By Dr Peter Emerson, de Borda Institute, author of The Punters’ Guide to Democracy (Springer, Heidelberg).
In the bad old days, people often resorted to violence: men on women, strong men on any weaker ones. Some thought a monarchy would be a good way of maintaining order, but that wasn’t too good either: nations were soon fighting each other, and some of them became tyrannies.
Then, about 2,500 years ago, the Greeks tried democracy. Initially, it was all very primitive of course: no women, no slaves, and the only voting system they had was binary. The logic was simple: when there isn’t unanimity, aim for a majority. Sounds OK, and it worked fairly well, not least because there were no political parties in those days; so Mr X could vote with Mr Y today, and against tomorrow, without either falling into groups of permanent antagonism.
The next step forward was in China, about 400 years later, when the emperor started to use majority voting amongst his ministers… and again, history relates, it worked quite well.
Now there are two types of binary voting: singletons like “Option X, yes or no?” (as in Brexit) and pairings, “Option X or option Y?” (as perhaps to be in the GFA: ‘NI to be either in the UK or in a united Ireland).
The first person to realise that majority voting has its limitations was the Roman, Pliny the Younger, in 105. The jury in a murder trial had three options on the table, A, B and C – acquittal, banishment, and capital punishment – but no option had a majority in its favour; this meant of course that every option had a majority against; (again: just like Brexit.) So Pliny suggested plurality voting, (which is similar to the first-past-the-post electoral system): you vote for one thing only, and the option with the highest score wins; not a majority perhaps, maybe just the largest minority. But – happy ending – the accused lived, albeit banished to an island.
Europe then went into the Dark Ages, and the first country to use plurality voting was China, in 1197, during the Jīn Dynasty. In a debate on their northern neighbour, Mongolia, there were three options on the table: war:alternating-policies:peace; and they voted, 5:33:46. Another happy ending: peace won. Alas, not for long; nine years later, a Mongolian assembly, a Khuriltai, elected Chinggis Khan and… you know the rest.
Some other countries followed, slowly: in 1894, New Zealand was the first to use multi-option voting in a referendum, and a few other countries have done the same since, like Finland and Sweden. Britain also did so but only once, in 1948, in a referendum in Newfoundland; the Irish have yet to even try.
So, back to the continent, the Dark Ages were over, and in 1268, Venice started to use approval voting: tick as many options as you like, and the one with the most ticks wins. In 1433 Cardinal Nicholas Cusanus proposed a preferential points system: with n options, a 1st preference gets n points, a 2nd choice gets (n-1), and so on, and the option with the most points is the winner; by historical accident, it’s called a Borda Count, BC. In 1770, M de Borda developed this into the Modified BC or MBC: in an n-option ballot, the voter who casts only m preferences gets only m points for a 1st preference, (m-1) for a 2nd choice, etc.
Next came the Condorcet rule, another preferential vote, and this identifies the option which wins the most pairings; is A more popular than B? is A > C? and is B > C. This was followed in 1821 by the Irish system (but British invention) of single transferable vote STV, which we all know from our elections. And then came the two-round system TRS, a plurality vote which, if nothing gets a majority, is followed by a majority vote between the two leading options. Of all these systems, only the MBC is non-majoritarian. At best, it can identify the option with the highest average preference, not the more preferred option of a majority, but the most preferred of everybody; it is inclusive, literally; it is egalitarian. Perfect, one would think, for a divided society.
* * * * *
So, Northern Ireland. When I arrived in Belfast (from Kenya) in 1975, I was often asked, “Are you ‘this’ or ‘that’?” But I’m either neither or both: I’m the child of an English Catholic mum, while dad was an Irish Protestant.
In 1985, Mikhail Gorbachev took over in Moscow; it marked the end of the Cold War, (or that one, anyway), whereupon Britain declared it no longer had a strategic interest in NI, and hence the (16 years, long overdue) Anglo-Irish Agreement. Ian Paisley shouted, nay screamed, “Ulster says, “NO!’” at the City Hall. One week later, six of us stood there in silence, with a banner: “We have got to say ‘yes’ to something.”
So what about an experiment in decision-making? Maybe we could use a non-binary preferential vote where people vote ‘yes’ to something, or ‘yes’ in their order of preference to a few things, but where NOBODY VOTES NO, Queens said a conference with unionists and republicans would be too dangerous. Corrymeela said too difficult. But the New Ireland Group under the late Dr. John Robb said, ‘t’would be good’… and hence the first of our People’s Conventions, on a Saturday in May 1986, an open public meeting in the Students’ Union.
