The forthcoming independence referendum now dominates the Scottish political scene. The airwaves are full of claim and counterclaim about what leaving or staying in the UK might mean. As a result everyday party politics is very much taking a back seat. But Thursday’s Dunfermline by-election for the seats vacated by the jailed former SNP MSP, Bill Walker, means that, fleetingly at least, it gets the chance to return to the limelight.
But what is happening in the party battle? It is often asserted that while the SNP and their friends in the Yes campaign may be significantly behind in the referendum, they continue to enjoy remarkable popularity when it comes to how people would vote in a Holyrood election. The impression is thus conveyed that the SNP are struggling to convert their popularity as a party into votes for independence.
In truth, this perception is a statistical illusion.
It arises for two reasons. The first is a systematic difference between the way that referendum vote intentions are commonly reported and the way in which Holyrood vote intentions are. The second is that the target for victory in the referendum is very different from what is required to win a Holyrood election.
As we have noted before, the common practice in newspaper reports of referendum vote intentions is to leave the don’t knows and won’t says – typically just under one in five of the sample – in the baseline on which the percentages of Yes and No voters are calculated. That inevitably depresses the headline Yes (and No) figure.
In the ten polls of referendum vote intention conducted since the middle of August, the headline Yes figure has been 32%. But if we take out the don’t knows and won’t says, it rises to 38%.
In contrast, when it comes to reporting Holyrood vote intention the don’t knows and won’t says – who can be just as numerous as they are in the case of referendum vote intention – are routinely omitted from the baseline. That helps push the SNP (and everybody else’s) tally up.
In four polls conducted in the last two months that have ascertained how people would cast their constituency vote in a Scottish Parliament election, the average SNP vote has been, well, yes, you’ve guessed it, 38%! In short, although not all Yes voters are SNP voters or vice-versa, the overall level of Yes support and that for the SNP are in fact much the same.
Even so, that 38% SNP vote somehow still looks a lot better than the 38% Yes vote. Why? Because in contrast to the position in the referendum it is still enough to put the SNP ahead.
At present the SNP’s nearest challenger, Labour stand on average at 35%, three points behind the SNP, while the Conservatives and the Liberal Democrats are a long way behind on 13% and 6% respectively.
To enjoy such a position after more than six years in power is indeed quite a remarkable achievement, and attests to a relatively high degree of satisfaction with the SNP’s performance in office. The Conservatives would certainly love to be in as equally strong a position in the battle for power at Westminster.
But it is not enough to win a referendum. A Scottish Parliament election can be ‘won’ simply by doing better than a multiplicity of opponents. Indeed, just 45% of the vote may be enough to secure an overall majority, as was demonstrated in 2011. However, in next September’s ballot anything less than 50% will simply not be enough. Securing a Yes vote for independence in 2014 will be much tougher than winning power proved to be for the SNP on that unforgettable night two and half years ago.