This week was perhaps one of the worst in terms of the Remain camp’s performance, and for the first time the Leave campaign appeared to take a considerable, but not yet decisive, lead.
While previous weeks have seen individual polls suggest that Leave may be in the lead by a relatively narrow margin, four of the seven polls conducted over the last week or so have suggested that the margin between the two has grown to between five and six points.
As a result, our most recent poll of polls has given the Leave camp a lead of 5.66 per cent. This would lie at the outer reaches of the normal two to three per cent margin of error for either side in such polls, suggesting that this may be a genuine lead rather than a statistical anomaly.
The decline in the number of people answering don’t know on average may also suggest that Vote Leave have been doing a better job of convincing undecided voters of the merits of their campaign. However, the wide variations between polls in terms of the number of people who answer don’t know makes this difficult to state decisively.
Furthermore, this week saw one of the first telephone polls to put the Brexit campaign in the lead, conducted by ICM.
In general, telephone polls have favoured the Remain campaign while online polls have been more generous towards Brexit. That a telephone poll now gives the Leave camp a considerable margin suggests a substantive shift in public opinion.
It is worth noting, however, that ICM’s online and telephone polls have been closer to one another than online and telephone polls in general.
While there was still one poll that suggested a five-point lead for the Remain campaign, from ORB, the polls in general point to a widening gap between the two camps.
However, psephologists have noted that there is usual a narrowing towards the status quo during the days before a referendum. This means that we may well see a revival for the Remain camp as polling day approaches.
Polling breakdown by region
In this week’s weekly breakdown of polling data from ICM, we examine the regional variations in the way that people say they may vote on membership of the European Union.
The data suggests a clear variation between the different regions and devolved nations of the UK. Scotland, for instance, is the most strongly in favour of Britain’s continued membership, with polls between April and May suggesting that between 50 and 60 per cent of the country want to remain in the EU.
By contrast, those polled in the Midlands have consistently been more likely than the rest of the country to say that they wish to leave the EU.
Perhaps the greatest comparative variability has been in Wales, which in some polls has been the most supportive of continued membership while in others has proven the most eager to leave.
It is possible that a particular spike around the end of April could be related to the impact of decisions surrounding the steel industry, but it is also worth remembering that the smaller sample sizes in these groups means that individual data points may not be as reliable.
Interestingly, while people in the South are more likely to suggest that they wish to remain members of the EU, there does not appear to be a stark divide in patterns of support between the north and the south of England.
Given the relative success of UKIP in the north of England at the last General Election, this may be considered surprising.
However, this may be partially attributed to the fact that the regional associations being used – “North” and “South” – are so broad. Further, it is far from certain how well different parts of these regions are represented in the data.
A more granular look at data from YouGov, conducted by University of Bristol academics, appears to show a correlation between the areas where people favour Brexit and the local authorities where UKIP performed best at the 2014 European Parliament Elections.
This does not necessarily mean that we can take the results of those elections as predictive. Referenda and elections are different by their very nature, and potential differentials in turnout could have a strong impact.
Nevertheless, examining the previous elections in 2014 could provide some understanding of the patterns of support for remaining in and leaving the EU across different regions in the country.
This article is part of a season of in-depth DeHavilland EU Referendum coverage. Click here to find out more and sample our definitive political intelligence services for free.
Madhav Bakshi is a Political Analyst within DeHavilland’s Editorial Team and leads on Energy policy. He is a graduate of King’s College London, where he studied International Politics.