*** Welcome to piglix ***

Opinion polling for the 2005 United Kingdom general election


In the run up to the general election of 2005, several polling organisations carried out opinion polling in regards to voting intention in Great Britain (i.e. the UK excluding Northern Ireland, which is usually excluded from such voting intention surveys). Results of such polls are displayed below.

The election took place on 5 May 2005. The previous general election was held on 7 June 2001.

Since each MP is elected separately by the first past the post voting system, it is impossible to precisely project a clear election outcome from overall national shares of the vote. Not only can individual constituencies vary markedly from overall voting trends, but individual countries and regions within the nation may have a very different electoral contest that is not properly reflected in overall share of the vote figures.

Therefore, the first past the post system means that the number of MPs elected may not reflect the overall popular vote share across the parties. Thus, it is not necessarily the party with the largest share of the popular vote that ends up with the largest number of MPs. (See details of the elections in 1951 and February 1974) Since 1935 no party has achieved more than 50% of the popular vote in a British general election. The voting system favours parties with relatively concentrated support: a widely distributed vote leaves a party at risk of getting a large vote share but doing poorly in terms of numbers of seats (as the SDP-Liberal Alliance did in the 1980s), whereas parties with localised votes can win seats with a relatively small share of the vote.

That said, in previous elections, approximate forecasting of results were achieved by assuming that the swing in each individual constituency will be the same across the country. This system, known as uniform national swing (UNS) is used by much of the media in Britain to assess and extrapolate electoral fortunes from opinion poll data, though there has been criticism that such predictions may be naive and unreliable, even from providers of such data.


...
Wikipedia

...