issue in isolation (which isn’t how people decide how to vote), and takes the answers at literal face value (which isn’t borne out by the predictive record of such questions).

For some rather, er…, punchier responses to the story and why the polling doesn’t tell us what the headline claims, see Verian’s Joel Williams and Southampton’s Will Jennings.

And with that, on with the show.

A seat to watch: what the MRPs say about Kemi Badenoch’s constituency

Over the summer, Political Betting ran a piece titled “Kemi Badenoch’s little local difficulty”. It pointed out how small her majority had been at the 2024 general election and how the combined Labour/Lib Dem/Green vote was well ahead of hers, leading to speculation about how at risk her own seat might be at the next general election.

Now that we’ve had a batch of five recent MRPs, how does the picture look in her seat? The resolution of these MRPs is sufficient to pick up factors such as a general Conservative recovery, an improvement in her national ratings, the incumbency effect for Conservative MPs, increased tactical voting and so on - but the MRPs can’t pick up a special personal vote for a party leader.

With that caveat in mind, here is how the figures look:

What to make of these figures?

First, the Conservative vote share is rather anaemic for a party that should be recovering from a historic thumping at the last election. None of the MRPs show the party’s support up on 2024 at all. In reality, though, Badenoch probably has a boost in her seat due to her visibility as party leader, so the real figures are likely to be better for her than those for her party.

Second, for all my praise last week of the similarity between the different MRP results, these figures show the amount of churn below the top lines. There’s a clear split in who the top two in this seat are. Possible tactical voters in this seat are not helped - at this stage - to figure out who the contest is between and who they might want to vote tactically for/against. This is the other boost for Badenoch in these figures: confusion which may keep votes for others split.

Both Reform and Labour supporters can dream, looking at these figures, that just maybe their party can win. And while Badenoch’s position is probably somewhat better than these figures suggest, for the reasons above, her seat certainly doesn’t look like one she can neglect while campaigning around the country.

This will be a seat to watch.

State of Play: a classic BBC political drama revisited

In the latest episode of my Political Fictions podcast, we’re heading back to the early 2000s, for a classic BBC political drama:

Mark and Cory begin their two-part series on the classic TV series State of Play by discussing episode one. It begins with Mark recoiling in horror at the low score Cory intends to give it, almost causing an irreparable rift in Political Fictions Towers. Thankfully, they manage to repair the relationship by talking about what the show tells us about politics in the early 2000s, praise the excellent cast (especially those who are able to stand awkwardly in the background very effectively) and undertake some seriously impressive letterbox-based investigative work.
You can view more about State of Play, including details of where it’s currently streaming, here.
Cory’s email newsletter is Paperback Rioter and Mark has a family of email newsletters.
Subscribe to our newsletter for some additional bonus content and to get alerted whenever a new episode comes out.
Our theme tune is “Monkeys Spinning Monkeys” by Kevin MacLeod/incompetech.com and licensed under the Creative Commons: By Attribution 4.0 License.

You can also listen to the episode on Apple Podcasts, Spotify, the web or YouTube.

Elsewhere from me…

If you spot a factual error…

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters.

The table is also online here, along with an explanation of its content. It is updated regularly throughout the week as new polls are released.

Next are the seat projections from MRPs, also sorted by fieldwork dates, listing those which are based on dedicated fieldwork (rather than models based on averages of other polls). Note that the seat change figures for the YouGov MRP are between their latest MRP (for Sky News) and the standard voting intention scenario in their MRP for Best for Britain earlier this year, rather than changes between the latest Sky News MRP and the previous Sky one from last year.

Finally, here is a summary of leadership ratings, with a graph from The Times:

Detailed numbers from the latest polls are updated regularly through the week and are online here.

For historical figures on both voting intention and leader approval stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

The first polls on Andy Burnham’s conference speech, and 9 other polling insights you shouldn’t miss

The following ten findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

We already have a batch of polling related to Andy Burnham’s first conference speech as Prime Minister, but let’s start with some polling from earlier in the year on one of the key issues in the speech, the triple lock:

The same but also different: the picture from five MRPs

Sun, 27 Sep 2026 17:00:47 GMT

Welcome to the 232nd edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

This time, I’m taking a look at the batch of new MRPs, and what the pattern across them all tells us about the state of politics.

This is followed by summaries of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

Where people got their news about the May elections from, including how leaflets compared with social media;
A tactical voting handicap for the Conservatives and Reform; and
Which party’s supporters use AI the most.

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

Before we get to all that, two little pieces of news about the troubles polling has been in, one good and one bad.

The good news came from the Daily Telegraph, the paper caught up in the Reform donors exposé as it had wrongly claimed a JL Partners poll had been carried out for it when instead the poll originated via Reform. But a few days ago, I spotted transparency about a poll’s origins above and beyond what’s been the norm in the past in the media: “The polling company Find Out Now (FON) has kindly shared a wealth of data from its own regular monitoring of public opinion about politics, research the firm funds itself.” Let’s hope that is more than a one-off.

The bad news came from IPSO, which describes itself as “the independent regulator for the UK digital and print news industry” and says “we uphold high editorial standards”. I covered earlier in the year a Daily Express story about the polls, with a headline that was not substantiated, indeed was contradicted, by the story that followed. The Express rejected my complaint and now IPSO has also decided not even to investigate it, despite it being such a clear case of misleading coverage and despite the Express’s own response to me strongly implying that it doesn’t think the IPSO Code’s requirements about headlines apply to it. IPSO did not even provide a reason for the rejection.

That’s not upholding high editorial standards; it is giving the green light to misleading headlines.

And with that, on with the show.

What five MRPs agree on, and where they differ

Four (four!) MRPs have been released in recent days. Given how coverage of political polling usually goes, though, while the MRPs have collectively had a good chunk of coverage, nearly all of it has been about one or another rather than looking at the full set. Yet of course pollsters and polling pundits repeatedly remind us not to pay attention to just the one poll. So let’s put that right, and look at all four.

Or rather, let’s look at five, because the fieldwork dates for the four new ones also straddle the release, a few weeks back, of a fifth from Stonehaven:

Where an MRP has excluded the Speaker’s seat from its calculations, I have added one to the ‘other’ column. The Electoral Calculus/Find Out Now MRP was carried out for Restore, but Martin Baxter has confirmed that the methodology was very similar to that used for their quarterly MRPs with PLMR and that the results are directly comparable.

Four things strike me about this.

First, the similarity of seat numbers across all the MRPs for Labour, the Lib Dems and the Greens. The political story is also the same for each party across the MRPs. Labour is consistently the largest party, though well short of a majority (a big net loss of seats, but a huge improvement on the situation under Starmer), the Lib Dems are on figures very similar to their general election result (which, given the party’s traditional mid-term dip in such things, is promising) and the Greens are performing around, or a little better, than their 2024 general election result (enough to give reasonable hope for making gains but also well down on their boom earlier this year).

Second, there is a big range for both the Conservatives and Reform, and these encompass rather different stories as to which is largest and so has the upper hand on the right.

The Conservative plus Reform total, however, varies modestly, in large part due to the significant number of seats that have Reform and Conservatives finishing in the top two regardless of which one is the winner. The big political significance is that the combined total, however it breaks down between the two parties, consistently comes in a long way short of what is required for a majority.

Third, in two respects, the MRP figures need treating with even more caution than usual. The pollsters are split over their picture for the SNP and, in addition, the resolution of MRPs is not high enough to reliably document Restore’s fortunes given its current levels of support and geographic concentration. Three of them ‘showing’ that Rupert Lowe would lose his seat is not meaningful; identifying the size of any personal vote for him in his own constituency is more than an MRP can do. (Nor should you pay too much attention to the MRP figures in the constituencies for the high-profile Conservative to Reform defectors as again, that’s beyond the resolution of these MRPs.)

Fourth, a lot of seats are very close in these figures. As an example, look at this analysis from YouGov:

Survation has a similar picture:

Or as More in Common put it:

Of the 631 seats contested by the parties, 302 (48 per cent) would be won by a margin of fewer than five points, 137 by fewer than two points, and 82 by less than a single point.

This cuts both ways regarding the utility of MRPs. Having so many seats be so close on the one hand makes life much harder for pollsters. On the other, it helps show the value of MRPs above simply having national voting intention polls.

People will of course pore over the details of individual seats, giving them more attention and reading into the figures greater precision than is warranted. Overall, however, the picture across them is sufficiently similar to give some reassurance that they are telling us a useful story.

And also, bearing in mind the political significance of the range in Conservative and Reform numbers, you really shouldn’t pay attention to just one MRP on its own.

Elsewhere from me…

If you spot a factual error…

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters. (The fieldwork for most of the ‘new’ MRPs started sufficiently far back that they have not made it into this table.)

The table is also online here, along with explanations about what is included. It is updated regularly throughout the week as new polls are released.

Next are the seat projections from MRPs, also sorted by fieldwork dates, listing those which are based on dedicated fieldwork (rather than models based on averages of other polls).

Finally, here is a summary of leadership ratings, with a graph from The Times:

Detailed numbers from the latest polls are online here, and both that and the graph are updated regularly throughout the week as new data comes out.

For historical figures on both voting intention and leader approval stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

How did people get news about the May elections?, and 9 other polling insights you shouldn’t miss

The following ten findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

Polling by Savanta for Ofcom, carried out shortly after the May elections but only just published in full, shows that, when it comes to getting news about those elections, the most popular source of information picked by people was

Why politicians should pay more attention to the polls (reprise)

Sun, 20 Sep 2026 17:02:08 GMT

Welcome to the 231st edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

With all the stories swirling around Reform, JL Partners and political polling, I am finally tearing myself away from talking about the regulation of polling this week - and instead re-running a previous piece reminding us why the solution to problems with polls isn’t simply to ban or ignore them. Rather, we need them to play a vital role in making our politics - and politicians - better.

This is followed by summaries of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

A trio of organisations and a pair of professors have been casting their eyes over the Liberal Democrats ahead of the party’s autumn conference;
What impact did the summer heatwave have on views on climate change? and
Which wins out between chocolate and cheese?

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

Picking up on my doubt last week about whether or not 35% of British train travellers have tried to order a takeaway for delivery to a railway station or carriage, thank you to reader Iain, who has been in touch to say he has done just that… in India.

Another quick pair of updates, this time on the controversies around Reform and polling. Politico has a further scoop - “Farage aide caught in donor sting attended Egypt retreat with JL Partners” - and has written about the problems with polling too: “Seeing the response of many polling experts in the last few weeks suggests that they are nervous about what is happening and keen to mitigate against it. Seeing the response of journalists, which has been basically non-existent, suggests the precise opposite.”

And with that, on with the show.

How politicians get public opinion wrong

Politicians have to pay attention to the public when it comes to election time. But how good are they at understanding the public in between elections? Image by John Mounsey from Pixabay.

I have written before about the benefits that come from widespread availability of high-quality political polling. We need more polls and more good coverage of polls.

However, what I’ve not dug into before is the research into how accurately politicians estimate public opinion on issues.

Thankfully we have academic research to tell us.

First, the bad news:

We present results from a study of 866 politicians in four countries.1 Politicians were asked to estimate the percentage of public support for various policy proposals. Comparing more than 10,000 estimations with actual levels of public support, we conclude that politicians are quite inaccurate estimators of people’s preferences. They make large errors and even regularly misperceive what a majority of the voters wants. Politicians are hardly better at estimating public preferences than ordinary citizens.
Walgrave, S., Jansen, A., Sevenans, J., Soontjens, K., Pilet, J.B., Brack, N., Varone, F., Helfer, L., Vliegenthart, R., van der Meer, T., Breunig, C., Bailer, S., Sheffer, L. & Loewen, P.J. (2023a), “Inaccurate Politicians: Elected Representatives’ Estimations of Public Opinion in Four Countries”, The Journal of Politics, 85(1): 209-222.

The bad news is not quite as bad as that makes it sound because, when asked whether the majority of public opinion was for or against various policies, politicians picked the wrong option ‘only’ 29% of the time. However, this included getting it wrong even when public opinion is far from evenly split on an issue:

They do not only get it wrong when the public is divided and hard to read (e.g., when there is a 51%:49% distribution). Seventy percent of the misplacements occur for policies with a relatively clear distribution of at least 60% (dis)agreeing citizens.

Moreover, the public scored only slightly worse than politicians, picking the wrong option 2.7 out of 8 times on average, compared with 2.3 times for politicians.

Turning to the other measure of accuracy (not which side is in the lead but what the percentage support is for each side), politicians on average were 18 percentage points out in their guesses. Again, this was only slightly better than the public, whose error came in at 21 percentage points.2

Further research, using the same mix of countries, gives a little more comfort about how (in)accurate perceptions are:

We show that politicians have a better understanding of public opinion when they think the issue matters to voters. Further, when an issue is personally important to politicians they more accurately estimate their party supporters’ opinions. The results confirm that politicians hold more accurate perceptions of voters’ preferences when they think it is important to do so but not necessarily when the issues actually are important to voters.
Butler, C., Walgrave, S., Soontjens, K. & Loewen, P.J. (2024), “Politicians are better at estimating public opinion when they think it is more salient”, Party Politics: 1–13.

Even so, it’s not a good picture.

The remedy for this?

Politicians should spend more time reading the polls.

So forward this newsletter to someone you know…

For a more general argument against banning polls, see my 2024 piece.

Is “Let Bartlet Be Bartlet” good advice or not?

