Why EU referendum turnout might actually favour Remain

When pollsters get together and talk about the EU referendum, it doesn’t take long before the conversation gets onto turnout.

The debate’s well explored – 1 in 4 sentences in this BuzzFeed article mentioned turnout – and has mostly concluded that the issue helps Leave. But I think the debate has underestimated something that helps Remain.

YouGov’s Freddie Sayers sets out well here the argument that turnout hurts Leave. As he says, Leave supporters are demographically more like people who vote, while Remain supporters – on average, younger people – look more like people who stay at home.

Polls already take this into account as far as possible. If someone says they’re not certain to vote, pollsters either discount them entirely or weight down their response.

The trouble is, polls before an election usually find that more people say they’re certain to vote than actually turn out. The ones who don’t live up to their word tend to be younger.

So, if this is replicated in the EU referendum, Remain may find that many of their younger supporters don’t actually vote, despite saying they would definitely do so, and so Remain might underperform their polls. Given that online (but not phone) polls currently find the race to be neck-and-neck, that could be crucial.

I don’t dispute this. But there’s another aspect that could be at least as important.

Currently, with two months until the election, around 6370% say they’re certain to vote.

Two months before the Scottish referendum, 7881% said they were certain to vote. In the last polls before the vote, 9495% said they were certain to do so.

Turnout in Scotland was 85%, so slightly more than 10% of those who said they would definitely vote in fact didn’t do so. But more people voted on 18 September 2014 than, two months before, had said they were certain to do so.

If something similar happens with the EU referendum polls, in mid-June we would see something like 75-82% saying they’re going to vote (and turnout would be around 70%).

One reason that could change the balance of the race is that stated turnout of Remain supporters has more scope to increase from where it is now than turnout of Leave supporters does. In ICM and Mori’s latest polls, 67-70% of Remain backers said they were certain to vote, while 74-80% of Leave supporters said the same.

So there are more people who support Remain but don’t currently think they’ll vote than there are who support Leave and don’t plan to vote. If turnout expectation increases, Remain’s support has more room to grow, without having to win over any undecideds or Leave supporters.

What I think this means in real terms is that Leave supporters tend to be more enthusiastic and already say they’re going to vote. Remain supporters are more grudging and haven’t yet decided to vote – but over the next two months a growing proportion of them might think it’s worth the effort.

When people start paying attention

But perhaps the Scottish referendum was so different from this one that we can’t learn much. Is it really likely that turnout expectation will increase for the EU referendum like it did in Scotland? There’s no way to be sure, but I think it’s a reasonable assumption (though I’m not claiming turnout will be quite as high).

Here’s how Google searches for “Scotland independence” (in Scotland) changed in the two months before the election.  The number of searches increased by a factor of around 16 between this point (two months before the vote, shown by the arrow) and the week of the election.

Referendum searches

It’s not like the referendum wasn’t in the news two months before the vote. That week saw independence-related stories about, among other things, tuition fees, the Law Society, celebrity endorsements, Jean-Claude Juncker, oil and gas taxes, and, of course, various polls. Yet most people weren’t googling it.

It seems fair to expect that public attention to the EU referendum will go the same way and there’ll be vastly more public interest in it in June than there is now. Most of those who aren’t particularly interested aren’t yet paying much attention.

Clearly, the index of Google searches isn’t a direct measure of interest in voting, and I’m not claiming that this is definitely going to happen. But if the Scottish referendum’s anything to go by, the last three to four weeks before the election are the time when most people seek out information. That would be the time we could expect to see more people saying they’re certain to vote.

None of this is to dispute that turnout is a threat to Remain. The polls may well be slightly overstating Remain’s relative score, since it’s probably the case that their numbers include some people who are deluding themselves – or the pollster – about their plan to vote on 23 June. If the polls were level by then, I would expect Leave to win.

But I think there’s, in general, less chance of the polls being level in mid-June because of the greater potential for an increasing turnout expectation to help Remain.

It’s not often I stick my neck out with a prediction on this site, but here’s one: stated turnout will increase between now and mid-June, to at least 75%. As I say, I’m not claiming this will definitely happen – but the likelihood that turnout expectation will increase significantly does seem to have been overlooked.

If it does happen, then, other events in the referendum being equal, that’s more likely to help Remain, giving it a slightly wider lead than it currently has.


Comments are closed.