EU Referendum polling disparities expose demographic divisions across UK

15 June 2016

The EU Referendum is really about the winners and losers of the processes of globalisation and the ways in which UK society has been shaped by them, John Curtice FRSA FRSE FBA FAcSS, Professor of Politics at the University of Strathclyde, Senior Research Fellow at NatCen and Senior Fellow of the ESRC’s The UK in a Changing Europe programme, told the first event on the Academy and Campaign’s Summer Programme calendar.

Speaking before an audience of more than 60 academics, pollsters and journalists, Professor Curtice discussed how the problems of the polling industry and the social tensions they expose will play out in the EU referendum on June 23rd.

“You might think this is a referendum about the UK’s membership of a relatively obscure international intergovernmental organisation. It’s not,” Professor Curtice said. “This is a debate about the kind of society the UK is, the kind of society we think it should be, and in particular the way that society has been shaped by the processes of globalisation.”

He commented on the stability of the polls throughout the campaign, with Remain consistently, though not significantly, ahead since the start. “In one sense we can think there’s been an awful lot of change, but actually there’s been very little change at all, perhaps until recently,” he said.

While phone polls have tended to put Remain ahead, he noted that there has been “absolutely no obvious long term trend.” As the balance of phone and internet polls has changed over time, disparities between them have become more stark. Of the nearly 150 attempts to measure the distribution of public opinion, virtually all initial polling was conducted over the internet, with only eight phone polls before mid-February. The shift towards increased phone polling provides “one of the reasons why in May people were saying Remain were making progress; they weren’t, there were just more phone polls,” he said.

Now, however, “The big headline is that after weeks and months of basically no change, it does look as though there has been at least some movement towards Leave in the wake of the advent of purdah.” Before this, “Leave were only able to play the ball and not the argument. Since then we have Leave making arguments and pronouncements to which Remain have to respond.”

He cautioned that perceptions of progress in the polls needed to be regarded with some circumspection, saying “It’s how far you are ahead of [the benchmark] 50 per cent plus one that matters, not the lead.” The “sudden panic” over the prospect of a Leave victory stems from a view that it has made gains since the end of May, when in fact it had never done particularly well in the polls to begin with, and internet polls have actually indicated from the start that the result would be “impossible to call.” If, however, phone polls start putting Remain under 55 per cent on a consistent basis, it would indicate a true shift in favour of Leave, he said.

Methodological differences between phone and internet polling partly account for this, as phone polls try to fit quotas, while internet polls consist of pre-recruited, non-randomized panels of people.

“The polling industry has basically operated on the assumption that we know we don’t get representative samples, but that partly by the way in which internet polls can stratify people and certainly by the fact that we can weigh the data afterwards, we can make these samples look representative. But the big message from the inquiry into what went on 12 months ago is that basically that assumption fell over.”

While the context in which a survey is administered and social desirability bias could explain differences between the methodologies, Professor Curtice contended that there is little evidence to convincingly support this. Rather, both face the same problem where some people may be easier to get hold of and some people may be more willing to participate than others. As a result, disparities in the polling are more likely due to sampling differences, such as whether internet participants are more politically committed, phone samples are truly representative, the extent to which polls weight education, age, degrees of social liberalism or conservatism, and the tendency to answer “don’t know” in surveys.

Although the common perception has been that ultimately undecided voters will break for staying in the EU, Professor Curtice said that, when pressed, roughly 30 per cent side with Remain, while 19 per cent favour Leave. “But very often pollsters are already taking this into account, so we shouldn’t assume that ‘don’t knows’ are suddenly going to change matters.”

The demographics of the referendum, therefore, expose an ideological divide that cuts across deeper social, political, cultural, and economic cleavages, revealing a “nation at unease with itself.”

He said that the structure of the referendum, in terms of who wants to stay and who wants to leave, is stable, with both sides making arguments that resonate with particular segments of the electorate based on demographic conditions. Nonetheless, “this is a referendum in which the electorate is torn.” People are wondering whether to go with the side on which “our bread is buttered economically, or with the fact that we’ve had far more immigration in the last 15 years and, by the way, there are a bunch of people in Brussels who shouldn’t be able to tell us what to do in the first place.”

The arguments put forth by both sides play into these divides, as people’s views on issues of immigration and the economy, and how these issues factor into their vote, differ according to education, socio-economic background, and age. It is, in the end, about globalisation.

“The basis of these social divisions is that on the one hand we have the people who are the winners from globalisation. They are the people who are in jobs, that do not feel challenged by the labour market, who don’t feel their job is going to go tomorrow because of people coming from the EU, and they quite enjoy working in a diverse environment.” These people tend to be younger and more educated, while at the other end of the spectrum there is a section of society, mainly older people and those with fewer educational qualifications, that feels disenfranchised and alienated from their own national identity because of immigration, and feel to be economically losers from the process of globalisation.

Professor Curtice concluded that the “outcome will determine whether or not we’ll put up with globalisaiton and embrace it,” or reject it on the grounds that it causes too much social division.

“It’s a close referendum, it’s a very divisive referendum, it’s very divisive for our political parties, it is divisive for our society, but it’s answering and addressing some very big questions about the nature of our society.”

Download the slides here.

Academy and Campaign summer event programme