A historic defeat


The 2019 election marks a historic low point in Labour’s electoral success. Long term trends in voters’ relationship with Labour were catalysed by a perfect storm in this election. The scale of this loss is unprecedented. The only comparable defeat for an Opposition was 1983, yet the Tories had been in power for only four years at that election, not ten years.

The seeds of this defeat stretch back over the last two decades, with the link between Labour and the communities we were founded to represent profoundly broken. The collapse of support across the North and Midlands, our continued marginalisation in Scotland, and losses in North Wales are the result of long-term changes in the relationship between our party and our voter coalition, including demographic change and cultural shifts.

Underlying weaknesses going into this election magnified our losses including the unpopularity of our leadership, the failure to have a clear position on the key issue of the day, Brexit, and a manifesto that did not allay existing concerns about Labour in office.

Labour lost support on all sides in this election. Compared with 2017 in net terms, Labour lost around 1.7 million Leave voters, and around 1 million Remain voters. We also failed to attract swing voters, winning over far fewer swing voters than at any recent election. In 2019 the Conservatives were more successful than Labour in turning out non-voters. They overtook Labour among 2017 non-voters gaining 2 million non-voters to Labour’s 1.8 million. These votes for Labour were offset by the abstention of 1.8 million 2017 supporters who did not vote in 2019.

Original research by Datapraxis shows that we lost all types of voters everywhere compared with 2017. Remain and Leave, young and old, working class and middle class, habitual and occasional voters. London is the only area that bucked the trend.

Our investigation shows that partly as a result of the loss of Labour voters,  age, education and place are the new electoral divides even more than traditional conceptions of class, with new voters, non-voters and abstentions playing a greater role in determining increasingly volatile elections.

We now face a mountain to climb. We should not underestimate the scale and nature of the change that is necessary within our Party. In order to win a general election again, we must change how we campaign and reset our relationship with the public.  


The findings in these chapters are based on an extensive review of independent evidence, including:

  • bespoke analysis of the result commissioned by us from Datapraxis, making use of exclusive access to a YouGov dataset of private polling responses from almost 240,000 people during the 2019 general election, as well as response-line data from the YouGov Profiles data warehouse (and including a semantic analysis which allowed us to abstract patterns in voters’ open text responses to polling questions.) 
  • early data and analysis from the 2019-23 British Election Study shared with us by Professor Edward Fieldhouse, Dr. Chris Prosser and Dr. Jonathan Mellon of the University of Manchester
  • a historical review of trends provided to the Commission by Greg Cook, Labour’s Head of Political Strategy from 1995 to 2019
  • an extensive review of relevant published data and research, including that highlighted in submissions to this Review