When do we want it? Now - The Guardian - 18th June, 2009

Dear Mr. Wills,

You say, "Plebiscites... offer the wealthy and powerful an opportunity to manipulate outcomes" and, if the vote is a straight yes-or-no, then that is indeed the case.  You continue, "That is what 20^th-century Europe teaches us."  I'm not sure if by this you mean the plebiscites of Hitler and Mussolini, the majority votes of Lenin and Stalin, or the referendums which the EU's Badinter Commission recommended for the former Yugoslavia; but I think it applies to all three categories.

The two-option majority vote has long been regarded as manipulable.  After all, in many instances, the question is the answer, and it's a pity that the 20^th century did not learn the lessons of the 19^th, when Napoleon started the rut.  In 1800, he re-imposed majority voting in the French Academy of Sciences, where they had been using a Borda Count.  The latter "is a unique method... to minimise the likelihood that a small group can successfully manipulate the outcome," (Professor Donald Saari).  Furthermore, it "is the best protection ever devised from the tyranny of the majority," (Professor Sir Michael Dummett).  In the same year, Napoleon held his first of three two-option referendums.

Given the furore over mps' expenses, many people are calling for changes to our system of governance: inter alia, some want a referendum on electoral reform.  The question, then, is what is the question?  pr-stv?  av+?  pr-list?  Or should the question be, first and foremost, what sort of referendum should we have?  A majority vote with just two options?  Or should we allow for some pluralism?

In 1992, New Zealand set up a commission which, having taken submissions, drew up a short list of five electoral systems.  They then held a multi-option ballot, with ams, av, fpp, mmp and pr-stv on the ballot paper.  The votes were counted in a variation of a two-round system, with the second round a majority vote between the winner of the first round and the status quo, mmp and fpp; some would say the ‘final' should have been between the winner and the runner-up, mmp and pr-stv.

What is true, however, is this.  In an obviously multi-option setting, any use of a two-option ballot is almost bound to be inaccurate.  It is as if the waiter in a restaurant asks me, "Do you want beef or cod?" when in fact I want an omelette.  Obviously, such a ‘beef-or-cod?' question is valid only for those who favour either one or the other, while those who fancy chicken, nut roast, or anything else, the question partially dis-empowers.