Dealing with turnout is very challenging
One of the common criticisms of pollsters in recent years is that they have a tendency to herd particularly as we get closer to elections. Well for sure that’s one thing that isn’t happening this time. With just a week and one day to go the above chart shows the Brexit party lead in the most recent polls and as can be seen there is a huge gap between the figures from YouGov and those from ComRes just out this morning.
I think we ought to acknowledge that this is a very difficult election to poll because everything’s dependent on turnout and if there is differential turnout between the parties then that could have a huge impact.
One factor that people not involved in politics perhaps don’t often appreciate that the elections that see the most on the ground campaigning are for local councils.
The last time the Euro elections in the UK were held alone, without simultaneous elections taking place on the same day, was in 1999 and the turnout was 24%. This year’s election comes three weeks after the locals and I’m expecting turnout somewhere in the mid-thirties. My reasoning is that there is a much greater interest in these elections because Brexit has dominated domestic politics coverage in the UK for the last 3 years.
All the recent polls have shown turnout figures considerably higher than that we have historically associated with elections to the European Parliament. For instance the latest ComRes had 60% of its sample rating their likelihood to vote at between 8 to 10. My view is that it’s not going to happen and am betting on overall turnout in the 30-40 range,
Maybe in the final polls to be published before the election a week tomorrow we will see some convergence but maybe we won’t.