Churn and gender gaps
Two things worth looking at elsewhere. The first is an article by Peter Kellner on churn. This is something I've been meaning to write about for sometime but never got round to, but Peter has now done the hardwork himself! I often see people taking about the next election from the starting assumption that Labour were at their low and so start from at least 30%, and that anyone who was going to vote Tory already did so they can't get above 37%. Neither of these is true - not least, because current polls show the Conservatives already winning the support of some voters they didn't have last time, and the Labour party already losing the support of some people who did vote for them last time. That's not to underplay how difficult it would be for the Conservatives to increase their vote share, or that the partial demise of the Lib Dems has provided Labour with a good boost in support, just that it's really not as simplistic as assuming Labour cannot lose any votes from last time, or that the Conservatives cannot gain votes... both are already happening to a great degree, it's just when we look at the headline figures we only see the net effects of Labour up and Conservative down. In both cases there is actually plenty of movement in both directions.
The second, with no modesty whatsoever, is something I have written over on the YouGov website using the same aggregate YouGov data, in this case looking at the Conservative gender gap or, perhaps, its absence. This is something that will not go away, every couple of months a journalist pops up writing a column about how the Conservatives are doing worse amongst women, normally illustrated by ripping one single poll out of context that appears to show a gender gap. Looking at the wider polls, it doesn't actually seem to exist. The aggregated monthly YouGov data in recent months has had the Conservatives on a solid 33% amongst men, and an equally solid 33% amongst women. No difference. The gap amongst women is bigger, but that appears to be because Labour do better amongst women and UKIP do worse.
At first site all the fuss about the Tory women problem is complete nonsense, but dig a little deeper and the Conservatives do appear to have a problem with some women. Specifically the Conservatives do worse amongst women than men amongst under 40s (and Labour the other way round). The reason the Tories don't do any worse with women overall is that as you move up the age ranges the pattern reverses, so that amongst over 60s the Conservatives do better amongst women than amongst men. I'm guessing the latter is because of UKIP (who seem to appeal to men more than women, and whose support is heavily skewed towards older people), while the former is presumably because the Tories do have some sort of problem appealing to younger women (or... logically equally likely... Labour have some sort of problem appealing to younger men).
