In other words, on controversial matters, how best can we identify “the will of the people” (or the will, the collective opinion, of their representatives)? Paradoxical though it may sound, you cannot identify a majority opinion by a majority vote, not least because that opinion has to be identified earlier if it is to be already on the ballot paper. You may ratify, perhaps; but you cannot identify.

In fact, in many situations, a majority opinion identifies not so much the will of the people (or of parliament); rather, it identifies the will of the few who set the question. Little wonder, then, that majority voting has been used in party caucus, parliament and/or referendum, by such notables as Napoleon, Lenin, Mussolini, Hitler, Duvalier, Máo Zédōng, and Saddam Hussein. Invariably, on such occasions, the question is the answer, and the only dictator who couldn’t dictate properly was Augusto Pinochet in his third national plebiscite.[1]

Other politicians also like majority voting. Croatia’s Franjo Tudjman wanted Croatian independence so, in 1991, he set the question (I paraphrase) – “Are you Serb or Croat?” Thus, at a stroke, he as it were disenfranchised any mixed partners and others who might have wanted to vote for a compromise. In 1997, Tony Blair wanted the Welsh to want devolution so, despite the nationalists who wanted independence included on the ballot paper, the question was devolution, yes or no? It ‘won’ by less than 1 per cent. In Wales, nobody died. The subject in general, however, could hardly be more serious: the example of the binary 2014 two-option referendum in Scotland was used by those in favour of separatism in Ukraine.[2]

Furthermore, majority rule (which is based on majority voting) was a cause of the so-called ‘Troubles’ in Northern Ireland; a significant factor in the Balkans, where “all the wars in the former Yugoslavia started with a referendum,” (Oslobodjenje, 7.2.1999); a major part of the 1994 genocide in Rwanda, where the slogan used by the Interahamwe was “Rubanda Nyamwinshi” (“the majority people”); and, going back further, a cause of the tragedy of bolshevism, indeed, the word itself meant majoritarianism, and the Bolshevik was a member of the supposed majority, bolshinstvo, while the Menshevik was of the minority, menshinstvo.[3]

The collective will

The democratic process should not be a means by which the people (or parliament) only ratify the will of the so-called ‘democratic leaders’ – a phrase which may be as oxymoronic as that of ‘democratic centralism’. Instead, the people (and/or their representatives) should be entitled to participate in the entire democratic process, i.e., in choosing the options, in drawing up a short list, and then in voting on a number of options, usually, on complex matters, between four and six of them.

There are very few contentious topics which must be reduced to a choice of only two options. Indeed, on any such issue in a pluralist democracy, there will invariably be a plurality of options ‘on the table’. The constitutional referendum in Egypt did not have to be a stark yes-or-no vote on only one option; alas, they did use this divisive methodology so, like night follows day, they divided; and then they fought. The 2011 UK ballot on the electoral system did not have to be a choice of only two options – first-past-the-post (FPP) or the alternative vote (AV)? Rather, as in New Zealand in 1992, it could have been a choice of five options. And even when a debate definitely does consist of only two options – as in Sweden’s referendum of 1955, when the question was, “Which side of the road shall we drive on?” – there may be more than three ways of voting: left, right and blank.[4]

Now if there are only two options on the ballot paper, A and B, then for those who do want to participate, there are only two ways of voting. When there are more than two options, however, various voting methodologies are possible. They include single-preference systems like plurality voting (which is like FPP) and two-round voting (as used in French elections), as well as the three preferential systems: the Borda count (BC), the Condorcet rule, and AV.[5]

In preference voting, a choice of three options offers six ways of casting a full ballot:

A-B-C, A-C-B, B-A-C, B-C-A, C-A-B and C-B-A.

With four options, there are 24 possible sets of preferences. With five, it’s 120, and so on.

If one or other of these methodologies is a more accurate way of identifying the will of those voting, then it should surely be used instead of majority voting. If, furthermore, this more accurate methodology is non-majoritarian, then the very basis on which so many western democracies is based – majoritarianism, i.e., majority rule by majority vote – is undermined. The Borda methodology – for which the full name is the Modified Borda Count (MBC) – is just such; and the rules for an MBC cover not only the voting mechanism itself but also the debate which precedes the ballot and the analysis which follows.

Consensus voting

If and when a problem arises, a debate in the appropriate forum – in parliament, on the web, in an independent commission or wherever – shall be called, in the full knowledge of who will chair, who can participate in the debate, and who can vote.

All relevant options, which do not contravene an agreed norm like the UN Charter on Human Rights, shall be allowed ‘on the table’, on a dedicated web-site and, in summary perhaps, on a computer screen. Every option must be a complete package.

Participants can ask questions, seek clarifications, propose variations, suggest composites or whatever. The speaker or, better still, a team of consensors as they are called, maintain a full list of options, editing all into a common format, and ensuring that the list is balanced and covers the entire spectrum of debate. To take a practical example, if the UN Security Council had deployed such a procedure when debating Iraq in 2002, the UK/USA option could have been labelled option A; France and Germany, who did not like one particular clause, the phrase “serious consequences” in Article 13 of Resolution 1441, could have drafted their own wording in an otherwise unchanged option B; other countries might have had their ideas as well: options C, D and E, say.

