Unlike many other democracies, the UK has no official body that collects and publishes official results straight after an election.

In Australia for example, the Electoral Commission publishes detailed results down to polling-place level for federal elections. Elections Canada provides similar levels of detail. In Germany, voters can examine maps and graphs of results provided by the Federal Returning Officer. But the UK’s Electoral Commission’s remit is more limited when it comes to collecting and publishing results: they are not required to publish detailed results but must produce “a report on the administration of” each “election or referendum”.

The Speaker’s Commission on Digital Democracy recommended that the Electoral Commission should set up a digital ‘results bank’, and they have published data on recent elections and referendums. But these files present the results for one election at a time.

Creating our own dataset, the Commons Library has now made general election results from 1918 to 2017 available to view in one place. For the first time, constituency results for this period can be compared in a single file, available online. This Insight explains how election results are collated, some of the problems with not having a systematic process for collecting and publishing results in a timely manner, and how we put our dataset together.

How are general election results collected?

When the polls close on election nights, counting officers in local authorities oversee the tallying of results. At the end of the night, or sometimes a little later, the returning officer declares the winner. Usually, the result, including the number of votes cast for all the candidates, is then added to the website of the local council organising the count. And that’s it. 

Data is not collated centrally for a swift publication of official results. The Electoral Commission doesn’t publish its analysis of how the election was administered, or the results, until much later. So, to produce a more timely overview of what happened across the country, journalists, academics and activists trawl through declarations and websites on election night and the coming days.

What data does the House of Commons Library use?

A team in the House of Commons Library checks council websites to collect a consistent and authoritative dataset of votes in each constituency at the general election. Because there is no requirement for councils to publish election results in a particular format, there is substantial variation in which data is included and how it is presented, including whether or not it is machine-readable (something the Government is aware of). Where we find discrepancies or data we need is missing, we check in with returning officers.

We then compare our data with news organisations doing the same thing, such as the BBC. Their datasets are completed more quickly than ours as they need to report the results on election night. As a result, their information on electorates (used for calculating turnout) does not take into account any changes immediately before the day of the election.

The resulting dataset is the best we have to discuss election results. That is, until the Electoral Commission and academics like politics professors Colin Rallings and Michael Thrasher publish detailed election results at a later date.

What is and isn’t in our data?

Our dataset is based on the information we receive from returning officers shortly after an election. However, returning officers sometimes need to correct the results they initially publish. As there is no central system for collating election results, we are not always aware of those changes. They therefore don’t always make it into our dataset immediately.

General election results in the UK are only systematically declared at constituency level (though some returning officers make them available at lower levels as well). So our dataset does not include results for different geographies, like wards or polling areas. Such results are often available in other democracies.

It is worth noting that the difficulties described above also apply to local and European Parliament elections; there is no central system to collate and publish these votes either.

Publishing election results

As there is no central resource for UK election results, organisations including the Electoral Commission and the House of Commons Library have so far published one datafile for each election. Academics have made efforts to combine election results from different general elections into coherent datasets, but these typically only cover up to 50-years of election results.

We have now combined these existing datasets into a single one; covering all general elections between 1918 and now. This data will be updated after each general election, providing a complete set of election results for constituencies online.

We have published the data as a spreadsheet and CSV file. It enables new analysis of electoral trends, such as the highest and lowest turnout in the past 100 years and the biggest and smallest majorities. This Insight discusses five such top five facts .

Further reading

About the author: Elise Uberoi is a researcher at the House of Commons Library, specialising in social and general statistics.