The latest episode of my Political Fictions podcast heads across the Atlantic for a dip into one of the political TV classics:

Mark and Cory talk about a giant of American political drama for the first time this week, as they tackle The West Wing. They talk about Let Bartlet Be Bartlet, the nineteenth episode of series one. In it, Mark wonders if The West Wing is actually the epitome of the “Boris Johnson school of politics”, Cory outmanoeuvres Mark into confirming we will record a Christmas Special about Love, Actually and they debate whether the central message of this episode has been misunderstood.
Cory’s email newsletter is Paperback Rioter and Mark has a family of email newsletters.
Subscribe to our newsletter for some additional bonus content and to get alerted whenever a new episode comes out.
Our theme tune is “Monkeys Spinning Monkeys” by Kevin MacLeod/incompetech.com and licensed under the Creative Commons: By Attribution 4.0 License.

You can also listen to our episode on Apple Podcasts, Spotify, the web or YouTube.

Elsewhere from me…

If you spot a factual error…

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters. It is all pretty static at the moment, with a small and consistent Labour lead in nearly all the polls.

The table is also online here, along with explanations about what is included. It is updated regularly throughout the week as new polls are released.

Finally, here is a summary of leadership ratings, with a graph from PollCheck:

The green line is for Zack Polanski.

Detailed numbers from the latest polls are online here, and both that and the graph are updated regularly throughout the week as new data comes out.

For historical figures on both voting intention and leader approval stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

A quintet of Lib Dem polling analyses, and 9 other polling insights you shouldn’t miss

The following ten findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

Ahead of the start of Liberal Democrat conference, we have a trio of organisations and a pair of professors with pieces out about Ed Davey’s party:

A crisis point for political polling

Sun, 13 Sep 2026 17:01:38 GMT

Welcome to the 230th edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

With more news breaking about the regulation of polling this week, we are at a crisis point for the current approach. So this week’s edition majors on this topic, for the fourth week in a row - and I’d encourage the many recently subscribed readers to read those three previous pieces for the full background (1, 2 and 3).

This is followed by summaries of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

The first early polling signs of the impact of the Reform donations scandal on the party’s standing;
New polling on attitudes towards Brexit; and
What people get wrong about London.

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

Before we get to the main business, a fond farewell to my 103-year-old reader who has just decided it’s time to scale back his reading a little. May you have many more years of other reading to come!

My usual reminder too: please do hit reply if you’ve spotted a poll, focus group or new piece of poll-based research that you think should be in contention for coverage in a future edition. Thank you!

And with that, on with the show.

More revelations, more questions about the future of polling

The past week has seen another media scoop about political polling as well as news of a regulator ruling against a polling firm. Plus I’ve caught up on an odd story about takeaways and trains which highlights some of the related issues too.

Let’s start with the scoop, and kudos to Politico for following up the Channel 4/Verbatim story about Reform, donations and polls with further examples of things going wrong with political polling.

Breaking down the details in its full story, first we have another example of the funding for a poll being kept back from the public, again involving the polling firm which featured in the Channel 4/Verbatim revelations, JL Partners:

JL Partners director [James] Johnson wrote an op-ed in the Telegraph earlier this year discussing a shift in public opinion against assisted dying without mentioning the poll cited in the piece was paid for by the Christian lobbying group CARE.

Then there are further examples of polls appearing in the media without appropriate transparency, and where they may have been funded by people with a particular political agenda but whose identity was concealed:

Johnson’s firm conducted polls on three occasions during the Tory leadership elections in 2022 and 2024 which received press coverage without disclosing who paid for them either in the media reports or in the company’s published tables.

It’s possible, in theory, that the pollster simply ran the polls itself, covering the costs internally, as a marketing move. But the evidence in the piece suggests this was not the case.

One of those polls appeared in the Daily Telegraph with no indication of its origin other than the pollster. (No client or funder is named in the newspaper’s coverage.) Moreover, Politico reports:

A person familiar with the Telegraph’s publication process said they didn’t have any record of being given any more information than what was published.

If that recollection is accurate, the newspaper ran a story without asking ‘where has this really come from?’.

Another appeared in the Daily Express (subject of a complaint on a different polling matter from me to IPSO), and the paper simply declined to comment to Politico on what it had known about the poll it too had published without giving any client or funder details.

Finally, there are more examples of the sloppiness about publishing data tables which I’ve commented on before. Politico reports that:

A recent Opinium poll carried by the website Guido Fawkes which highlighted Londoners would rather have an entrepreneur than an ex-minister as mayor did not state the poll was commissioned by London mayoral hopeful Festus Akinbusoye. While this information was included in an Evening Standard write-up of the poll, it was only published belatedly by Opinium…
In another case in January 2023, the Liberal Democrats commissioned Savanta to ask voters who was responsible for the problems the country faces … some of its results were briefed to The Observer newspaper, which brought it under the British Polling Council’s transparency rules. Despite this, Savanta did not upload the full tables - confirming the client, sample size and details of the questions and answers - until Politico contacted the firm this week.

Missing tables are a regular issue. I fairly regularly email polling firms to chase up the non-appearance of data tables within the British Polling Council’s two-day deadline, for example. I don’t think there’s anything nefarious about these cases, but still, it shows that the rules are not being zealously followed.

Moreover, the other problems that have come to light make this sloppiness more problematic as, innocent though it may be, it still creates a pattern of delayed or missing transparency within which nefarious people can seek to deliberately obscure information.

Likewise, the widespread assumption (which I’ve made too) that when a client is not mentioned by a polling firm, the firm has paid for it itself in order to promote or assist its work, is often true. But as the recent stories show, this also has created a space which can be exploited by the less well-intentioned.

As I told Politico:

This is a moment of reckoning for the polling sector. Who has arranged for a polling firm to carry out a poll and who has paid for the poll need to become standard information released at the time a poll is published.

British Polling Council rules against Whitestone Insight

To add to all that, the pressure group Campaign for Dignity in Dying have published a piece headlined:

Anti-assisted dying polls breached rules, British Polling Council finds
The British Polling Council has upheld a complaint about Whitestone Insight. Polls intended to influence assisted dying debates in the House of Lords and Holyrood did not meet the industry’s professional standards.

It goes on to explain that:

In March, the polling company Whitestone Insight published findings from two polls on assisted dying involving 1,004 adults in Scotland and 2,080 adults across the UK. The polls included questions commissioned by anti-assisted dying campaign groups, such as Care Not Killing and Not Dead Yet. Findings from the polls were reported in the media and cited by Parliamentarians opposed to law change…
Rule 2.3 of the BPC’s Objects and Rules states clearly that when polling results enter the public domain, its members must publish the wording of “any prior questions on the same or related topics” that were put to respondents.
It appeared that Whitestone Insight repeatedly failed to do this and acted without appropriate transparency. Instead, it published questions from its polls separately, which meant there was no way for people to know respondents in the polls had already been presented with other questions on the same subject. This is known as priming and doing this without full transparency is exactly what rule 2.3 is designed to prevent.

The most famous example of how priming people can skew results came in the second episode of Yes, Prime Minister, in which Humphrey Appleby explains to his more junior colleague Bernard Woolley how the ordering and wording of polling questions can be used to get the answer you want:

This was not just fanciful satire; Ipsos replicated the fictional questions in some real polling and found that yes, the results do get skewed.

Have you ordered a takeaway to your train seat?

Are they heading to Bristol Parkway to deliver a pizza to the 09:53 from London Paddington? Photo by Rowan Freeman on Unsplash

Finally, I’ve been catching up this week on an odd little story from The Independent in late August:

Domino’s launches pizza delivery service for train passengers…
A majority (57 per cent) of travellers thought that food on offer at stations and on board UK trains was lacking.
A third (35 per cent) said they had attempted to get a takeaway delivered while on board a train - but only one in five (20 per cent) managed it successfully.

35%?!

Now, I often say to others when they have expressed incredulity about a polling result that a great thing about polling is that it not only tells you about people who are just like you, it also tells you about other people who are different from you.

What’s more, a bit of careful reading starts to make that 35% seem a little more plausible to me.

The figure comes from a publicity campaign organised by the agency Hope&Glory and which featured Domino’s, Lumo and The Trainline.

The latter got the 35% figure wrong, claiming it was 35% “of Brits”. Given not everyone uses trains, that would have been a sky-high figure indeed.

But as The Independent piece implies, and as The Trainline confirmed to me, the poll filtered out people who have “never travelled by train”, who were 10% of the sample, leaving 90% as train travellers. That 35% is 35% of the 90% who have travelled by train.

That 90% number, though, seems on the high side as the 2025 National Travel Survey for England puts it at 61% using surface rail at least once a year - though of course some people will have used the train more than a year ago (and the poll was of the UK, not just England).

Then the takeaway ordering question has two nuances to note. It includes people who have tried to order takeaways while on a train for delivery to a “train carriage” and also to a “train station”. Moreover, there is no time limit on how far in the past someone has tried to do this and they may have done it just the once.

Even so, the idea that over a third of train travellers have tried to order a takeaway like this strikes me as… surprising.

Not only because of my own personal experience of not seeing, hearing or smelling this activity, but also because of my failure to notice the likely consequences of such ordering being a regular occurrence: seeing people receiving the takeaways, regular abandoned takeaways on trains or platforms, delivery drivers waiting to meet a train, complaints at some stations about increased litter or obstructions from delivery bikes, the train YouTuber doing a piece about the station that has the most takeaways delivered to it, the traveller tips or marketing promoting the best delivery firm to use, etc.

It all, still, seems implausible. Perhaps you are about to fill my inbox with emails about how you got prawn crackers delivered to Bristol Parkway last Friday night. Perhaps I’m about to be tagged in social media posts showing folk tucking into their pizzas on the West Coast Mainline. Perhaps… but I’m still sceptical.

Yet in several key respects the poll passes muster. It is from an established, widely used polling firm - Censuswide. The data tables are online for all to see. The firm is a member of the Market Research Society (MRS) and the British Polling Council (BPC).

Why then mention this food-meets-transport poll in a newsletter about political polling?

Because it shows the limitation of polling transparency. It can still leave you a long way short of enough information to judge whether or not a poll is accurate. This is one reason why some British Polling Council members are keen that BPC membership isn’t seen as a quality mark, because obeying the rules, desirable though that is, isn’t a guarantee of a poll being conducted well using a high-quality sample.

Moreover, while you might hope that a national newspaper reporting on such a poll would do some digging to check out the numbers, that isn’t what we find in The Independent’s story. Rather, its wording is extremely similar to the wording used by The Trainline and Lumo in their own marketing material, strongly suggesting a high degree of churnalism by the newspaper. Moreover, there is no surrounding information added by the newspaper, such as reports from a train station operator of their frustrations at extra litter on platforms or an accessibility lobby group complaining about disabled parking bays getting blocked by delivery bikes. In other words, none of the additional information you would expect to come to light if the poll were right and if the newspaper also had (the time and money to) put in some extra research to check out if the story stands up.

So it’s a story that shows the limitations of polling transparency and of how polls get covered by the media, both of which are problems for getting political polling out of its current pickle.

Elsewhere from me…

If you spot a factual error…

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters. They have settled into a pattern of generally showing Labour a small handful of points ahead of Reform and the Greens have slipped back from being consistently ahead of the Liberal Democrats.

The table is also online here, along with explanations about what is included. It is updated regularly throughout the week as new polls are released.

Finally, here is a summary of the latest leadership ratings, sorted by pollster name. The established pattern here is of Burnham having the best ratings, with both Polanski and Farage not having recovered from the previous falls in their ratings. They have not fallen further, but nor have they bounced back.

The table is also online here with additional details, and is updated regularly throughout the week as new data comes out.

For historical figures on both voting intention and leader approval stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

Is there a political cost to Reform for the controversies over their donations?, and 9 other polling insights you shouldn’t miss

The following ten findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

The full impact of a (claimed) political scandal on a party or politician can take months to properly play out in the polls. But we do now have some early signs of

Extra questions about the polls in Reform’s funding scandal

Sun, 06 Sep 2026 17:01:32 GMT

Welcome to the 229th edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

This time, for the third week in a row, I’m writing about the regulation of polling. I was going to move on to a different topic, but then the Thursday evening edition of Channel 4 News aired. It puts three polls in the spotlight, but as I write below, there are good reasons to think there are other polls which should also be in the spotlight - and that the newspapers involved failed to heed the lessons from previous incidents over how they report polls.

This is followed by summaries of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

The latest on which issues people say are the most important;
How attitudes to members of the Royal Family vary sharply by age; and
A metric on which Kemi Badenoch matches Nigel Farage.

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

My eyebrows were half-raised this week by news from IPPR that young people are “five times more likely to have no friends than a decade ago”. In this case that five-fold increase is from 1% to 5%. I think it’s fair to say that an increase of four percentage points is rather less dramatic than most people would take from that headline, but the IPPR’s Ellie Harris made some good points in response to me.

Also, a quick digression into US podcast land before we get to the main business. I suspect many readers of TWIP used to enjoy listening to the FiveThirtyEight podcast. Three of its former hosts - Nate Silver, Clare Malone and Galen Druke - are back with a new show, Still Counting.

It was Clare Malone’s departure that I missed most from the previous podcast (sorry Galen Druke, and as for Nate Silver - well, there’s a lot to be said about how he’s evolved over the years). I will be giving it a try - and for once the below-the-line YouTube comments contain some wisdom, with lots of praise for Clare Malone’s return.

And with that, on with the show.

How newspapers covered the three Reform-linked polls

After two weeks of writing about the regulation of political polling, I was expecting to turn to a different topic this time. But then events intruded, so it’s back to that topic.

