I’m Ailsa Henderson, Professor of Political Science and Head of Politics and International Relations at the University of Edinburgh. So, Ailsa, what do we know about how voters are responding to the campaigns so far? Yeah, there’s a couple of things we can look at. We can look at the messages that seem to be resonating. It clearly has been a risk-based campaign on the part of those wanting to remain within the EU. They’re either warning of all the things that can happen, if the UK leaves the EU. And, in that sense, I suppose it is - it would be reminiscent of what was happening in the independence referendum in Scotland.
But we do know that some of those messages are resonating with voters. The ‘Remain’ side does appear to have won the economic argument, with voters clearly concerned about the economic consequences of the UK leaving the EU. And that holds for just the general state of the economy, but people’s concerns about unemployment, but also house prices, and so on. So, in that sense, that perhaps is the ‘best’ argument, or the strongest argument that ‘Remain’ has happened to find. In terms of arguments that appear to be resonating on the other side, it’s clearly migration.
And they clearly do seem to have put forward this notion that, if the UK leaves the EU, that migration will decrease, with almost two-thirds of voters believing that to be the case. So, it does appear to have narrowed down to key arguments - economic ones on the part of one side and arguments about migration, and control of migration, on the other. We also know things about levels of support. And we know that it’s quite tight at the moment, with the polls of polls showing, kind of, 51% - 49% for ‘Remain’, but a lot of those results are within the margin of error for polls.
And then the other thing we know is about demographics of support - the kind of voters that back one side or another. So, what kinds of similarities or differences have been emerging between public attitudes to the EU in the different parts of the UK? Yeah, we see differences around demographic groups, but also regional groups. So, demographics - we know that age
matters: young voters more supportive of staying in, older voters more supportive of going out. It’s about two-thirds in either direction. And we get the same kind of support across education levels, for example. In terms of regional break-downs in support, Northern Ireland is where support for staying in is strongest - and that’s not surprising. And, then, after that, it’s kind of London and Scotland are most supportive of staying in the EU. And then other parts of England, Wales - so, other parts of the south of England, Wales and Northern England are, on balance, more supportive of ‘Leave’, but the margins are really small. So, that’s one way we can look at it.
The other way to look at it is whether predictors of support, or the demographics of support, are varying across different regions of the UK. And one thing we know, for example, is that Conservative supporters are more likely to back - are less likely to be ‘Leave’ supporters in Scotland than they are in England. So, we see different interactions between partisan support and regional variation in support as well. What indications do we have of what the turnout might be at the EU referendum? Yeah, well all polls ask a ‘How likely are you to vote’? question. And, right now - and it’s on a zero to ten scale, with ‘ten’ meaning ‘I’m absolutely certain to vote’.
And, before - you know, they ask these questions right before elections - it’s sitting at about 65, 66% at the moment. But, that doesn’t necessarily mean that turnout will be 65, 66%. So, if we were looking at, say, the Scottish Parliament elections that happened last May, the stated likelihood of voting was also 66, 65% - and actual turnout was only 55. And the same is true if you look back to 2015 - actual turnout was 10 points lower than the intended turnout in the polls.
So, if we believe what happened last time, in the 2016 Scottish elections, on the 2015 UK elections, then turnout will be around 10 points lower than the percentage of people who are saying that they’re absolutely certain to vote. That’s one thing. The other thing is, however, we know that turnout is higher when it’s a close contest. And it wasn’t a close contest, in Scotland, in May, and so that might have depressed turnout more. But the other thing we know, that’s cutting across that - so, it’s a close contest, you might expect higher turnout - but the other thing we know is that voters do suffer from fatigue. There’s been a lot of elections recently.
More, depending on where you’re living in the UK. But, there is a real issue of voter fatigue and turnout. So, that might cause turnout to fall. Voters sometimes use so-called ‘second-order’ elections ‘ that’s local elections or elections to the European Parliament - to give a verdict on the government rather than to address the actual question or election at hand. How likely is that to be the case in this EU referendum? It’s kind of hard to say, because the government’s divided - or the governing party is divided in terms of which side they want you to back.
In referendums, we can typically distinguish between a referendum that has been called by a government and the government is clearly backing one side and they want you to support them, and independence referendums are usually like that. So, the government says ‘we want independence - we think independence is a good idea - will you back us in this referendum?’ That’s one kind of type of referendum we get. The other kind if when the government says ‘well, actually, we’re neutral on this issue’. So, referendums on electoral reform, for example, referendums that follow recommendations of civic assemblies - citizens’ assemblies.
You know, those are instances where, there will be proposes put forwards towards the electorate, and the government says ‘we are neutral - you make up your mind’. Formally, the government - we’re in the former situation. You know, the government is saying, the Prime Minister is saying ‘we want you to vote to remain within the EU.’ But the Conservative Party is so divided on this issue, it’s hard to know whether this is really a referendum on the government. Certainly, it would appear to be a referendum on the Prime Minister. But, one of the ironies of this referendum is that it - you could say that it was called predominantly to deal with internal divisions of the Conservative Party.
It was seen as a way of managing those divisions. And it’s hard to see any result solving those divisions - other than a clear, clear win for one option rather than another. It’s hard to see how the referendum result will solve those divisions. I mean, the other reason why you could say that the Prime Minister called the referendum, or advocated a referendum, was to see off support leaking to UKIP. And, so, perhaps, arguably, it will have been a success, from a Conservative strategic standpoint, on that front.
But, I think, in terms of being a way to move on from longstanding divisions within the Conservative Party on Europe, and the UK’s role in Europe ‘ I think it’s questionable whether the referendum will actually satisfy that. The polling for the UK’s 2015 General Election was infamously inaccurate. How confident can we be that the opinion polls are better this time?
Yeah, it’s a good question. In a sense, polling for a referendum is different than polling in an election, because you’ve got this multi-option world in an election that you just don’t have in a referendum. And the polls were quite good on the 2014 referendum. Hard to track it in a situation in which opinion is changing. Polls were more accurate for the devolved elections than they were for the UK election in 2015. That said, there’s real - there’s obviously real cause for concern - largely down to the fact that we’re getting very different results depending on the way the polls are conducted. So, a lot of polls are done online. And a lot of polls are done over the phone.
And we’re getting completely different results. We’re getting closer contests, or closer results, for the online polls. Whereas, the telephone polls have typically been showing a strong lead - a stronger lead for ‘Remain’. And, so, we’re wondering about that. Partly it’s to do with the sampling method for online polls. And we do know that online panels over-represent those who are very engaged - in other words, our expectations about likely turnout are higher than would be the case. We know they are also more likely to attract younger people. However, I think one of the - there’s been a lot of reports down on this by other people, and one thing they’re noting is that telephone polls often don’t ask about ‘don’t knows’.
They force people to choose between ‘I want to vote to leave’ or ‘I want to vote to remain’. And when you give people the option of a ‘don’t know’, you increase the proportion of ‘don’t knows’ - you decrease the proportion of people who will actually give a preference. So, it could well be that, when pushed to shove, those ‘don’t knows’ are actually tipping over into ‘remain’, which could mean that the polls are more accurate of what will happen on the day, when people have to choose between two options, because you can’t vote ‘don’t know’ on your ballot. That assumes, however, that all those people are going to vote.
And I think equally likely is that people who genuinely don’t know how they are going to vote, as Referendum Day approaches - well they’re maybe not going to vote ‘remain’, maybe they’re going to stay at home. So, we’ve got a lot of unanswered questions at the moment about poll quality.