(7) In reality, the collapse of the Lib Dems, initially to the benefit of Labour (and to some extent the Tories) masked the fact that ‘big2’ vote share has been declining since the split between Labour and the Social Democrats in 1981. Mark Pack’s wonderful polling database shows, this combined share isn’t normally at its lowest at election time, so it may be that the smaller parties, collectively speaking, are squeezed once again.
(8) But where are these parties polling in the first place? Pollsters have tended to show relatively similar spreads between Labour and the Tories. But in terms of the overall level of the big two, they show considerable variation. The flip side of this coin is that the other three main parties also vary considerably. As I’ve previously discussed at some length, this has been a well-known Lib Dem issue for some time, but the same can be said for UKIP and the Greens. Taking an average of each pollster’s last three polls, we find the big two anywhere from 60% to 70%, UKIP between 12% and 23%, the Lib Dems between 7% and 12% and the Greens between 3% and 7%. Take your pick!
(9) Swingback, the historical tendancy for governments to regain support and oppositions to lose it as elections approach, is a controversial topic, with a variety of views about the extent to which “it’s different this time”. In terms of the spread between the two largest parties, the examples above (all elections since 1979 where the government was trailing, i.e. not 2001 or 2005) show wide variation, but the year to the latest full month (i.e. 17 to 5 months before this year’s election) has shown almost exactly the same pattern as the average of all of the above. In other words, this is a texbook swingback. It is true that it’s entirely due to Labour losses rather than a more normal mix of Labour losses and Tory gains, but in terms of the spread it is right on schedule.