UK pollsters should follow the firm that created the industry and move on from party vote shares
Arguably Gallup has the right approach for the future
The screen grab above is from the Election 2016 page of Gallup – the firm that created modern political polling in the 1930s. Its busy with lots of data, analysis and often excellent insights but one thing that you won’t find are voting intention polls.
After a lacklustre performance with its voting numbers at WH2012 the firm took the strategic decision to drop that aspect for WH2016 and focus on elements that can often give better pointers to election outcomes.
-
What do independent voters really think of Trump? Are views of Hillary really on the decline after taking a fair bruising in the first phase of the effort to get the nomination? Could Sanders really have an appeal?
The key political questioning by Gallup and many other US pollsters is is on perceptions of favourability which I now regard as the best leader rating format.
One of the the responses you get from UK firms about getting adequate and suitably random samples for vote share polling is the cost. There’s a lot in that. We’ve got used to having a mass of poll data on the cheap.
What I can’t understand is the huge reluctance of pollsters and their media clients to include proper leadership findings in a consistent format so that you can monitor changes from poll to poll. The record here is abysmal. These are far more important than doubtful voting figures for an election more than four years away.
To me the current most significant UK polling trend is the comparatively small proportion of LAB backers who are ready to give Corbyn top ratings on leadership. We had the same during the entire period of Ed Miliband’s leadership and these, surely, were strong pointers that all was not well with the red team.
I for one should have highlighted this much more.
I write while on the train to London for today’s polling inquiry event. I’ll be Tweeting and writing about this later.