Obama adviser’s ‘catastrophic’ modelling made Tories confident for 2017 election
The scale of problems in the Tory election campaign has been revealed in internal data suggesting the party was on course for a comfortable win just days before Theresa May’s humiliation at the ballot box.
The confidential information prepared the week before the June 2017 election suggested the Conservatives were on course to win 371 seats, with Labour expected to slump to just 207 MPs.
The internal modelling, revealed in a new book on the election, shows that the information – used by Tory HQ to plan Mrs May’s campaign – was spectacularly wide of the mark in key battleground seats.
The snap election, called by Mrs May to capitalise on soaring opinion poll leads and the desire to secure a majority to get her Brexit plans through the Commons, resulted in a hung Parliament with just 317 Tories securing a place at Westminster.
The book, The British General Election of 2017, says that modelling done by ex-US president Barack Obama’s former adviser Jim Messina was wide of the mark as the race went into its final straight.
The book states: “A week before polling day, Jim Messina’s analysis had Labour to fall to 207 seats, with the Conservatives on course to win 371.
“Even then, there were some concerning signs, especially in London. Battersea and Croydon Central were already identified as probable losses to Labour.
“Both Putney and Kensington were seen as very tight – with just one point between the parties in both – and other seats in London such as Ealing Central and Acton, which should have been marginal, were clearly going to stay Labour.
“However, one of the other big surprise results of election night – Canterbury – was more of a shock, since at this point the Conservative modelling had it to be held with a majority of over 20%.
“But the party still seemed on course to take almost 30 Labour-held seats — mostly in the North and the Midlands.
“The Conservatives also expected to make seven gains from the Liberal Democrats, almost wiping out the party in the House of Commons.
“Expected Conservative gains from the Lib Dems at this point also included Sheffield Hallam and Leeds North West, both of which were indeed lost by the Lib Dems – but to Labour.
“However, in addition to over-estimating Conservative success in England and Wales, Messina’s modelling also under-estimated the forthcoming Conservative advance in Scotland; at this point, it was predicting just five Conservative gains from the SNP instead of the 12 that did occur.”
The book’s co-author Prof Philip Cowley, of Queen Mary University of London, told the Press Association: “This isn’t, as far as we know, the very last bit of modelling they did, but it’s near enough to the election to be a decent sign of how catastrophically wrong their targeting was, and it’s the only bit that’s been made public.
“It shows just how misguided the Conservative campaign was in much of the country – and how they were putting resources into seats they ended up being miles from winning, while thinking they were safe in seats they ended up losing.
“The scale of some of the errors are just astonishing and shows how ill-informed the party’s field campaign was.
“The Conservatives put a lot of faith in Jim Messina’s analysis to drive their targeting.
“It is clear that was misguided.”
The book, which Prof Cowley co-authored with Prof Dennis Kavanagh, is published on Thursday.