There will be a General Election in the UK during 2010. That is predictable - the date can be guessed at (most people think that May 6th is the most likely, being local election day) - but otherwise this could be the most unpredictable election for some time. A number of factors are at play -New boundaries will take effect. There are few constituencies that will see no change. Electoral data from previous elections is only available at constituency level. Therefore it is difficult to compare directly the old and new seats. Colin Rallings and Michael Thrasher have produced projections of change which are heavily relied upon. (C. Rallings & M. Thrasher, The Media Guide to the New Parliamentary Constituencies, (Plymouth: LGC Elections Centre, 2007)) A very sophisticated model is used, but exact data on a ward by ward basis is not published.
In British elections the concept of "swing" has featured heavily in analysis. (A wikipedia article explains the concept in more detail here). Essentially the formula is
Swing = ((A2 - A1) + (B1 - B2)/2)
where A2 is the percentage vote for Party A in election 2 (most recent election), A1 is the percentage vote for Party A in election 1,and similarly for Party B.
This works well when there are only two parties contesting a seat - but the vote has become more fractured over time. The next election is likely to see even more independents standing.
Many predictions for individual constituencies are based on a uniform swing across the country - which of course never happens.
There has been a crisis of confidence in the main political parties partly as a result of the expenses scandal, but which has deeper and older roots. This will make predictions very difficult for individual constituencies.
There are a number of computer websites that allow you to "predict" the results such as
As the General Election nears, this blog will (as it did with the 2008 Congressional Elections in the USA) give you the background to the key seats.
