Which campaigns are doing best?

Welcome to the 113th edition of The Week in Polls (TWIP), which takes a look at whose campaign is performing best, according to the polls.

Then it’s a look at the latest voting intention polls followed by, for paid-for subscribers, 10 insights from the last week’s polling and analysis. (If you’re a free subscriber, sign up for a free trial here to see what you’re missing.)

First though, this week’s raised eyebrows come courtesy of Wes Streeting who first wrongly dismissed the accuracy of polls (see the truth here), and then anyway tweeted polling figures he likes.

---

Been forwarded this email by someone else? Sign up to get your own copy here.

Want to know more about political polling? Get my book Polling UnPacked: the history, uses and abuses of political opinion polling.

---

How are the different campaigns performing?

In the end, it’s only the votes that win seats that count.¹ You don’t get any prizes for being well liked by others.² Which is why it’s the voting intention polls that are summarised in this newsletter each week.

It is also because people’s answers to so many questions are coloured by partisanship. It’s often that people like a policy, person or state of affairs because of the party its associated with, and so questions that at face value are about something else end up being just other ways to discover overall levels of support for each party.

That’s why when it comes to questions about whether a party is doing well or badly in a general election, I prefer to look not at the raw data but at how a party’s score compares with its voting intention. If it’s at 40% voting intention and 30% say their campaign is good, that’s an underperformance of 10 points. By contrast if only 20% say another party’s campaign is good, but it’s running at 8% in the polls, then that’s a 12 point over-performance and, relatively speaking, a better result.

And so this graph:

It’s not a surprise that the two campaigns over-performing the most, the Reform and Lib Dem ones, are for parties who have risen in the polls. I suspect Labour will be pretty happy with a generally close to zero score (save for the Diane Abbott week), as a campaign that just safely ticks over with alarms or excursions is a win in their book.

As for the Conservatives, well… despite their low poll ratings, their campaign consistently rates even worse. That’s not a good sign.

What, though, about the party leaders? Frustratingly, many polls who ask leader questions don’t consistently ask the question and/or don’t ask it of that many leaders. Which is why, with apologies to the Greens who are omitted, I prefer the BMG data set for its breadth and weekly consistency :

Leader approval type questions produce widely varying results depending on the wording used. The trends though are consistent, so although other pollsters find significantly more positive or negative figures depending on their wording, the relative orders - such as Sunak doing very poorly relative to the others - and trends - such as Davey improving during the campaign - are fairly consistent.

Again, I suspect Labour will be happy with the Starmer picture. On some measures his standing has improved during the campaign. But most importantly for him, he started out leading a party that looked favourites to win and as long as he doesn’t fall in the eyes of the public then he’s not putting at risk that status.

Voting intention

But as I said, often these other questions are in large part synonyms for what party people are going to vote for.

So on voting intention, here’s one graph that captures how the government is doing:

A good campaign for MRP modellers?

In many ways, it has been a good election so far for those who produce MRP models. (If you’re new to this, my MRP explainer is here.) They are appearing in greater numbers than before and they are dominating news coverage, often being the headline or frontpage story.

Although amongst polling experts there has been quite a debate about the pros and cons of MRP overall, and some models in particular, it’s fair to say that generally the individual constituency results are being treated as a high quality source of information and given significant prominence in the media.

That veneration, and the difficulties of having just so much data to use, means those individual seat vote share projections are not getting much scrutiny.

(Some pollsters are expressing caveats about any of us paying too much attention to individual constituency figures. However, I do feel that risks being a bit like not merely having your cake and eating it but running a bakery while declaiming the sins of cake eating. If you don’t want us to look at your individual constituency figures, then don’t publish them, promote them, jazz them up with fancy graphics and entice us to look at them. Instead you could just give us seat totals for the country.)

There is, to my eyes, lots of eyebrow raising detail in the figures we’ve had from various models. Whether it is examples such as the Conservative vote due to rise in seats in Liverpool, by-elections winners predicted to come third or the massive variations in the vote share figures in places such as Queen's Park & Maida Vale or Sheffield Hallam, it is hard to see how we don’t end up with at least one MRP model looking rather underwhelming in some significant respects.

A topic to return to definitely when we have the results to compare with the models (and my own expectations).

---

National voting intention polls

Lots of polls! One consistent story: the Conservatives a long way behind, and heading down, not up. Much more variation over support for Reform and where that is relative to the Conservatives.

On that, Matt Goodwin’s PeoplePolling poll is very much the outlier compared with other polls carried out around the same time. Although he’s described his findings as ‘ahead of the curve’, it isn’t clear what methodological foundation there is for thinking that if his figures are different from other pollsters, those other pollster will change in the future to come into line with his, rather than for thinking that his figures are off. (Especially as his own graph of previous and current Reform ratings with different pollsters doesn’t show other pollsters coming into line with him but rather his figures becoming less similar to those of others.)

For more details and updates through the week, see my daily updated table here and for all the historic figures, including Parliamentary by-election polls, see PollBase.

---

Last week’s edition

Understanding tactical voters, with the help of a new polling technique.

---

My privacy policy and related legal information is available here. Links to purchase books online are usually affiliate links which pay a commission for each sale. Please note that if you are subscribed to other email lists of mine, unsubscribing from this list will not automatically remove you from the other lists. If you wish to be removed from all lists, simply hit reply and let me know.

---

What is cutting through?, and other polling news

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