Colin Irwin Mon 04 Feb 2019 updated: Mon 11 Feb 2019
In my series of blogs on Brexit I have suggested that Brexit needed a peace poll (1), that much of the polling done on Brexit was partisan and misleading (2), and that as Brexit was creating deep divisions in UK society pollsters should use conflict resolution best practice to analyse their data (3). With this point in mind I have completed a Brexit peace poll pilot to illustrate how this can be done.
Firstly the questions in a peace poll should be agreed and drafted with the cooperation of the parties to the conflict. In this case that should be the Parliamentarians elected to the House of Commons. But for the purposes of this pilot I have simply taken the relevant items from the House of Commons Order Paper No. 239 Part 1 that lists the Governments European Union (Withdrawal) Act, and all the amendments proposed by Parliamentarians (page 26-38) for selection by the Speaker on Tuesday the 29th of January.
From a conflict resolution/negotiations perspective this Act and amendments can be loosely characterised as being ‘substantive’ elements of an agreement or ‘procedural’ elements for getting to an agreement. Using Google Surveys I was able to test nine solutions for resolving Brexit against each other, with a tenth question asking the informant if they would vote ‘leave’ or ‘remain’ if a referendum was held today (Table 1 in pdf below). I would then be able to compare the opinions of Leavers and Remainers on these issues, and see if a compromise could be found anywhere, that they might be able to agree to. The same was done for nine procedural issues (Table 2 in pdf below).
From Table 1 we can see Leavers do not want to stay in the European Union at 79.9 per cent ‘unacceptable’, while Remainers do not want to leave the EU without an agreement at 71.8 per cent ‘unacceptable’, and Leavers do not want a referendum to leave or remain in the European Union at 81.8 per cent ‘unacceptable’. So nothing to agree to there at this time. But if we take a look at the other end of this five point scale, at what Remainers and Leavers consider to be ‘essential’ (Table 3 in pdf below) then we get a slightly different picture. The number one priority for Remainers is ‘a permanent customs union for trade with the EU, strong relationship with the single market, shared institutions and alignment on rights and standards’ at 47.6 per cent ‘essential’. The same item is third on the Leavers list at 17.5 per cent ‘essential’ but significantly it is only 20.2 per cent ‘unacceptable’ so perhaps something can be done with this.
Other options include a Canada-style deal and a Norway-style deal and they are possibly ‘doable’ but they presently require a Northern Ireland backstop which Leavers want removed at 45.4 per cent ‘essential’ while Remainers consider it ‘unacceptable’ at 26.8 per cent. On the other hand the ‘permanent customs union’ approach does not need a backstop so perhaps this is ‘the lesser of the evils’ in this case given its otherwise more general ‘acceptance’ by Leavers’ at 30.8 per cent.
With regards to the procedural issues (Table 4 in pdf below) it is interesting to note that Remainers do want them ranging from a high of 45.5 per cent ‘essential’ for a vote on any deal agreed to by MPs, to 42.7 per cent for MPs to vote on various deals (an Indicative Vote) to ‘ruling out a no-deal scenario and preparing for a People’s Vote with an option to remain in the European Union’ at 41.9 per cent ‘essential’. When phrased in this way Leavers consider this form of referendum to be 55.6 per cent ‘desirable’ so if the Brexit process has to go ‘down this road’ then perhaps this is the way to go. Interestingly, but not surprisingly, Leavers are far less enthusiastic about the various procedural options on offer as they want Brexit and presently it is the law.
This Google survey of 100 interviews was collected between February 2 and February 4. It is only a pilot costing a very modest £160. Clearly a larger survey is needed with input from the Parliamentarians who wrote the draft law and amendments tested here. I am a little concerned that my “Remain’ or ‘Leave’ question got such a high ‘Remain’ response so I do not think that result or the general result for the UK population as a whole should be cited here. But comparing the opinions of Remainers and Leavers as ‘indicative’ of what is happening to opinion is valid and I doubt if a more thorough survey would come to conclusions very different to the ones I have drawn here.
Although the pilot only has an N=100 sample I am used to working with small samples around the world as I am generally working on conflicts where such samples are often very hard to get. The thing then is to know what one can draw conclusions from and what one can not. With this point in mind although the overall sample may not be as good as we would like by taking out the most polarised groups (in Northern Ireland Protestants and Catholics for example and in Brexit Britain Remainers and Leavers) we can compare the differences between these two groups with some certainty.
I was thinking of running the pilot again for political party breakdown to compare Labour and Conservative supporters. But the results would not be so definitive in this case as some Conservatives are Remainers and some Labour Party supporters are Leavers. So to do this I think we really do need a much better sample.
Also there is a problem with the Google Programme. I had to include “I prefer not to answer’ in every question up front so the 'No Answers' are a bit high. It is possible that quite a lot of Leavers said they ‘Prefer not to answer’ in question one and that is why the Leave result is a bit low. This problem can be overcome by using ’stock’ questions on this issue that have been tested by various polling companies and that are know to work well.
So I think the poll should be run with a better sample and questions and that comparisons should be made between political parties as well as Remainers and Leavers. Inevitably the Conservatives and Labour Party supporters will be closer on all the issues tested here than the Remainers and Leavers.
Finally I should add that the level of ‘unacceptable’ for Protestants for the Power Sharing option that became the Belfast Agreement was 52 per cent (Table 5). But I suspect that Conservative ‘unacceptable’ for a permanent customs union would be less than 20 per cent so resolving Brexit is a ‘walk in the park’ compared to doing the Belfast Agreement! And the other options such as a Norway-style deal are almost as equally acceptable. But I did not want to say too much about this from the pilot as these similarities are all within the margins of error. So someone really should run these polls again to the best possible polling standards with a bigger sample!
The results for the Brexit Substance Pilot poll can be viewed here:
And the results for the Brexit Process Pilot poll here:
And the data files for both polls are available here:
Finally a free book on peace polling is available here: