Whatever the outcome, the 2019 general election will be one of the most decisive polls in British history. Like the Liberal landslide of 1906 which led to the foundation of the welfare state, the Labour victory of 1945 or the 1979 election which introduced free-market Thatcherism, the 2019 election is likely to determine the nature of the UK for decades to come.
The 2019 election is also highly uncertain, and few pollsters are confident of the likely outcome. There are several reasons for this. Although Boris Johnson’s Conservatives are well ahead in the polls, the memory of Theresa May's melting majority in the 2017 election looms large. A second important reason for the uncertain outcome is the presence of three or more parties with a substantial share of the vote. In the UK’s first-past-the-post system the complex relation between seats and votes becomes more volatile when at least three parties have similar vote totals. Past Liberal surges produced few seats for the party but could have turned into a landslide if their vote had been a few percentage points higher.
The unprecedented nature of the election makes predictions precarious, but it is possible to make scenario projections. To attempt to assess how the current polls translate into seats, I have run an election model under two scenarios. The models show that if opinion polls stay at the same level, the Conservatives should win a majority in the Commons, even if a Remainer electoral pact and tactical voting against the Tories is successful.
The first scenario takes the results of the 2017 general election and adjusts support in each constituency for changes in the polling since then. The change is assumed in this scenario to occur uniformly across constituencies in England and Wales. Scotland and Northern Ireland are treated separately. The Brexit party is mapped to the Ukip vote share in 2017 and is assumed to take three votes from the Tories for every one it takes from Labour
Below are the results of the first scenario:
Projected Seats 2019
The results are surprisingly clear. The table shows that even with only 37 per cent of the vote, the Tories would win the election with a 95 seat majority, which could rise to 127 if the party is backed by Brexit supporting parties.
In this scenario the Conservatives would win 69 seats from Labour, even when facing the Brexit party, while losing only 14 (ten to the SNP, three to the Lib Dems and one to the Brexit party). The Brexit party is projected to gain two seats (Hartlepool and the Rother Valley) from Labour.
The news that the Brexit party will not stand in any seats with sitting Tory MPs does not change the result of this scenario. The Tories would still win the seats they held in 2017, apart from in Thurrock, which they might have lost to the Brexit party. In seats held by Labour in 2017, where the Brexit Party still intends to stand, the above results are unchanged because we have already assumed that the Brexit party will siphon off Tory voters.
Most of the Tory gains in this scenario are in the North of England, with further gains in the Midlands, London and the South East. The main reason for the Conservatives' gains is the rise in the Lib Dem vote which takes votes away from Labour in seats they hold. This allows the Lib Dems to gain 16 seats in total, but also helps the Tories take many more seats from Labour.
Whether this scenario, based as it is on uniform swings, is realistic will depend on local factors and on tactical voting. In our second scenario, tactical voting is investigated by assuming that the Lib Dem/Green/Plaid Cymru electoral pact works perfectly (in the sense that a party's lost voters transfer to the right party for the pact). More importantly, our tactical voting scenario assumes that a quarter of Lib Dem and Green voters decide to support Labour candidates in seats the Tories may take from Labour.
The Lib Dems, Greens and Plaid Cymru are generally too weak to make much difference in the 60 constituencies (49 in England and 11 in Wales) covered by their pact. In Wales most seem aimed at protecting existing Plaid Cymru seats and in two cases this may preserve PC control (Llanelly and Ynys Mon). The net impact is likely to be marginal at best.
Informal tactical voting is potentially much more important. The rise in the polls of the Lib Dems (and to a lesser extent the Greens) reflects their appeal to Remain supporters, but this increase in votes could easily be distributed unevenly across constituencies through tactical voting rather than with the uniform swing assumed above. The assumption here is that a quarter of Lib Dems and Greens vote Labour to prevent local Tory victories.
The prediction in this scenario is that the Tories would lose 21 lost seats compared to scenario one. Their majority would then fall to 53, but with DUP and Brexit party support their effective majority could be over 100. The intervention of the Brexit Party is much less important than generally assumed.
Neither does the Brexit party’s decision change the scenario result in the 21 constituencies in which Remainer tactical voting prevents the Tories from winning. These were all seats which Labour held in 2017 and hence areas where the Brexit Party will continue to stand. In a sense the Brexit Party are standing down in the wrong areas. It would have much more impact if they stood down in Labour constituencies which the Tories have the potential to win.
Will these scenarios reflect reality on December 12? This depends chiefly on what happens to party support nationally. The polls have been consistent over the early days of the campaign and the Tories are determined not to repeat their vote collapse during the 2017 campaign. Differential turnout may also change the result. In some areas working-class Leave voters are turning against democracy on the grounds that their vote is ignored anyway. If this led to a large fall in turnout in Leave areas the outcome might change, although in exactly which direction is difficult to predict. Barring major changes in the polls or dramatic differences in turnout the likelihood is however a comfortable victory for the Tories.
Dr Graham Gudgin is the co-author with Professor Peter Taylor of ‘Seats, Votes and the Spatial Organisation of Elections’ republished in 2012 by the European Consortium for Political Research as an ECPR Classic.