Over 200 came, pretty well everyone from Sinn Féin to the political wing of the UDA, all except the DUP. They sat in one big, tiered circle; they paused for silence; they listened as the late John Hewitt read his poem, Anglo-Irish Accord, written especially for the occasion, and then they started to debate the constitutional status of NI.
Anything and everything – as long as it complied with the UN Charter on Human Rights – was allowed ‘on the table,’ and we soon had twelve options. Next, each was examined in separate workshops. And then, after lunch, back in plenary for the vote: all to cast one or more preferences… and NOBODY TO VOTE ‘NO’. Sure enough, we thus found the consensus of those present: “NI to have devolution and power-sharing, within a tripartite Belfast-Dublin-London agreement,” a mini GFA, just 12 years ahead of its time.
* * * * *
It was all so easy: preferences, points, add ’em up, and there’s the winner. Why hasn’t this been done before? I asked myself. “Do some research, my boy,” I was advised… and, ah ha, it had been done before, in 18th Century revolutionary France, when democracy had been hotly debated; where members of l’Académie des Sciences had looked across The Channel at a very binary Westminster, the western world’s only known democracy in those days, and they’d said, “Mon Dieu, c’est incroyable!” If you tried to get the average height of all the MPs or TDs by asking a binary question – “Are you tall or small?” – the answer would be wrong! In a similar fashion, you cannot identify the common good, the collective will, la volonté général, in a binary, “Are you left- or right-wing?”
As a result, l’Académie invented two multi-optional, preferential, decision-making voting procedures: Le Marquis de Condorcet proposed an analysis all the pairings; and Jean-Charles de Borda’s MBC, a NOBODY-VOTES-‘NO’ points system. Now if you were running the six-nations rugby tournament, you could use a knock-out system (majority voting), but the current league system is much fairer. In most years, the champion (or Condorcet winner), the team which wins the most matches (or pairings), is also the team with the best points difference (or MBC). Yes, both systems are very good, and in many sports leagues, the team which wins the most matches (pairings) is also the team with the best goal difference (points); l’Académie debated, and chose the Borda rule.
But, 1789, the revolution. L’Académie was changed to l’Institut Français. It had a new boss. He didn’t like this consensus stuff, so went back to majority voting. He then held a majority vote referendum, chose himself to be the only candidate, and thus he became l’empereur, Napoléon, on 99% plus of the vote. Je suis un ‘democratic dictator,’ he might have said. Actually, the first autocrat to ‘dictate properly’ and get 100% was an Irishman – in 1818, Bernardo O’Higgins became El Supremo in Chile. The extraordinary thing, however, is that today, in France, very few people know anything about Jean-Charles de Borda. Indeed, in 2006, in Coutances in Normandy, this Irishman spoke to a hundred or so French folk about a Frenchman they’d never heard of; it’s an Irish story, if nothing else.
There’s another extraordinary thing. In their wisdom, the authors of the Belfast Agreement decided that we the voters must be allowed to cast our preferences in elections – which is as it should be – PR-STV; but, in complete contrast, not in decision-making – which is not what it should be at all. When choosing our representatives, we can vote for men and women, Catholic and Protestant, whomsoever. But when it comes to decision-making, be it the MLAs in the Assembly or ourselves in any future referendum, questions shall be either/or, yes-or-no, for-or-against. It is a peace agreement which rules out compromise; yes, quite extraordinary. And ideas like a W-I-S-E federation, of Wales-Ireland-Scotland-England, are not even on the table!
Time for some rugby. Ireland beats England. England beats Wales. So which team is the best if Wales beats Ireland? It’s the same in majority voting, if we take pairings on three options, A, B and C, we might find that A is more popular than B, which we write as
A > B;
we may also find that B > C,
and that C > A, so
A > B > C > A > B > …….
and it goes round and round for ever. Le Marquis de Condorcet’s famous ‘paradox of binary voting.’ So that’s what can happen with pairings, while with singletons, as we saw, there might be a (Brexit-type) majority against everything!
So what’s the craic in politics? Imagine a committee of a dozen: one is the impartial non-voting chair; of the 11, 5 want A, 4 choose B and 2 like C. So there are majorities of 6, 7 and 9 against A, B and C respectively.
And if the
5 have 1st-2nd-3rd preferences of A-B-C
4 ” B-C-A, and
2 ” C-A-B,
then
A > B 7:4
B > C 9:2, and
C > A 6:5.
If they are debating these three options, the debate and then the votes will be:
choose the more popular amendment; and
adopt or reject it. .
