DEBUG: blog_post
07 Nov, 2017

We will never get Leave voters to listen to our case while some Remainers call for the 2016 referendum just to be set aside. So Remainers need to drop the bad arguments that are used to justify that call. This article is the last in a series of five by Committee Member Michael Romberg showing why those arguments have no force.

“Only 37% voted for Brexit”

Or should we say that 65% of the electorate voted for Brexit? That figure comes from adding the 28% of the electorate who did not vote to the Leave vote (37% of total electorate). All we need to do is to assume that the non-voters had forecast correctly how the vote was going and so did not see the need to vote themselves.

It is just as reasonable as the view of those Remainers who go on about how Brexit only attracted 37% of the total electorate and the referendum result is therefore invalid because the non-voters must have been happy with the status quo.

What about the 2017 general election? 31% (15m people) of the electorate did not vote. That is more than the 14m who voted Conservative. Should we have no government, because that was the most popular vote? Should we count the non-voters as supporters of the status quo and give Theresa May back her majority?

Life does not work like that. It is a basic rule of voting that the views of non-voters do not count. The result was 52:48 for Leave.

“Only a Simple Majority Voted for Brexit”

Referenda will usually be about change. It is reasonable to say that those wanting a change should pass a higher hurdle: there are costs to change; there are recognisable risks and unforeseen risks to any change; those who are happy with the status quo will not be energised to vote. So the burden of proof should be with those wanting action. Many countries that have referenda require supermajorities for all or some, eg those calling for constitutional change. 

The arguments are not that persuasive. After all, once the question has been asked No-Change is an option just like Change, so why not treat them equally? Why should some votes in practice count for less than others? And why do we believe that the electorate is incapable of assessing the costs and risks of change in referenda when they routinely do it in general elections? 

We need to think moreover of a downside of a requirement for a threshold or a supermajority. No-Change voters who think it unlikely that the Change campaign will cross the threshold may stay at home rather than vote against. One result might be that the Change campaign do actually win a supermajority which they would not have done if complacent No-Change voters had gone to the booths. Another unfortunate outcome would be that the Change campaign might win the vote but not by enough to pass the threshold. That would engender a sense of grievance. 

"Each country in the UK should have a veto".

This is one of those proposals that people are putting forward because it would have given the desired result in June, rather than because it is a good idea.

Let’s start with the principle. The UK is not a confederation of sovereign states. In international law all states are equal. To some extent that makes sense. You may not invade another state no matter how much bigger or richer you are. To some extent it is daft – and does not happen: does anyone think that the voice of Tuvalu (Polynesia, population 10,000) is or should be equal with that of the USA in trade talks?

The EU of course is an organisation of sovereign states. And when it comes to treaty change, each state has an equal vote and a veto.

But the UK is a unitary state. We have organised devolution by territories which we sometimes call countries, but they are not sovereign states. Indeed, Scotland in the recent independence referendum decided against becoming a sovereign state.

EU membership exists at the UK level.

Our Parliament does not have a quadruple lock for legislation requiring majorities for MPs from each country.

Of course, we may organise our referenda however we like. We could have a rule that says that no referendum may pass unless Leicestershire votes in favour. That would be as good a rule as the quadruple lock that some people are now calling for.

And people calling for a quadruple lock are assuming a bias against change. But there could be a bias against no-change: “Only two of the four countries voted to keep things as they are so we should change”.

You need to ask, how would the quadruple lock be democratic? Look at the figures for parliamentary electorates (2015): England – 37,400,000 Scotland – 3,900,000 Wales – 2,200,000 Northern Ireland – 1,200,000 Then suppose that all of Great Britain and half of Northern Ireland voted in favour, while half of Northern Ireland plus 1 voted against. Under the quadruple lock the motion would fail. Yet the votes would have been 44,100,000 in favour 600,001 against. Would that be a fair outcome? How does it relate to the principle of one person one vote (which is worthless if votes are not of equal value)?

We must ask ourselves “would this rule be good and fair in all circumstances (or is it just that l would have liked the result it would have produced in June)”?   


The 2016 referendum was a valid way of reaching a decision. The nature of the decision has been misinterpreted.  Leave had no plan. So the 2016 referendum is a binding decision to produce a plan. We are not committed to implementing whatever plan is later produced. Once the terms of Brexit are known there should be a further referendum with the option to Remain.