Referendum or Opinion Poll?

Brian Souter's privately funded "referendum" in Scotland on the repeal of Section 28/2A of the Local Government Act reached its conclusion this week in an inevitable flurry of recrimination about its significance, if any. The Scottish Executive, in dismissing it, described it as a "flawed opinion poll". This is entirely a misuse of the term.

What is an opinion poll? It is, simply, an attempt to measure the opinions of the public of a certain question by questioning a representative sample of the population. Note that key phrase, representative sample. A phone-in 'poll', where a self-selecting group volunteers to participate, is not an opinion poll, because it does not and cannot make any attempt to make its sample representative. (We prefer to call this sort of exercise a 'voodoo poll' - very occasionally it may justify itself as a marketing tool, but it has no validity at all as a measure of public opinion, even on the rare occasions when some effort is made to restrict multiple voting.) Similarly, a census, referendum or any other exercise which attempts to elicit an answer from the entire population is not an opinion poll because, although those given the opportunity to respond are representative, they are not a sample.

The Souter exercise, therefore, is clearly not "an opinion poll", flawed or otherwise. There have been opinion polls on the question, both in Scotland and across Great Britain, and they remain the best available measure of public opinion on Clause 28, even if a couple of months out of date. MORI Scotland Clause 28 poll, January 2000. They are also, of course, a great deal more convenient, quicker, and cheaper than Mr Souter's private "referendum".

So what is the point of holding referendums at all, when you can take opinion polls? Painful as it is to admit it, opinion polls of course have some disadvantages. First, because they only interview a sample, they are subject to a margin of error. Although in a properly conducted poll this margin should be small, and for almost all purposes for which the polls are legitimately used such a level of accuracy is quite sufficient, they are clearly not a substitute for a democratic vote when the electorate is entitled to expect an exact result; the improbability of the wrong side winning a decision by opinion poll would not excuse the theoretical possibility.

Second, and more significantly, the opportunity for participation in a referendum is itself a part of the point of holding a referendum, which is intended to confer or deny extra democratic legitimacy to a government proposal. Even if an opinion poll could be guaranteed to predict with absolute accuracy what the result of a referendum would be, it would not be a substitute for holding it, any more than it could be a substitute for holding a general election (on which topic see Isaac Asimov's short story "Franchise".)

But most significant of all, consider what happens before a referendum. There is a full scale election campaign lasting days or weeks. Both sides have an opportunity to put their case, and the public have leisure to consider it and to make up their minds. In Britain, at least, there would be spending restrictions, regulation of the television coverage to ensure a balanced debate, and the probability that even the most biased newspapers will make some concessions towards considering the opposing point of view. None of this is true when an opinion poll takes a snapshot of public opinion, even on an issue which is well established in the public mind and which many of the population hold strong views on. Nor has the debate on Section 28 in Scotland in recent months been on such balanced and considered terms.

So, in these terms, we can see that Mr Souter's exercise is not a referendum, either. To be purely pedantic, it couldn't be because it is not a referral of a decision to the people by the decision-making body - Mr Souter is purely a private citizen with no official standing. But even allowing the term to be stretched to include soundings of opinion that have no official status, it has lacked the balanced campaign period and opportunity for deliberation that characterises a genuine referendum in a democracy.

If it is neither a referendum nor an opinion poll, what has Mr Souter spent his money on? It should become obvious if we drop the attempt to think of the turnout and the result in percentage terms, and rely purely on the numbers. Almost 1.1 million Scots returned ballot papers indicating that they opposed repeal of the Clause. What Mr Souter has is a petition, a device with a long and honourable history in the British democratic tradition. Of course, a petition by secret ballot is a somewhat unusual idea, but it seems to make perfect sense. It has no executive power - the Scottish Parliament is entirely within its legal and constitutional rights to ignore it completely if it wishes. But since Mr Souter has more 'signatories' to his petition than the combined list vote at last year's elections of Labour and the Liberal Democrats, the two parties making up the Scottish Executive, ignoring it might be unwise, to say the least.

See previous commentary columns - Section 28 & Labour slumps in Ayr