C'mon. You knew it was coming. And I'm giving you lots of flag, so you can pass on by this post if you can't bear reading it. But. I did tell you so.
In a post written on March 1, I said:
"At this stage of the game, I'm going to stick my neck out a bit further. I think the Tories are going to win, with a very small overall majority. And Labour will actually lose a few seats overall. Not to the Tories. But to SNP (in the main) and to UKIP (a bit; possibly one seat; possibly Grimsby; but lots of votes; which may hand some seats to the Tories)."
Now, this post, and a succeeding one also written on March 1, were essentially wondering out loud why all the pollsters were saying #GE2015 was a tight race between Labour and the Tories, when all the sub-polling data and evidence suggested otherwise. Data and evidence which, among other things, made very clear that the election was about jobs and fear of Labour.
Now, hang on, I hear you say, polling doesn't 'suggest,' Geoff. And nor do pollsters. It's a science. There is no interpretation. No guesstimation. You ask some folks a question and publish the answer. Um. Actually. That's wrong. Polling is all about bias and interpretation. Although the pollsters don't call it that. They call it weighting and sampling.
It's all about how many people you ask. How you choose the people (remember the argument about how ICM chose the audiences for the BBC?). What questions you ask. How you follow up. How you then interpret the answers, in terms of variation between regions of Great Britain, etc.
There is now argument raging among pollsters themselves about how and why they got their predictions so wrong - here and here. Indeed, they've even asked their professional group organization to set up an independent inquiry.
The results of that inquiry might well be interesting to read. But, I suspect they will be woven in impenetrable mock-geek language, in order to avoid the inescapable truth that their member pollsters got it wrong, because they are either biased, or not very good. And I don't mean simply left-wing or right-wing biased. I mean making-the-wrong-assumptions biased.
For sure, most polling firms in Great Britain are left-of-center. But they are, for the most part, pretty much straightforward and honest. Much more so than the polling firms in the US, which are becoming renowned for tilting all over the place.
Indeed, even one of The Telegraph's senior political editors, James Kirkup, was forced this morning to apologize for getting his predictions wrong.
So, what 'bias' am I talking about? Well. The primary one is that no-one believed what was before their eyes. Either didn't believe it, didn't want to believe it, or quite plainly, did not have the nous to understand it.
I don't entirely blame them. I didn't believe it at first, either. Ok, ok, Geoff. You're brilliant. But, get to the point. Look. I'm entitled to my moment in the sun. Ok? Pause. Right. Enough of that.
They didn't believe that they were seeing a surge to the Tories, because the surge was predicated on the underlying assumption that, by and large, the people of Great Britain didn't blame the Tories for austerity. They blamed Labour for the Great Recession. And besides, they just got over both really quickly.
Forget all this guff about caring people. We do not live in a caring age. Not even among young people. The mantra is: if I'm ok, I don't care; where is my next selfie? It sounds harsh. And it is. But the truth became clear to me from polling done for The Guardian at the end of last year. Polling which suggested that young people in Great Britain had just moved on from the Great Recession.
So? Well, as a child of the Seventies, I had concluded that the primary effect of the Great Recession and the following austerity would be a resurgence of socialism among the young. I was wrong. But I wasn't the only one. It was the primary predication for the failed approach of Ed Miliband. We both made the mistake of assuming people cared. At least I woke up to the awful truth.
Once I had absorbed the import of that polling, I took a wholly new view of upcoming polling and the associated data and anecdotal evidence. Which is why I was able to write the two posts I mention above. And why. Heck hem. My prediction was better than those of the people getting paid to make predictions. Which is a hint, someone.
As for them, and including right-wing commentators like James Kirkup, they read the same information as me (I am assuming). But they simply chose not to believe what they were reading. I'm guessing they all said to themselves: this can't be right; we're not that cynical; surely we haven't just 'got over' the Great Recession, austerity and their tragic human consequences and costs that quickly? Surely not? Hmm. I know. I'll adjust my weighting. My interpretation.
Don't believe me? I know. Why not take a poll of them ... ??
Oh. By the way, if you wonder what I mean about underlying this and anecdotal that, I think this post gives the best idea of what I look at, and what I'm talking about.
Let's end on a prediction. Anthony Wells, CEO of YouGov, will be looking for a new job. He also runs UK Polling Report. And is pretty much regarded as the dean of British polling.
It was quite clear to me as far back as last summer that he was skewing his weighting and sampling in favor of Labour. Brazenly.
Now, I'll try really hard to engage in humility for the rest of the weekend. But I really am open to job offers ...
Posts
