Who Labour lost in 2019 – and why

One of the most comprehensive public post-mortems of an election defeat ever has just been published. The 2019 Election Review from Labour Together, commissioned jointly with LabourList, includes fresh, previously unpublished research conducted by my organisation Datapraxis.

During the first three months of this year, we conducted a deep-dive into our unique dataset of hundreds of thousands of polling responses from YouGov, aiming to find out which voters Labour lost in 2019, where, why, and how it might win again. We were able to build on work we conducted during the election to map the message and electoral battlegrounds, and to advise Labour and other progressive campaigns.

The election review research was revealing and cathartic. Understanding what went wrong is just the beginning of getting things right. Data cannot tell the whole story, but it can correct misunderstandings and reveal blind spots. This article provides a short overview of some of our most important and startling findings.

Labour lost people on all sides. Our data indicates that it lost around 1.9 million who voted Labour in 2017 and Remain in 2016, as well as 1.8 million who voted Leave in 2016 and Labour in 2017. Crucially, it also lost almost a million referendum non-voters – a mostly younger and often overlooked group, whose turnout for Labour in 2017 played a vital role in in delivering the unexpected hung parliament result, but who then became gradually disillusioned.

Labour gained only around a million swing voters from other parties in 2019 to compensate for its losses, and only 450,000 directly from the Tories – far fewer than in the previous two elections. Most of these were Remain voters, who were in any case far more receptive to Labour’s identity and messages than Leavers. Once again, Labour’s main gains came from a new group of around 1.8 million mostly young 2017 non-voters.

But in 2019, our data indicates that the Conservatives succeeded in turning out around two million 2017 non-voters, doing a significantly better turnout job than Labour this time, and vastly outstripping the Tory turnout effort in 2017. This group looks very different to those who Labour turned out. They are older, and almost half of them voted Leave in the referendum. But the rest were for the most part habitual non-voters, who had also sat out the referendum, but whose clearest characteristic was an overwhelming and intense negativity towards Jeremy Corbyn.

These estimates are based on 2019 YouGov vote recall data from 206,060 respondents, the largest sample available in the UK, using a combination of YouGov’s nationally representative schema and Datapraxis political weighting. The data includes many people with lower levels of political interest or inclination to turn out, who pollsters often struggle to reach. We believe its quality and scale helps to understand turnout, abstention and patterns in the key battlegrounds more deeply than more traditional smaller-sample research methods.

We found that Labour ended up losing around 1.1 million Remain voters to Remain parties in 2019, plus around 600,000 to abstention and 240,000 to the Tories. It lost around 800,000 Leave voters to the Tories, over 200,000 to Remain parties and around 600,000 to abstention, but less than 200,000 to the Brexit Party (whose vote collapsed during the campaign, as its supporters switched to the Conservatives to “get Brexit done”). The majority of its lost referendum non-voters, again around 600,000, went back to not voting rather than switching to other parties.

Even I had expected the majority of the 1.8 million 2017 voters Labour lost to abstention in 2019 to be Leave voters. But our data suggests that they were split more or less evenly between Remain, Leave and referendum non-voters.

The results could have been even worse: more than one in three of people who said at the start of the 2019 campaign that they wanted to vote for the Lib Dems, the Greens or Plaid Cymru ended up voting instead for Labour, either tactically or through persuasion, increasing its vote share by millions. And almost half of the 1.4 million Labour Leave voters who the party ended up retaining had been undecided or voting for other parties at the start of the campaign. Without regaining these voters, the defeat would have been even more seismic.

Our analysis of the seats Labour lost found that here, too, the Tories’ greatest gains were from the two-million strong group of 2017 non-voters they turned out. Labour Leave voters seem to have made up a slight majority of the party’s lost voters in the 40 seats which were lost most heavily; but this was not true of the 20 narrowest marginals. Losses from all categories of voters played a role in the battleground defeats.

To investigate the contested reasons behind the defeat, we combined a wide range of YouGov-licensed datasets with message-testing we conducted during the campaign, and a semantic analysis of tens of thousands of open-ended responses about people’s views of Labour, segmenting the results by how people say they voted. Our findings were stark and clear. 

The most predictive factors we found for switching away from Labour in 2019 were the view that a Corbyn-led government was the most serious danger facing the country; voting in the 2019 European Parliament elections (which Labour hardly contested, but became a conveyor-belt for defection to Leave and Remain rivals); views of leadership; and political self-identification – with Labour mostly retaining a minority who describe themselves as fairly or very left-wing, but losing more centrist voters.

We found that the three main factors behind Labour’s losses were antipathy to the leadership (often associated with perceptions of weakness or incompetence, a shift to the far-left and/or antisemitism); Brexit-driven switching; and concerns about the affordability or credibility of the manifesto (although individual policies were mostly quite well supported). These three factors were interwoven, but the leadership factor was the most significant and consistent across all groups. Tragically, even many of the 2017 voters Labour hung onto in 2019 disliked Jeremy Corbyn by the end. 

Only among less than a million defectors to the Tories or the Brexit Party does the main driver of switching seem to have been Brexit – and even for them, leadership and the other factors loomed large. Most 2017 Labour voters who ended up abstaining in 2019 said that factors other than Brexit were more important to them, and this was even true of the abstaining Leavers.

Datapraxis will publish a full report on our findings in the coming weeks, complementing and underpinning Labour Together’s 2019 Election Review (which was itself drawn up by a commission involving leading figures from across Labour’s broad church). Most importantly, we have begun conducting detailed work on how Labour can rebuild a winning coalition, bringing people together from across the scattered voter tribes of Britain.

In his article launching the review, Ed Miliband wrote: “The Herculean task of winning the next election will require vision, imagination, discipline and unity. Every part of our movement must be part of it. We hope our report can be a starting point not just for winning an election but what really matters: a radical transformation of our country.”

It is time for Labour to turn its face to the future. This renewal project will take years and involve many hard choices. But we can understand and address the causes of the loss – and there are plenty of reasons for hope. Leadership demands values and ideas, but it also requires listening to the people, learning from them in all their diversity, and understanding where those who seem further from you are receptive or resistant. Politics should be a living dialogue with the voters, not a top-down monologue. I believe Labour can win the next election without compromising its courage or values. It will take an extraordinary swing, but this is a volatile and extraordinary time.