The core of the Channel 4/Verbatim story about Reform this week is political finance. But political polling plays a prominent role too, as set out in Krishnan Guru-Murthy’s introduction to the story this week on Channel 4 News:

Earlier this year James Orr, the party’s head of policy, and Dan Jukes, a close aide to Nigel Farage, arranged for a US company to fund three key influential opinion polls. Those polls made headline news and were seen by Reform UK as hugely beneficial to them. But the public was never told that they were commissioned by Reform UK.

Those polls appeared on the front pages of The Times and The Daily Telegraph, and indeed some of that polling made it into this newsletter too.

One part of this story is still very much under dispute: whether these amounted to covert foreign funding for Reform which broke the law. Although both James Orr and Dan Jukes have left their roles in Reform, with Lee Anderson saying, “They won’t be coming back,” Reform have been defending themselves too:

While Reform UK denies all wrongdoing in relation to the Channel 4 hit job this evening, as a matter of course, the Party is launching an internal investigation regarding the circumstances surrounding the broadcast.

Having worked through all the evidence made public, I have my own views about what happened. You may well too. However, whoever is right - Channel 4/Verbatim or Reform - either way, how these three polls were presented to the public was very problematic. Even if Reform are cleared of all allegations, there would still be serious questions for the polling industry and the media.

Before I get into that, if you want to get up to speed with the story, I’d recommend watching the Thursday night Channel 4 News show as that not only includes the broadcast package with undercover reporters in full, it also I think has a better introduction and summary of the allegations that even Channel 4’s own email write-up.

Turning to those polls, they were conducted by JL Partners, widely covered in the media and paid for by foreign money. The political finance allegations are that this constituted illegal foreign donations to Reform, something the party denies. But those polls were also presented to the public as not being polls that involved Reform at all.

It is legitimate for a party to commission a poll and then publish it, seeking media coverage as a result. But when this has been done by others, both the party and the media outlets are up front that the poll has come from the party. The issue here is that the origin of the polls was obscured.

It is also wrong to assume that a properly conducted poll is dodgy just because an interested party funded it. (A point I’ve often made!) But likewise, knowing the true provenance and context of a poll is relevant to judging whether or not it has been properly conducted.

Let’s get into the specifics of the three polls.

The first mentioned in the report is one from JL Partners run by The Daily Telegraph on the evening of 16 January. The newspaper has removed the story from its website, but it is still available in the Internet Archive and is still given prominence in the newspaper’s subsequent story “How The Telegraph forced Starmer into local democracy U-turn”.

The original poll story itself said:

Labour majorities on 10 councils would be wiped out if cancelled elections were to go ahead, according to polling for The Telegraph … The poll by JL Partners, carried out for this newspaper, looked at 25 of the council areas where elections are set to be postponed.

Notice the use of “for” in both the sentences I’ve quoted. I took it at the time that this was simply a poll that the newspaper had commissioned from the polling firm. That would be a normal situation and a normal reading of the language. But it would also be wrong - whether you believe Channel 4/Verbatim’s version of events or you believe Reform’s. Neither version has the newspaper paying for the poll.

The data tables on the JL Partners site were described (and, at the time of writing, still are) as “Polling of local authorities in which the 2026 council elections are due to be delayed, as covered in the Telegraph”. Again, no alternative source for the poll is proffered.

In the undercover filming that was broadcast, James Johnson of JL Partners is presented as saying:

I think there is a little bit of, like, informal rule bending that we did with our last one. So like, you know, when we published those results, we should have disclosed who the client was.

He goes on to talk about how the poll was organised, comments which likely will be at the heart of the political finance inquiries. Yet both what is presented in the film, and also any likely defence on his part, are quite some way from the ordinary meaning of what was published: “polling for The Telegraph”.

This is perhaps a linguistic nuance to wrangle over in the use of “for” rather than “by”. Either way though, I think it’s fair to say that the provenance of the poll was not clearly presented by the newspaper, even based on what the newspaper knew at the time.

James Johnson goes on to make disobliging comments about how polls not paid for by the media are often presented as being commissioned by the media and without questions being asked about the provenance of polls. As this is a side-angle to the main story about political finance and Reform, the reporting and footage released so far at least do not go further into that or into what journalists and editors might have to say in defence - such as whether the issue is really their lack of probing or a pollster firm’s lack of transparency.

We do, though, also have the second poll from the report, again one which appeared in The Daily Telegraph. This one was published on the evening of 9 April and again has been removed from the newspaper’s website but is still available in an internet archive:

Two-thirds of voters want Starmer out
Nigel Farage seeking to exploit PM’s unpopularity in May’s council elections

This story simply talks about a “new poll, by JL Partners” without any further details of the poll’s origins. Again, the normal reading of this is that either it is a poll commissioned by the newspaper, or perhaps one done by the polling firm as a freebie to get some publicity for itself.

Likewise the data tables from the pollster read that it is “JL Partners polling shared with The Telegraph”.

But according to the investigation, after the poll appeared Reform’s James Orr secured overseas funding for the poll. Leaving aside once more the details of the funding and whether or not they amount to illegal foreign party funding, the situation seems to be that a pollster carried out a poll, it was published in a newspaper and then only afterwards was a payment from someone else agreed to cover its costs. That is an odd sequence of events.

Then there is the third poll which featured in the Channel 4 film, this one appearing on The Times website on the evening of Sunday 31 May. It also was published on the front page of the newspaper the next day. The online story is still up on The Times’s website:

Trade union members are now as likely to support Reform UK as Labour, according to “damning” polling that their own leaders say proves “the working class have abandoned” Sir Keir Starmer’s party.

(The headline is different from that shown in the Channel 4 film, although it is not that unusual for print and online headlines to vary.)

The story provides a bit of normal detail about the poll:

JL Partners polled a representative sample of 1,002 trade union members between May 14 and May 19.

But again it is presented as if it were a poll carried out either for the newspaper or for the polling firm’s own self-publicity purposes. The latter is also what the poll’s data tables suggest, calling the poll “JL Partners polling shared with the Times”.

Except that… digging into the temporal archives a little I see that there has been a small but significant change in wording on the JL Partners website.

It used to say on the downloads page for data tables that this was:

Polling trade union members on their current voting intention and affiliation with political parties for the Sunday Times

It now says:

Polling trade union members on their current voting intention and affiliation with political parties shared with Sunday Times

Note the change from “for” to “shared with”.

Yet according to the investigation, Reform’s James Orr talked about “that poll we got on Times front page last week” and he then tried to bring in money to pay for it, securing it from the same foreign individual. That is the same unusual sequence as with the second of the Telegraph polls.

(I should also add that I covered this poll in The Week in Polls, and ahead of that JL Partners had responded quickly and fully to my queries about the poll’s weighting.)

All this raises quite a few questions, particularly as the British Polling Council’s rules state that:

2.1. All data and research findings made on the basis of social or political polls conducted in the United Kingdom by member organisations that enter the public domain, must include reference to the following:
Client commissioning the survey;…

The BPC has, wisely, not gone with the flow of its usual policy of not publishing rulings, and has publicly said:

With respect to the Channel 4 News investigation of polling apparently commissioned by Reform and conducted by JL Partners, it is clear there have been breaches of British Polling Council Rule 2.1. According to the programme, this may have been deliberate. We are seeking further information from the BPC member concerned and expect them to make information available after further consideration.
The Market Research Society has a wider remit than the British Polling Council and we note that the matter is being referred to the Market Research Standards Board. We will await the outcome of that enquiry.

The Market Research Society have added:

MRS is aware of the Channel 4 reporting involving an MRS Company Partner. The matter is being referred to the Market Research Standards Board (MRSB) for review in line with its established complaint procedures

Both JL Partners and the two newspapers have responded to the story. The pollsters say:

Despite there being no concern about the accuracy of our work, the report did reveal weaknesses in our procedures. For that reason JL Partners is reviewing those procedures in order to eliminate possible grey areas and ambiguities and strengthen our safeguards for the future.

The media group which includes The Daily Telegraph meanwhile said:

We are reviewing our future relationship with this polling company, as we believe we have been misled. As a matter of precaution we have suspended the relevant articles while The British Polling Council conducts a review.

As for The Times:

[It] was not made aware of Reform’s involvement in commissioning and arranging payment of the poll.

The newspapers could have learnt from previous incidents

However, it is worth noting in closing that both newspapers have some form for running stories about polls that come with issues over the origins of the poll.

Late last year I covered in some detail how The Times ran a poll supporting the government’s line on an issue, but without mentioning that the poll had been paid for by the government, that it had been given to the paper by the government and that key data in the poll undermining the government’s position was omitted. Its eventual correction did not contain any direct apology or explanation for why it omitted to mention that it was a government poll.

Or in the previous calendar year there was the massive front page story in The Daily Telegraph featuring a poll. The pollster had to disown the way the polling was presented and everyone admitted they didn’t know where the money for the poll had come from. And then the same sort of questions once more came up a few days later with further polling data.

Of course, regular readers will know how I’ve often covered problems with how other media outlets cover polls too - including Channel 4 refusing to release poll data after their then CEO made questionable claims in a speech.

So my previous two editions of this newsletter, warning about how the current regulatory structure for political polling in the UK may not be robust enough to withstand future scandals and may need to evolve, are turning out to be rather prescient, and much more quickly than I expected. (What, I wonder for example, would the BPC have done if the pollster involved had been one of its members but not a member of the Market Research Society and so it couldn’t leave the heavy lifting of the regulatory response to the MRS?)

But I think it’s fair to say that one aspect of the story - not as important of course as the question of whether senior figures in a political party deliberately broke the law to secure foreign money - is that media outlets have a track record of not being rigorous enough in checking and publishing the full origins of polling data they run.

Tim Shipman in The Spectator puts it rather bluntly:

It is for the police and Electoral Commission to determine what happens to him [which may be a reference to James Johnson or Nigel Farage] and the Reform duo. But it is quite something for the Daily Telegraph to falsely claim they commissioned the poll they ran and then ditch JL Partners when they got caught with their trousers down.

In amongst everything else, let’s hope that this time around journalists and editors learn the lessons that weren’t learnt after previous, much lower-key, incidents.

How much wider does the story go?

Particularly as there is one other question left hanging: none of the reporting suggests these three polls were the only problematic polls.

The arrangements behind the polls are not presented as something specially created for this one (fake) rich family. How many other polls were published in similar circumstances, but involving real people rather than fake ones?

The polling aspect of this scandal may yet spread further.

Elsewhere from me…

If you spot a factual error…

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters. It is too early to expect to see any movements in this or the leader rating tables following the Thursday night Channel 4 broadcast. Note though that the Opinium poll carried out Wednesday to Friday has Reform scoring highest on the proportion of people who think it has had a bad week (31%, with the other parties ranging between 19% and 27%).

A small error in this table has been corrected since this edition first appeared. Thank you to Matthew for spotting it.

The table is also online here, along with explanations about what is included. It is updated regularly throughout the week as new polls are released.

Next are the seat projections from MRPs, also sorted by fieldwork dates. Because they are published infrequently, some of the entries below are quite old.

Finally, here is a summary of the latest leadership ratings, sorted by pollster name.

The table is also online here with additional details, and is updated regularly throughout the week as new data comes out.

For historical figures on both voting intention and leader approval stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

Which issues people say matter most, and 9 other polling insights you shouldn’t miss

The following ten findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

The latest Ipsos monthly tracking data is out showing us which issues people say are most important:

Market Research Society upholds complaint against pollster

Sun, 30 Aug 2026 17:01:42 GMT

Welcome to the 228th edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

This time, I’m looking again at the regulation of political polling - and the regulator I didn’t cover last time, the Market Research Society.

This is followed by summaries of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

Reform voters are most likely to think that the majority of people share their views when they don’t;
New polling of Labour Party members; and
What people think of public service broadcasters such as the BBC.

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

One other piece of news related to polling transparency: I’m quoted this week in a news piece about Reform’s polling operation, an obscure pollster and missing invoices. As I say in that piece, “Reform are doing themselves no favours with the lack of transparency over the campaign spending.”

And with that, on with the show.

The Market Research Society

Last week I wrote about how recent polling scandals in the US should warn us against complacency over how polling is (self-)regulated here in the UK.

Probably the part that works best is the basic transparency over polling questions and related detail. It’s not perfect. It is not uncommon for it to be a matter of weeks and a couple of emails from me before poll tables appear for a poll that has had coverage in the media. (I currently have one email query awaiting resolution and a second email query to send shortly.) But generally, we can expect a level of transparency that is not matched in the US - and to be fair to pollsters, my emails often get prompt replies even over weekends.

There was a handy reminder recently of the importance of such transparency with a test poll from YouGov showing the importance of question wording when asking about views of the government:

Is government approval at 19% or 33%? Wording matters, and hence knowing the wording matters for knowing what to make of polling results.

But to keep the piece to a reasonable length when writing about all this last week, I only talked about the British Polling Council (BPC). There is, however, another relevant regulator, the Market Research Society (MRS):

The Market Research Society (MRS) is the UK professional body for research, insight and analytics. We recognise 5,000 individual members and over 500 accredited Company Partners in over 50 countries who are committed to delivering outstanding insight. As the regulator, we promote the highest professional standards throughout the sector via the MRS Code of Conduct.

Its Code of Conduct is considerably more wide-ranging than the BPC’s own rules, reflecting the different nature of the two organisations.

While the BPC’s rules concentrate on disclosure of information about polls, the MRS goes more widely, including the potential catch-all of principle number 11 for its members:

Protect the reputation and integrity of the profession.

Many of the polling firms you are used to hearing about regarding political polling in this country appear in both the BPC and MRS member lists. YouGov and Survation, for example, appear in both. But there are some exceptions, with the pattern being that the more narrowly focused an outfit’s activities are, or the newer it is, the less likely it is also to be a member of the MRS. More in Common and Convergent, for example, are BPC members but not MRS members.

