The issue is as old as referendums themselves. What is the best way to ask the question? Is it to pose a proposition, and ask people to answer Yes or No – or to offer two (or more) alternatives and invite people to choose which they prefer? Britain’s government has set out its preferred wording for the coming vote on the UK’s relationship with the rest of Europe. It is a straightforward Yes/No question: ‘Should the UK remain a member of the European Union?’
As a pollster my instinct is to avoid Yes/No questions. There is a mountain of literature about acquiescence bias. Some people prefer to vote Yes, to avoid causing offence, especially on issues on which they don’t feel strongly. The process is sometimes subconscious. But the result is that slightly more people may reply Yes than they would to the same proposition asked in a different way. So, occasionally, more people will say that they like A when the question is, ‘do you like A’, than they would if asked, ‘do you prefer A or B?’
Not surprisingly, supporters of Brexit smell a rat. They suspect David Cameron of tilting the coming referendum in favour of a vote to stay in the EU. Maybe they are right about the Prime Minister’s motives. However, I doubt whether the fears of the No camp or any hopes of the Yes camp will be borne out by reality. This is because a referendum is different from an opinion poll. It is not a question sprung on an unsuspecting voter. It offers a choice at the end of a campaign in which both sides have the chance to make their case. By referendum day, people generally know which side they are on, regardless of how the question is worded.
That’s not just supposition. Consider the five major Yes/No referendums the UK has held over the past forty years. Only one has produced a decisive Yes majority – the two-to-one vote to remain in the Common Market (as it then was) in 1975. One other produced a narrow Yes majority, on Scottish devolution in 1979 – but the vote fell short of the numbers required for devolution to take effect. The other three Yes/No referendums produced No victories: on Welsh devolution (1979), changing the system for electing MPs (2011) and last year’s referendum on Scottish independence.
It’s a similar story when we look at Yes/No referendums in other countries. Consider these examples:
- In 1982 Spain’s socialists came to power. They had opposed the previous government’s decision to join NATO, and promised a referendum on the issue. In office, Felipe González, the new Prime Minister, campaigned for a vote to stay in NATO, though much of his party remained hostile. In 1986 the country voted Yes by 57-43 per cent to remaining in NATO.
- Twice, in 1980 and 1995, Quebec voted No to seceding from Canada, the first time by 60-40 per cent, the second by 51-49 per cent.
- In 1999, Australia voted No, by 54-46 per cent, to a plan to become a republic.
In the case of the Australian and second Quebec referendums, polls pointed to Yes majorities until late in the campaigns. Likewise, in Scotland in 1979, polls reported big Yes leads for much of the campaign: the gap narrowed dramatically in the last ten days.
It’s possible to scour the history of referendums round the world and find plenty of examples of both Yes and No victories, without detecting any sign of a consistent acquiescence bias of the kind that pollsters try to avoid.
There is, though a different pattern, which is common to the nine Yes/No UK and overseas referendums listed above. It is that none of them resulted in a change to the status quo. Only in Scotland in 1979 did more people vote for change rather than keeping things as they were, but by too small a margin for change to take effect. The two other Yes victories, in the UK in 1975 and Spain ten years later, were votes for the status quo, as were the six No victories.
That’s doesn’t mean it’s impossible for referendums to lead to change. Scotland voted by a big margin for devolution in 1997 – and that was NOT a Yes/No vote. (People were asked whether ‘agreed’ or ‘disagreed’ with the devolution plans.) But in that case, as with a number of countries that voted down the decades to join the EU, the referendum essentially ratified a broad consensus for a new way forward. In general, victory goes normally to those who are perceived to offer reassurance rather than risk. The best way to obtain a majority for change is to persuade voters that it is safer than the status quo.
In the end the precise question wording is largely irrelevant. The UK’s future relationship with the rest of Europe will depend mainly on which future looks safer and less hazardous to most voters: remaining a member of the EU, or the prospect of life outside it.