How to Solve the Scottish Referendum Question Controversy

Scotland’s upcoming independence referendum has been in the news in Britain. The Scottish government wants to hold the referendum in 2014, but UK Prime Minister David Cameron has said that Westminster holds ultimate control over the wording and timing of any legally binding referendum and wants to hold the referendum sooner.

Another point of contention is whether the referendum question should include two or three options. The SNP government in Scotland is open to a three-question (status quo, independence, or “devo max“) referendum, while the Conservative-Liberal Democrat coalition government in the UK wants a two-question (in or out) referendum. The apparent worry from Westminster is that a three-option referendum could split the unionist vote and allow independence to win with a bare plurality (say, 40% for independence and 30% each for status quo and devo max). Here is a debate among British political prognosticators about what will happen.

The solution to the problem is simple: hold a three-question referendum by Condorcet voting. While I generally favor approval voting for single-winner political elections, this referendum is a perfect case where Condorcet voting should yield the best result. Allow each voter to rank all three choices, independence, no change in Scotland’s constitutional status, and full fiscal autonomy, and then count up which option defeats the other two by a majority. That option becomes law.

Here’s an example. Suppose 25% of the population ranks the options this way: Status quo (SQ)>Fiscal autonomy (F)>Independence (I); 5% ranks them SQ>I>F; 10% ranks them I>SQ>F; 25% ranks them I>F>SQ; 25% ranks them F>SQ>I; and 10% ranks them F>I>SQ.

We then compare each option to each of the others. Status Quo beats Independence 55-45%; Fiscal Autonomy beats Status Quo 60-40%; and Fiscal Autonomy beats Independence 60-40%. Even though Fiscal Autonomy gets only 35% of first-place votes (identical to Independence), it emerges as the clear majority consensus winner.

Condorcet or pairwise majority voting therefore avoids the splitting-the-vote problem and is more likely to select a consensus winner than is plurality rule. Of course, a Condorcet winner does not always emerge; sometimes there is a cycle. If that is the case, the referendum law could state that the status quo wins by default.

A Condorcet election could be win-win for all sides involved. If devo max wins, as it probably would, the SNP can take credit for a massive constitutional change that satisfies long-held Scottish aspirations, while still campaigning for ultimate independence; the majority of Scottish voters will get the compromise alternative that appears to satisfy them best; English taxpayers would benefit from no longer having to subsidize Scotland; Scotland’s economy would probably benefit in the long term from having to reorient itself toward private enterprise; and Simon Jenkins makes an excellent case for why devo max would be a political winner for Cameron’s Conservatives.