So the MRS’s standards provide additional, albeit incomplete, cover for political polling.

My one previous direct experience of it was not great. Many years back the MRS rejected a complaint from me that a pollster undermined the reputation of the profession by asking questions which included false information about a politician. The MRS’s argument, at least back then, was that such message testing was acceptable, and they didn’t think that safeguards such as subsequently telling the sample that they’d been told untrue things were necessary. While I have sympathy with the idea of letting people research the impact of untrue claims, I think it is reasonable to have safeguards alongside that. Whether the MRS’s views on this have changed in the interim is a topic I hope to return to.

But for the moment, its recent ruling about a pollster is worth considering:

MRS MARKET RESEARCH STANDARDS BOARD (MRSB) STATEMENT
MRS received a complaint about Savanta an MRS Company Partner. The complainant alleged that, in conducting the data collection activity complained about, Savanta did not ensure that participants were not led to a particular point of view. The complainant also alleged that, in conducting the data collection activity complained about, Savanta did not ensure that responses and/or data collected are capable of being interpreted in an unambiguous way.
Having conducted its investigation, MRSB found that Savanta did not take reasonable action when undertaking data collection to ensure that participants were not led to a particular point of view, in breach of rule 28 (d) of the MRS Code of Conduct. MRSB also found that Savanta did not take reasonable action when undertaking data collection to ensure that that [sic] responses and/or data collected are capable of being interpreted in an unambiguous way, in breach of rule 28 (e) of the MRS Code of Conduct.
In reaching this conclusion, MRSB notes that Savanta fully co-operated with the MRSB investigation.
MRSB concluded that it was appropriate that Savanta was formally advised of the breach in accordance with paragraph 43 c (i) of the MRS Company Partner Complaints Procedure.

Compared with the BPC’s remit and approach, this ruling shows how the MRS goes further: it polices a broader set of standards and, unlike the BPC, it publishes its rulings. It also shows that the MRS is, despite the failure of that previous complaint of mine, willing to get into broader questions about how a poll has been conducted. That is welcome.

But the ruling is also rather cryptic. Reading it, I know that I may have come across some polling results from this firm which have misled me due to the way they were conducted. But which poll and in what way? Something I think I know may be wrong, but I don’t know what.

I asked the MRS why they don’t publish more details and they explained to me that:

1. It would breach confidentiality between our members and their clients. We typically do not have regulatory agreements in place with the client organisations commissioning research (given these are not necessarily MRS members) and therefore we are not in a position to involve them publicly in a complaint.
2. Companies and individuals which sign up to the MRS Code of Conduct make a public and voluntary commitment to a set of professional standards. The MRS Company Partner Complaints Procedure is focused on allowing scrutiny of the research sector’s work and ensuring improvements are made where they need to be.

There is an internal logic to those reasons, though the result is that other people - like you or me - will be left in the dark. (I did also ask Savanta if they wished to say anything further about this ruling.)

This is not, for example, the approach which press regulators take. For all their other controversies, their rulings are published and corrections made. You may not see the correction, but if you do see a ruling or a correction you then do know what was wrong about what you may have previously read. You are not left in the dark in the way that MRS rulings leave you.

It’s also not the approach taken elsewhere, such as in this recent ruling against a pollster from Research Association New Zealand (RANZ). (Thank you to my reader Murray for highlighting it to me.)

I think it’s reasonable to conclude that this situation in the UK is an unsatisfactory one - and one that pollsters should want to see changed for the long-term protection of their reputation.

Elsewhere from me…

If you spot a factual error…

Voting intention and leadership ratings

As More in Common’s Luke Tryl has explained, that pollster this week started prompting for Restore in its voting intentions questions, as did BMG. Where pollsters report figures for them, I include Restore (and also Your Party) in the full version of PollBase. I don’t include them, currently, in the summary table below because of the low level of support found for them (repeatedly in low single digits in the polls that have asked about them in the last couple of months or so). It was the same with the Greens and Reform, who only started appearing in the summary table when their support started being consistently higher than that. If another party starts consistently scoring higher support, then I’ll add them in too.

Here are the latest voting intention polls from different pollsters, continuing the recent pattern of Labour being generally, though not universally, in the lead over Reform:

The table is also online here, along with explanations about what is included. It is updated regularly throughout the week as new polls are released.

Next are the seat projections from MRPs, also sorted by fieldwork dates. Because they are published infrequently, some of the entries below are quite old. But note that there is new data this time and also how we’re starting to see Labour regularly ahead of Reform on seat numbers:

Finally, here is a summary of the latest leadership ratings, sorted by pollster name.

The table is also online here with additional details and is updated regularly throughout the week as new data comes out.

For historical figures on both voting intention and leader approval stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

Reform voters are the most likely to wrongly think their view is the most common one, and 9 other polling insights you shouldn’t miss

The following ten findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

A new YouGov poll “examines how many Britons think their own views are similar to those of the wider public, both in general and specific cases, whether they are right, and how groups differ in their perceptions.” Here is what it finds:

A warning from the US for UK political polling?

Sun, 23 Aug 2026 17:01:52 GMT

Welcome to the 227th edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

This time, I’m taking a look at recent US polling horror stories, and what we can learn from them for political polling in the UK.

This is followed by summaries of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

How the favourability ratings of political parties compare;
New data on whose support Labour has lost and to whom; and
Whether people think there is two-tier policing in the UK.

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

And with that, on with the show.

A warning from the US for UK political polling?

I have no idea why I was offered this photo when I searched a stock photo website for “warning”. Perhaps it is a meta-warning about not trusting search results. Photo by fr0ggy5 on Unsplash.

Three different US pollsters have been in the news in the last few days for all the wrong reasons.

First, there were Median Strategies, which published a trio of polls before being rumbled as a fake. Their claim (perhaps genuine, perhaps an after-the-event excuse) is that the release of fake polls was “a short-term social experiment examining how purported polling information could enter and spread through the political information ecosystem without independent verification.” The Guardian has an interview with the person responsible.

The take-up for their fake polls was relatively limited, particularly compared with the stories exposing the fakery, as one US polling expert pointed out:

That so often isn’t the case, but let’s all take heart from the truth sometimes being more widespread than the myth.

Second, The Public Sentiment Institute published their own post-mortem on a polling miss in which they admitted to changing the voting intention answers of some respondents so that the results of the poll more closely matched what they thought (or wanted?) the result to be. The campaign which benefited from these changes was also paying the pollster, something that hadn’t been disclosed. More details have been written up by the American polling aggregator site FiftyPlusOne (which is run by some of the former Five Thirty Eight crew, though not Nate Silver). However, even their very good write-up leaves me puzzled about the thought process behind the pollster voluntarily publishing such a self-condemning document.

Third, and very much the least of the trio but still less than ideal, Patriot Polling has been suspended from that polling aggregator for refusing to answer questions about how their polling is done.

One reasonable reaction looking at this from the UK perspective is to be thankful for, and perhaps even take a little pride in, the fact that the UK polling industry is in much better shape. This trio of stories is not that unusual in the history of US polling. Here in the UK, however, the industry has a much more reputable track record, helped by the culture of transparency embodied in the British Polling Council’s rules. What we take for granted in knowing about each poll is not by any means automatically the case with US polls.

There is one pollster currently regularly covered by the media who is not a BPC member, Lord Ashcroft. But in practice he follows the BPC’s rules too and his polls are as transparent as those of BPC members. Similarly, the title “most controversial pollster” has passed among a few firms over the years, but in my experience whoever has held it has been has always been pretty good at answering my questions and following the BPC rules. The only exception, many years back, was BPIX, though as I understand it that was more a question of a legitimate shoestring operation run by academics falling short rather than anything worse. Not ideal by any means, but still a whole lot better than the absence of transparency that far more regularly happens in the US.

But there is a second possible reaction too, which is to take these US stories as a warning against complacency.

I wrote a detailed piece last year about ways in which the regulation of political polling in this country could, and should, be improved:

One point I made in that piece is that although the transparency standards are good and widely followed, compliance is often rather lax. That is, the required publication of data tables only happens after someone (often myself) contacts a pollster to query where the tables are.

On average, I reckon it’s about once every six weeks that I send a message like that to a pollster. To be fair, the responses are good - data tables do appear, usually promptly - and it’s not uncommon for there to be a half-decent reason for the previous non-appearance of the tables (e.g. a client having released the poll at a time the pollster wasn’t expecting). Even so, finding that tables haven’t been released when they should have is not a rare occurrence; it’s a regular thing. The news cycle on a poll regularly comes and goes before the data is published.

As I wrote last year:

It is all a bit sloppy. The two-day deadline is not being followed with the zealousness with which many other sectors of the economy follow their own regulatory deadlines.

This is not a major problem, but it is perhaps a warning sign.

On re-reading my piece from last year, one additional thought also occurs - which is that the British Polling Council in my experience is pretty good at dealing with complaints raised with it, and indeed often resolves things quickly and informally first.

But what it doesn’t do is publish its rulings. That’s not a major problem, particularly as the core of what it deals with is complaints about polling data not being published, and so if the outcome is that the data has been published then something has happened publicly following a complaint.

It does though mean we don’t have the benefits of a public record of how pollsters are doing. Who are the ones that the BPC most often rules against, for example? Given the importance of polling, that seems a reasonable thing for people to know. And what’s the pattern in the complaints, and what lessons might be drawn from that? Again, not a major issue, as from what I know, the BPC does self-reflect. Again, though, I think it is fair to point out how much less robust this is than in other sectors, where the idea that a complaints body doesn’t publish its rulings would be anathema.

And to point the finger at myself for a moment, the steps I take before deciding whether or not to add a new pollster to my tables in this newsletter, to PollBase and so on, are pretty rudimentary compared with what FiftyPlusOne does. So far, I think that’s justifiable given the relative risks we each face. In the future though…?

So as with the BPC, it may be a system that works, but it’s also a more relaxed and less robust system than we get in many other areas. Which is why we shouldn’t simply react with complacency to US stories such as the recent ones. Yes, those don’t happen here. But is the system really robust enough to guard against what might happen in the future?

Worrying about that is, I think, a better reaction to the news from the US than being smug about the comparison between the US and the UK.

Podcast: Jeffrey Archer’s ‘First Among Equals’

The latest episode of my podcast Political Fictions is out, this time taking a look at Jeffrey Archer’s tale of four rival politicians with their eyes on 10 Downing Street:

Cory and Mark are joined by Shane Whaley of the Spybrary podcast this week. Shane has picked for us to watch First Among Equals, a prime time ITV political drama from the 1980s. As its episodes total up to 494 minutes, this time we’re dipping into the first episode only. Take a listen to see if we - and you - want to watch the rest of the series. The series is based on the book by Jeffrey Archer, once a highly praised and successful novelist whose books were festooned with quotes praising his literary abilities. With the advantage of hindsight, however, we ask if Jeffrey Archer’s work is best compared to Anthony Trollope or a McDonald’s Filet-o-fish?
Watch the first episode of First Among Equals, and indeed the rest of the series, on YouTube here and second-hand copies of a DVD version are available from the usual places. The original novel is also widely available.
Cory’s email newsletter is Paperback Rioter and Mark has a family of email newsletters.
Our theme tune is “Monkeys Spinning Monkeys” by Kevin MacLeod (incompetech.com) and licensed under the Creative Commons: By Attribution 4.0 License (creativecommons.org/licenses/by/4.0/).

You can also listen to our episode on Apple Podcasts, Spotify, the web, or YouTube.

If you spot a factual error…

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters. Lots of small fluctuations but no clear trends for any party in the last week.

The table is also online here, along with explanations about what is included. It is updated regularly throughout the week as new polls are released.

Next are the seat projections from MRPs, also sorted by fieldwork dates. Because they are published infrequently, most of these are somewhat dated.

Finally, here is a summary of the latest leadership ratings, sorted by pollster name. Although some of the latest polls apparently show notable movements in the net scores, they are often in contradictory directions. Conclusion? Again, it’s all noise without clear trends in the last week.

The table is also online here with additional details and is updated regularly throughout the week as new data comes out.

For historical figures on both voting intention and leader approval stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

How party approval ratings compare, and 9 other polling insights you shouldn’t miss

The following ten findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

Each week I cover leader net approval ratings, but we do occasionally get party approval ratings too and new figures have recently come out from YouGov and Ipsos. What do they tell us?

Are the polls overestimating support for Reform?

Sun, 16 Aug 2026 17:01:58 GMT

Welcome to the 226th edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

This time, with the Clacton Parliamentary by-election result now in, it’s a good time to look at how the polls perform when it comes to Reform’s vote share.

This is followed by summaries of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

The coalition preferences for the supporters of different parties;
What people think of the Burnham government’s priorities; and
How attitudes towards climate change have shifted through the recent heatwaves.

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

Before all that, a couple of updates on last week’s piece about research into whether polls systematically underestimate support for parties on the right. First, there is a companion piece of research about measuring polling error, which makes the case for using Mean Absolute Error (MAE). Very attentive readers will know I’ve tended to prefer the Root Mean Square Error (RMSE) which places greater weight on larger errors. The piece also points out that “research using this approach has found that polling errors have remained relatively stable over time. The performance of polls in the last decade broadly aligns with what would expect based on polling records from the 1960s onwards.”

Second, I played down the significance in the UK context of the average error found by the research. One of the co-authors has, however, highlighted its significance in the US context.

