Nate Silver has a nice appraisal of the voter swing models that many news organizations are using the connect poll results and hypothetical election results. Many British news sites, such as the BBC, The Daily Telegraph, The Guardian, and the UK Polling Report, have swing models where one can tweak the percentage of the vote that each party receives and it will output the arrangement of seats by looking at the margins in the past election and extrapolating what would happen if the numbers were to move by x% in any given direction.
Now, the notion that swings aren’t uniform should be rather intuitive to those who have paid any attention to U.S. elections. After the 2008 election, the New York Times did a very good job illustrating in map form the lack of uniformity in the swings with almost all the counties in Indiana showing a 15+% swing toward Obama vs. Kerry, whereas Arkansas was mostly over 10% in the opposite direction, with most of the counties in Minnesota and Arizona showing a swing of under 5%. So the fact that such uniform swing models are likely to be inaccurate and have been inaccurate in the past is not surprising.
However, what Nate does is show some of the possible effects of these inaccuracies in the biasing of the projected results that they output, and his conclusion is that they give Labour, the party of the current government, significantly more cushion than they’ll likely have, even with a relatively good outcome. This is especially pertinent as the leader of the Liberal Democrats, Nick Clegg, has made it clear that he thinks that a continuing Prime Ministership for Gordon Brown would be illegitimate should Labour come in third in the popular vote.
It should be noted that Nate’s model only corrects for regional variations in swing insofar as they track according to party lines, but as the 2008 U.S. presidential election showed, party lines are not the only determinant of vote swings. There is also an element of “local flavor” that made Obama seem so much more favorable to those living near the Great Lakes and so much less favorable to those living in Appalachia and the Ozarks and that should very much be a wild card in this election, seeing I’m not sure what regional attitudes look like in the UK right now.
What I can say is that this looks to be a bad election for Labour and only time will tell whether we are looking at a Conservative government or a hung parliament on election day. Regardless, it will probably be the most interesting UK election in a long time and one that will underscore the structural problems with a parliamentary system based on a single member district plurality voting system.