Do the Conservatives really have a problem with support among young people?

Welcome to the 104th edition of The Week in Polls, which takes a turn towards brevity after last time’s behemoth edition digging into MRP results. This time, it’s a dip into four different recent poll findings.

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.)

As ever, if you have any feedback or questions prompted by what follows, or spotted other recent polling you’d like to see covered, just hit reply. I personally read every response.

---

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.

---

Four polling insights from the last week

Do the Conservatives really have a problem with support among young people?

You might think that the only thing to wonder in answering this question is how robust a word to insert into ‘of course they ______ do’.

But take a closer look at this graph from Robert Colvile:

Sure, there is a huge age difference that has opened up since 2010. So much so that we are probably not far off people starting to rewrite Pulzer’s famous 1960s quote about class to read, “age is the basis of British party politics; all else is embellishment and detail”.¹

Sure, the Conservatives are massively behind with the 18-24 year olds.

But look again. Their current standings are only slight down on 2019. It’s not a collapse in support among young people that has taken them from the heights of a big election win to the depths of a huge electoral hole.

The big collapse is at the other end of the age spectrum.

Which is one reason why I think Robert Peston made a mistake in not following his own advice when writing:

The most over used phrase in politics is “existential threat”, but the results of this Savanta poll for #Peston is surely that for the Tory Party: “The Labour Party has increased its utter dominance over the Conservatives among 18 - 25 year old's, with a record high 47-point lead”.

As the graph shows, that’s missing the point that apparently disastrously low levels of support are compatible with winning a general election, as 2019 showed.

There are lots of good reasons to be wanting to be popular with younger voters. However, as history shows, doing poorly among younger voters in the present is a weak predictor for doing poorly at general elections in the future.

One Mayor race, two very different polls

We’ve finally had two polls for the West Midlands Mayor contest, where Conservative Andy Street is trying to hold on. They paint very different pictures of the state of the contest as postal voters are starting to vote:

A 14 point for Labour with Redfield & Wilton but a 2 point lead for the Conservatives with Savanta.

What gives?

Not only do the two polls give very different pictures, the spread in lead is greater than we’re seeing for national voting intention polls (though the spread in the latter is less noticeable as it’s not a spread that varies who is in the lead).

The figures are not quite incompatible because, as Patrick Sturgis points out:

Given around a third of the sample is dropped (don't knows/won't vote) the implied confidence intervals on these vote shares will be wide and quite possibly overlapping (if calculated properly).

Even so, it’s likely one pollster will have some bragging rights, and the other some explaining to do, come results day.

There’s such a paucity of Mayor polling outside London that there isn’t much of a pollster formbook to consult either. Moreover, the Mayor polling from London is the other way round. While in the West Midlands, Redfield & Wilton has Labour doing much better than Savanta, in London their latest Mayor poll gives Labour a much smaller lead than Savanta.²

The likelihood to vote figures in both polls are a little surprising. As Paula Surridge points out, 2019 Conservatives are shown as less likely to vote in the Mayor contest than 2019 Labour supporters, which runs counter to the usual turnout patterns and the relative age profiles. But that’s a finding common to both polls and so does perhaps tell us something is going on with demotivated Conservatives.³ Even if it does, however, that doesn’t explain the difference between the pollsters.

Savanta suggests higher turnout than Redfield & Wilton.⁴ Which if that goes with a pattern of 2019 Conservatives facing more of a turnout challenge, then that might help explain their better findings for Andy Street’s party.

Savanta also has a higher proportion of people recalling that they voted Conservative in 2019. People increasingly recall their previous vote wrongly as time goes on, so past vote recall can’t simply be compared with the actual previous result. So while we can’t simply conclude which pollster is right, that does perhaps also help explain the divergence in their results.

Underneath it all, it looks as if though the big difference between the two pollsters is a pretty simple one: how many Conservatives versus Labour people each found in the first place, rather than what their weightings and turnout adjustments then did to them.

Which means we’ll have to wait for more polls or the result to have a clearer idea of who is right.

Redfield & Wilton also has a Tees Valley Combined Authority Mayor poll out, showing Conservative incumbent Ben Houchen in a dead heat with Labour.

UPDATE: Savanta has now corrected its figures after a problem with the past vote weighting, giving Labour a 3 point lead. That is still significantly less than Redfield & Wilton’s 14 point lead.

Record levels of tactical voting on the way?

The Economist runs analysis from the ‘gold standard’ British Election Study (BES)⁵ suggesting that, as Owen Winter puts it, “One reason the Conservatives might do worse than uniform swing implies: tactical voting. In BES data, I found that the number of voters considering more than one progressive party is at its highest in recent history”.

Measuring willingness to vote tactically via opinion polls is notoriously difficult, as I wrote about in Polling UnPacked. Slightly different question wording produces very different results.

But even if we should be cautious about levels, this BES data has the advantage of long-run consistency, showing up trends.

The Economist adds:

Constituencies that have held by-elections since 2019 show evidence of tactical voting. In by-election seats targeted by the Lib Dems, the Labour vote fell by an average of 7.5 percentage points; in seats where the Lib Dems were not going for victory, Labour’s vote rose by ten percentage points. This may partly be because Labour chose not to campaign hard in seats where the Lib Dems were stronger. But it is also likely to reflect voter behaviour.

One other thing to note in this graph: how low the Labour/Conservative considerers are. Which is a clue that the route back for the Conservatives is a tough one as there isn’t a big group of Labour/Conservative switchers to pitch for (and who, in most contests, would count double for the government - a Lab/Con switcher is both one on the Conservative pile and one off the Labour pile).

A strange poll question… or a portent of a big polling disaster to come?

Lord Ashcroft’s latest polling paints a familiar generally grim picture for the Conservatives as seen in other polls. In this case it includes the Conservatives now being behind Labour on defence and national security, a 42%-38% lead for scrapping the government’s Rwanda plan and Labour leads on both best Prime Minister trust on the economy.

But there was also this:

According to that question, 45% of people either are satisfied with the current government or prefer it to a Labour one, rising to 81% among 2019 Conservatives. That would point to a pretty healthy voting intention showing for the Conservatives, which isn’t what the polls are showing. (Lord Ashcroft’s poll doesn’t ask a traditional voting intention question, but recalculating his figures into a more normal set of results gives a large Labour lead, as shown in the table below.)

So either the wording of this question produces quirky results or, if the polls have a disaster on the scale of 1970, then this is going to be one of the clues that we’ll all be thinking we should have paid more attention to.

---

National voting intention polls

It is not obvious from the latest polls why there is some talk of a general election being called soon after the local elections:

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

MRP vs MRP: what happens when you compare two of them?

---

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.

---

Pollster finds only those over 70 more likely back Conservatives than Labour, 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.