In defence of YouGov (or why Labour supporters shouldn’t dismiss them so quickly)

There’s a post I’ve been seeing a lot of people share on social media, expressing frustration over mainstream media coverage of public reaction to Tuesday night’s debate between Boris Johnson and Jeremy Corbyn. I wanted to take a look at it, and clear some things up:

The criticism is simple: the polls with loads of voters showed a massive Corbyn lead. The poll with the smallest sample was basically a statistical tie, but had Johnson in the lead, but look which poll the broadcasters are reporting on!

It’s frustrating of course — but it’s also overly simplistic, it’s a poor tactic to gather support for Labour, and, crucially, it’s missing a trick. Here’s why.

1. Not all of these polls are the same

It’s worth starting this by pointing out that anyone can set up a Twitter or website poll and the broadcasters can’t necessarily gather all of them.

Second, anyone can take part in a Twitter or website poll: the electorate is self-selecting and crucially, activists share links to polls in their networks to encourage their allies to vote in it. If one side is more effective at that than the other, that skews the result.

It’s harder to do that with an opinion poll carried out by a polling company: they don’t just ask you the question, they ask a load of other questions that give them a better idea of who they’re talking to. They then group you together with other people like you and weight the sample to make it more reflective of the country. Let’s say a company asked 900 men, 90 women and 10 people identifying their gender in a different way the same question. They wouldn’t publish the raw results: they’d analyse their sample and be able to scale down the response among men, and scale up the response in the women and non-binary groups. A Twitter poll can’t do that: it can’t correct for over- and underrepresentation of different groups who vote in it.

So on that score, the YouGov poll is more reliable than the Twitter polls.

2. Choose your battles

I remember the last couple of weeks in the Indyref campaign, when a poll showing Yes ahead for the first time turned everything upside down. The main Westminster parties went into panic mode, and threw everything they had at the campaign (including almost the entire Parliamentary Labour Party marching through Glasgow).

So on balance, maybe the response by some Yes campaigners to spend the last Sunday before polling day marching on Pacific Quay in protest at BBC bias was ill-judged.

Firstly, that was a day they could have spent ringing around their friends and family, or knocking on doors talking to voters who hadn’t yet made up their mind.

Secondly, I think activists of all political colours can fall into a big trap: they can think that everyone is as into all this as they are. When you start thinking like that, you don’t easily see how your message can land with people whose political engagement begins and ends at the ballot box. And sadly, complaints that the media are biased against your cause, now matter how justified they may be, are often met with far more cynicism than you may realise. That march probably repelled some undecideds on the very weekend when they absolutely needed to be attracted.

As frustrating as it is that the mainstream media aren’t giving your side a fair hearing, you have an opportunity to go around them, and use social media to reach your loved ones to get to them before tomorrow’s newspapers hit the shelves.

The good news is that your Facebook and Twitter accounts are a great way to spread your message. The bad news is that you are using them to complain about Laura Kuennsberg instead of cheering your preferred party leader. Ignore her, ignore Peston and the others, and focus on giving people the news you feel they’re not getting!

3. That YouGov poll is better news than you think

And the news they’re not getting is better than you realise. YouGov itself isn’t reporting the debate as a win for Johnson, but are pointing out that viewers were split. Give people the message straight from the horse’s mouth rather than filtered through a media site you don’t trust.

Besides, remember the sampling I told you about? YouGov now publish the details, breaking down the results according to what else they know about the respondents, and this breakdown makes for interesting reading.

YouGov found that if you voted Tory in 2017, you were more likely to think Johnson won. If you voted Labour or LibDem in 2017, you were more likely to call it for Corbyn. Leave voters went for Johnson. Remain voters for Corbyn. Even the generation gap kicked in: under-50s were more likely to call the debate for Corbyn; over-50s for Johnson.

YouGov also asked viewers who they expected to win before the debate started. Those who expected Johnson to win were more likely to say they’d been proven right. If you expected Corbyn to win, you were more likely say you’d been proven right.

The problem with deciding who has won a debate is that it’s subjective. And in this case, if you were already more likely to agree with one of the participants (or more like to disagree with the other one — these are not necessarily the same thing), or expected them to win beforehand, you were more likely to decide that they had won.

So here are the killer facts Corbyn supporters might like to focus on: firstly, a majority those respondents who didn’t know who they thought was going to win decided that Corbyn had won.

And more importantly, Johnson picked up a majority of voters who have already decided how they’re going to vote. Corbyn picked up the majority of voters who haven’t. They were the voters both men needed to reach, and Corbyn reached more of them than Johnson did.

The media very rarely go into the detail of an opinion poll, but for me, the important points can be found there. So if I were a Labour or Corbyn supporter, I wouldn’t be dismissing the YouGov poll, but highlighting the key detail. It’s better news for you than you think.