Last week saw the official close of nominations for the 650 Westminster seats. Both UKIP and the Green party are no longer contesting hundreds of seats throughout the country. UKIP is standing aside from 255 out of the 632 seats in England, Scotland and Wales. And the Green Party is standing aside from 171 of them. The Liberal Democrats are also standing aside in Brighton Pavilion to help the Green party leader Caroline Lucas.
What effect will this have on the other parties? Will Labour and the Liberal Democrats benefit from this partial electoral alliance. Do UKIP's actions make any difference?
We can look at this with a new electoral model. I built this model at Electoral Calculus explicitly to handle the behaviour of voters whose natural party is not standing this time. What will these people do? One thing is sure: they won't necessarily vote the way that their party leaders are expecting or suggesting. Some people will, but many others won't.
We can use the polls to estimate how voters will "break" when their preferred party is not standing. For UKIP voters, three-quarters will go to the Conservatives with the rest split between Labour, Lib Dem (yes, really) and other parties. For Green voters, there is less poll data, but I estimated they would be more likely to go to Labour (or the SNP in Scotland) and the Lib Dems than to the Conservatives.
The average vote in 2015 for UKIP, in seats where they are not standing this election was around 4,700 votes. The equivalent figure for the Green party was 900 votes. Since UKIP and Green supporters do not transfer their allegiance to the same party, the net effect on the final outcome will be relatively small. So in most seats, it will not make a difference to the winner.
But in very closely contested seats, the absence of the Green or UKIP candidate could be enough to tip the balance between the two major parties. I have analysed in how many seats that could be a factor. The analysis is complicated by the fact that the set of very close seats is not fixed in stone. We know which seats were very marginal at the last election, but they are unlikely to be the marginal seats this time round. We can use polls to estimate which seats are close now, but the list will change every time a new poll comes out.
To get round this problem, I ran a set of different outcome scenarios. These scenarios spanned all the likely and less likely results ranging from a Conservative landslide to a majority Labour government. In each scenario, I worked out which seats were close and whether they would be changed by the absence of UKIP and/or the Green party. The change of seats can then be averaged over the various scenarios.
On average, the Conservatives gain nine seats overall from the absence of minor parties, mostly at the expense of Labour who lose seven seats. The Liberal Democrats and the SNP lose one seat each, on average.
So the net effect is not massive, but it is a noticeable boost for the Conservatives. This reinforces the view that the collapse of UKIP has caused an involuntary realignment (or consolidation) of the right-of-centre parties, which has been a big benefit to the Conservatives. The left-of-centre parties have not achieved anything similar, and the absence of 170 Green party candidates has made little difference, other than saving the Green party £85,000 in lost deposits.
In terms of the situation today, it is dangerous to appear precise, because the list of close seats will change by polling day. So there is a large health warning that these seats may not be the actual ones to swing because of missing UKIP candidates. But some seats will swing that way, and these are real-world examples of those types of seat.
On current opinion polls, there are four seats directly affected, which is below the average. There are three gains for the Conservatives:
And there is also one seat in Scotland which the Conservatives would have hoped to gain, but the absence of Green votes tips it back into the SNP camp:
And also notable is another Scottish seat where the absence of the Greens helps the SNP, but the absence of UKIP helps the Conservatives more. This one is held by the deputy leader of the Scottish National Party:
There are additionally another eight seats which the Conservatives were originally predicted to win with wafer-thin majorities, but which become a little safer because of the absence of UKIP. These include two Lib Dem seats of Norfolk North (Norman Lamb) and Carshalton and Wallington (Tom Brake).
Again, there is a health warning that the model does not include the incumbency effect or other local factors, so results in particular seats will vary from that shown.
But the overall pattern is clear. The absence of UKIP from many seats helps the Conservatives both by tipping the balance in some seats, and increasing their majority in others.