Still no change. In contrast to the first part of the UK election campaign, especially following the initial TV debate, the opinion polls for the last couple of weeks have been pretty steady. The Conservatives are well in front, somewhere in the mid-30s and maybe gaining slightly, while Labour and the Liberal Democrats are fighting out second place (with Labour probably just ahead) in the high-20s.
The BBC’s “poll of polls” this morning has Conservatives 35%, Labour 29%, Lib Dems 26% and others 10%. On a uniform swing — a dodgy assumption, as we know — that would be almost a tie in seats: 272 Labour, 270 Conservative, 79 Lib Dems and 29 others.
To the extent that the assumption of uniform swing breaks down, that will almost certainly work against Labour. Nate Silver at fivethirtyeight.com, who has a much more sophisticated (although largely speculative) model, is currently tipping 312 Conservative, 204 Labour and 103 Lib Dems — still short of a Tory majority, but getting very close.
The momentum in the polls seems to be with the Conservatives, and we know that the late polls have underestimated their vote in recent elections. On the other hand, the polls are almost certainly understating the support of the far-right — the anti-EU UKIP and the fascist BNP — and those parties are most likely to be taking votes from the Tories (and possibly from Labour in some areas).
So this still looks like a very close election, with a Conservative government the most likely outcome but quite probably having to contend with a hung parliament. Given the unusually large number of postal votes, it might not be until tomorrow evening (eastern Australian time) that we know what the new House of Commons will look like, and actually forming a government could take weeks.
This is something of a new experience: the last few British elections — going back to Margaret Thatcher’s first victory in 1979 — have been pretty clear cut. Even the surprise result in 1992 ended up being quite comfortable, with the Conservatives winning 41.9% and 336 seats. (Adam Carr, of course, has all the figures.)
The last close finishes were the two 1974 elections: in the first, the Tories actually won more votes (37.9% to 37.2%), but Labour won four more seats. In the second, Labour won a narrow majority (319 out of 635) with 39.2% of the vote — that was reduced to a minority through defections and by-elections, and the government eventually fell on a no-confidence vote in March 1979.
As 1974 shows, it’s possible in Britain — as it seems not to be in Australian federal elections — for close elections to go to the opposition. Labor also won government by a wafer-thin majority in 1964, as did the Conservatives in 1951 — despite trailing Labour in the popular vote.
The relationship between votes and seats has always been arbitrary.
Perhaps the most relevant precedent for this year is 1983. It was a Tory landslide — they won 42.4% and 397 seats — but there was an interesting contest for second place. Labour, with 27.6% of the vote, won 209 seats.
The SDP-Liberal alliance — forerunner of today’s Lib Dems — were only just behind, with 25.4%, and they won 23 seats. (No, that’s not a typo: it’s 23.) This could be the year that Britain wakes up to the need for a change in the electoral system. Or it could just be 1983 all over again.
There’s now an update at fivethirtyeight.com – http://www.fivethirtyeight.com/2010/05/uk-election-day-scenarios.html – in which Silver confirms my view that the polls have underestimated the Tory vote in the past, to the tune of an average 1.6% over the last three elections (and a whopping 6.4% in 1992). He explores various scenarios for how the polls might be wrong, and concludes as follows:
“As of this writing, betting markets now give Conservatives about a 38 percent chance of claiming a majority. I don’t know that there’s much arbitrage either way there, but I’d be slightly inclined to take the Tories on those odds.”
Me too.
What are the odds on the introduction of proportional representation voting before the following general election?
the history of PR campaigns in the UK is that none of them have worked. (the evidence is all around us). -because, the promises were always ‘…sometime in the future’ for either exploring it, or considering the reports of the explorations.
if the SDP^H^H^HLiberals^H^H^H^H^HDems want to get it this time, they will have to extract the kind of commitment which has NEVER been given before, that a specific form of PR will be presented to the house, in a way which makes it unlikely either chamber can block it.
otherwise, its monkeys dreaming of coconuts.
I should of course have said that the presentations of proposals for PR made to date have been done in ways which made it inevitable they would be blocked. The Scots model of a combination of direct member and balancing cross-constituency MPs might now be more tenable. Once a member of the UK political scene has adopted a model, its less plausible to say ‘it absolutely doesn’t work’ although I am told the compromises in government up north are now starting to tell a bit.
I guess it takes a number of long drawn out counts and hung parliaments to light a real fire under a PR campaign.