Important bit of new polling from @deltapoll.bsky.social: ask people for the top issue facing the country and immigration is second, at 49%. But ask what’s most important for *them* and it drops to fourth, at just 21%. That gap might suggest a way out of Labour’s mess…

While it’s true that Labour has become unduly obsessed with the threat from Reform while it loses voters to its left, it’s also true that there are far more Lab-Reform marginals than Lab-Lib or Lab-Green ones. And in those seats, votes lost to Reform count double.

Painfully recognisable to anyone who has ever been involved in left-wing politics.

This from @owenwntr.bsky.social is excellent. There’s a limit to how much you can learn from precinct-level data, but this piece goes right up to that limit.

We were lucky enough to have our YouGov/The Economist poll in field when the US bombed Iranian nuclear facilities. You can see the partisan realignment in real time

Telling paragraph on culture at the ONS from the Devereux Review.

Right now on political prediction markets you can place bets on Newsom, AOC, Buttigieg, Shapiro, and Harris to be the Dem nominee in 28 and more than double your stake if it ends up being any of them. Good deal?

Polls often get criticized unfairly. But in #NYCmayor, they had an unambiguously bad night. Mamdani averaged 28.5% of 1st-place votes in the last 4 independent polls of the race. He currently has 43.5%.

People usually assume that when a popular politician adopts an unpopular policy, the politician will become less popular. But it’s just as likely that the policy becomes more popular instead.

A personal update: yesterday was my last day at Deltapoll. I’m going travelling for a couple of weeks before starting my new gig, where I will continue asking people for their opinions about politics and trying to figure out when they’re lying.

It was so obvious for so long Labour would form the next government that by the time they took office they should have had a clear, costed plan and a proven strategy for communicating it. They had neither.

Do the people suggesting that FIFA is going to pull the World Cup from the US because of its contempt for migrants’ rights remember where the last World Cup was held?

Lower than Truss.

Everybody is (rightly) talking about the grim numbers for Labour here, but it's also dismal for the Tories. Still toxic with supporters of all other parties AND they have lost more of their own base than any other party of opposition.

Keir Starmer's net favourability rating has dropped 12pts in a month to -46, his lowest level ever, including a 34pt drop among Labour voters All Britons: -46 net rating (down 12 from 13-14 Apr) By 2024 vote Labour: -5 (down 34) Lib Dem: -13 (down 12) Conservative: -76 (up 1) Reform: -94 (down 5)

This means Starmer (-34 net favourability with YouGov), Farage (-38), and Badenoch (-38) are all about as (un)popular as Prince Harry. Which of them will be the first to sink to Meghan's level? Will any of them catch Prince Andrew? This is the only way I can get myself to care about the royals.

This is remarkably similar to the Projected National Share from the locals (within 2 points for all parties). Maybe the end of the three-way tie?

I really like the second graph here, but I don’t agree that the first one is “awful”. It’s a lot more useful for answering questions like “Which party won in my seat?”

He's right. It is possible to underestimate voters' knowledge.

A brilliant parenthetical quote in a brilliant piece by @duncanrobinson.bsky.social. www.economist.com/britain/2025...

The problem with most machine-based random number generators is that they’re not TRULY random, so if you need genuine randomness it is sometimes necessary to link your code to an external random process like a physical noise source or the current rate of US tariffs on a given country.

The growth in support for Welsh independence is a real and interesting trend, but it's worth noting that YouGov consistently finds support to be ~10 points lower than R&W (who conducted this poll).

A negative example of a dynamic many centre-right & centre-left politicians seem to have forgotten - political leaders can change public opinion, they do not always have to follow the polls.

I hear this is a great gig!

I’m just a girl standing in front of a boy telling him it’s not true he has become more right-wing… it is me who has become more left-wing 🫠

In need of some weekend reading? Week in Public Services is out! @cassiarowland.bsky.social on purposeful activity in prisons @stuarthoddinott.bsky.social on NHS targets & GP partnerships model And me on address data in public service delivery & what's happened to disadvantaged pupils' funding

It will be really interesting to see how, if at all, these numbers change once the announcement filters through to voters. Public opinion on some issues is highly responsive to parties' positions. Is Net Zero one of those issues?

Not just because most of the British public, including 76% of Conservative voters and 52% of Reform(!) voters support net zero.

It's hard to think of a strategy the Conservatives could adopt that plays into the Lib Dems' hands more than abandoning Net Zero. But give Kemi a week or two - I'm sure she'll come up with one.

To celebrate St Patrick's Day, Deltapoll asked Brits to name their favourite famous Irish person. Who do you think was mentioned the most? Guesses below, I'll share the answer later on.

I wrote about why Reform is at a turning point in the polls. Whether it gains or loses a few percentage points of support has a massive effect on its seat prospects. Take a look!

NEW: The actions of Trump and Vance in recent weeks highlight something under-appreciated. The American right is now ideologically closer to countries like Russia, Turkey and in some senses China, than to the rest of the west (even the conservative west). My column: www.ft.com/content/3046...

A British election factoid I hadn't noticed until I was updating my lecture slides - in 2024 only 15% of constituencies were won by a majority of votes. That is less than half the previous low of 34% recorded in 2010.

Political science people: is there a standard metric for how concentrated/dispersed a party’s national vote is across constituencies? I imagine you could use the variance-to-mean ratio or some kind of Gini coefficient, but is there a convention?

Whenever I hear calls for cuts to the foreign aid budget.

In this poll Labour are 'losing' more votes to the Lib Dems than to Reform and the same number to the Greens. (I say losing but it is probably partially a tactical unwind at this point). The Reform lead is at least in part an enthusiasm gap very few 2024 Reform voters are undecided.

German public TV just published this poll of first-time voters. A massive lead for socialist Left party on 27%. Left + SPD + Greens on 54% among first-time-voters. Those voters were also slightly less likely to vote AfD than national average. So much for the “far-right youth”.

In modern history, what is the least popular that party leaders have ever been to their OWN voters? Second place was the GOP in 2009, where their net approval with Republican voters was only +6. Currently Dems are -8, shattering all previous records.

The 2029 UK general election will be decided by which party leader fails to do an episode of Off Menu.

This is a great piece on a topic that we pollsters would do well to think about more deeply. As usual, the solution starts with drawing DAGs!

Maybe not surprising, but certainly striking - more than a third of AfD supporters rate their trust mainstream TV news as 1/10. From a great thread of Deltapoll/TBI data on the AfD ahead of Sunday's federal election in Germany.

Great piece of data-driven journalism in @theguardian.com today. Without migration, Europe's population would fall off a cliff in the next 75 years. Any argument for tightening borders needs to contend with that reality.

The fact that there are still only 4 Conservative MPs here makes me sceptical that this site will ever be like old Westminster Twitter.

Rare polling question where men are LESS likely to claim to have read something.

Nice bit of number crunching by the Economist. www.economist.com/finance-and-...

Absolutely. And all economies should have above-average growth.

This tweet has more retweets than the number actual votes cast

Always nice to see important research validated by a second source.

We talk a lot about temperamental electoral coalitions, but I'm not sure I've seen a better example than Sinn Féin. It won in 2020 thanks to a motley collection of voters with views on immigration from right across the spectrum. Then, when immigration became salient in 2023, it all fell apart.

I (finally) wrote about why Sinn Féin lost the Irish election. Check it out! (Spoilers for the impatient: yes, it was immigration - but no, the party should not have pivoted to the right.)