DEBUG: blog_post
Sam Gyimah's three-choice referendum
15 Jul, 2019

Almost the right idea

London4Europe Committee member and former Home Office civil servant Michael Romberg writes that the one pro-Remain pro-Referendum candidate in the Conservative leadership election was almost right. No-Deal, the Deal and Remain must all be on the ballot paper. But there is better way of structuring a multiple choice referendum: two stages. Round 1 chooses the best Brexit, Round 2 sets that one Brexit plan against Remain.


Our Chair, in his recent message, backed in broad terms the ideas put forward by Sam Gyimah in the Conservative leadership contest: Remain and a referendum. Unsurprisingly, Gyimah failed to obtain enough support to be nominated. His ideas live on in his own writing. We can continue to promote them while improving on his proposal.


No-Deal must be on the ballot paper

Sam Gyimah was right on that. No-Deal is a lousy option. It has been misrepresented by its proponents - the latest incarnation is called "A clean managed Brexit". That has been criticised by Professor Steve Peers, Dave Henig of the UK Trade Forum, Peter Foster (Europe Editor of the Daily Telegraph) and Ivan Rogers

But we have to face the fact: No-Deal is the preferred option for most Leave voters - 59% in May. No referendum could hope to have legitimacy and acceptance if the leading Leave option was not on the ballot paper. It would be like making the choice No-Deal or Norway, and saying to Remainers that that was close enough to what we wanted.

Moreover, if we wish the widest range of MPs to support a referendum we must give each of them a reason for backing it, including those who see No-Deal as the way forward.

We have to win the argument against No-Deal, not prevent its case being heard.


It is difficult to structure a multiple choice referendum

Sam Gyimah had put forward two options:

  • a transferable vote
  • two questions: first "Leave or Remain?". Then, if Leave: "Deal or No-Deal?".

Neither quite works.

Transferable votes tend towards second-best options that no-one actually wants. Note that Theresa May’s Deal is not a compromise but a hard Brexit. There is no compromise option available any longer (Norway?), if there ever was. 

Asking Leave or Remainas the first question invites all possible Leave votes to be added together. That would repeat the problems of 2016. That could be resolved by asking both questions on the same day so "Leave" would refer to the two options on the ballot paper. But then a voter who preferred Brexit 1 to Remain and Remain to Brexit 2 would have to guess the outcome of the second question before marking the ballot for question 1.


Fewer problems: "Backward Induction"

Only economists could have come up with such a bad name for such a good idea!

You work out what the final question is. Then you work backwards step by step to see what is needed to enable you to answer that final question.

Brexit is discretionary. Of the universe of possible Brexits we can leave on only one. So the final question is: shall we Leave the EU on this one Brexit plan or shall we Remain? In order to know what the best available Brexit plan is you need a prior voting round which considers two or more Brexit plans and chooses the best.

There are several advantages to this approach:

  • it brings real clarity to the decision-making process
  • it reflects how decisions are made. Let’s go to the cinema this evening! We both agree, though I wish to see a drama, you a romance. We look at what's showing, decide whether there is a film we might both wish to see, look at showtimes. It is only then that we finally decide whether to go ahead. If our look at the programme had shown that there was nothing we both wanted to see then we would not be committed by our earlier decision. But if "Cinema/ not cinema?" had been a binding decision at the outset we might end up stuck with seeing a film we did not wish to see.
  • it would be quite obviously different from 2016, not just a re-run
  • there could be more than two Brexit options in Round 1 - the problems of transferable votes are much reduced if all options tend in the same direction. That should therefore allow the widest degree of support from MPs who favour different Brexit options (Norway Plus, Customs Union)

It does mean going to the polls twice - I can already hear Brenda from Bristol. But the French do that routinely in run-off elections. People do care about Leave/Remain and Leavers care about the form of Brexit.

Some hold that Remainers might game the first round by voting for the least attractive Brexit. But it would be hard to know which that was. Each form of Brexit has some adherents amongst Leave voters.



I know they hate it. But we can continue to push the idea of a referendum to Conservative party members (and everyone else). In addition to the arguments of principle we can point to an option for how to conduct the referendum that would:

  • ensure that no-deal - and other Brexit options - were put before the electorate even though Parliament would block their direct implementation without a referendum; and
  • reconcile the party to its millions of pro-Remain voters.

We have brought together all the blogs that we and certain other organisations have published on a referendum in this compendium.



The London4Europe blogs page is edited by Nick Hopkinson, Vice-Chair. Articles on this page reflect the views of the author, not necessarily of London4Europe.