Referendums hover over British politics like a thunderous sky, shaping and reshaping debates, dividing parties and then artificially unifying them, changing history and yet never quite doing so. Referendums are a device that leaders cling to for comfort before they threaten to destroy them. Considering there have only ever been two UK referendums, the plebiscite feels like a strangely ubiquitous device, the mere prospect of one causing as much excitement as fear.
At least we know there will be one in Scotland. The formal announcement of it is a historic event, although whether the outcome will change very much is a different matter. Polls suggest that those who argue in favour of the union will win and probably by a substantial margin, but this is a campaign that will last longer than the seemingly never-ending presidential contests in the US. Predictions at this point are premature to say the least.
There is a widespread assumption that the SNP has been outmanoeuvred by David Cameron in agreeing to a single question on independence without the fudge of a second question on so-called devo max, ie more powers for Scotland within the UK. In reality the SNP leadership had no choice. Cameron realised that Alex Salmond's bargaining strength was limited as the SNP is not itself an advocate of devo max. "We're not going to co-operate because we cannot get a second question on a policy that we are opposed to" is hardly a credible position when the SNP is getting the power to hold a referendum and at a time of its choosing.
The timing of a referendum often influences its outcome, one of many reasons why referendums are a hopelessly flawed method for making decisions. In this case the future of the UK will depend partly on whether it seems likely by the autumn of 2014 that the Conservatives are on course to win a general election the following year. Such a prospect would be quite a card for Salmond to play in a country where the Conservatives are close to non-existent.
If he cannot play that card – and even if he can – he may well be forced to hold a referendum knowing that he will lose. The greatest political conjuror of recent times will have run out of tricks. I had wondered whether he would find a way of not holding it if the polls looked bleak. This is what leaders tend to do if they fear they will lose a referendum, which is another fatal in the system. On the whole, leaders only offer referendums if they are confident of winning.
In this case, Salmond had no choice. His party would have never forgiven him if he had sought to keep all options open any longer. But as with all referendums, a defeated leader has get-out clauses, another fatal flaw. Already I can hear Salmond state in the aftermath of defeat that as there was no second question about devo max, he will begin negotiations to acquire greater powers. When I was in Edinburgh recently, senior SNP figures told me that if they were to lose the referendum they expected calls for independence to start up again within a few years. I am sure they are right.
Look at the evidence from the UK referendum on Britain's continued membership of the EU, or Common Market as it was in 1975. The referendum was held in the summer of that year, with an emphatic victory for the Yes vote. By 1981, and in opposition, the Labour party was pledged to leaving the EU without a referendum, a policy that formed part of its 1983 manifesto. Yet Labour had been the governing party when the referendum was contested.
Indeed, the obsession with a referendum on Europe has never subsided, but has switched from Labour to the Conservatives. In a new twist, it is probable that all three main party leaders in the UK will emerge from the exhaustion of the poll in Scotland to pledge a referendum on Europe. As an added act of perversity all of them will not want to make such a pledge and will dread the consequences of holding the wretched poll if they win the next general election.
David Cameron will make the commitment to hold his party together, counter the threat from Ukip and keep the Tory-supporting newspapers as enthusiastic for his cause as possible. But he knows the dangers of an in-or-out referendum. The promise of one will keep his party almost contented, but the holding of one could split it fatally. If there were to be a majority in favour for withdrawal, he would face a crisis of unfathomable proportions especially if he had led the campaign to stay in. No wonder he keeps vague the precise question he might pose, and indeed leaves somewhat ambiguous the question of whether he will hold one at all.
Still, it looks as if he will be forced into a similar position as Harold Wilson was in February 1974 when the Labour leader promised to "renegotiate" Britain's terms of membership and put the outcome to the voters in a "binding" referendum. Goodness knows what might emerge from another renegotiation but I doubt if anyone who voted in 1975 can remember anything that Wilson and his Foreign Secretary, Jim Callaghan, claimed to have re-negotiated then.
Meanwhile, it is highly likely that Ed Miliband will back a referendum, on the grounds that it would be impossible to go through an election campaign explaining why he does not plan to "consult the people" over Europe. With Nick Clegg desperate to secure the "trust" of voters, and having once advocated an in/out referendum, I doubt he would campaign on his plans "not to consult the people" even though as the most pro-European of the trio he will be acutely aware of the dangers arising from such a course. Considering how few referendums are held it would be healthier and more honest to stop offering them altogether.