A referendum in which voters put three options in order of preference is perfectly sensible, but STV is not the way to count it (your report, 27 March).
STV is a good system for choosing a set of three, four or more representatives, but not for a single choice. This is because it can lead to the elimination of an option that would command a majority in each possible two-way referendum.
For example
, consider the case where first preferences are: status quo 35 per cent, more powers 30 per cent, independence 35 per cent. Here STV leads to a run-off between the two extremes, despite the fact that the middle option is preferred to either alternative by 65 per cent to 35 per cent.
The best way to decide a multiple-choice referendum was pointed out more than 200 years ago, by the Marquis de Condorcet, a contemporary – and intellectual equal – of our own Adam Smith and David Hume.
He pointed out that such preference votes implicitly tell us the voters' preference in each possible two-way referendum. For example, if the number of voters that put A above B is greater than the number that put B above A, then we can deduce that A would beat B in a straight vote.
The winner is the option that commands a majority in each possible two-way referendum. When the options have a clear ordering, as in the present case, there should always be a clear winner.
If a three-way referendum is thought too complicated, the same comparisons can be achieved by a two-question referendum, just as we had in 1997.
Thus a new constitutional referendum could ask:
1, Should the parliament have more powers?
2, If it has more powers, should these be within the Union, or should Scotland be independent?
Either of these, equivalent, approaches would give an outcome that commands a majority when compared with each of the alternatives. This seems a pretty basic requirement if the outcome to be accepted by the public as legitimate.
(PROF) DENIS MOLLISON, Inveresk, Musselburgh, East Lothian
Commentators and some politicians claim it is only right that there should be a review of Holyrood's powers at this stage, after ten years of devolution. But it is only eight and two-thirds years since the Scottish parliament commenced.
Is it not more likely that the number ten fits in more suitably by reference to the number of months since the SNP assumed minority power? Is there anyone, even with only half an eye on the political scene, who believes there would have been a commission appointed on the lines proposed, had, say, first minister Jack McConnell led Labour to a 53 to 38 victory over the SNP? Considering the 2007 Labour manifesto proposed no changes to the devolution settlement, the answer must be "Not a snowball's ...".
D R MAYER, Thomson Crescent, Currie, Midlothian
On the subject of the so called Constitutional Commission, you say: "Sir Kenneth has to ensure the work of the commission is politically neutral, open and inclusive" (Leader, 26 March). It can never be politically neutral as it does not take into account the subject of independence. It can never be open as it will not allow anyone to discuss the option of independence, and thus on both these accounts it can never be inclusive, so it falls at all three hurdles, even before it has started.
ANDY ROSS, Tomonie, Fort William
The full article contains 588 words and appears in The Scotsman newspaper.
