DEBUG: blog_post
Would Remainers game a three-choice vote?
13 Mar, 2019

Probably not

A worry about Remainers gaming a three-choice referendum might push us to a sub-optimal method of conducting one. London4Europe Committee member and former Home Office senior civil servant Michael Romberg writes.

There is substantial support for no-deal

December 2018 polling by Yougov for the ESRC Party Members Project asked about a choice between Brexit on Theresa May’s deal and Brexit on no-deal. Electors as a whole divided equally with 31% supporting each option; Conservative voters divided 51:33 in favour of no-deal; Conservative Party members divided 64:29 in favour of no-deal.

No-deal fills many people – including most MPs - with horror. But given the scale of support for no-deal, it seems likely that only a three choice referendum could settle the Brexit question. Otherwise the Brexit dream will be kept alive:  “If only the people had been offered the one true Brexit…”. We have to win the post-referendum period too.

The right way to structure a three-choice referendum

One problem with a three-choice question is that the way in which the question is structured, the order of the questions, the weighting of answers will determine the result. Economists would say that the right way to conduct a three-choice referendum is to work backwards from your final decision.

I decide to move home because my flat is too small. No matter how firm my intent, that decision is provisional until I have actually found somewhere else. I look at all the flats on offer and choose this one as the best available. Only then can I make a final decision between the best available flat and staying where I am.

So there should be two voting rounds held on separate days. Round 1 would choose the best available Brexit; Round 2 would set the Brexit champion against Remain. Economists call this – alas – Backward Induction, and it is set out elegantly in a blog by Professor Alan Winters of the UK Trade Policy Observatory and less elegantly in one of my blogs.

Only some Remainers would even wish to game the first round

An objection is that Remainers might game the first round by choosing what they see as the Brexit most likely to lose to Remain in the second round.

Well, some might. But others would vote for what they thought would be the best Brexit if Brexit there had to be. Others would not vote – “a choice between Brexits is nothing to do with me”, “both are awful”.

It would be too difficult to game the first round

In order to choose the Brexit that was most likely to lose to Remain you would need to know which that was. At the moment the polls are not clear which that is - depending on the alternative the Remain lead would be 44:28 or 45:35. That's too close to be sure.

Gaming is harder to do than it might seem. Think of those Conservatives who infiltrated the Labour Party’s leadership election to vote for the candidate who would make Labour unelectable. They voted for Jeremy Corbyn. I suggest that did not look such a clever move in the 2017 election campaign.

Possibly by the time of the first round the opinion polls would show absolutely clearly which Brexit option would be most likely to lose against Remain. Think how clear it would have to be to be certain that no campaigning could affect the outcome. Theresa May entered the 2017 campaign with a strong and stable double-digit lead over Labour.

If it was so very clear which Brexit proposal was best placed to beat Remain then Leavers might vote tactically in favour of it even if it was their second choice, thus neutralising any Remain gaming.

It would not matter if Remainers gamed the first round

First Past the Post encourages gaming (tactical voting). If you are a Green or Liberal Democrat there are only a few dozen seats where you are in with a chance of winning. So those parties’ supporters can either “waste” their vote or choose tactically to support a least worst candidate. If Liberal Democrats vote Conservative to keep a Labour Government out we do not accuse them of gaming the system.

More importantly, each Brexit option has real support amongst Leavers – that is why we would be having a three-way choice. Whichever way Remainers who were gaming the system voted they would be voting for what some Brexiters say is the best Brexit.


We can be sure that there would be a “betrayal” narrative if Leave lost the 2019 referendum. However there would be no basis for claiming that Remainers had gamed a two-stage vote to be a part of the betrayal narrative. The risk of gaming a vote would not justify a different method.





Articles on this page reflect the views of the author and not necessarily of London4Europe