In addition, you may have noticed a typo in last week’s edition. In case it wasn’t clear, when I wrote of a “disproportionate interest politicos in the US take in the details of US politics (myself included)” I of course meant that first ‘US’ to be ‘UK’. Sorry for any bafflement caused by that and I’ve corrected the online version.

And with that, on with the show.

Are the polls overestimating support for Reform?

Photo by Ryan Grice on Unsplash

Last week I talked about evidence showing that overall polls tend to slightly underestimate support for parties of the right. The Clacton by-election result prompts a return to that question in a different guise: the record of by-election opinion polls in attempting to get Reform’s support correct.

There was one proper poll for the Clacton by-election, from Survation. There was also a super-small sample follow-up from the same firm.

The full poll put the Reform vote for Nigel Farage on 73%, the super-small sample put it at 79% and the combined sample across both gave 73% again. Reality, however, gave Farage 63%, a full ten points lower than the full poll.

In fairness, what Survation got right was the big picture: Farage a long way ahead and Count Binface by far the most popular of the others, on course for the best-ever result for a joke candidate.

But still, ten points is ten points.

Of course, some, most or all of that could be explained by the poll being carried out well ahead of polling day. It may have been spot on at the time the poll was carried out, with people simply changing their minds a bit by the time they voted. Or it may have been out. We’ll never know for sure.

In that sense, the Survation poll was rather like the Greater Manchester Mayor polls which I described as right but lucky in an earlier edition of TWIP. There are understandable reasons for polling well ahead of polling day in such contests, and the pollsters have got the overall picture right. So yes, the pollsters have been right. But they’ve also been lucky that there hasn’t been a big late swing of the sort that turns early fieldwork from a handy convenience into the source of a big polling miss. The gap between polling and voting also means that the differences between the vote shares in the polls and the actual results can’t be clearly ascribed to polling error, as they may instead have been due to people changing their minds. If the polls have been wrong, that’s been obscured from us.

There is though also another issue, which is how Reform’s vote share in by-election polls has compared with the actual result in the four Parliamentary by-elections since Nigel Farage returned to lead the party:

Runcorn & Helsby - 1 poll (Lord Ashcroft) - Reform vote share 1 point too high;
Gorton & Denton - 3 polls (Find Out Now, Omnisis and Opinium) - Reform vote share 7 points too high in one, dead on in another and 1 point too low in the third;
Makerfield - 5 polls (Survation twice, More in Common / UCL Policy Lab, Convergent and Opinium, ignoring the half-reported unidentified additional poll mentioned by The i) - one 6 points too high, two 5 points too high, one 4 points too high and one 2 points too high; and
Clacton - 1 poll (Survation) - 10 points too high.

Note: in calculating the errors in the Reform vote share I have compared the published vote share figures from pollsters, which were rounded off to the nearest whole number, and the by-election vote share, rounded off similarly. If unrounded numbers for each poll were used against the unrounded result, it is possible that the occasional figure would be 1 point different. But the rounded figures are the published ones and so the reasonable ones to use in this context. Had the Reform candidate polled 11 fewer votes in Makerfield, all of those errors would have been 1 point higher (as his vote share would have rounded down to 34% rather than up to 35%).

It’s only four by-elections so the evidence base is small and fragile. But we have eight polls putting the Reform vote share higher than it turned out, one poll being dead on and one poll putting the Reform vote share too low.

That is starting to be suggestive of a more general issue with by-election polls and the Reform vote share.

It also fits with what happened in the 2024 general election. Looking at polls whose fieldwork was within a week of polling day, i.e. started on or after 27 June, there were 25 polls of which 20 were on the high side for Reform, three dead on and only two underestimating Reform. If you restrict this just to the polls carried out in the three days before polling day, it becomes 8 too high, one dead on and one underestimation.

As some of these polls were published to one decimal point and some rounded off to the nearest whole number, I’ve taken whichever form the pollsters chose and compared that with the equivalent format for the Reform vote share.

Again, the limitations of this evidence are important to emphasise, especially the fact that only one of those 25 polls (from Matt Goodwin’s PeoplePolling) was more than 3 points out on the Reform vote share - which is a pretty darn good performance by the polls.

But it does all mean we have a growing pattern in Parliamentary elections - both general and by-elections - of Reform’s support when the votes are tallied being less than it was in the polls in advance of polling day. This pattern needs to be treated with caution and caveats, and it’s not a certain prediction for the future either.

It is though a pattern that’s there and one worth paying attention to.

Elsewhere from me…

If you spot a factual error…

Borrowing an idea from Stuart Ritchie and others, if you spot a factual error in an edition of this newsletter, I will give you a free 12-month subscription to the paid-for version (or an extra 12 months on your existing one). Factual errors qualify but grammatical errors or disagreements over interpretation do not, and my decision is final. Messages about both of those are always welcome, though.

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters, and more polls than not in the last week put Labour ahead of Reform. Whether that will last or whether this is a short-term new-PM bubble, we’ll have to see of course.

The table is also online here, along with explanations about what is included. It is updated regularly throughout the week as new polls are released.

Next are the seat projections from MRPs, also sorted by fieldwork dates. Because they are published infrequently, most of these are somewhat dated.

Finally, here is a summary of the latest leadership ratings, sorted by pollster name. We’ve now got a well-established pattern of Andy Burnham ahead, and with Nigel Farage and Zack Polanski forming the bottom two, and neither of their ratings (yet?) recovering in any meaningful way.

The table is also online here with additional details and is updated regularly throughout the week as new data comes out.

For historical figures on both voting intention and leader approval stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

How coalition preferences vary by party, and 9 other polling insights you shouldn’t miss

The following ten findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

Interesting new polling is out from JL Partners about how coalition preferences vary between supporters of different parties.

Do polls underestimate support for right-wing parties?

Sun, 09 Aug 2026 17:01:59 GMT

Welcome to the 225th edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

This time, I’m taking a look at new evidence about the accuracy of opinion polls, focusing on the long-running debate over whether polls systematically exaggerate support for parties of the left.

This is followed by summaries of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

Which party leaders are getting noticed on social media?
A big fall in Rupert Lowe’s ratings with the public; and
How many, or rather how few, people have been following the Jason Arday story (and if you have been then, like me, you may be puzzled by quite how small the fundraising total was on his original marathon fundraising page).1

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

Before all that, an update to last week’s edition, in which I wrote:

My eyebrows were raised this week by the Daily Express’s headline, “Andy Burnham dealt huge blow as new bombshell poll reveals who people truly want as PM”. Yet as the story went on to reveal, his ratings in this poll are better than those of any other party leader and better than the initial ratings of Starmer, Johnson and Sunak. (His ratings also beat Liz Truss, although those came later in her premiership.)
Some bombshell.

The paper has now replied to me, standing by its story, arguing, firstly, that an improvement in Kemi Badenoch’s ratings justifies the headline and, secondly, that “headlines should not be read in isolation”.

On the former, it does not bring my eyebrows back down as an improvement in her ratings while Burnham newly entered the ratings in first place is hardly a “huge blow” to Burnham. And who is meant to be revealed as the person “people truly want as PM”? If it’s the person who tops the poll… well, that’s Burnham again.

On the latter, while it’s true that headlines may simplify a story that follows (and mine can do so too, of course), that’s not a justification for having a headline that says the opposite of the story that follows. Indeed, the code from the relevant press regulator (IPSO) warns against “headlines not supported by the text.” Such a response therefore doesn’t give reassurance about how seriously the newspaper takes the code.

So let’s see what IPSO make of my follow-up complaint to them.

One other preliminary piece of news. Due to the disproportionate interest politicos in the UK take in the details of US politics (myself included),2 I suspect the big polling miss in the Michigan Democratic primary will have been noticed by a fair few readers. So this explanation from is worth quoting:

Michigan’s primary polls missed by 12 points on Tuesday night. That is, in fact, better than the typical primary poll. The pre-election average had Abdul El-Sayed up about 13, but he barely skated by with just a 1-point win. When Elliott looked at Senate and governor primaries going back to 2014 earlier this year, he found the average poll was off by 13 points on the margin between the top two candidates—and about 20% of the time the candidate trailing in the polls won outright. Evergreen reminder: Polling primaries is hard!

Thankfully there isn’t a strand of political polling in the UK that has as patchy a record as primary polling has in the US. Polls for local government, mayoral and party leadership contests, for example, all have their own challenges, as I’ve covered in previous editions. But in all of them the record of pollsters is rather better. US primaries very much stand out as being the exception in the polling universe.

And with that, on with the show.

Do polls underestimate support for right-wing parties?

Long-time readers will be familiar with my basic approach to political opinion polls:

The evidence is they are pretty accurate (US primaries aside, see above);
Many apparent problems with them are due to how they are reported (hello Daily Express, see above);
The answer to other problems is almost always to have more polls and pay more attention to them; and
The genuine occasional polling disasters are not enough to outweigh all the benefits they bring.

But aside from the occasional dramatic polling miss, there is also a longer-term pattern that causes pollsters concern. In the UK, Martin Boon has been particularly associated with highlighting the problem - which is that over time and across countries, voting intention polls tend to overestimate the left and underestimate the right.

A new working paper on this issue has just been released by Stuart Perrett, Will Jennings, Jonathan Mellon, Christopher Prosser and Christopher Wlezien.

As they summarise it, with my emphasis added:

Pollsters have underestimated support for notable right-wing electoral outcomes, including Donald Trump’s vote share in 2016, 2020, and 2024, the UK Conservative Party in 2015 and 2024, and Leave in the 2016 EU referendum.
Do these outcomes reflect isolated polling failures, or a persistent partisan asymmetry in polling error across democracies?
We provide the first comparative analysis of partisan polling error. Drawing on over 38,500 election-poll-party observations across 32 countries and 372 elections between 1945 and 2025, we use a Bayesian hierarchical modeling framework to estimate latent election-party polling error.
We find that, on average, support for parties on the left is overestimated, while support for parties on the right is underestimated. This pattern is evident across countries, party classifications, and vote intention polling dating to the 1940s.

In their dataset, polls the month before an election on average overstated the left by about 1.4 points. The right tends to be understated, by a slightly smaller amount, while for liberal parties there is almost no persistent bias.

That’s an important finding, though it also prompts the question of how big or small 1.4 points is. In one sense, it is certainly a big number because:

Across elections the average directional bias in the right-left lead is equivalent to approximately 45% of average absolute polling error.

Something that explains nearly half the average error in such voting intention polls is not small.

However, if the polls are close enough for a 1.4-point error to matter, they are close enough anyway to be cautious about who is ahead. Moreover, the 1.4-point average has a lot of variation behind it.

So this is an important finding for helping identify ways that pollsters can improve their performance. Yet it is also a small factor to bear in mind when looking at polls in a very close contest (because there are so many other ways polls can be out too and you can’t rely on a poll to the nearest percentage point or two anyway). So my four basic rules stay intact.

You can read the research in full here.

A murder mystery set in Parliament, written by a pioneering MP

The latest edition of my podcast Political Fictions features a murder mystery set in Parliament, written by the first female Minister of Education, Ellen Wilkinson:

Mark and Cory are joined by Elinor Taylor for our first book! It is Ellen Wilkinson’s The Division Bell Mystery, a political thriller disguised as a golden age detective story.
It is written by one of the country’s first female MPs and is full of convincing and vivid political detail about what it is like being a bag carrier for an important minister.
The plot’s setting is accurate right down to some key architectural detail on which it all hinges.
But is it also a good murder mystery? Find out as they discuss ridiculous murder plots, why politicians should beware of the weekend, and what this novel tells us about class, sex and empire.
You can find The Division Bell Mystery on Goodreads here.
Cory’s email newsletter is Paperback Rioter and Mark has a family of email newsletters.
Subscribe to our newsletter for some additional bonus content and to get alerted whenever a new episode comes out.
Our theme tune is “Monkeys Spinning Monkeys” by Kevin MacLeod (incompetech.com) and licensed under the Creative Commons: By Attribution 4.0 License (creativecommons.org/licenses/by/4.0/).

You can also listen to the episode on Apple Podcasts, Spotify, the web, or YouTube.

Elsewhere from me…

If you spot a factual error…

Borrowing an idea from Stuart Ritchie and others, if you spot a factual error in an edition of this newsletter, I will give you a free 12-month subscription to the paid-for version (or an extra 12 months on your existing one). Factual errors qualify but grammatical errors or disagreements over interpretation do not, and my decision is final. Messages about both of those are always welcome, though.

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters, confirming that Reform’s lead in the polls has given way to the party generally being in second place.

The table is also online here, along with explanations about what is included. It is updated regularly throughout the week as new polls are released.

Next are the seat projections from MRPs, also sorted by fieldwork dates. Because they are published infrequently, most of these are somewhat dated, but we do now have the first MRP of Andy Burnham’s premiership. I have also trimmed the list to remove any MRPs from before the start of this year.

Finally, here is a summary of the latest leadership ratings, sorted by pollster name.

The new BMG figures included in the table are Farage’s worst since he took over as Reform leader in 2024, as are those from Ipsos. The More in Common figures are his joint second-worst result (with the worst coming in the More in Common poll a few days earlier). The Opinium figures are his third-worst, with the other two having come in the last month. In other words, the Reform leader’s ratings have or had fallen to their worst level, though there may be some signs now of a small bounce back.

The table is also online here with additional details and is updated regularly throughout the week as new data comes out.

For historical figures on both voting intention and leader approval stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Count Binface does not feature in any of that, so here is his latest data from Ipsos:

Related to this, by 32%-25%, people have a negative rather than a positive opinion of joke or satirical candidates in elections, according to YouGov. Conservative and Reform 2024 voters are the most negative.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

Which party leaders are getting noticed on social media?, and 9 other polling insights you shouldn’t miss

The following ten findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

Andy Burnham’s use of social media has been generating plenty of commentary, but are the public noticing? Here’s what new data from More in Common shows:

Pollsters were right, but also lucky, in Greater Manchester Mayor contest

Sun, 02 Aug 2026 17:01:28 GMT

Welcome to the 224th edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

This time, it’s another of my ‘how did the pollsters do?’ round-ups as we now have the reality of the Greater Manchester Mayor result against which we can judge how the pollsters performed. With a bit of luck too, I’ll manage to send it at the right time, unlike last week’s mistaken late-night Saturday send. Sorry for anyone who was suddenly confused as to what day it was!

This is followed by summaries of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

Attitudes towards paying for social care;
Some better polling for Nigel Farage compared with Count Binface; and
How belief in opinion polls stacks up against belief in God.

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

My eyebrows were raised this week by the Daily Express’s headline, “Andy Burnham dealt huge blow as new bombshell poll reveals who people truly want as PM”. Yet as the story went on to reveal, his ratings in this poll are better than those of any other party leader and better than the initial ratings of Starmer, Johnson and Sunak. (His ratings also beat Liz Truss, although those came later in her premiership.)

Some bombshell.

(I have raised the headline with the newspaper, though only shortly before sending this newsletter. So I’ll update you next week on any response.)

Thank you to More in Common, Public First and Trajectory Partnership for answering my various queries this week. (The latter confirmed the point I highlighted last week about one of their crosstabs, and hence the difference between some of my figures and Wikipedia’s.)

Congratulations and good luck also to PJ Stock of the PollCheck website, who has joined Survation.

And with that, on with the show.

Greater Manchester’s mayoral polls: right verdict, risky timing

Photo by Dayne Topkin on Unsplash

Let’s cut straight to the data, with five polls from four pollsters published in the race to succeed Andy Burnham as Mayor of Greater Manchester:

An online version of this table, which is easier to zoom in on and resize, is here.

In many ways, this was a good performance by the pollsters. Four of the five polls got the top two correct, and the one poll that had this wrong was superseded by a later poll from the same pollster that got it right. All had Labour easily beating Reform in a second round. If you followed the poll headlines, you got the right overall story.

The second round vote shares were all also reassuringly close to reality, though YouGov’s error (four points on each, an eight-point difference in the lead) would have looked rather worse had the contest been closer.

On first round vote shares, the picture was much less impressive, but there’s a strong alibi for the pollsters - which is that voters may have changed their views between the polls being conducted and the votes being cast. Aside from the very early Focaldata poll, there’s a plausible story running through the other polls of a becalmed Reform campaign going nowhere, the Greens fading and Labour on the rise. It’s more of a stretch to extend those trends to envelop the Focaldata poll too, leaving it instead looking more of an outlier. But it was far enough ahead of polling day, and of the other polls, that we don’t really know.

The Labour first round score, 47%, is six points higher than even the highest Labour score in any poll, so there’s room for a bit of doubt about the accuracy of the polls. Overall, though, it may well be that the public’s mood shifted during the campaign and the polls were reasonably accurate for when they were taken. That would also fit with the coverage of the campaign: the Greens talking up their early chances, but their momentum seeming to run out.

So, once again, the polls told us pretty much the right story and there’s nothing much for the pollsters to worry about?

Yes, but.

It’s become the norm for pollsters to generally carry out their polls for such contests well in advance of polling day, with Find Out Now’s second poll being the one exception this time.

It’s important to acknowledge and welcome the spread of polling to Mayor and council contests in recent years. But it has come with a habit of polling well ahead of polling day. Which is a risky move because voters can change their minds late in an election. Even though many postal votes are cast well in advance, late swing is still a thing.

Post-mortems into polling misses are littered with the lessons of polls being carried out too far in advance of an election. It was a prime cause of previous headline-grabbing polling misses at the Presidential level (US) and the general election level (UK), following which the ‘final prediction’ polls moved to be closer to polling day.

So far, the welcome spread of Mayor and council polling has not resulted in a similar embarrassment. Yet it would be a hard argument to sustain that late swing is never and can never be a thing in such contests.

There certainly are benefits in terms of convenience and publicity in polling earlier. (The earlier you poll, the more time there is to get coverage of your results.) But, as I pointed out in both my 2024 and 2025 round-ups of mayoral poll performances, this is a risky approach too.

Pollsters have - so far - been lucky that they haven’t been caught out in a big way. That luck may last. But then again…

Also of interest…

Aside from how the pollsters did, these two other data points caught my eye.

First, the changes since the local elections:

Second, how the preferences flowed between the first and second rounds:

Summer reading for polling enthusiasts

Polling UnPacked is available from Bookshop.org (independent bookshops), Waterstones, Amazon (including Kindle version) or direct from the publisher.

Elsewhere from me…

If you spot a factual error…

Borrowing an idea from Stuart Ritchie and others, if you spot a factual error in an edition of this newsletter, I will give you a free 12-month subscription to the paid-for version (or an extra 12 months on your existing one). Factual errors qualify; grammatical errors or disagreements over interpretation do not and my decision is final. Messages about both of those are always welcome, though.

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters. Last week I wrote that “all in all, Reform’s poll lead has not gone yet, but it is getting very close to going.” This week’s figures allow us to say that, yes, it has now gone. The general picture - for the moment at least - is of Labour tying with Reform or being ahead, and the movement in the polls has continued beyond the initial Burnham polls from each pollster.

It is also worth adding up the Conservative and Labour scores in the table below. They don’t exceed 50% - which makes the recent fashion for talking about a return to traditional two-party politics, fuelled by Labour’s rise and Reform’s slippage, some way off the mark.

The table is also online here, along with explanations about what is included. It is updated regularly throughout the week as new polls are released.

Next are the seat projections from MRPs, also sorted by fieldwork dates. Because they are published infrequently, most of these are somewhat dated.

Finally, here is a summary of the latest leadership ratings, sorted by pollster name. There is now enough data to say with confidence that Burnham has the best net ratings. In addition, neither Farage’s nor Polanski’s ratings have substantively recovered from the hit they took after controversies earlier this year. Whether they have bottomed out or are still on a longer downward trend, though, is less clear; the data is a bit noisy.

The table is also online here with additional details and is updated regularly throughout the week as new data comes out.

For historical figures on both voting intention and leader approval stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

Where would you find £20 billion?, and 9 other polling insights you shouldn’t miss

The following 10 findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

I rather liked this recent JL Partners question which gave people the chance to pick a mix of tax rises and spending cuts to fill a hypothetical £20 billion hole in government finances:

Have we really had an unusually large number of PMs recently?

Sat, 25 Jul 2026 19:29:10 GMT

Welcome to the 223rd edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

This time, I am going for something short and polling adjacent: a graph showing how our turnover in Prime Ministers has varied over time.

This is followed by summaries of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

More on Net Zero after last week’s edition: the views of MPs and the problems that Kemi Badenoch may be storing up for the Conservatives;
Dan Neidle dissects a poll about wealth taxes; and
One of the worst polling question wordings I’ve seen in a very long time.

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

You will see that the voting intention table below now includes figures from Trajectory Partnership, a member of the British Polling Council (BPC). A description of their methodology is here.

As sometimes happens, my table and PollBase dataset differ slightly from the figures in Wikipedia. In this case, for the May Trajectory poll (the one before the latest), the numbers in Wikipedia come up several points lower in total than expected due, I think, to an error in the Trajectory data table which gives the unweighted total for ‘all likely to vote’ as 1,170 when it should be 1,075. I have used the latter figure.

And with that, on with the show.

Have we had an unusually large number of Prime Ministers recently?

Picture by Ben Dance / No 10 Downing Street and used under the CC BY-NC-ND 4.0 licence.

It’s become a commonplace comment that we have got through a lot of Prime Ministers recently.

So much so that people as varied as Lord Ashcroft, CNN and the Institute for Government are just a few of those who have recently done pieces on the question of whether the country is now ungovernable.

But is the number of recent Prime Ministers really that unusual? Let’s look at the data:

The blip in the line near the end - a brief drop from 6 to 5 - is because David Cameron left office in mid-July while Starmer left office later in July. Therefore, there’s a brief period in July 2026 when the tally ticks down due to Cameron dropping out of the data before Burnham’s replacement of Starmer ticks it back up again.

Three things strike me about the graph. First, yes, the graph does pick up in recent times.

But, second, the current bump up is not particularly notable in the long historical perspective. (This is another excuse for me to highlight how wild the 1920s were and how little attention they get in current political discourse even though they were a previous period of economic stress, international worries and a traditional two-party system under huge strain.)

Moreover, third, if you want to explain the recent rise in the graph, it’s as important to look at how low the graph was from the mid-1980s for around three decades. If you want to explain why churn is now higher than it was then, part of the story has to be about why it was so low in that three-decade run.

Simply ascribing the current higher churn to recent changes (e.g. rise of social media) can never be the full picture as such factors don’t explain why previously the figures were so low. And certainly beware any pundit who speculates on why we’ve had a higher churn of PMs recently while treating that previous lower churn run as if it was the norm. It wasn’t; it was an outlier in its own way too.

Summer reading for polling enthusiasts

Polling UnPacked is available from Bookshop.org (independent bookshops), Waterstones, Amazon (including Kindle version) or direct from the publisher.

Elsewhere from me…

If you spot a factual error…

Borrowing an idea from Stuart Ritchie and others, if you spot a factual error in an edition of this newsletter, I will give you a free 12-month subscription to the paid-for version (or an extra 12 months on your existing one). Factual errors qualify; grammatical errors or disagreements over interpretation do not and my decision is final. Messages about both of those are always welcome, though.

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters and we now have two further polls without Reform in the lead: Deltapoll have a Labour/Reform tie, while newer pollster Trajectory have Labour two points ahead, with these adding to the three-way tie in Lord Ashcroft’s poll. (Trajectory also had Labour ahead back in March while the other recent Survation poll, slightly older than the one in the table, had the parties tied.)

All in all, Reform’s poll lead has not gone yet, but it is getting very close to going.

The table is also online here, along with explanations about what is included. It is updated regularly throughout the week as new polls are released.

Next are the seat projections from MRPs, also sorted by fieldwork dates. Because they are published infrequently, most of these are somewhat dated. This time, there is a new More in Common MRP.

Finally, here is a summary of the latest leadership ratings, sorted by pollster name. Andy Burnham has jumped into the best net figures though we don’t yet have data from his premiership from most of the pollsters in the table.

The table is also online here with additional details and is updated regularly throughout the week as new data comes out.

And to give those numbers a little more context:

For historical figures on both voting intention and leader approval stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

Is Kemi Badenoch setting the Conservatives up for failure?, and 9 other polling insights you shouldn’t miss

The following 10 findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

YouGov’s Patrick English has been looking at the potential electoral consequences of Kemi Badenoch’s anti-Net Zero and anti-ECHR stances. He

Polling on climate change and Net Zero

Sun, 19 Jul 2026 17:00:48 GMT

Welcome to the 222nd edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

This time, does polling support those who are pushing Andy Burnham to water down the government’s targets for action on climate change?

This is followed by a summary of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

New polls are out for two big city Mayor races;
Polling on a possible Conservative-Reform seat deal; and
How attitudes to radical change have themselves changed.

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

And with that, on with the show.

What the public thinks about climate change and Net Zero

Photo by Nicholas Doherty on Unsplash

“Andy Burnham urged to stick to net zero targets if he becomes PM” - such headlines, in this case from The Guardian, have been common in the last few days. There’s a genuine policy and political tussle going on over whether to change policy or not.

What’s curious about much of this coverage is an implication that taking action on climate change is a political gamble, and that politicians seeking popularity should think of diluting, or dropping, their support for such action. Of course, there’s rather more to the policy topic than what the polls say, but the usual political framing of the issue is often as if Net Zero is a controversial, vote-losing policy.

The evidence from the polls, however, suggests otherwise.

As puts it:

Weird thing is most testing will show you that one of the most vote losing positions Labour could adopt is explicit anti-Net Zero signalling.
This is because climate [change] still disproportionately matters to Lab/Green voters while Lab/Reform voters don’t care either way, so you drive voters to the Greens while not bringing back (or over) any from Reform or the Tories.
In this sense, the Labour coalition is pretty together on it. It’s the Conservative vote that’s more split between what’s left of the Goldsmith environmental types and new wave Net Zero skeptics

One of his examples of the evidence for this is this graph:

There was also new data from YouGov this week, showing the continuing public support for taking action over climate change, albeit with a question in a generalised form:

Note how Reform 2024 voters are a massive outlier here; chasing after them takes you away from everyone else.

Or as Politico reports:

Recent record-breaking hot weather in the U.K. has made a chunk of voters more worried about climate change and impatient for help from the government, according to polling conducted for Politico by London-based public opinion polling company Public First …
The consistent finding as the country heats up, said Amy Norman, director of energy, climate and environment at Public First, is that it’s less a “panic about climate” and “more a frustration that Britain is not able to cope particularly well with hot or disruptive weather.”
The political opportunity for Burnham, she said, is about showing “practical competence” and delivering ways to protect people and public services — which Brits believe are not equipped to deal with the heat.

More generally, there’s also this from More in Common:

65 per cent of the public say they are very (25 per cent) or somewhat (41 per cent) concerned about climate change. Just 16 per cent are unworried.
Even among current Reform UK voters, who show the lowest concern, a greater proportion are worried (46 per cent) than unworried (32 per cent).

