London won the first battles, and the ''No'' side will probably win the referendum in September - but it is going to be a long war.
The opening shot was fired by Chancellor George Osborne in London, who declared an independent Scotland could not negotiate a currency union with the rest of the United Kingdom. With only one-tenth of Britain's population, Scotland is just too small to demand an equal say in how the pound is run. Besides, why would London want to keep the responsibility for Scotland's huge and rather dodgy banking sector?
Alex Salmond responded by threatening to repudiate Scotland's share of the national debt if London wouldn't agree to a currency union but the conclusion was obvious. Scotland could go on using the British pound if it wanted (like Panama and East Timor use the US dollar), but it could have no formal link.
Next was the president of the European Commission, Jose Manuel Barroso, who warned that it would be ''difficult, if not impossible'' for an independent Scotland to join the European Union. Other EU members that don't want their own autonomous regions to secede would almost certainly block Scotland's membership - Barroso was talking mainly about Spain and the separatists in Catalonia - and one veto is enough.
Independence was looking more complex and expensive by the minute - and then Standard Life spoke up. The Edinburgh-based company is the largest pension provider in the United Kingdom, managing about $US350 billion ($NZ410 billion) in assets and employing 5000 people in Scotland.
Ninety per cent of Standard Life's four million UK customers do not live in Scotland, however, and it warned that it might have to leave if the Scots voted for independence. A poll subsequently revealed that 36% of Scottish firms would consider leaving following a ''yes'' vote.
It was a cold shower for the Scottish National Party (SNP), and the number of people planning to vote ''yes'' in the referendum dropped to 32%, while the ''noes'' remained unchanged at 57%. Lots of Scots would like independence if it doesn't cost them anything but they don't want it badly enough to risk any major changes.
Unless something changes quite dramatically, the final vote will be 60-40 or more against independence.
So what are the Scottish Nationalists really hoping to achieve? Originally, Mr Salmond planned to build support for independence through a long period of successful government within the UK but the SNP's landslide victory in 2011, in the depth of the recession, stoked unrealistic hopes among his militants and forced his hand. Nevertheless, he probably knew he was going to lose this one.
That's how it worked in Quebec in the 1980 referendum, which the separatists lost 60-40. The idea of leaving Canada and striking out on their own frightened the French-speaking majority in Quebec too much at the time. But it did put the question on the table, and it never really went away again.
Mr Salmond will know the history of Quebec separatism well, for it is the best analogy to his own situation. He will be aware that the second referendum, in 1995, came within a hair's breadth of succeeding. And he will have noticed that the separatist Parti Quebecois is still around, is likely to win the provincial election due on April 7 - and will almost certainly call a third referendum in the next few years.
It's what English-speaking Quebecers call the ''neverendum'', but it actually does end eventually. You only have to win the referendum once. After 34 years of this, the ''Rest of Canada'' really doesn't care any more, so there will be no pleas to Quebec to stay this time, no special offers to sweeten the Confederation.
The ''Rest of the United Kingdom'' is already there: the English, in particular, seem distinctly unmoved by the prospect of Scottish independence. This may be because Scotland has much less of the UK's population than Quebec has of Canada's (one-tenth versus one-fifth), and because Scotland is at the far end of Britain, whereas Quebec is in the middle of Canada. So maybe it will only take two referendums in Scotland.
They should pray that this is so, because the four-decade, three-referendum scenario is pretty grim. In Quebec, it caused the most spectacular case of ''planning blight'' in recent history. The perpetual uncertainty about Quebec's political and economic future drove the corporate headquarters out (they moved to Toronto) and the immigrants and the investment went elsewhere. The population numbers in Canada's two biggest provinces tell the story.
In 1980, the year of the first referendum, there were 6.5 million people in Quebec and 8.5 million in Ontario, and the ratio had been steady for most of the century. There are now 8.2 million people in Quebec - and 13.4 million in Ontario. Montreal had always been Canada's biggest city but Toronto is now more than 50% bigger.
Salmond must know that this is where he is taking Scotland. He presumably thinks it is worth it.
- Gwynne Dyer is an independent London journalist.
