Clacton, Heywood & Middleton
Yesterday we had two by-elections. Claction was an emphatic, walkover win for Douglas Carswell and UKIP, which was largely what everyone expected - he was particularly well regarded as an MP and the demographics of the seat could not have been better for UKIP. Heywood and Middleton was more of a surprise, many expected the seat to be a relatively easy Labour hold when in fact UKIP came within 2 points.
After every by-election you essentially see the same comments reading far too much into them, and I make the same blog post saying that by-elections are extremely odd events and you can't read too much into them: they have low turnout, are in a single seat that will not be representative of the wider country, are far more intensely fought than normal election and, crucially, do not have any impact on who the government will be the next day. If by-elections behave like the national polls, they tell us nothing new. If they behave differently, it's probably because by-elections themselves are very different.
That doesn't mean they aren't extremely important, as they are. They help shape the political narrative and public opinion. While Carswell's victory was broadly expected and costed in within the Westminster village, its getting huge coverage on the media now and I expect it will result in a boost for UKIP in the national polls. If you follow UKIP's support in the opinion polls over the last few years you'll see it's a pattern of spikes in support from positive publicity, mostly as a result of electoral success (local elections, strong by-election showings, European elections), each time settling back at a slightly higher level. For a challenger party the big challenges are getting coverage, being taken seriously and being seen as a viable choice rather than a wasted vote. This will help them a lot - I'd expect a spike it in the polls, and it'll be interesting to see which other parties they draw support from. On that issue, it will also be interesting to see how Labour react to the closeness of the result in Heywood, currently support for UKIP has disproportionately come from former Conservative voters and the Labour party seem to have regarded UKIP as their enemy's enemy, but they also have the clear potential to draw support from more Labour demographics.
As this is a polling blog I should save my last comment for the polls. The two polls in Clacton, conducted by Ashcroft and Survation, were both conducted more than a month before polling day, so they cannot in all fairness be compared to the final result (opinion in Clacton could easily have changed in the interim period), for the record though they were both pretty close to the actual result, certainly they got the broad picture of a UKIP landslide correct. The two polls in Heywood and Middleton (conducted again by Ashcroft and Survation) are more worrying. They were conducted about a week and a half before the election - so there was time for some change, but not that much (and many would have voted by post before polling day). Both showed a nineteen point lead for Labour when in reality they ending up squeaking home by two points. In both cases the polls both overestimated Labour support, and underestimated UKIP support.