There has been a reduction in support for Net Zero policies, but as polling by Ipsos found earlier this year:

The share of the UK public who say the country needs to reduce carbon emissions to net zero sooner than 2050 has nearly halved since 2021, according to a major new study.
29% of the public now say the UK should achieve net zero before the government’s 2050 target – down from 54% in 2021, when this question was last asked.
The proportion who feel the UK either doesn’t need to reach net zero by 2050 or shouldn’t have a net zero target at all has risen from 9% to 26% over the same period.
But despite this declining sense of urgency, a significant majority (64%) still believe the government’s target for net zero should be at least 2050, if not earlier.

Likewise, YouGov’s latest Net Zero tracker shows support for trying to hit the target running at 63%-29%, with 2024 Labour voters even more in favour: 80%-16%.

Science and economics of course matter more than polling in deciding what is right to do on climate change. But it is notable that when media coverage talks of pressure on politicians to change policy, it often leaves out how popular the polls tell us it is to be committed to action on climate change.

PS You may be wondering ‘but what about that polling by Merlin Strategy for Looking for Growth? Didn’t that show rather different attitudes towards Net Zero?’ It’s true that poll found 2024 Labour voters backing cutting bills over Net Zero by 84%-16%, but look at the question wording:

Which of the following do you think a new Labour Prime Minister should prioritise?
Cutting the cost of living and energy bills, or
Making sure the UK is on track to hit Net Zero

Note both how (a) you can wish to prioritise one of those options while also wanting the other to be done, and (b) the first option offers up benefits to the respondent in the wording while the second doesn’t (e.g. it doesn’t mention tackling rising insurance due to climate change), which risks tipping answers towards the first.

Summer reading for polling enthusiasts

Polling UnPacked is available from Bookshop.org (independent bookshops), Waterstones, Amazon (including Kindle version) or direct from the publisher.

Elsewhere from me…

If you spot a factual error…

Borrowing an idea from Stuart Ritchie and others, if you spot a factual error in an edition of this newsletter, I will give you a free 12-month subscription to the paid-for version (or an extra 12 months on your existing one). Factual errors qualify; grammatical errors or disagreements over interpretation do not and my decision is final. Messages about both of those are always welcome, though.

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters. The overall picture is of a pick-up in support for Labour ahead of the change of Prime Minister, with Reform continuing to drop. Previous wobbles in those trends have been just that and not (yet?) changed them. The one trend that has likely changed is a bottoming-out of the fall in Green Party support.

The table is also online here, along with explanations about what is included. It is updated regularly throughout the week as new polls are released.

Next are the seat projections from MRPs, also sorted by fieldwork dates. Because they are published infrequently, most of these are somewhat dated. This time there is a new MRP from Stonehaven.

Finally, here is a summary of the latest leadership ratings, sorted by pollster name. Once there has been a formal change of Prime Minister, I will swap in the new person for Starmer.

After last week’s indecision, the extra data this week means we can say that Nigel Farage’s ratings are continuing to fall. However, as that has been a long-term trend, it is not clear whether the most recent events have had any particular impact so far.

The table is also online here with additional details and is updated regularly throughout the week as new data comes out.

For historical figures on both voting intention and leader approval stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

What does the public think of Andy Burnham’s Manchester plans?, and 9 other polling insights you shouldn’t miss

The following 10 findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

Soon-to-be Prime Minister Andy Burnham is promising a No.10 North in Manchester, relocating part of the current London-based 10 Downing Street operation. What does the public think of this? Ipsos have been polling just that,

Count Binface outpolls Nigel Farage, sort of

Sun, 12 Jul 2026 17:01:28 GMT

Welcome to the 221st edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

This time, let’s see if the polls can help us make sense of what the political significance is of an anthropomorphised bin that passes as an alien and changed its name after a US legal threat.1

This is followed by a summary of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

Who do people view as sleazy?
Has the data started changing over concerns about immigration?
How Rachel Reeves scores only one point worse with Labour members than Ed Miliband

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

One preliminary little bit of news to highlight to you: there’s a new polling aggregator in town, from Calum Weir. It includes some tempting features to look at accumulated crosstab information from multiple polls.

In addition, the latest quarterly update to my polling database, Pollbase, is now out. As well as containing data from the latest quarter that has just gone by it also includes:

Backfilled Stonehaven MRP data for this Parliament, with thanks to Luke Betham for the missing data.
Additional Gallup data from the 1951 Parliament from the Bodleian Library’s set of Gallup/BIPO reports in the Conservative Central Office archives. Thank you to Marcus Collins for sharing his copies of this data with me.

Grab the new and old data here.

And with that, on with the show.

Nigel Farage, Count Binface and the opinion polls

Binface vs Farage

Ipsos have given us the evidence-based bar chart we all wanted:

This polling result builds on Binface’s previous decent polling figures compared with political leaders. Of course, some important caveats apply beyond this only being the one poll. One is that it is a poll of Britain, not Clacton. Another is that it does not follow the usual Ipsos methodology for voting intention polls, including that this poll only surveys those up to the age of 75, while the usual Ipsos voting intention polls do not have an upper age limit on their sample. Nor is there the usual turnout modelling, squeezing of don’t knows etc. Ipsos also appear not to have got into past vote recall for previous Count Binface supporters.2

But we do have the crosstabs, which show a big variation between those who say they voted Remain and those who voted Leave in 2016. Among Remainers, Binface’s lead extends to 46%-10% while among Leavers he slips behind (18%-46%).

It is perhaps notable that under half, 46%, of Leave voters say they prefer to see Nigel Farage win, with 27% picking neither and 9% picking don’t know. Leave voters are also split in their support for Nigel Farage when asked by Opinium whether or not they find his defence of his actions convincing - they split 39%-31% against him, though those figures are more favourable than the 73%-13% split among Remainers.

What people think of Farage’s actions

Ipsos also finds that:

16% think Nigel Farage is right to resign as an MP and stand for re-election to Parliament in the resulting by-election. 16% think he should have continued as an MP. 54% think he should resign and leave Parliament.
74% state the Standards Commissioner should be investigating whether Nigel Farage broke rules by failing to declare a £5 million gift received from a Reform UK donor before his election.
73% of the public believe the investigation should continue even if Farage wins the upcoming by-election, compared to 15% who think it should not continue in these circumstances.
32% think other main political parties should not stand candidates in the Clacton by-election, whereas 31% believe they should stand.

However, on that last point, YouGov paint a different picture. Asking the different question of whether “the other major political parties are right or wrong to boycott the Clacton by‑election triggered by Nigel Farage?”, YouGov has people picking right over wrong by 37%-25%.

A 12-point lead worded one way (YouGov) or a mere 1-point lead worded the other way (Ipsos) suggests that public opinion is open to persuasion still on this point.

Opinium meanwhile highlight a cynical realism among people with leads for both the view that Farage was right to call the by-election and that he’s doing it for his own ends:

While more people think Farage was right to call a by-election (44%) than wrong (31%), the public remains sceptical about his motivations. Two in five (42%) believe the move is primarily a political strategy to strengthen his position, while just 19% think it is mainly about allowing voters to judge his actions directly. The polling also shows that majorities find the allegations surrounding financial support unacceptable, are unconvinced by Farage’s defence, and believe Clacton should not re-elect him.

The Conservative swing voters?

The Conservatives finished second in Clacton in 2024 and have previously won the seat. Moreover, the crosstabs in polls such as those above typically show that Reform voters give the most pro-Farage answers, that Labour, Lib Dems and Greens cluster with the most anti-Farage answers and that Conservatives are somewhere between the two.

That suggests a key role for current and former Conservatives in deciding Nigel Farage’s fate. More in Common, however, have a more nuanced segmentation, splitting people into three groups:

The first group is those who already dislike Farage, and they’re hardening in their opposition. Overall, the public has grown more critical since the story first broke: the share saying the £5 million gift reflects badly on Farage has risen from 48 per cent in May to 57 per cent in July…
The second is his base - those who voted for him in 2024, when Reform’s national support was half of what it is today. On this scandal at least, it looks like these voters haven’t moved. Three-quarters of 2024 Reform voters (74 per cent) say it hasn’t hurt their view of him…
Nationally, the group that might worry Farage most beyond his immediate by-election are - the movable middle - voters who might consider Reform but could be put off by scandals like this. This includes many 2024 Conservatives who Reform will need to run up the score with, in a General Election - but half of these voters now say that the Farage scandal has damaged their view of him.

However, at least some YouGov data is less promising for Nigel Farage, even among his base. This pollster has it that, by 60%-12%, people think Reform UK’s leader has not been honest about his financial affairs. That 12% rises to 40%, though only 40%, among 2024 Reform voters. 39% of them say ‘don’t know’, which suggests a greater potential brittleness for Farage’s reputation among his core supporters than in the More in Common data.

Is this all being overhyped?

For those who follow politics closely, this story has grabbed their attention. But what about the wider public?

Opinium asked: “It has recently been reported that Nigel Farage is being investigated by Parliament’s Standards Commissioner over whether he should have declared gifts and other support he received before becoming an MP, including a £5 million gift from businessman Christopher Harborne. Nigel Farage has denied wrongdoing, saying the gifts and support were unrelated to his political activity with Reform UK. How much, if anything, have you heard about this?”

Overall, 78% say they have heard of the story, with 33% picking “heard a lot” and 45% “heard a little”. Only 15% say they have not heard of the story until getting this question. In similar results, YouGov have 17% saying they are not aware of the story about donations made to Nigel Farage.

Returning to the Opinium data, those who say they haven’t heard of it are more likely to be women than men (18% versus 11%, though it’s common for men to be less likely to admit ignorance in such polls) and are more likely to be younger than older people (25% for 18-34s versus 8% for the 65+).

Also of note is how little partisan split there is in answers by voting intention, so this isn’t a case of Reform supporters trying to downplay the issue by saying they’ve heard less of the story than they really have.

Beware the vox pops

Even setting aside the usual risks with vox pops, the Clacton ones seem particularly unhelpful as a guide to what people are actually thinking:

One of the other places Jim and June appeared was in The Sun and another was The Independent.

Good luck, pollsters

Summer reading for polling enthusiasts

Polling UnPacked is available from Bookshop.org (independent bookshops), Waterstones, Amazon (including Kindle version) or direct from the publisher.

Elsewhere from me…

If you spot a factual error…

Borrowing an idea from Stuart Ritchie and others, if you spot a factual error in an edition of this newsletter, I will give you a free 12-month subscription to the paid-for version (or an extra 12 months on your existing one). Factual errors qualify; grammatical errors or disagreements over interpretation do not and my decision is final. Messages about both of those are always welcome, though.

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters. There has likely been a small continued fall in Reform’s vote share, with more polls than not in the last couple of weeks showing its support down on the previous comparable poll from that pollster. But the latest news about Nigel Farage does not (yet?) look to have altered the picture.

Labour’s bump in the polls may perhaps have come to an end, as its run of consistent such comparable increases in late June has now become a more mixed picture. With the Conservatives also generally up, we’ve seen a small revival of traditional two-party politics, but just a small one, and it’s only the occasional poll that has them both on more than 20%.

The table is also online here, along with explanations about what is included. It is updated regularly throughout the week as new polls are released.

Next are the seat projections from MRPs, also sorted by fieldwork dates. Because they are published infrequently, most of these are somewhat dated. This time there is a new MRP from Electoral Calculus and Find Out Now for PLMR.

Finally, here is a summary of the latest leadership ratings, sorted by pollster name. Once there has been a formal change of Labour leader, I will swap in the new person for Starmer.

The very latest Opinium poll shows Nigel Farage’s ratings taking a big hit, but the other poll with fieldwork from this month - from More in Common - doesn’t. More data needed.

The table is also online here with additional details and is updated regularly throughout the week as new data comes out.

For historical figures on both voting intention and leader approval stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

Which people and parties do people view as sleaziest, and 9 other polling insights you shouldn’t miss

The following 10 findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

There is new data from YouGov on how sleazy people think various leaders and parties are. It’s handily summarised in this graphic:

The George Gallup myth: how a lucky guess became a legend

Sun, 05 Jul 2026 17:01:45 GMT

Welcome to the 220th edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

This time, I return to some polling history, with the truth about pioneering pollster George Gallup’s famous polling success. The full story is rather different from the myth that he so successfully spread.

This is followed by a summary of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

Public attitudes towards how political parties are funded;
How Britain’s attitude towards the USA has changed; and
What heat waves do to our attitudes towards climate change.

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

And with that, on with the show.

George Gallup: self-publicising polling pioneer

One of my favourite pieces of research for my book about polling - Polling UnPacked: The History, Uses and Abuses of Political Opinion Polls - was examining the full story of George Gallup. As the readership of this newsletter has more than doubled since I last wrote about the Gallup myth, here again for your enjoyment is how reality differed from the hype, updated since last time.

If you know a little bit about the history of polling, you probably know something like the following. In 1936, a magazine - The Literary Digest - asked its readers to write in to say who they were going to back for President of the United States. The result was a predicted Republican landslide.

But George Gallup also ran the first modern scientific poll at an election. He said the Democrats would win in a landslide. Not only that, he predicted how wrong the magazine’s figures would be before they came out.

The result? Gallup was right about the election and right about the magazine’s error. Modern polling had shown its superiority and reader self-surveys died off as a way of calling elections. The Literary Digest was reduced to running a headline, “Is our face red!”.

In this story, Gallup is the scientific pioneer who got it right. Yet it’s also a story that is as much, if not more, public relations puffery from Gallup as it is truth.

Even sources that are as sceptical of PR and spin doctors as the Wikipedia editing community give you an uncritical version of Gallup’s self-aggrandisement. (I clearly don’t have enough Wikipedia editors among my readers as that’s still as true now as it was when I pointed it out three years ago!)

