In general terms, I would place greater trust in a weighted opinion poll than in a self-selecting survey.
In more specific ones, what do YouGov’s weighted polls of Conservative members tell us compared to this site’s self-selected surveys?
Both we and YouGov asked more or less the same question at more or less the same time during the last Conservative leadership contest – namely, which candidate party members would vote for.
Ballot papers were issued to Party members during early July 2019.
Our survey published just after the ballot opened on July 5 scored Boris Johnson on 67 per cent and Jeremy Hunt on 28 per cent.
A YouGov poll published a day later found Johnson on 74 per cent and Hunt on 26 per cent.
Our second survey on July 11 had Johnson on 72 per cent and Hunt on 28 per cent.
Our third and final survey on July 21 put Johnson on 73 per cent and Hunt on 27 per cent.
The actual ballot percentages were: Johnson 66 per cent, Hunt 34 per cent.
So the closest finding to the result was our survey of July 5, and my presumption is that many Party activists voted early.
Nonetheless, our proprietor’s words apply, as ever – that a poll is a snapshot, not a prediction.
Our subsequent surveys were as close to the YouGov poll as makes no difference – and you will see that both our survey and the YouGov poll underestimated Hunt’s score by between six and eight points.
Next, to a question that neither we nor YouGov can measure against a real event – namely, who will be the next Conservative leader.
We asked in our last survey: who should be the next leader of the Conservative Party after Boris Johnson?
YouGov has asked in a poll for Sky News: if Boris Johnson stood down, who would you like to replace him?
Those seem roughly comparable questions.
We published on December 27, and our top two were Liz Truss on 23 per cent and Rishi Sunak on 20 per cent.
YouGov published yesterday evening, and its top two are Rishi Sunak on 33 per cent and Liz Truss on 25 per cent.
Part of the reason for the difference is likely to be the “others”.
YouGov tried putting Jeremy Hunt (eight per cent), Michael Gove (six per cent), Sajid Javid (five per cent), Priti Patel (four per cent) and Dominic Raab (four per cent) into the mix.
For each of those, we had Hunt on eight per cent, Javid on five per cent, Gove on four per cent, and Raab and Patel on two per cent each.
We also put in eight other potential candidates: of these, Penny Mordaunt got nine per cent and Steve Baker eight per cent – marginally higher ratings than the Cabinet Ministers that both we and YouGov listed.
At any rate, Sunak and Truss are clearly the front runners, as far as both YouGov and ConHome are concerned…in an election that may not take place in this Parliament.
Next, our last survey’s Cabinet League Table found Boris Johnson in negative ratings for the third time, for the second time in successive months, and on his lowest ever score (-34).
YouGov asks the party members it polled whether they believe that he is doing well or badly as Conservative leader. Sixty one per cent say well, 38 per cent badly.
These findings point in different directions, and not even a shift in the Prime Minister’s favour, which our survey picked up in its later returns as it became clear that there would be no new restrictions post-Christmas, appears to explain the difference.
According to the YouGov poll in a question that we have not asked (at least yet), “Do you think Boris Johnson should stand down as Conservative leader?”, 59 per cent say he should remain, 34 per cent say he should stand down and seven per cent don’t know.
YouGov polled 1,005 Conservative members for Sky News between 30 December and 6 January. Here is the full write-up.
Our last survey went out on December 23, we got just under 800 replies, and we wrote the first result up on December 27.
As far as I know, YouGov is the only pollster to poll Tory members, although it doesn’t so regularly. There’s a history of our panel sometimes turning up results in the same ball park.