Debate over Labour’s electoral performance under Ed Miliband and Jeremy Corbyn has been rekindled by the publication of Labour Together’s 2019 election review. The review’s commissioners argue that several trends lie behind the 2019 defeat, including dramatic changes in voting cleavages, a loss of support among the working class, and ongoing changes to the geographical distribution of Labour’s vote – including further losses to the SNP in Scotland.
The report also claims that a swing away from Labour in its heartland seats in England and Wales in 2017 foreshadowed the loss of many of these seats in 2019. Other commentators have pointed towards the loss of the ‘red wall’ as a pivotal moment in Labour’s electoral history. But there is an important issue which is underplayed by Labour Together (in an otherwise comprehensive report) and in the subsequent debate.
A complete analysis of the 2019 election result requires discussion of the complexity of the UK’s voting system, and the way it translates voting patterns into results. In this article we point out an important, yet underappreciated, outcome of the UK’s first-past-the-post system: while Labour’s national vote share fell in each of the general elections between 1997 and 2010, it is not the case that support for Labour fell between 2010 and 2019 – either nationally or in the lost ‘red wall’ seats.
In fact, across the 50 seats lost in the North, Midlands and Wales in 2019, total votes cast for Labour increased by nearly 30,000 so that Labour’s vote share was 39% in 2019 – unchanged from 2010.
Labour did not, therefore, lose the red wall seats because of a dramatic collapse in its vote over the last decade. Instead, the Conservatives made large gains in these constituencies at the expense of the Liberal Democrats, increasing their vote from around 670,000 in 2010 to over a million in 2019, compared with Lib Dem losses of around 300,000 votes over the same period.
As illustrated in the above chart, the majority of Lib Dem losses occurred between 2010 and 2015. While both Labour and the Conservatives made modest gains, UKIP saw much bigger gains. This does not imply that voters deserting the Lib Dems went to UKIP: instead, the 2010–15 period saw flows of voters between all parties, with the net result that UKIP gained and the Lib Dems lost. In 2017, it was UKIP’s turn to collapse. This time, both Labour and the Conservatives made substantial gains. Finally, in 2019, Labour lost voters, taking its vote share in the red wall back to the low point of 2010.
Similar shifts in party support are observed for England and Wales as a whole: Labour gained over two million votes between 2010 and 2019, while the Conservatives added three million votes and the Lib Dems lost almost the same number. Labour ended the decade with a vote share in England and Wales that had increased from a low of 28.5% in 2010 to 34.3% in 2019. In 2017, Labour’s vote share of 42.4% in England and Wales approached the record of 44% set under Tony Blair in 1997.
As illustrated in the chart below, swings in the lost ‘red wall’ seats between 2010 and 2019 were similar to those across England and Wales as a whole, with two main exceptions. First, Conservative gains in 2017 (as UKIP collapsed) were stronger in the ‘red wall’. Second, while the England and Wales vote swing in 2019 was characterised by Labour losses alongside gains for the Lib Dems and Greens, Conservative gains were stronger and Labour losses greater in the ‘red wall’ seats.
Aside from the effect of shifts in support for smaller parties, first-past-the-post elections can also produce apparently paradoxical outcomes because of the way votes are distributed among parliamentary constituencies. This is a significant contributing factor to Labour’s net loss of 16 seats between 2010 and 2019. As noted by Will Jennings, and illustrated the figure below, many of Labour’s success stories have occurred in safe seats so that Labour’s vote is becoming more concentrated in fewer seats. In contrast, the Conservatives have added votes more uniformly, allowing them to increase their seats in England and Wales by 54.
Much has been made of the reasons for the loss of the red wall. Commentators have highlighted the role of socio-economic and cultural factors in driving Conservative gains. If we compare the 50 seats that Labour lost in the North, Midlands and Wales – seats including Burnley, Lincoln and Wrexham – to the 50 that returned Labour’s highest vote shares in 2019 – seats such as Birmingham Ladywood, Liverpool Wavertree and Manchester Gorton – we find that the latter suffer from significantly higher levels of deprivation, but also have younger populations who are more likely to have gone to university and more likely to be working. These demographic shifts have been widely noted and discussed – including in the Labour Together report.
But these shifts cannot be considered in isolation from the voting system. It is not the case that Labour has lost support overall in England and Wales – Labour’s vote share has increased. But in first-past-the-post systems, there is little to gain for a party that increases its vote share in seats it already holds, regardless of the socio-economic or cultural characteristics of the voters it attracts.
Keir Starmer undoubtedly faces a challenge in winning enough seats to form a government at the next election. But addressing that challenge requires getting the diagnosis right. Labour’s popularity among some groups it attracted in the nineties and noughties may have waned, while support is growing among other groups. But increasing support is no guarantee that a political party will gain seats in first-past-the-post voting, let alone win an election. Any Labour electoral strategy needs to take this into account.
Appendix: ‘red wall’ seats lost in 2019
Definition: seats lost by Labour in 2019 in the North, Midlands and Wales:
Barrow and Furness
Bolton North East
Wolverhampton South West
Crewe and Nantwich
Heywood and Middleton
North West Durham
West Bromwich East
West Bromwich West
Wolverhampton North East
Penistone and Stocksbridge
Vale of Clwyd