BMG’s latest EU Referendum tracker shows remain seven points ahead of leave.

The poll of 1481 adults living in Britain asked voters ‘Should the Should the United Kingdom remain a member of the European Union or leave the European Union?’, which is identical to the question that featured on the 2016 EU Referendum ballot paper.

Including those who are undecided, BMG’s latest poll put Remain on 49%, leave on 41%, with 10% selecting don’t know.

Examining our long-term tracking data for this question, we can see that until late last year, support on both sides was fairly stable. Between November 2016 and November 2017, both sides were effectively neck-and-neck, each on around 45% support. Since then there appears to have been a small yet consistent shift towards Remain.

The biggest Remain lead was in December 2017, where we recorded a ten-percentage point margin.  Fieldwork for this poll was conducted during a particularly turbulent period during the negotiations after the Government and the EU failed to reach agreement after the Democratic Unionist Party refused to accept concessions on the Irish border issue. And while their lead has fluctuated since then, remain have consistently polled ahead of leave in 2018.

 

A ‘soft’ lead?

Although top-line polling suggests that Remain are now consistently and slightly ahead of Leave, there are certainly questions over how ‘soft’ or ‘reliable’ this lead is.

It is important to note that whilst the proportion who say that they support leaving has trailed consistently since the turn of the year, it also has been resolutely consistent. Leave support is currently 43%, which is only three points below the 46% support it achieved in November 2016, a time when the Leave-side was routinely recording marginal leads over Remainers.

This would suggest that the Remain lead has not come from ‘switchers’ – those that voted to Leave in 2016 but have now changed their mind – instead the current Remain lead appears to come from those who did not vote in 2016 and those who don’t tend to vote in general. It has also come from those who were unsure post-referendum.

Consistent with our findings over the last two years, we find around nine in ten of those who voted both Remain and Leave in 2016 would still vote the same way. No minds changed there then.

However, those who stayed at home in 2016 say that they are overwhelmingly in favour of remaining in the EU. 63% of those that did not vote in 2016 say they support remaining, with just 21% saying they would vote to leave.

With record turnout at the EU Referendum in 2016, and little-to-no evidence that turnout would rise dramatically for a ‘people’s vote’ (or at least that traditional non-voters are very unlikely to turn out in large numbers), the polling suggests that another referendum on EU membership, asked in exactly the same way, is most likely to produce the same result, not a different one.

 

An accurate methodology

This poll shares the same methodology as BMG’s pre-Referendum polling, which consistently reported Leave ahead in the run-up to the EU referendum in 2016, and also called the correct outcome.

This approach now has the added benefit of an additional weighting variable; i.e. how respondents voted, if at all, at the EU Referendum in 2016. This is the only major change to our approach since. Although, at the time BMG favoured it’s telephone poll, having published the phone and online polls simultaneously, which can be found here (see table 9 on page 89), BMG has continued to favour the online methodology since the referendum in order to ensure that the results are both accurate and comparable with an approach that accurately called the final result.

 

Methodology, fieldwork dates, and a full breakdown of these results can be found here.

For a more detailed breakdown of results from this poll, or any other results from our polling series, please get in touch by email or phone.

polling@bmgresearch.co.uk

@BMGResearch

0121 333 6006

 

 

 

Robert Struthers – Senior Research Executive

 

 

 

Dr Michael Turner – Head of Polling