In so far as there has been a “story” in terms of polling since the last election, it has largely been dominated by the precipitous decline of the Liberal Democrats, and the remarkable rise of UKIP. Within a matter of months after the General Election, the Lib Dems went from a comfortable low-20s poll figure to barely hanging on in double figures (and haven’t managed to hang on at all in recent months); UKIP, meanwhile, rose slowly and steadily from the low single figures to the high teens, occasionally even breaking out into the twenties.
It’s perhaps surprising, therefore, that their fortunes in terms of predicted seat numbers are actually looking very similar right now. Here’s the mean number of seats gained by the parties over 1000 simulated elections, using as their basis the most recent UK Polling Report Polling Average (at the time of writing, Con 31%, Lab 34%, LD 8%, UKIP 17%, Green 4%).
As with the line graphs I’ve been posting recently, the error bars represent roughly 95% confidence intervals, and the horizontal black line is the winning post. We’ll ignore the predicted result of this election for now (for the record, 72.6% of simulations resulted in a hung Parliament, and in all of those the only mathematically feasible coalition was Lab-LD), and focus on the UKIP and Lib-Dem bars.
As you can see, they’re effectively level on about 40.5 seats (UKIP 40.7, LD 40.5). The interesting part here is that the UKIP error bar is significantly larger than the Lib-Dem one; this gives us a clue as to how the Liberal Democrats are managing to hang on to about 71% of their seats despite polling at about 35% of the level of support they enjoyed in 2010. It’s simply this: the Liberal Democrats, as a very long-lived third party in a first-past-the-post system, have learnt how to play the game extremely well. Their support is focused into a few seats where they know they can win, and they are therefore considerably more immune to swings in public opinion than almost any other party.
(The Greens are managing something very similar – despite polling less than 1% nationally in 2010, they concentrated their support so effectively in the Brighton Pavilion seat that they managed to return Caroline Lucas as their MP – and like the Liberal Democrats, they are largely unaffected by changes in opinion. The downside is that they are also unlikely to make any kind of breakthrough anywhere else, despite currently polling at a level of support four times higher than in 2010.)
UKIP, on the other hand, are enjoying a lot of support nationally, and accordingly did very well in the European elections which are held under a form of proportional representation; however, as they have not managed to concentrate this support anywhere, their predicted Parliamentary breakthrough is merely impressive for such a relatively young party, rather than the overwhelming landslide for which they are presumably hoping.
Electobot: The Blog 