What we learned from the leaked internal memo on Labour’s MRP analysis

LabourList has an advance copy of Left Out: The Inside Story of Labour under Corbyn and our editor is trying to read it as quickly as possible before its publication. But in anticipation of its full release, (Sunday) Times readers have been given a couple of sneak previews and the authors have shared a revealing internal party memo. It details analysis predicting a dire performance for Labour in the general election, which the document itself described as “extremely sobering reading”. Here’s what we learned…

The picture painted was much worse than the actual vote

Now, don’t get me wrong – last December was a dark day for Labour and has rightly led to much soul searching. But the analysis revealed in this leaked memo painted an even worse picture. The internal multilevel regression post-stratification (MRP) analysis carried out using marked register data by Labour in August, emulating those conducted by pollster YouGov ahead of the general election, predicted that the party would return just 138 MPs.

This would have been the party’s worst defeat since 1918. The MRP study projected that Labour would lose most seats, 23, in the North West of England, as well as 18 each in London, Yorkshire and the Humber and in Wales, and that the party would see 14 majorities overturned in the West Midlands. Ultimately, Labour lost 60 seats last December – losing 13 in the North West, retaining 49 seats in London, conceding six in Wales, nine in Yorkshire and the Humber and nine in the West Midlands.

Under this model, we would’ve been missing some big beasts

Had the 2019 general election have held true to the model, there would have been some significant names missing from the Parliamentary Labour Party. Casualties would have included the then Labour Party chairman Ian Lavery, Jon Trickett, the current Shadow Home Secretary Nick Thomas-Symonds, deputy leader Angela Rayner, Shadow International Development Secretary Preet Kaur Gill and Labour’s women and equalities spokesperson Marsha de Cordova, among others.

It identified a “massive misallocation of resources”

The memo produced last year warned that, based on the research, the leadership team would be criticised immediately after the election for a “massive misallocation of the resources of the Labour Party”. It said the party had been basing its strategy on the assumption that it would take a “major step forward” but this study suggested the election would in fact be a “significant step backward”. The report told the leadership that it would be judged after the vote to have wasted funds, “pouring money into seats we didn’t expect to come anywhere near winning, and starving seats which ended up being very close”.

Wales demonstrated this well: the strategy from Labour was to go on the offensive in Aberconwy, Arfon, Carmarthen West and South Pembrokeshire, Clwyd West, Preseli Pembrokeshire, and the Vale of Glamorgan. Yet the MRP showed that 18 seats in Wales were in danger of being lost – with 15 projected to go to the Conservatives, two to the Lib Dems and one to the Brexit Party. The document reports that Welsh Labour general secretary Louise Magee had expressed “concern about the absence of any defensive seats on the list in Wales”. In the end, her concern was well placed.

Equally shocking was the fact that, based on the MRP study of key seats that Labour was then targeting ahead of the general election, the Lib Dems were predicted to see a bigger improvement than Labour in every single one. The report concluded that this “brings home the scale of the threat from the Liberal Democrat surge which we face”, which brings us to our next point…

The biggest concern was the Lib Dem threat

The report described the threat posed by the Lib Dems as “by far the most significant message that this analysis provides”. It likened the effect on Labour’s vote share as similar to that of the 2005 general election, telling the leadership that the Lib Dems had the ability to “take votes from us and to attract from among swing voters support that we otherwise might have been able to rip from the Tories”.

A list of 18 seats forecast to be taken from Labour by the Lib Dems included some pretty hefty majorities. Manchester Withington, a seat held by Labour with a majority of 29,875 votes in 2017, was forecast to fall. Similarly – according to the analysis – the Lib Dems would have seized Leeds North East, overturning a 16,991 Labour majority, and Cardiff South and Penarth, which Labour won by 14,864 in 2017.

The report said that Labour needed to have “clear strategies for (i) blunting their appeal directly in seats where they are directly in competition with us for victory, and (ii) squeezing their vote and limiting their ability to attract votes”. In the end, the Lib Dems obviously did not win the 44 seats that the model projected – instead, they returned just 11 MPs to parliament. The memo did not account for Jo Swinson’s unpopularity, nor could it take into account their off-putting Brexit revocation policy. But its strong warning was heard loud and clear by some figures at the top of Labour.

… which had an impact on Labour’s campaign strategy

Many of the seats projected to be lost to the Lib Dems were those that had voted Remain in the 2016 EU referendum. The list of casualties included ten Labour constituencies in London, which had been won by Labour with healthy margins in the previous general election and also voted overwhelmingly to Remain in 2016. The poll predicted the party losing places such as Bermondsey and Old Southwark (12,972 majority in 2017), Hampstead and Kilburn (15,560) and Vauxhall (20,250), for example.

While stressing the threat of the Lib Dems, who pledged in 2019 to cancel Brexit if they won, the memo also said that the results of the MRP analysis were “broadly consistent” with the idea that Labour voters who backed a Leave outcome in the 2016 referendum would be more likely to stay Labour even if it became a more pro-Remain party. Conversely, the document suggested that those Labour voters who backed Remain in the referendum would be more likely to switch to a pro-Remain party, and said that there were “obvious political consequences which flow from such an understanding”.

The report concluded: “We will need a much stronger focus on preventing the splintering of the pro-Remain element of our party’s electoral coalition, and on attracting such voters from the Conservatives.” Labour’s 2019 manifesto, unveiled by the party in November, committed to giving people the “final say on Brexit” by securing deal within three months and the submitting that deal to a public vote alongside an option to Remain in the EU.