The ConservativeHome members’ panel seems to appear to a number of commentators and professional pollsters as a discomfiting anomaly: a broken clock that is, somehow, right more than twice a day. Quite a bit more, actually.
As people who perform badly in it are quick to point out, our survey is not a scientifically-weighted poll. But it has a habit of lining up with them – not to mention with the eventual results.
During the last leadership contest, for example, the views of our panellists broadly tallied with those of the party members surveyed by YouGov – whilst getting a little closer to the actual outcome.
This time out, we were the amongst the first to detect a grassroots surge for Penny Mordaunt; she came second to Ben Wallace in our first special survey, and then topped the poll in our second one after he decided not to stand.
At first, it looked as if there was a gap between our findings and those of the pollsters after Opinium put Rishi Sunak in front. But then YouGov came out and, like us, found the Trade Minister routing all comers, topping the list of preferences and winning every head to head.
This is now wrong, as YouGov have results v similar to ConHome.
Political scientists critical of ConHome’s survey might need to consider whether they’ve happened upon a kind of ‘Ann Selzer’ formula, or whether we’re seeing the website able to shift perceptions of candidates. https://t.co/4Nc7dXYF6F
— Alan Wager (@DrAlanWager) July 13, 2022
Nor do the similarities end there. Like us, they find another relatively unknown factor, Kemi Badenoch, in second place – although she runs Mordaunt much closer with our panel than with YouGov’s – and Rishi Sunak in third place.
This is perhaps explained by a rightward or eurosceptic slant amongst the members who take our survey, which would also explain why we found Suella Braverman in fourth place whilst YouGov place her sixth.
Some of the head-to-head findings also match up pretty closely: we had Mordaunt vs Truss at 51/33, YouGov at 55/37; vs Sunak at 58/31, YouGov at 67/28; and vs Tugendhat at 61/23, YouGov 64/26.
Again, the two where we found candidates running the Trade Minister closer than YouGov were the right-wing candidates: we had vs Badenoch at 46/40 against YouGov’s 59/30, and vs Braverman at 50/37 against YouGov’s 63/25.
Did we simply detect a latent surge for Mordaunt before anyone else? Or did our initial survey results shape the early stages of the race? It will be difficult to prove it either way – and when it comes to the question of whether the survey is worth paying attention to, it doesn’t really matter.
But Mordaunt’s surprise performance isn’t the only story in town. There’s also Rishi Sunak’s poor showing with the members.
Of the candidates still standing, YouGov find the Chancellor losing to Badenoch, Mordaunt, Truss, and Tugendhat, and within the margin of error against Braverman; we have him losing to Badenoch, Braverman, Mordaunt, and Truss, but beating Tugendhat.
Again, the divergence could be explained if our panel skews rightwards – we also have Badenoch beating Sunak by a much bigger margin than YouGov, for example.
But the broad picture is still of a candidate who looks very badly prepared for the second round, and that will doubtless weigh on the minds of MPs as they try to work out who is best place to defeat whoever emerges as the candidate of the right.
Of course, it’s early days yet. There is still a week of MP voting to go before we see who’s whittled down to the final two, and the results could yet surprise everyone.
But whether we’re detecting the landslides or setting them off, the ConservativeHome Members’ Panel seems to be (infuriatingly, I’m sure) on the money so far.