27 October 2011

The Scottish people will soon be given the chance to vote on withdrawal from the United Kingdom. The plan is to hold a multi-option referendum, not a straight yes/no choice. How will this work?

This week’s rebellion by Conservative backbenchers may have been thwarted, but a referendum on withdrawal from the Union is still likely to take place this Parliament. Not the European Union (though if treaty renegotiation is back on the agenda this cannot be ruled out), but the Union of England and Scotland, which has lasted a full three centuries and counting.

And just as the Tory Eurosceptics were proposing a three-option referendum, it appears that the SNP’s planned referendum in Scotland will not be a straight yes/no choice on independence but will instead offer voters a choice between independence, the status quo and some stronger version of devolution (dubbed “devo-max”).

The precise shape of these constititional options remains to be determined – even the status quo is likely to have changed by the time of the poll, assuming the current Scotland Bill passes in some form. But another important question concerns the mechanics of a multiple choice referendum: how do you determine which constitutional model has the backing of the people?

New Zealand offers an interesting model. Next month, a complex multi-option referendum on electoral reform takes place. On polling day, voters will be asked first whether they favour a switch from the current MMP system, and second, which of four alternatives they prefer. Subsequently, if there’s a majority in favour of change, a second referendum is expected to be held in which voters choose between the status quo and the most preferred of the four alternative systems. Could this approach work in Scotland? Probably not, unless all the parties put forward their own variant of devo-max. The SNP have also already ruled out a two-referendum approach.

Almost certainly, the issue will have to be settled in a single poll. In that case, one approach would be a simple “first past the post” vote, where voters select their preferred option, and the most popular is then implemented. But this could see the UK dissolved on the preference of a minority of voters (if independence gained 40% of the vote, with each of two “unionist” options gaining 30% each). Conversely, the status quo might triumph despite a desire for greater devolution, if the two decentralising options cancel each other out. The uncertainty – and potential controversy – surrounding such an approach makes this unsatisfactory for all sides.

If all options are to be presented side by side, then the alternative vote could work best, with voters ranking options in order of preference. This system (used in 1977 to take the slightly less historic decision on Australia’s choice of national anthem) would ensure that the winning option had gained a majority of votes either as first or second preferences, thereby increasing the legitimacy of the result.

Or, one might split the referendum into more than one question. This appears to be First Minister Alex Salmond’s plan, to ask voters first if they favour independence and then if they favour devo-max. But what will happen if both garner majority support?

In California, if two ballot propositions contradict one another, but both gain over 50%, the simple solution is that whichever option has gained the stronger mandate, in terms of vote share, is passed. Alternatively, one might look to Switzerland, home of direct democracy. There, if a referendum is proposed on a particular subject (for instance by one particular party or group), the majority in parliament can set out an alternative solution (the contre-projet) to the problem at hand. The ballot paper then contains three questions: a yes/no choice on each of the two proposals, and a tie-breaker, asking which of the two is preferred. The tie-breaker comes into effect if both options gather the necessary “double majority” (of voters and cantons).

Scotland itself holds an interesting precedent. The 1997 devolution referendum asked voters first whether they backed the creation of a Scottish Parliament, and second, subsidiary to question one, whether they agreed to the devolution of (very limited) tax-varying powers. Could this be a replicable model?

Perhaps. Specifically, a first question could ask whether Scots backed the transfer of a defined range of fiscal and social security powers, which would mark a move to devo-max. And then second, voters could be asked whether they also (assuming question one were passed) backed the transfer of all remaining UK-wide powers, including over foreign policy, defence and so on. An affirmative result here would then pave the way to independence (though further negotiation with the British government would be needed on the terms of divorce).

If this were the logic followed, then a 51% majority for independence would suffice to dissolve the UK, even if the devo-max option in question one had garnered 99% support. This was in fact the claim made by Alex Salmond this week, though unsurprisingly, the case is not accepted on all sides. Either way, what is clear is that even agreeing upon the referendum process will prove controversial, with the referendum campaign itself likely to be one of the most bitterly contested political decisions made in this country in many years.


I think a 3-option referendum is a great idea, it offers better insights into how the voters feel about the EU membership. Very insightful post!

Add new comment

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
By submitting this form, you accept the Mollom privacy policy.