Polling & Predictions & Eggnog

Quite a short blog tonight on why it is going to be too hard to call this election at this early stage but not to give up on the Polls you will read over the course of the next six weeks.

Polls are only a snapshot of the people surveyed at the time. They are not a prediction of what will happen on Election Day ! There has been plenty of criticism of the Polls in the last few elections when they appear to have got it wrong. In fact they have been generally ok and some have been good!

Perhaps 2017 is a good example. Many Labour supporters point out the Polls had Labour miles behind when May called the election and as the Poll of Polls below shows that gap by the end was narrowed to a couple of points. Instead of this showing the Polls were wrong what it does show is the Polls reflecting the change in mood as the election progressed and quite a few pollsters were about right in that last week. They saw it before the Tories did. There is always a margin of error of +/_ 2-3 points anyway.

2017 Poll of Polls

Remember Labour lost this election and returned with only a handful more seats than when they lost under Gordon Brown in 2010. If you believe many on the left of the party you would almost be led to believe they won because of the amazing closing of the gap by Polling day. I will return to why this seemed to be the case and why it would be dangerous for activists to assume it will just happen like this again.

My good friend Andrew Hawkins at ComRes writes a much better blog about why it is so hard to predict much at this election here https://www.comresglobal.com/the-hardest-election-to-poll/

As Andrew and others point out there are no longer just the two parties to take into account this time round. In 2017 we returned to largely two party politics with the Tories and Labour sweeping up 80% of the votes. If we look at the pattern of voting intentions in 2019 we can see this is vastly reduced by the entry of the Brexit Party and the return of the Lib Dems as a force – both because of Brexit!

This graph shows why it is such a gamble for both Labour and the Tories To hold this election. They are both being hurt by the Brexit and Lib Dems. The other major risk is simply that voters are increasingly motivated by Brexit for their allegiances at the moment and not their old party loyalties. This was breaking down over time anyway but Brexit has accelerated this process. So the Tories hope to pick up some heavy Leave Voting Labour seats but will expect to lose the gains they made is Scotland. Labour will probably lose some seats to the Lib Dems. At the moment on these figures it is hard to see where they make the 50-90 gains they need to make to have an overall majority.

Usually we use these polls and then extrapolate to give our selves a uniform swing and national prediction. I don’t think we can do this at this election. I am sure we will have to do seat by seat analysis to get to the bigger picture.

This is where I will try to work out what might happen in the Midland marginals – the ones I know best and specifically what might happen to the 10 seats in Leicestershire. I am working on the basis that the only potential marginal will be Loughborough – a seat that was 50/50 at the referendum and as a bellwether seat has been held by the largest party or government for quite some time!

When all the candidates are known and we have some better national polling data I will have a better idea of what the result *might* be on polling day!

Finally some people always dismiss the polls they don’t like. First polls are an art and a science. I wish I had a £1 for every time I heard somebody say – well they never asked me! They won’t. On a sample of just over 1000 people from across the country with weighting and other measure the Polls predict the likely views of people on the day of asking. Second those on the Left who always dismiss anything by Yougov because ‘it’s owned by a Tory’ are I am afraid wrong and it’s a pointless argument anyway. No reputable polling company destroys its reputation by being wrong. Political polling is a minute fraction of their work.

We are fortunate that if you want to fully understand polling we have a former constituent and good friend Matt Singh who is an expert. if you get the chance follow Matt on twitter @MattSingh or find him at his website http://www.ncpolitics.uk

And the Eggnog. As Andrew Hawkins points out above – you need to have drunk plenty by now to be daft enough to make a prediction !

One Reply to “Polling & Predictions & Eggnog”

  1. Speaking purely as a remainer I wish those opposed to Brexit whether it is via Labour’s Lexit followed by a referendum (people’s vote) or the LibDem’s revoking of article 50 would not tear into each other and unite to oppose the Conservatives. We must Abandon tribal loyalties and vote tactically .

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s