So, if A is the motion while B and C are the amendments, it’s like this:
(B v C) v A,
in which case,
(B v C) v A = B v A = A;
but if B were the motion, while C and A were the amendments,
(C v A) v B = C v B = B,
or again, it could be
(A v B) v C = A v C = C.
In other words, with binary voting, if there’s a paradox, the chair who chooses the order of voting can get any outcome she wants. This is terrible. And if there isn’t a paradox, she can suggest other options, create a paradox, adjust the order of voting, and again, get whatever outcome she wants! This is divide and rule, democratic divide and rule.
No wonder the position of chair is so important. And no wonder those who are chairpersons don’t want to even talk about preferential decision-making, and compromise, and all that. Politics, they say, is the art of compromise. But, on the whole, politicians ignore the science. At a recent meeting with the Assembly and Executive Review Committee with the de Borda Institute, no-one wanted to talk about let alone question majority voting.
To a large extent, the Troubles were caused by majority rule. But no-one wants to talk about it. Again, it’s quite extraordinary. Similarly, in the media and academia, far too many professionals just blindly accept binary voting, as if it is not only democratic, and almost as if it is the only methodology available, just as it was 2,000 years ago. Everything else has been modernised. Lots of new devices have been invented. Quite a few multi-option voting procedures have been devised. But along with the spade, majority voting remains unchanged. And it is ubiquitous. It is even in Article 97 of the Constitution of North Korea.
What should happen, of course, is like this: everything is allowed, not only ‘on the table’ but also on the (perhaps short-listed) ballot paper. In the above example of A, B and C, the 11 members could debate all three options in turn; and then take just one preferential MBC vote, in which case option B wins on a score of 24, while option A has 23 and C gets 19. It’s close; but it’s clear. It’s transparent. It’s unambiguous. The maths can get a bit cumbersome, sometimes, but that’s what computers are for.
Accordingly, when a multi-party chamber like the Assembly is debating the budget, for example, let every party (which so wishes) put their proposal on the table. A big party might want to go solo, that or cooperate with a smaller party; while smaller parties might opt to team up with one another; so in our Assembly, there might be five options in contention; in the Dutch parliament, with 15 parties represented, there would probably be up to eight proposals. Debate each in turn; allow any proposal to be altered, or even composited with another, but only if the proposer(s) agree.
Granted, it will take a little time. Better that, surely, than the Brexit fiasco: a vote, followed by a debate (democracy normally works the other way round)… what’s more, that debate is still on-going.
* * * * *
The New Ireland Group first used electronic preferential decision-making in another People’s Convention in 1991, when Michael D was our guest speaker. Again, the participants were cross-community. They cast their preferences; they waited for a nano-second or few; and then, on the big screen, up popped the answer: their collective will.
Today, the technology is even better, and preferential decision-making can be done with everyone using their mobile phones. It works. Just this last year alone, I’ve used it in universities in Ukraine, Mongolia, Taiwan and China; while in previous years, I’ve lectured on the MBC in academic institutes and/or conferences in Austria, Bosnia, Canada, England, Germany, Ireland, Scotland, South Africa, Spain and the US… but hardly at all in our very binary NI.
Conclusion
“Shall we drive on the right- or left-hand side?” is a duality. So too is the query, “Shall we have red or white wine for dinner?”
But you can’t have a vote on “Yīn or Yáng?” – it’s not a duality; it’s almost a unity. The 1991 referendum in Croatia – in effect, “Are you Serb or Croat?” – was also not a duality; a question with that implication should never have been posed; it started a war. So too the question – “Are you Russian or Ukrainian?” – (is now, perhaps, but) was not a duality; it led to yet another war.
Brexit was another nonsense. “The UK in the EU?” was one of four options; we could just as easily have been in the EEA, the CU (Customs Union) or WTO; alas, we had only one singleton on the EU, and 52% said they didn’t want it; but we still don’t know what they actually wanted!
That vote should have been multi-optional, as stated in my press release of February 2016, four months before the ballot. Sadly, the press ignored it; indeed, the media rarely discuss voting procedures for decision-making. Academia is almost as bad.
And as for our politicians? Well, in 2004, I managed to get a consensus in Belfast City Council. The question was, would these’uns like to adopt a methodology which would facilitate consensus decision-making with those’uns. And, yes, I achieved what some might have thought was impossible, a consensus. Neither side wanted to be in consensus with the other side, so there was a consensus… against consensus.
Further info
Find out more about the de Borda Institute at deborda.org.