If, at the end of the debate, a verbal consensus proved to be elusive, the chair could have called for a vote on these five options. Each of the 15 nations could then have cast its preferences for one or more of the options listed. The preferences cast would have been translated into points, and the option with the highest points total would have been the outcome. The rules for a five-option count are as follows.

If a nation casts only one preference and says nothing about the other options, it exercises 1 + 0 + 0 + 0 + 0 points.

If another country casts two preferences, it gives its favourite 2 points and its second choice 1 point, so to exercise 2 + 1 + 0 + 0 + 0 points.

And so on.

But the delegate which casts all five preferences exercises 5 + 4 + 3 + 2 + 1 points.

The mathematical incentive, therefore, is for every nation to cast a full ballot.[6] The democratic consequence is far more important: in effect, those who cast full ballots not only recognise the validity of others’ aspirations but, by implication, they also accept the overall democratic result. No one state votes against another; no one nation may veto another; instead, albeit in their order of preference, each recognises their neighbours’ aspirations, and all are obliged to come to an accommodation.

Some options in the count will get a score above the mean, others below. (The chances of a five-way tie, when 15 voters are each casting up to 21 points for the various options, is minimal.)

If the winning option is way out in front, it may be regarded as the collective will. If it is not so far ahead, maybe consensus would be a better term. If the margin of victory is even smaller, then best possible compromise might be a more appropriate description. And if the leading option is only just above the mean, then obviously, other options will have gained a similar level of support, in which case there is no consensus and the debate should be resumed.[7]

There is one further possibility: the two leading options may be, as it were, neck-and-neck. Well, in a five-option scenario, it is highly unlikely that all five options will all be totally mutually exclusive of all the other four. In this situation, therefore, the consensors shall consider the possibility of a composite, incorporating any compatible aspects of the runner-up option into the winning option.

This MBC – which is more in line with what M. Jean-Charles de Borda actually wanted when he first advocated the BC in 1784 – measures what is, in effect, the average opinion; and an average, of course, involves everybody, not just a majority. It is, therefore, more inclusive.


Power-sharing is a necessity is many jurisdictions, and the list gets longer and longer as the years go by: Belgium, Bosnia, Burundi, Cyprus, Iraq, Israel/Palestine, Kenya, Kosova, Libya, Northern Ireland, Syria, Zimbabwe, etc..

But even in ‘normal’ societies, democracy should be for everybody, (not just 50 per cent and a bit). So, in a representative democracy, not only should parliament represent all the people but, in like manner, its government should represent the entire parliament. Power-sharing should be the norm, not just in Switzerland and conflict zones, but everywhere.

Why, then, do politicians still use majority voting, an ancient and one of the most inaccurate measures of collective opinion ever invented? The answer is probably because it allows them all, both democrats and dictators, to determine the question, i.e., to control the debate. So why do so many political scientists and journalists etc. also support the use of this “blunt” instrument? Majority voting is ok if the question is not contentious – shall we now have a coffee break? Let’s have a show of hands, please – but when the subject is complex and/or when passions are high, the MBC is much more accurate and, therefore, much more democratic. It is also ethno-colour blind.

If it is to be used in conflict zones, in the UN Security Council and in conferences on climate change (where again, no one should have a veto), maybe we Greens should first start to use electronic preference voting ourselves.



[1]           Mugabe also lost a referendum, but he had made it non-binding.

[2]           The similarity was the binary nature of the questions. The main difference of course was that, while there was prior agreement between London and Edinburgh, there was nothing between Kiev and Sevastopol and/or Donetsk.

[3]           The vote took place in London in 1903. The Bolsheviks got 19 votes, the Mensheviks 17, and there were 3 abstentions. So in fact, neither side had a majority. Furthermore, when Russian elections were eventually held in 1917, an absolute majority was won by the Social Revolutionaries.

[4]           Accordingly, any committed democrat who on this question was actually indifferent could vote blank and thus, as it were, go with the flow; over 40,000 did exactly that.

[5]           Otherwise known as the single transferable vote, stv, instant run-off voting, irv, or preferential voting, pv, in the British Isles, North America and Australasia respectively

[6]           In an mbc on n-options, voters may cast m preferences, where n ≥ m ≥ 1; points shall be awarded to (1st, 2nd … mth) preferences cast according to the rule (m, m-1 … 1).

[7]           For any debate, standing orders should state just what levels of overall support qualify for which description, and, most importantly, what threshold is required for enactment.

Cookies on our website allow us to deliver better content by enhancing our understanding of what pages are visited. Data from cookies is stored anonymously and only shared with analytics partners in an anonymised form.

Find out more about our use of cookies in our privacy policy.