Forecasting the European Election in the East of England

You rarely see serious forecasts of the European elections on a regional (constituency) basis, mainly because the proportional voting system makes it much harder than “first past the post” systems. But I thought I’d take a look, spurred on by an enquiry via Twitter regarding any advice I could give on voting tactically against a particular party. You can probably guess which one.

So firstly, let’s look at the voting system we’re given. The East of England constituency covers over 4 million voters in Bedfordshire, Cambridgeshire, Essex, Hertfordshire, Norfolk and Suffolk. We each vote for our favoured party, and the votes result in seven MEPs being elected, in a way which ingeniously aims to reflect the proportion of votes given to each party. This is called the d’Hondt method, which sounds very Eurocratic, but dates back to Thomas Jefferson.

Each party nominates up to seven candidates (in case their party gets almost all of the votes!), in an order of preference. When the votes are in, a table is drawn up which shows how many candidates (if any) from each party are elected. The table shows the vote share gained by each party in the first column, then their vote share divided by 2 in the next, their vote share divided by 3 in the next, and so on. This is the table for the East of England after the last European Election in 2009:

The seven MEPs elected were simply the seven biggest shares in the table, i.e:

So you can see that the first three candidates on the Conservative list were elected, along with two from UKIP, one from the Liberal Democrats and one from Labour. The battleground is always for the last seat or two, which in this case saw the second UKIP candidate and the third Conservative candidate beating the first Green candidate. That’s why this time around, the Green Party is saying that if it can only increase its vote share from 9% to 10%, it should win one of the seats.

OK, so how’s it going to go this time? There don’t seem to be any reliable regional opinion polls, so we have to look at the national polls, and impose them on our local voting patterns. Firstly, let’s look at how the East of England vote compared to the national vote in 2009:

So we see that the East of England was more disposed towards the Conservatives and UKIP than the country as a whole, and less disposed towards Labour. Now let’s take an average of recent national opinion polls for 2014, and include the same regional bias:

Obviously this is very crude, but it’s the best I’ve got to go on. Now let’s put those votes into the proportional representation table, as above:

We see that there are two changes compared to 2009, with a third UKIP and a single Green candidate being elected at the expense of the third Conservative and the single Liberal Democrat candidate. The safe candidates are the first two Conservative and UKIP candidates, and the Labour candidate. The battleground, as ever, is for the last couple of seats, with the third UKIP and the Green candidates winning out over the second Labour and third Conservative candidates. The existing Liberal Democrat MEP is well off the pace, finishing not only behind those four, but also behind the fourth UKIP candidate.

If you wanted to vote tactically to stop the Green candidate, you’d probably be best off voting Labour. If you wanted to stop the third UKIP candidate, you’d probably be best off voting Labour, while encouraging non-Labour voters to vote Green.

I have no great expertise in the European elections, so please feel free to pick holes of any size in this analysis if you know more than I do.