In 2011, Ruth Ellen Brosseau was a bartender in Ottawa, Canada’s capital. Some of her regulars were political activists for the New Democratic Party and when the federal election was looming, they twisted her arm to stand as a paper candidate in a no-hoper constituency in a heavily French-speaking area of Quebec. She didn’t campaign (just as well, since her French wasn’t very good), she didn’t even visit the constituency. During the election campaign she went on holiday in Vegas.
2011 was a breakthrough election for the New Democratic Party in Canada. They supplanted the Liberals as the main opposition, taking 103 seats when they had previously held only 36. Among those 103 MPs was a very surprised Ms Brosseau.
The 2017 general election in Britain also looks likely to be a mould-breaker. The Conservatives currently look set to make sweeping gains. If so, uniform national swing (UNS) is going to be of limited value. It works well when considering smallish movements in the polls. The bigger the swings, the more unevenly distributed those swings will be. In 2015, the swing from Labour to the SNP in Scotland was 23.9%, but the swing to the SNP reached 39.3% in Glasgow North East (and only 10.9% in Edinburgh South, which Labour held onto).
Even smaller swings are usually unevenly distributed. In 2015, Labour obtained a 0.4% swing from the Conservatives, but this concealed substantial variations – the Conservatives obtained a 3.9% swing towards themselves in Vale of Clwyd, while Labour benefited from a 6.4% swing in Ilford North.
As at 18 April, when Theresa May announced the election, three different polls found that the Conservatives had a 21% lead over Labour, representing a 7.5% swing from Labour to the Conservatives (though separately Opinium found only a 9% lead). If that projected 7.5% swing to the Conservatives is replicated at the general election, we might easily see some seats with no swing to the Conservatives and others with a 15% swing.
Overlaying that, the EU referendum has upended previous loyalties. The Prime Minister is seeking a mandate to deliver Brexit and the Lib Dems are seeking votes from opposing it. Labour is seeking a policy on it. It is likely that this will make the effects in different constituencies lumpier than usual, as some voters switch allegiances in order to back the party they judge will best deliver their preferred referendum outcome.
So the election will be wild. The most obvious consequence is that no one will really be clear which seats are in play and which seats are foregone conclusions. With the sort of leads that the Conservatives are enjoying, they will be looking to take seats in which their party membership is not strong and where they will not have the intensively-gathered information that they have accumulated in the seats vital for gaining power.
Meanwhile Labour need to decide where to try to construct a firewall. A 7.5% adverse swing sees Labour lose 67 seats to the Conservatives. Labour could not sensibly seek to defend all of these (and would be daft to try on current polling). They will need to focus their efforts.
But they will also need to keep an eye on seats that fall to a greater than 7.5% swing. 135 Labour seats are vulnerable to a 15% swing to the Conservatives (some of these are vulnerable to other parties on smaller swings as well) and, as I note above, if some seats have a less than average swing, others will see a greater than average swing.
I haven’t begun to talk about the Conservative-held seats that Labour should be taking aim at. Right now, those don’t look like a priority.
The risk in these circumstances is always that the defensive party is too optimistic. It does Labour no good to keep the adverse swing down to 6% if the Conservatives only need a 3% swing to take a seat. Meanwhile, if a seat that is safe up to an 8% swing gets a 10% swing, that’s two seats lost where one might have been saved. But it is very hard to tell a sitting MP that he or she is not going to be supported.
The Lib Dems were nearly wiped out in 2015 because they were too optimistic in such circumstances, despite being pretty disciplined about these calculations. This stuff is hard, emotionally but also strategically.
So Labour have some excruciating decisions to make about prioritising. With a membership of hundreds of thousands, they have the troops to mount a campaign but they need to deploy them effectively. This is going to take detached judgement, ruthlessness, discipline, focus and unity. These are not qualities that Labour are currently noted for. I expect Labour will either be far too optimistic or, perhaps more likely, that it will never get as far as drawing up a defensive strategy and leaving every constituency for itself.
My expectation, therefore, is that Labour will probably do significantly worse than uniform national swing suggests, as they fail to keep the seats that they are actively defending and see greater than average swings in some seats that they haven’t actively defended that could have been saved. As to which seats those are, I don’t know either. No one knows anything.