It’s the story I used to believe too, until I started researching and asking simple questions such as, ‘what vote shares did Gallup’s polls show?’ and ‘how did they compare with the result?’

Let’s start with how Gallup predicted how wrong The Literary Digest would be. Gallup did his own poll and, claiming to know how the magazine would err, then produced adjusted figures predicting its error. That sounds impressive. Yet as I wrote in that book, Polling UnPacked, Gallup was lucky in getting so close:

Gallup produced his figures six weeks before The Literary Digest survey had yet started, and the magazine’s gathering of responses for its poll lasted several weeks. Therefore, for Gallup to be so close required a large degree of luck, as the support for candidates can change over time.
Even had Gallup’s polling been perfect and his understanding of The Literary Digest’s flaws been spot on, there was no way he could be sure that reality would not change between his poll and the conclusion of that of the magazine. It was not so much a brilliant prediction as a lucky guess. Data made it an informed guess. But a lucky guess it still was, turned by luck and a flair for publicity into apparent brilliance.

Turning to the overall election result, Gallup did correctly predict a clear victory for FDR, something the magazine had failed to do.

But he predicted the Democratic candidate winning 54%, which was seven points off his actual winning vote share of 61%. A seven-point polling miss can also be seen as a pretty big miss (though in this respect Gallup set an early tradition for polls - missing by a lot doesn’t bother people that much if both the wrong figures and the real ones show a landslide win).

Nor was Gallup the only pollster with modern methods to be polling the election. In fact, he wasn’t even the one with the best results as I recount in my book:

He had two rivals who, like him, had developed polling expertise for business research and also applied it to politics: the Fortune survey, masterminded by Elmo Roper (whose commencement of polling for the magazine pre-dated Gallup by a few months), and Archibald Crossley, who masterminded surveys for Hearst Publications. Ironically, Crossley was formerly of The Literary Digest. He had then gone on to pioneer random sampling by telephone to generate radio audience data.
These two rivals of Gallup both also got the 1936 election result right. But Gallup, with his flair for publicity, scooped the attention despite Fortune being closer on the vote shares.

Roper was ahead of Gallup both in adopting modern polling for public consumption and in the accuracy of his 1936 election prediction.

The reality then was that Gallup’s prediction of how wrong The Literary Digest would be was lucky, his own figures were not particularly close - and others got closer than he did. But he was the best self-publicist.

There’s a darker side to the Gallup myth too, which I dig into in more detail in the book. As academic research has shown, he presented his polls as the voice of America, but his methods deliberately downplayed the views of women and of Black Americans. The roots of that may have been in relative turnout at elections (with physical intimidation and legal obstacles reducing turnout among Black Americans, for example). But the down-weighting of their views went wider than that and featured in polls that were presented without caveat as representing the views of all Americans. What he called the voice of America was too much the voice of white, male America.

Of course, Gallup does deserve credit for his 1936 breakthrough. He correctly predicted the winner. His methodology was much better than The Literary Digest’s, and the 1936 US election did mark a turning point in how we try to measure public opinion ahead of an election. He provides a human interest encapsulation of a genuine change for the better.

But in doing so, he’s also become a figure of myth, a myth that doesn’t withstand scrutiny of the evidence.

You can find out more in Polling UnPacked, which is available from Bookshop.org (independent bookshops), Waterstones, Amazon (including Kindle version) or direct from the publisher.

Elsewhere from me…

If you spot a factual error…

Borrowing an idea from Stuart Ritchie and others, if you spot a factual error in an edition of this newsletter, I will give you a free 12-month subscription to the paid-for version (or an extra 12 months on your existing one). Factual errors qualify; grammatical errors or disagreements over interpretation do not and my decision is final. Messages about both of those are always welcome, though.

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters. There has continued to be a small improvement in Labour’s poll position, with the party now regularly in second place, behind Reform by single digits. The Greens have also slipped, starting to appear in many polls behind or level with the Lib Dems rather than ahead of them.

Note: the Opinium vote share changes are from their previous poll using the same methodology. There was also an intervening one with a different methodology (prompting for Restore).

The table is also online here, along with explanations about what is included. It is updated regularly throughout the week as new polls are released.

Next are the seat projections from MRPs, also sorted by fieldwork dates. Because they are published infrequently, most of these are somewhat dated.

Finally, here is a summary of the latest leadership ratings, sorted by pollster name. Once there has been a formal change of Labour leader, I will swap in the new person for Starmer.

The table is also online here with additional details and is updated regularly through the week as new data comes out.

For historical figures on both voting intention and leader approval stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

Public attitudes to how political parties are funded, and 9 other polling insights you shouldn’t miss

The following 10 findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

The Electoral Commission has some great new research out on public attitudes towards the funding of political parties, which I particularly like because rather than just using the superficial answers from a quickly completed poll, it uses

Brexit regret runs deep, but support for rejoining the EU is soft

Sun, 28 Jun 2026 17:01:30 GMT

Welcome to the 219th edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; share it with your friends and colleagues:

Share The Week in Polls

This time, as it is a decade since the Brexit referendum, I’ve returned to the topic of Europe, the subject of one of my most popular posts back in 2024. That post …

How the polls did at the Makerfield by-election

Sun, 21 Jun 2026 17:01:10 GMT

Welcome to the 218th edition of The Week in Polls (TWIP), a polling newsletter that always welcomes new readers. So please don’t keep TWIP a secret; please share it with your friends and colleagues:

Share The Week in Polls

This time I am taking a look at how the polls performed in the Makerfield by-election.

This is followed by a summary of the latest national voting intention polls, seat projections from MRPs and the most recent party leader ratings.

Subscribers who pay for the full edition can also read ten insights from the latest polls and analyses, including:

A gender gap in support for Reform;
New research on whether opinion polls alter election results; and
What the public thinks of social media’s impact and regulation.

If you are not yet a paid subscriber, you can read all ten by starting a free trial now:

Get FREE 7-day trial

For more polling news as it comes in ahead of next week’s edition, you can also follow The Week in Polls on Bluesky.

My eyebrows were raised this week by the Daily Express’s coverage of its Brexit poll from Whitestone Insight, mentioning that in one question 45% said they would vote to rejoin the EU in a referendum, prefacing that with the word “only” but omitting to mention that in the same question it was 38% saying they would vote against (with the rest being don’t know, and so on). The story does make some fair points about what the poll shows about the possible softness of that 45%, but failing to mention the 7-point lead at all? Hmm.

(The full poll details are here. Yay for the British Polling Council’s transparency rules.)

In unrelated news, my co-author on a pamphlet about the Liberal Democrats, Jim Williams, has a cat-themed wordplay game out with Gibsons, Categories. The final version appears to ignore my playtesting feedback, which I can only assume is why it has turned out so well.

And with that, on with the show.

How did the polls measure up against reality in the Makerfield by-election?

The original version of this table, which is easier to zoom in on, is here.

The opinion polls published before the result came in all showed Andy Burnham ahead, with Reform in a clear second place, between three and 12 points behind him.

Andy Burnham did indeed win, with Reform in a clear second place, but his winning margin was 20 points. So what should we make of the performance of the polls?

Here is how they measure up, based on the root mean square error (RMSE) of party shares for each of them, using the poll closest to polling day in the case of Survation:

Convergent 3.0
Survation 3.3
Opinium 4.8
More in Common/UCL Policy Lab 5.0

To give those scores some context, the best pollster on this measure at the 2024 general election (Norstat) scored a little under two, and a score of three was bettered by only three out of 20 pollsters at the last general election.

So pretty good in many ways and kudos to the newcomer for coming out on top. As I wrote last week about the firm:

Although some may be tempted to be cynical about the poll giving Labour the largest lead also being from a firm founded by two former Labour staffers, it’s worth bearing in mind that how this poll performs will have a significant influence on the new firm’s reputation.

Reputation enhanced.

However, Andy Burnham’s win was much bigger than the polls were pointing to. Does that mean the polls were wrong?

Perhaps, but perhaps not as there are three main contenders for explaining the difference:

had questioned in advance the past vote weighting used by the polls. His theory was that it was faulty, depressing Labour’s share by mistake.
(Survation’s Damian Lyons Lowe also highlighted a similar point before polling day.)
This theory matches up beguilingly well with what happened, but some caution before concluding it is correct is required, at least in advance of most post-election number crunching. That is because the methodologies used by several of the pollsters had been used in previous by-elections without producing the same problem. The number of data points here is limited so this isn’t a slam dunk rebuttal either. Rather, it’s an open question.
Turnout was, hooray, unusually high in the by-election, actually increasing on the general election rather than, as almost always happens, falling. The last time a Parliamentary by-election saw raised turnout was way back in 1987. Turnout is hard for pollsters to model and not infrequently pops up as at least a partial cause of polling errors. Again, more number crunching is needed to properly assess this theory, though Opinium’s initial post-election analysis points to this being an issue.
None of the polls were carried out close to polling day. In general elections, pollsters have long learnt from previous polling misses the importance of carrying out their final marquee prediction poll as close to polling day as possible. Otherwise, a post-poll, pre-polling day change in people’s views can make even an accurate poll look foolish. Despite the rise of postal voting, this is still a risk, and it is an even bigger risk for by-elections which often see very large movements in party support just before polling day as those who have seen, or calculated, the internal data from parties regularly attest.
So it’s possible that polls weren’t wrong, but rather that some people changed their minds in the last few days. It is also plausible that there was some momentum in the campaign away from Reform and towards Andy Burnham given what we saw in candidate approval ratings in the polls and events such as the BBC’s Question Time.

Hopefully we’ll get further post-election analysis from the pollsters themselves which will help shed light on which combination of these three theories best explains what happened - or indeed if there is a fourth factor to add to the mix.

But perhaps the most important point about the performance of the polls is a different one from all the above.

We’ve had ten Parliamentary by-elections in the last Parliament and in this for which there was at least one opinion poll made public. In nine of the ten, the final poll got the top two parties right, and in the tenth the poll was only partially wrong, showing two parties tied for second when reality put them several points apart. (Although the record in getting the winner right is much less good, due in particular I think to third factor mentioned above.)

In other words, as a guide to who you may wish to vote for tactically, the final poll in a by-election has a very strong track record. With tactical voting, in a variety of forms, likely to be widespread in British politics for the foreseeable future, getting the top two right in a seat is, perhaps, even more important than getting vote shares or leads right. Getting the former right informs voters and helps them get any tactical voting decisions right.

That is worth emphasising given the popularity of comments such as this, from , which happened to fall into my inbox just when I was looking for a comment to keep for this edition:

Constituency polling is notoriously unreliable.

Constituency polling is not without its problems, and future editions of this newsletter will no doubt tear apart various constituency polls in the future.

But getting the basic picture right so often makes constituency polling in by-elections and high-quality MRPs at general elections, I’d suggest, a useful guide for voters wishing to figure out how to vote.

Peter Kellner has a more robust version of my last paragraph in his piece, drawing on the Hungarian Uprising of 1956.

As before, apologies to those interested in the two Parliamentary by-elections in Scotland. As there were no polls published for those contests, there isn’t much to say about them in a polling newsletter.

UPDATE: After finishing but before sending this newsletter, Luke Tryl from More in Common shared with me this re-analysis of their data which also suggests that past vote weighting may have been a factor.

The Keir Starmer we never got

Over on my newsletter about Parliament, A Lord’s Eye View, I’ve got a new piece out about a welcome announcement from the government which gives us a glimpse of what a successful Starmer premiership might have looked like:

If you spot a factual error…

Borrowing an idea from Stuart Ritchie and others, if you spot a factual error in an edition of this newsletter, I will give you a free 12-month subscription to the paid-for version (or an extra 12 months on your existing one). Factual errors qualify; grammatical errors or disagreements over interpretation do not and my decision is final. Messages about both of those are always welcome, though.

For people who love politics, on screen and off

If you follow politics and love political drama - on screen, on the page, or on stage - Political Fictions is the podcast for you. Each episode, I join Cory Hazlehurst to explore the portrayals of politics in fiction. Subscribe to our email list and we’ll let you know when each new episode drops:

Subscribe to Political Fictions

Voting intention and leadership ratings

Here are the latest voting intention polls from different pollsters. It was a less good week for Reform, so perhaps their decline in the polls has not come to an end after all? Also seeing some less good polls recently were the Greens, as several - though by no means all - of the recent polls have put them back down in fifth. Again, a trend to watch.

The table is also online here, along with explanations about what is included. It is updated regularly throughout the week with new polls.

Next are the seat projections from MRPs, also sorted by fieldwork dates. Because they are published infrequently, most of these are somewhat dated but there is some new Stonehaven data this time.

Finally, here is a summary of the latest leadership ratings, sorted by pollster name. After the previous surge in his leadership ratings, Zack Polanski’s net ratings have slipped back to regularly placing him in third or fourth place. More on that in TWIP #207.

The table is also online here with additional details and is updated regularly through the week as new data comes out.

For historical figures on both voting intention and leader approval stretching back to the 1930s, including Parliamentary by-election polls, see PollBase.

Catch up on the previous two editions

My privacy policy and other legal information are available here. Links to purchase books or other items are usually affiliate links which pay a commission for each sale.

Quotes from social media messages are occasionally lightly edited for punctuation and clarity.

What does the public think of social media bans?, and 9 other polling insights you shouldn’t miss

The following 10 findings from the most recent polls and analyses are only for subscribers who pay for the full edition, but you can sign up for a free trial to read them straight away.

With possible social media bans and other regulation in the news, we have a bundle of polling from YouGov on what people think about social media’s impact and about potential further regulation of it. The simple headline finding is that