Monthly Archives: May 2017

Let’s talk about polls


Polls. Polls. Let’s talk about polls: Are the Tories really ahead still in them?

I think this is a pertinent question on a number of counts. Firstly – a disclaimer. I used to work in a position (for a fundraising agency) that required me often to analyse data. Analysing data is something that is very subjective – one can skew it in order to get an answer that is favourable to oneself, and/or to a client.

People talk a lot about opinion polls, naturally, in the run-up to a GE. The results of opinion polls are also dependent on the questions being asked. As posted on my Facebook page last week, an utterly biased poll was being sent out by #YouGov, with ┬ásome very leading questions. Then one gets on to the people surveyed – how can we be sure that this is genuinely a ‘cross-section’ of society? We only have the polling company’s word for that, and if they are being employed by Conservatives, they will most likely choose the option that appeases their client the most.

Think this doesn’t happen and that this is some conspiracy? Think again. I know that the Tories (for example) at the last election hired at least one telephone fundraising/marketing company to work on their behalf, Return Fundraising. How do I know this? Because they told me so when they attempted to ‘head-hunt’ me (a move motivated by the demise of a former employer). They also told me at the time that they did the canvassing for the MP Jacob Rees-Mogg.

When call-centres do this sort of work, their scripts will be scrutinised to the nth degree so that, subliminally, the answer will always favour the client. Of course people have their own opinions; but there is much that can be done, in terms of the language used, the questions asked, to subliminally affect the answers in favour of clients. Anyone who works in fundraising and marketing knows that this is the case, and in a sense, that is ok. As much as politics, it’s about the art of persuasion – but let’s be very clear also that an answer elicited for a survey/charity donation doesn’t always hold once the conversation is over and the phone put down. In fundraising, it’s known as attrition. So: pollsters may have an opinion at the start of a call, they may hold to that opinion or have their minds changed during the course of the call (because one of the curious things I found out when telephone fundraising is that even on the end of the phone, people are social animals that often want to be seen as agreeably nice people to the faceless voice at the other end of the phone), but when that phone is put down, they can often change their minds. In that sense, if one is polling people with semi-formed or wholly uninformed opinions, there is a high probability that those opinions are malleable during the duration of a call, but potentially not beyond that.

As well as all of this: how current is the data we are being presented with? How many thousands of people are being canvassed? Is it a genuine cross-section of society, or targeting areas where the answer is more or less a done deal? Who has commissioned the polls (this is important)? Which media outlets are these polls being distributed to, and what is their editorial agenda? (In the UK, this latter point is crucial) With those papers, would it potentially upset their investors/advertisers if the polls projected an outcome they didn’t want? Who are the polling companies being run by (as in who is their CEO), and what is their political agenda? For example, YouGov has as its CEO Stephan Shakespeare, who formerly ran ConservativeHome – immediately there is cause for scepticism there over political neutrality in their polling.

The media are also key in polling in the UK, perhaps more than in most countries, because they are undoubtedly (on the results of previous elections and the influence of Murdoch) able to sway elections. The Sun and Murdoch are the big beasts here as regards this, because to the best of my knowledge there has been not one PM since 1979 that hasn’t had Murdoch backing. Essentially what we’re looking at there is corporate propaganda filtered through politics at the highest level. Murdoch’s other investments (let’s say, like his investment in Genie Energy, as well as Sky News) would be subject to huge unbiased and potentially unfavourable scrutiny if the potential PM was not a Murdoch acolyte.

And then there’s the public. Up until a few years ago, there was little interest in politics among the British public, and often an embarrassment in saying who they were voting for (especially on the more right-wing side of things). This is in marked contrast to, let’s say, Ireland, where, for better or worse, people are incredibly vocal about their side of any given issue, thus enabling more polling accuracy.

This is not even getting into the increasing influence of organisations like Cambridge Analytica, who collate evidence from social media and target voters accordingly. We’ve seen how successful this was in the #Brexit referendum and in the election of #Trump.

So while not dismissing polls entirely, let’s be aware that they are (a) not unbiased at the source i.e. the company that is commissioning the polls and the company that is working on their behalf (b) dependent on media source and editorial bias (c) open to interpretation (d) time-dependent (a week is a long time in politics, etc) and (e) subject to change. It’s that latter point that the Labour Party and other non-Tory parties have to really make the case for.