Lord Ashcroft has produced another round of marginal polling – full details here. His last batch of polls revisited the Con-held ultra-marginals with Labour in second place. This time they deal with some Conservative held marginals with slightly larger majorities and revisit some Labour held marginals. The latter aren’t of much interest at the moment (with Labour ahead in the polls there is no realistic chance of the Conservatives gaining seats from Labour, so until and unless the Conservatives start showing a lead over Labour in national polls polling Labour-held Conservative targets is a little bit academic, though I may return to them in a future post), for the immediate future the Con-held marginals are more relevant.
Lord Ashcroft’s earlier polling of Con-marginals took the eleven Con-Lab seats with majorities under 2 percent. This round of polling took the seats with majorities between 2 and 3 percent, eight of them. The average swing across these seats was 6 points from Con to Lab, the equivalent of a 5 point Labour lead in the national polls. A little larger than in national polls at the moment but, as with Lord Ashcroft’s previous waves of polling in Conservative -v- Labour seats, not that different.
As usual with Lord Ashcroft’s polls this wasn’t a poll of a group of marginal seats, it was eight fully fledged polls of individual constituencies and looking at the individual seats spits out a few interesting findings. Lord Ashcroft used the two stage voting intention question for the constituency poll, first asking people a generic voting intention question and then asking people to consider their own constituency and the candidates likely to stand there in an attempt to squeeze out tactical or incumbency effects. Normally this has a huge effect in seats where the Liberal Democrats are in contention, and very little effect in seats where they aren’t. The effect of the two stage question here was illustrative – in four seats the Conservatives did very slightly better in the second question (what we’d expect from the incumbency effect). In Warrington South (a three way marginal) and Bedford (which has a strong local Lib Dem presence) the second question boosted the Lib Dems. In Stroud the second question boosted Labour, perhaps because the ousted Labour MP David Drew is seeking to return at the next election (Patrick Hall is also seeking to return in Bedford, where the second question also showed a slightly bigger swing to Labour). Also worth noting was the healthy performance by the Greens in Stroud, up on 12%.
A final observation: Southampton Itchen, one of the four Labour held seats, showed a swing of 0.5% from Lab to Con, putting the Conservatives and Labour neck and neck on 33%. This seems unlikely, while John Denham is standing down in Southampton Itchen the last time Ashcroft polled the seat in May it had an eight point Labour lead. A more likely explanation for the rather odd result is probably that suggested by Lord Ashcroft himself in his commentary – that Southampton Itchen has a substantial university population (students and staff) who wouldn’t have been around when the poll was conducted.