I’ve blogged several times about how YouGov’s poll ratings for the Liberal Democrats are consistently lower than those from other pollsters (in part, it looks, because YouGov consistently gives the Conservatives higher support amongst female voters than other pollsters), but what about ICM, which consistently has some of the highest ratings for the Liberal Democrats?
Last week Mike Smithson looked at this question on Political Betting, highlighting the different wording used by the different polling companies in their voting intention questions.
However, I think Mike missed one of the most important differences. ICM (and only ICM) helps concentrate people’s minds on the local situation with its voting intention question:
The Conservatives, Labour, the Liberal Democrats and other parties would fight a new election in your area. If there were a general election tomorrow which party do you think you would vote for?
None of the other companies use wording which talks about your area, local or anything similar. But we know from other evidence that the more people’s minds are focused on the local choice, the more likely they are overall to pick the Liberal Democrats.
For example, this came out in the recent mega-YouGov poll of marginal seats across the UK, where the focus in those polling questions on the local contest produced a much better result for the Liberal Democrats than the national polling, without any local mentions, which YouGov was carrying out at the same time.
As to which is the most appropriate wording? Well, voters in a general election are presented with a constituency ballot paper that names their local candidates, so ICM’s gentle nudge towards focusing people’s minds on the local choice (albeit one with national ramifications) looks to me quite sensible. Of course, as we get nearer a general election polling day and voters’ minds become more focused on the choice of local candidates (rather than simply national political party preferences), you’d therefore expect other polling companies to see their results converge on ICM’s, which indeed is what has happened at recent general elections.
Follow @libdemvoice.org on Bluesky
Like us on Facebook
Subscribe to our feed
Sign-up for our daily email digest
What happens after a party wins a landslide? - The Week in Polls
Today's Bells on Sunday
Divine Comedy: I've Been to a Marvellous Party
Lochee Park - 1973 #dundeewestend
Urgent action needed on Welsh rivers
Mike Peters
John Waller
Simon R
Mick Taylor
4 Comments
Very plausible.
I wonder if YouGov’s practise of following up a brand index survey with political questions predisposes participants to think of themselves as passive consumers when it comes to the political choice. And therefore to choose not only parties which prefer passive citizens, but also the bigger brands.
From memory, ICM tend to be one of the polling firms who get closest to predicting the result.
Joe, that’s an interesting point about the BrandIndex surveys. Not all YouGov’s political polls are done that way, though (I’m a member of their panel). Has anyone looked into whether the ones that are attached to BrandIndex polls produce statistically different results than those that are not?
You highlight the different questions the pollsters ask – but might not the methods the pollsters use to select their sample population also be a factor. Or is there now universal agreement on sampling methodology?
Many years ago, I discovered that the polling organisations restricted their sampling to certain, specified constituencies – presumeably to ensure proper representation of Labour & Conservative voters. I assumme three party politics has changed all that – but, nevertheless, might not different sampling methods account for the different results?