Brexit Institute News

A Better Measure of the Will of the People

A Better Measure of the Will of the People

Peter Emerson (de Borda Institute)

In the Brexit referendum of June 2016, a majority of 51.9% of UK voters opted to leave the EU. However, this vote was in effect a vote of ‘no’ to EU membership rather than a positive ‘yes’ to any specific alternative. Since then UK politics have been consumed in an argument over the meaning of that ‘no’.

Majority voting can indeed be hopelessly inaccurate. Briefly, you might be able to find the collective will if, in the vote, (nearly) everybody says what they want. But you cannot find it if some voters say only what they do not want. Therefore, and because it was basically a ‘yes-or-no?’ question, the Brexit referendum did not and could not identify the will of the people. Little wonder, then, that post-referendum, no-one knows what ‘no’ means.

If and when a subject is complex, as the UK’s relationship with the EU obviously is, people will have many different opinions. The only way they will all be able to say what they want, therefore, is via a multi-option ballot, the sort which was used in 1992 in New Zealand: a five-option referendum on their electoral system; or in Guam in 1982, a six-option plebiscite on their constitutional status. (See the author’s Defining Democracy, 2012, p.164.) Preferential voting offers a way out of the impasse.

From Majority Voting to Preferential Voting

Majority voting was first used by the (male and rich) citizens in ancient Greece and the (equally male but richer) ministers of the Imperial Court in China’s Former Han Dynasty. When the subject was complex, however, such binary votes were seen to be inadequate. Thus, in CE 105, Pliny the Younger suggested a form of plurality voting in which there should be a plurality of options, with the voters able to choose any one of them. The next person to think seriously about these matters came after the Dark Ages: in 1199, Ramón Llull proposed preferential voting; doubtless, given last year’s binary referendum in Catalonia, this Spanish Catalan is now rolling in his grave. The next ‘inventor’ was Nicholas Cusanus, writing in 1435.

However, the first analysis of preferential points voting was done by Jean-Charles de Borda in 1770. These were heady times in France. L’Académie des Sciences knew that l’ancien régime was coming to an end; they realised that la volonté général (the general will) could not be identified in a majority vote; so one member, de Borda, suggested preferential points voting while another, Le Marquis de Condorcet, proposed a comparative system. In 1784, L’Académie adopted the Borda Count (BC) and it worked quite well. Then came the revolution. Shortly afterwards, l’Académie had a new president, who decided that he’d had enough of all this consensus nonsense, and they should go back to majority voting. He, by the way, was one Napoléon Bonaparte.

Thereupon the science of social choice vanished for another century. The next person to talk about it was the Englishman, Charles Dodgson. He should really have written a book about it but no; he wrote Alice in Wonderland instead. He was, of course, Lewis Carroll. In 1942, a Scot, Duncan Black did write a book. In his introduction, he said he did not plan to summarise the history of social choice science, because he did not know there was one. Alas, in the wake of Scotland’s recent binary referendum, his corpse is also gyrating.

The most recent ‘inventor’ of the Modified Borda Count (MBC), is Irish. Things get invented when things go wrong, and in Northern Ireland, majority rule had gone seriously wrong. So the current author proposed a preferential points system, which he put to the test in a public meeting of over 200 persons, in Belfast, in 1986, still eight years before the IRA cease-fire. (The New Ireland Group’s People’s Convention, held in Queen’s Students’ Union, six months after the signing of the Anglo-Irish Agreement; a full report is available in the Linenhall Library.) Over 200 persons were present, representing groups ranging from Sinn Féin to the Ulster Defence Association, all except the DUP. Participants could propose any constitutional arrangement they liked (as long as it complied with the UN Charter of Human Rights); they finished up with ten options; they cast their preferences; and thus they found their consensus: ‘NI to have devolution and power-sharing within a joint Belfast-Dublin-London tripartite agreement.’

A Preferential ‘People’s Vote’

But back to Brexit. In February 2016, this Institute said in a press release that if the question is to be ‘yes-or-no?’ (‘remain-or-leave’?), then the answer will be ‘no’. Instead, we argued for a three-option ballot: voters should choose between EU membership, partial membership through the European Economic Area (EEA), or full independence trading on World Trade Organization (WTO) terms. (See Press Release No 7 here.) Given that the ‘EU or no’ question was 48% to 52%, the outcome of an ‘EU or EEA or WTO?’ ballot might well have been 48 to 26 and 26%, which would have meant that ‘remain’ was the winner. In other words, the choice of voting system can be crucial.

Sadly, the press release was ignored; the June 2016 referendum was held; campaigning was divisive; and Britain is now more divided than ever.

So what now? Well, logically, given that the people rejected the David Cameron offer, they should be offered whatever deal is eventually struck between the UK and the EU based on Theresa May’s Chequers proposal or whatever comes out of the latest negotiations. If it is to be a ‘yes-or-no?’ type question, however, it might produce yet another ‘no’, and the process could continue ad nauseam.

So the vote should be multi-optional, something like: “Do you want the UK…

  1. under the WTO terms; or
  2. under the EU-UK deal based on the Chequers proposal; or
  3. to ‘remain’ in the EU?

England is now a deeply divided country, and tensions are high. Therefore the resolution of this matter, be it in parliament and/or a second referendum, must be seen to be fair.

A one-option ballot – ‘a) yes-or-no?’ ‘b) yes-or-no?’ or ‘c) yes-or-no?’ – or a two-option ballot – ‘a) or b)?’ or ‘b) or c)?’ – would be inadequate and inaccurate. The best answer would be a preferential ballot under an inclusive voting procedure, the MBC, so to identify the option with the highest average preference. Because an average involves everyone. This would be the most accurate measure of the will of the people.


Peter Emerson is Director of the de Borda Institute, Belfast (