“What would happen if you reran the UK’s “in/out” EU Referendum today, having seen the news headlines immediately following the announcement of the result?”
I am sure many have asked the question. As you will surely know: It is not possible to answer with anything even approaching a rough degree of certainty. However, I have been running opinion polls on the Brexit question for the last year, and have carried out three across June (twice before, once after the announcement of the result). While opinion polls are far from perfect, they do still help in understanding rough trends.
Below are the results of an opinion poll carried out immediately prior to the Brexit vote; followed by a poll carried out in the days immediately after.
First is a straight ‘Before & After’ comparison of the two polls. Below that is a little greater detail on each of the polls. Additionally, there is data gathered over the last year on ‘the Brexit’ question, to add further context to the results.
Note of Caution: I am not publishing this to suggest there should be another referendum. I have simply been tracking this for the last year, and find the massive change in results very interesting.
Before & After
Below is a comparison of an opinion poll carried out over the weekend prior to the ‘Brexit’ vote, compared to the same poll carried out over the week following the announcement of results:
Summary: The poll results show a big shift from the ‘Undecided’ group to the ‘Remain’ group. The size of the ‘Leave’ group changes, but only by four to five percentage points.
Each poll here was anonymous, carried out among adults in the UK, via the Internet. There were 2,018 responses gathered in the ‘before’ poll. There were 1,092 responses to the ‘after’ poll. The data is weighted based on the Internet population of the UK, and that weighting is based on demographic data from 1,485 respondents in the ‘before’ poll, and 829 respondents in the ‘after’ poll.
More Detail: Result Immediately Prior to the Vote
Here were the results of the poll, carried out over the weekend immediately prior to the referendum, including additional notes:
- The ‘undecided’ group is quite high. This was a UK-wide poll, with no filter clarifying whether respondents were registered to vote. Ie, it is representative of adults in the UK—not just voters.
- Importantly: ‘Remain’ polled at 37.2 percent, the ‘Leave’ answer polled at 32.5 percent. You can see from the bars that each had a margin of error of +2.6/-2.5. Ie, the ‘Remain’ result was predicted between 34.7 and 39.8; the ‘Leave’ result predicted between 30.0 percent and 35.1 percent. So, according to the snapshot poll, ‘Remain’ would probably win, but there was a chance ‘Leave’ would prevail.
More Detail: Result Immediately After the Vote
Here is the same poll, carried out in the days immediately after the vote was announced, and newspaper headlines and politicians across the world reacted. The data here was gathered between June 24-29, 2016. Again, I have included additional notes:
Three notes there:
- ‘Remain’ has leapt enormously.
- ‘Leave’ has dropped, but only by four or five percentage points.
- The ‘undecided’ group has dropped hugely—from 30 percent to 15 percent.
Based purely on this poll, those believing the UK should remain a member of the EU outnumbered those who believe we should leave the EU by a margin of 2:1 after the referendum result was announced and newspaper headlines and politicians had responded to the result.
More Detail: Data Gathered Over the Last Year
Below is a summary of most of the Brexit polls I have carried out over the last year, to add context and further illustrate the large change in results. Most were carried out among just over 2,000 respondents; one or two were carried out with just over 1,000 respondents. The early polls I chose to omit the ‘Undecided’ group. The referendum question changed late last year, from a ‘Yes/No’ question to ‘Remain/Leave.’
- Among these polls, there is a large swing from ‘Undecided’ to ‘Remain’ following the announcement of the results.
- The ‘Leave’ group has dropped, but had polled at similar levels previously.
- The ‘Undecided’ group has dropped significantly.
Above are simply opinion polls. As we all know, opinion polls do not necessarily predict the actual result of elections or referendums, they simply offer a snapshot of opinion among a group of respondents. The respondents above were in the UK, but not necessarily registered to vote.
Dan Barker is a consultant based in London & Continental Europe. From Software Engineer to CMO, he has worked across more than 100 brands—from retailers to publishers to finance institutions. He writes about data, marketing, politics, ecommerce, and other related areas.