By George Kerevan
Nate Silver is a youthful American opinion poll analysist who correctly predicted the outcome in every state, in the last US presidential election. Nate has grabbed headlines on this side of the pond by predicting the Yes camp has no hope of winning next year's Scottish referendum. Is he right?
For a start, while Nate wins points for his 50 out of 50 score last year, we shouldn't get carried away with his psephological genius. His attempt to call the 2010 UK general election was an embarrassing failure. He predicted the Lib Dems would get 29.1 per cent of the vote, but they actually got only 23 per cent. Nate failed to see the collapse of the Clegg bandwagon. Lesson: opinions can change in the heat of a campaign.
That said, every poll since January has put the Yes camp behind. Averaging the polls, stripping out the "don't knows" and massaging for who is actually likely to vote on 18 September, 2014, the Yes voters are running at around 39 per cent. That's better than some of the selective headline figures in the newspapers but still significantly behind 61 per cent for the No camp. It would need a swing of 11 per cent merely to even the odds.
The independence campaign has seized on the fact that one polling agency, Panelbase, consistently shows a higher Yes vote than other pollsters such as YouGov and Mori. Using the Panelbase numbers, the Yes people claim it would only take a 4 per cent swing to win. But choosing the particular poll that suits your purpose is a rather delusional way to fight a campaign.
True, rogue polls are a fact of life and Panelbase might be right. But if you average all the referendum polls (which is more scientific than picking the one poll that suits you) then the required swing for independence is a lot more than 4 per cent. I will not be popular in the Yes camp for saying this. However, more wars are lost by ignoring the enemy's strengths rather than by over-estimating them.
For a Yes voter such as myself, the issue is: how do we shift the polls in our favour? Nate Silver argues that in constitutional referendums the norm is for undecided or wavering voters to drift back to the status quo during the campaign. I'm not sure that is true. The evidence in the 1995 Quebec referendum was that the "certain" Yes vote climbed steadily in the last two months before polling day, as the debate heated up. That vote was lost only by the narrowest margin.
On the plus side, the Yes camp has greater credibility with the voters. One recent poll from Panelbase for the pro-independence Wings over Scotland website shows that only 28 per cent of Scottish voters believe Alistair Darling always or mostly tells the truth about independence, while 47 per cent reckon he never or rarely does so – a net rating of minus 19. This suggests to me the electorate is not taken in by Darling's baleful "Project Fear".
Above all, the latest polling evidence, from the same poll, shows clear majorities for Holyrood to control welfare spending (60 per cent), oil revenues (53 per cent) and taxation (52 per cent). That tells me Scotland is half out the Union door. The one policy people fret about handing over to Holyrood remains defence, supported by only 35 per cent. (Somebody should tell voters that David Cameron has dismantled the British Army and scrapped all the maritime reconnaissance planes that used to protect Scottish fishermen.)
Nevertheless, for the Yes camp to triumph, it will have to win over two in every three voters who are undecided or wavering in their preference. That is sobering but not impossible, provided we focus. Where should the Yes camp target?
First, traditional Labour supporters. Polls from different agencies agree that between 13 and 16 per cent of Labour voters are open to voting Yes. Support for independence is already highest in working-class housing areas.
Second, women voters, who are the weak link in the campaign. Despite the fact that 31 per cent of women say they are "Scottish" not "British" (compared with only 28 per cent of men) this does not translate into Yes votes. Women – I'm talking heroic averages here – will vote on the economy rather than identity. Women as a whole are not convinced by the economic case for independence. We need to talk jobs, taxes and welfare.
Finally, our youth. Nate Silver is correct in pointing out the anomalous fact that the national movement has failed to engage the imagination of Scotland's young people – the very generation who are being sold down the proverbial river by London's obsession with austerity. Yet young folk should be the easiest to reach and mobilise, using modern social media.
There's more than psephology in all this. The Yes campaign is playing by traditional general election rules, and that won't work. Contemporary electioneering is based on "triangulation": getting close to the opposition parties in policy, in order to nibble their marginal voters. In the UK, such marginal voters are predominantly social and economic conservatives who identify with institutions like the monarchy. Typical marginal voters are therefore the least likely to vote Yes.
Fortunately, the referendum "undecideds" are not traditional marginal voters. The new "undecideds" are fed up with the status quo – remember those majorities for Holyrood to control tax, oil and welfare. But they need to feel comfortable they are making the right decision in leaving the UK. That is why – after all – they are undecideds and not No voters.
Isolated and atomised, the undecideds won't take the risk of independence. The only way to move this group is to convince it that everyone is moving together. The referendum "undecideds" – predominantly working-class Labour voters, women and young people – will move into the Yes camp only if "others like me" move too.
That demands the Yes campaign start acting like a grass-roots insurgency rather than a campaign to sell soap powder.
Courtesy of George Kerevan and the Scotsman