As we start to edge towards the home strait of the electoral cycle, new polling tells us some interesting things.
The conventional wisdom has become, owing to its consistent poll lead since early 2012, that “Labour is on course to win”. Meaning that, even if it means winning as part of a coalition, it would be hard for it to lose the election from here.
This is rather dangerous thinking, for two reasons.
The first is that it overlooks a statistical fallacy, in extrapolating a poll lead in a hypothetical election tomorrow directly out to an equal poll lead in eighteen months’ time. A lead now is patently not the same as a lead then.
If you ask someone the question “if there were an election tomorrow, who would you vote for”, this is not the same as calculating the expected value of their future answer in 2015, for the simple reason that there is not an election tomorrow, nor will there be, barring the “political lightning” of an unexpected no-confidence vote.
No, day-to-day polling is an indicator of future voting intention, nothing more. For example, it omits the effect of all governments deliberately manipulating their administrative and economic cycles so that the necessary bad news mostly falls in the middle of the electoral cycle and the good news towards the end (obviously, having too much bad news at the end is suicidal). Hence, the economic and other government programmes tend to get a deliberate boost in the final year before an election.
It seems reasonable to conclude that this could be at least one reason why, as pointed out here at Uncut at the start of the year, the polls almost always narrow approaching an election.
Back then, we suggested that the 11% lead enjoyed by Labour would, on previous trends, tend to edge downwards to approximately zero over the final two years of the parliament.
So far, during the nine months January to October (approximately the first third of that remaining period), that model has held: the YouGov lead which was 11% in January has now tailed off to 5%*.
Now, we cannot read too much into this just yet, apart to remark that it’s consistent**; obviously there will be fluctuations deriving from short-term factors, and this election could be an outlier, of course.
But it’s also reasonable to expect this trend to continue, given that this is what’s happened in most other elections studied.
Bottom line: while there may be fluctuations (particularly around next year’s Euro elections), don’t expect too much of that already-shrinking 5% comfort blanket to be left by May 2015. Oh, and that’s if UKIP support stays at its current historical high level rather than returning to the Tories; an extremely debatable point.
The second reason to be wary of the conventional wisdom is Miliband’s personal polling.
None of the main party leaders has been doing well recently, in that all have net negative approval ratings. At the time we started writing our book,Labour’s manifesto uncut, we noted Mililband’s April approval rating: this was fairly poor, a net rating of -27%***. According to YouGov’s latest figures, it is now a little worse (-30%). Nick Clegg’s is awful (-48%) and has stayed so (-49%).
Cameron’s rating, meanwhile, has surged upwards, from -21% to -11%.
In other words, whereas Miliband was generally felt by the media to have “won” the battle of the party conferences, this polling is telling us a different story.
But if your headline party polling is good, what is the point of worrying about leaders? After all, it’s the party that counts in a parliamentary system, right?
Not really. Whether we like it or not, our modern, media-driven politics is increasingly “presidential”. As we approach an election, it is not merely “do I like this party?”. It’s “can I see this person as Prime Minister?” It is natural to expect, therefore, that poor personal polling for a leader would start to exert some downward pressure on the party’s own polling as we approach election day.
Now, as Chapter 2 of the Uncut book argues, there are positive, constructive ways to address this, the principal one being delivery on the promise of party reform, showing the world that Miliband can make things happen, not just talk about making things happen.
But it is not just that delivery on party reform could help. The reverse is also true: failure to deliver would likely push things the other way and send an already poor personal polling sliding towards disastrous. Miliband would then be ruthlessly painted by Tory press and CCHQ as someone who (a) talks but doesn’t deliver, and (b) is powerless in the face of resistance from union leaders. And damage to personal polling at this stage of the parliament, as we have just noted, would further hurt party polling, already on target to converge down to a zero-point lead by 2015.
All in all, a pretty unedifying scenario.
Interestingly, there seemed to be a marked scepticism among the Twitterati last weekend to John Rentoul’s suggestion that Miliband is preparing for a two-term, not a one-term, opposition. But it’s really not that outlandish at all. Whilst he may not necessarily see this as the probable outcome – and even if he did, he could not reasonably admit it except to the closest of advisers – it would be frankly foolish not to entertain the possibility.
There is still all to play for: given the strong possibility of hitting May 2015 with one of the two main parties having a very minor poll lead over the other, things are far too close to call.
But in short, the conventional wisdom that a Labour loss is unlikely from here is dangerously wide of the mark. Anything can, and probably will, happen.
This post first posted at Labour Uncut and selected for PoliticsHome's Top Ten Must Reads
*This number, from Sunday, is chosen to be fairly conservative and avoid more extreme fluctuations (yesterday’s YouGov poll lead was down to 1%). Also, for those of you rightly concerned that this reflects only YouGov, and therefore is subject to pollster-specific error, we should note that Labour’s lead in the BBC Poll of Polls has also fallen from 12% on 15 Jan 2013 to 6% on 2 Oct 2013, i.e. exactly the same fall over the last nine months.
**in fact, it has fallen a little further that a “straight line depreciation” over two years would imply. Not that a straight line is necessarily likely – the convergence effect could in reality be concentrated more at the beginning or more the end – but it at least gives us a starting point.
***the % of respondents which think the party leader is doing a good job as leader minus those who think they are not.