Why should I believe the opinion polls? The Scotland evidence.

Scotland held a referendum in September 2014 over their independence from the United Kingdom. An eventual “No” vote by 55% of the population was the outcome, but were the opinion polls any good at predicting this? And what does this tell us about our EU referendum vote – based on current polling?

I hope you find this article interesting. If so, please share with all your friends via social media!

“No” was predicted, but the undecideds all went one way

If the 18 polls carried out in September 2014, only 2 predicted a “Yes” vote. The remainder all predicted “No”. The polls were very close near the end of the campaign.

So why was the victory so much bigger than the polls predicted?

As with any opinion polls about voting intentions there are a number of undecided voters. And if you look at the figures for the polls (link here to excellent Wikipedia page) it is clear that almost all the undecided voters landed on the “No” vote when standing in the polling booth.

So what does this mean for the current opinion polls about the EU Referendum?

You’ll probably be aware that the latest polls for the EU referendum are much closer than those in the Scottish referendum. It’s bloomin’ close…

Looking over the last ten polls we’ve got 6 wins for Leave, 3 wins for Remain and one dead-heat.

But the undecided vote is massive. It looks like it’s averaging toward about 10%

How will these undecided voters act?

I’m going to draw a parallel between the “No” vote in Scotland and the “Remain” vote in the EU referendum. Both votes are for remaining in a union, and both votes are risk-averse in nature (i.e. not embracing the uncertainty of a brave new world…)

So my prediction is that most of the undecided voters will side with “Remain” on polling day.

If my prediction is correct, we can make adjustments to the polling data by assuming that all “Undecideds” will vote “Remain”. Following this through, the shape of the last ten polls change significantly.

Rather than 6 results supporting leave, it’s now 2. Rather than 3 results supporting Remain it’s now 8.

So what does this mean?

In my opinion, this evidence strongly supports a victory for Remain. If voters act in a similar way as they did in Scotland, then the result of the EU referendum will be for us to stay in the EU.

Before David Cameron pops the champagne, it might be worth considering some of the issues with this hypothesis.

  1. Scotland leaving the UK was a bigger decision. A new country would have been formed, and the underlying risk of uncertainty was higher. This will introduce bias, and perhaps a smaller number of undecideds will vote remain.
  2. The U.K. is roughly 10 times the size of Scotland, and has many more regions with different identities and goals. Simplistic extrapolation of the Scottish data may lead to erroneous hypothesises.
  3. The 2015 general election polls were almost universally wrong. The silent voters the pollsters missed could be missed again. Any theory using the polls as a baseline is therefore flawed.

In summary

Extrapolation of Scottish polls suggest victory for Remain camp due to undecideds converting to the perceived lower risk vote.

If you have found this article interesting or helpful, please share with your buddies on social media…

And good luck to both sides for Thursday…

Author: Oliver Gearing

Oliver Gearing is a beatboxing, acoustic looping, 38 year old father of 3. Sounds like an electrocuted version of Elbow

Leave a Reply

Your email address will not be published. Required fields are marked *