It only takes one poll and suddenly the blogosphere is a lather with New Year predictions for the Holyrood elections in May. The boys at Better Nation got into a bit of a spat over it all. Jeff applied his shiny new election predictor file to the findings from the TNS-BMRC poll for the Herald while Malc H preferred a less scientifically rigorous application. James was left to hold the jackets. Meanwhile, Lallands Peat Worrier presented the gender differential on voting intentions in stark relief and it didn’t make pretty reading for the SNP.
The analysis would of course be incomplete without the burd getting out her divining rods and giving her crystal ball a rub.
The treatment of the findings deserves comment. There has been a wee bit of jiggery pokery in order to create a phenomenon called the committed voter. Effectively all those who refused to answer or didn’t know how they intended to vote were removed from the sample and the findings re-calibrated. This “tweaking” resulted in Labour’s findings on the first constituency vote leaping from 31% to 49%. But it is unusual to effectively disenfranchise don’t knows from the process at this early stage in the campaign. A bit naughty really of the Herald to report these findings as the main ones.
The burd also notes that the committed voter tallies are what Better Nation used to come up with an overall 3 seat majority for Labour. What would be the result if the whole sample findings were poured into the election predictor? Something far closer to what the result in May is likely to be, I reckon.
Let’s also consider the methodology. The Scottish Opinion Survey is one of the few polls to be conducted face to face, in people’s homes; most others are done by telephone or completed online. Does this account for the apparent collapse in Conservative and Liberal Democrat voting intentions?
The burd grew up in Galloway when Ian Lang was laird of all he surveyed. For 18 years he was the area’s MP, but in all that time, few ever admitted openly to voting Tory. No one ever boasted of it, the most that could be coaxed was a sheepish sotto voce admission that maybe they had, at one point, voted Conservative. Could a similar factor be at work to explain the very low polling of both the Tories and the Lib Dems? I mean, who in the current circumstances, would openly want to admit, face to face, to a stranger no less, that one might be considering voting for the parties preparing to dismantle the welfare state and inflict huge public spending cuts on us all?
But that does not explain why the predilection of the bashful would appear to be for Labour at the expense of the SNP. Especially when BBC Scotland’s poll on spending cuts last autumn showed that voters blame the previous UK Labour government for our current economic woes.
Yet, the SNP is actually at roughly the same level as before the 2007 election. The gap between them and Labour is not insurmountable, and as is often the case with polls, the don’t knows are sufficiently large in number to make it all to play for. On the constituency vote, Labour leads by 10% but nearly 1 in 5 of participants are undecided. On the regional vote question, which was phrased fairly clumsily, Labour’s lead is only 9% with over 20% still to make up their minds.
Moreover, if you look at the key target voter groups that, in the burdz humble opinion, constitute Scotland’s squeezed middle, namely women, 35 – 54 year olds and C2s, there is also still hope for the incumbent Scottish government. On the voting intentions of women, the picture does seem pretty bleak for the SNP. The direction of travel is all wrong with the gap widening. What is going on here? Why is the SNP’s problem with wimmin growing? Not sure. It needs more cogitation and deliberation before blogging on it but it is clear that the SNP needs to focus some energy onto fixing this before it becomes a chasm that costs them the election.
The future seems less bleak when looking at the other groups. Of those aged 34 to 54, 34% intend to vote Labour and 19% SNP on the constituency vote, with 32% voting Labour and 19% SNP on the regional vote. It’s not great but a brighter picture emerges when the intentions of those aged 45 to 54, who are more likely to actually vote, are considered. Labour’s lead over the SNP narrows significantly. Incidentally, if we look at the age group most likely to vote, the over 55s, the lead is down to 6% and 5% respectively.
Classic switcher territory also indicates there is still all to play for. Amongst C2s, Labour polled 29% to the SNP’s 23% on the first vote and 25% to 24% on the second vote. These poll ratings matter because previous election analysis shows that it is skilled manual labour voters who are most likely to move between parties and in marginal constituencies, their votes could make the difference between crucial losses and gains.
Is Labour in poll position to win the Scottish election in May? Maybe. But the fat lady – or rather the wee dumpy wumman in her pinny – hasn’t yet sung for the SNP. And beware those low ratings for the Conservatives and Liberal Democrats. Early reports of their demise may be grossly exaggerated.