This dataset presents UK general election results by constituency from 1918 onwards. It draws on several sources, described below, which may contain errors. For each constituency, the dataset includes the number of votes and vote share for different political parties, as well as the electorate and turnout. Turnout was calculated as valid votes divided by the electorate.Jump to full report >>
The data is presented in two forms: an Excel spreadsheet where each election is presented on a new tab, and a CSV file where all elections are presented together.
The spreadsheet includes more information than the CSV file. The CSV file classes as ‘other’ all parties except for the Conservatives, Labour, the Liberals (later Liberal Democrats) and nationalist parties in Scotland and Wales. The spreadsheet gives more detail about smaller parties, depending on their performance at each election and the source data. Note that candidates standing as National Liberal and Conservative were categorised as Conservative in the source data for the 1955 and 1959 elections. Before then, they were classed as National Liberal in the spreadsheet and as ‘other’ in the CSV file.
The spreadsheet also includes notes with additional information for specific general elections at the bottom of the relevant tabs. These notes include details for people standing for parties categorised as ‘other’ who were elected. They also identify Speakers winning elections, where data was available. Speakers seeking re-election usually stood as ‘Speaker’ after 1955 but were elected as party MPs until then. By convention, the main political parties do not oppose them.
Data for Northern Ireland was not always recorded consistently in our sources. We have added more detail in the spreadsheet. This is not reflected in the CSV.
Constituency boundaries were changed following boundary reviews in 1945, 1948, 1955, 1974, 1983, 1997 and 2005 (Scotland)/2010 (rest of the UK). Comparisons through time work best for the periods between these changes, during which constituencies had the same boundaries (but note that after the 1918 General Election, Ireland was no longer represented in the UK Parliament). To enable such comparisons, the CSV file includes an indication of the boundary set used for each election, denoted by the first and last election for which this set of boundaries was used. Within each set, each constituency also has a constituency code that works as a unique identifier (also included in the spreadsheet). These codes overcome the problem that constituency names are sometimes spelled in different ways. Note that the codes are consistent within but not across boundary sets.
The data includes the following boundary sets and identifying codes:
The spreadsheet locates English constituencies in the counties they belong to, as they were at the time, as well as in the current English regions. These regions are mostly made up of counties and the external boundaries of these groups of counties have remained largely the same since 1918. There were, however, some changes in 1974, when local government structures in England were reorganised. The most significant changes were in Lincolnshire and Peterborough. It is also worth noting that the Yorkshire and the Humber region is not made up of historical counties in the same way that other regions are. Moreover, London has continued to expand. As a result, some of the constituencies and counties as they existed prior to 1974 do not fit neatly into the current regions.
The spreadsheet gives definitions of the location of counties in regions for each period in the notes at the bottom of the tabs. In the CSV file, constituencies are located in current regions only.
Compared to individual constituencies, the regions are more stable units for comparisons over time as their boundaries have changed significantly less.
1918 to 1945: two-member constituencies, uncontested seats and university seats
Prior to 1950, general elections operated slightly differently. Three features of these early elections are reflected in the data.
Some seats returned two members. Unless specified otherwise, the total number of votes cast for each candidate has been distributed among the parties. For more information, see the guide to Field’s data available from the UK Data Service. Note that Field found a number of errors in the results for two-member seats, so these need to be used with some caution. Note also that the calculation of votes for each candidate means that the total vote count may be different from that found in other sources.
There were seats where there was only one candidate, often the sitting MP. In these uncontested seats, no votes were cast. They are highlighted in orange in the spreadsheet. Winning MPs are indicated with a -1 (or -2 for two MPs from the same party).
Graduates of certain universities were represented in university seats. As these graduates are not concentrated in geographical areas, these seats are not located in counties and regions. More details on the Single Transferable Vote stages in the university seats is included on a separate tab in the spreadsheet.
This dataset continues to be developed and feedback is welcomed. Please contact firstname.lastname@example.org if you have any comments or suggestions. While we welcome feedback, we cannot respond substantively to each suggestion.
1918-1949: W. Field, British Electoral Data, 1885-1949 (UK Data Service)
1918-1992: B Walker, Parliamentary election results in Ireland 1918-1992 (Ireland & Northern Ireland)
1918-1949: FWS Craig, British Parliamentary Election Results 1918-1949 (University seats)
1950-1951: FWS Craig, British Parliamentary Election Results 1950-1970
1955-1992: Prof D Dorling, British General Election Results 1955-1992 (UK Data Service)
1974-Feb: Politics Resources, UK General Election results February 1974 (Web Archive)
1997-2017: House of Commons Library elections data
House of Commons Library, UK Election Statistics: 1918-2019
Commons Briefing papers CBP-8647
Authors: Philip Loft; Elise Uberoi; Edward Hicks