Election History and Commentary

This page first posted 2 May 1997, last updated 19 December 2019

Election history since 1900

The table shows all the British general elections since 1900. The "Lib" column includes the original Liberal Party, the Liberal/SDP Alliance and the Liberal Democrats. The "Con" column includes Ulster Unionists up to 1970. Vote shares are on a UK basis, which will differ from the GB-basis used elsewhere.

Source: British Political Facts 1900-1994, David Butler and Gareth Butler, Macmillan (1994). With corrections.

The 2019 Election

The 2019 election gave a big majority of 80 seats to the Conservatives who ran under the slogan "get Brexit done". The Conservatives gained seats in the midlands, Wales, and the north of England as Labour voters deserted Jeremy Corbyn who ran a hard-left and Brexit-ambiguous campaign.

Electoral Calculus predicted the outcome satisfactorily, showing a final prediction of 351 Conservative seats. This was closer to the correct result than any other major forecaster. There was a small amount of polling error, but the campaign polls were broadly in line with the outcome.

See the Track Record of the Electoral Calculus site since 1992.

The 2017 Election

Despite the pollsters, the Conservatives were only narrowly ahead of the Labour party under Jeremy Corbyn. The Conservatives lost their overall majority, to end up eight seats short. Labour made gains throughout England, though both the Conservatives and Labour gained seats in Scotland from the SNP. Smaller parties faced a loss of votes, though the Liberal Democrats managed to increase their seats nonetheless.

Electoral Calculus mis-predicted the outcome, and had predicted a Conservative majority along with most other pollsters and commentators.

The 2015 Election

Despite the pollsters, the Conservatives won the 2015 election to get a small overall majority in the House of Commons. The Labour party under Ed Miliband lost almost all their seats in Scotland and failed to make progress in England and Wales. The Liberal Democrats, who had been part of the 2010-2015 coalition government under Nick Clegg, lost most of their seats as well as two-thirds of their voters deserted them.

Electoral Calculus mis-predicted the outcome, and had predicted a hung parliament along with other pollsters and commentators.

The 2010 Election

Electoral Calculus predicted the outcome correctly, and had the Conservatives' number of seats correct to within 10, but overestimated the Liberal Democrats.

The 2005 Election

Labour's poll standing slumped to within less than 1% of the their support in 1992 (when they lost), but they were still more popular than the Conservatives, so won more seats. Their majority would have been greater except that there seemed to have been "tactical unwind" of Liberal Democrat supporters refusing again to vote tactically for Labour.

Electoral Calculus predicted the winner correctly, but overestimated the majority. This was caused both by opinion poll error, regional swing and local factors.

The 2001 Election

Prior to the election campaign, there was a popular perception that Labour would indeed win, but with a significantly reduced majority. A figure of 100 seats was frequently mentioned.

Readers of this site would have seen a very different story - we were predicting a large majority for Labour even before the start of the campaign. The news was broken to the rest of the nation by The Sun newspaper on 3 May 2001, which predicted a majority in excess of 200.

In the event, Labour's majority did decrease, but only by 12 seats, to 167. Their share of the popular vote declined by two points to 42%.

This time, the pollsters did not get it exactly correct. Their final polls forecast a Labour lead over the Conservatives of 16%, but in the event the actual margin was only 9%. Although this was about 4% lower than in 1997, it did not translate into very many seats changing hands. This may have been due to differential turnout (Labour voters in safe seats staying at home) and tactical voting.

The 1997 Election

However that majority was bought with a modest share of the popular vote - Labour won a majority of 179 with 44% of the votes.

Although this victory was striking, it was possible to predict it in advance. Unlike the Conservative victory in 1992, when the professional pollsters mis-estimated the swing by 9%, this time the campaign and exit polls were pretty accurate. It was possible to predict the size of the Labour victory from the polls and an analysis of each constituency.

Using these techniques, I publicly predicted the result of the 1997 election as a Labour majority of 180 (Cambridge Evening News 30 April 1997, and BBC Radio Cambridgeshire 1 May 1997). Although the analysis is not complex, the result nevertheless came as a surprise to many people. It needn't have.

The Next Election

Before the votes are cast and counted, it is not possible to state with accuracy how the public's view of the parties will develop. Will the Conservatives continue to dominate the Labour party?

The best scientific answer to these questions will come from the opinion polls. In the period after the last election, there will probably not be much opinion polling published. As polls are produced, you can see a history of new opinion polls. As they are published in the months ahead, you can use them to give a guide to the majority, using our make your own prediction feature, or the Battlemap tool.