Post is reproduced from Huffington Post

New polling after the British referendum shows that it’s not the leave voters who regret their choice – it’s those who stayed home. Despite media declarations of widespread regret from leave voters, only 5% would vote to remain if another referendum was held. Non-voters, however, do seem to regret their decision, with 37% saying they would vote to remain if given another chance.

The economic turmoil following the vote did very little to persuade leave voters to change their mind. While the economic expectations of remain voters have dropped like the British Pound to the US Dollar, leave voters live in a different economic reality. Our polling shows them becoming increasingly optimistic in their economic outlook as Brexit went from possibility to reality.

We have been tracking the expectations of the British public on the economic future of Britain and their families. We specifically conducted polling two and one week prior to the vote on the British exit from the European Union (i.e., Brexit), and one week after the vote. The polling was conducted with the mobile polling application Pollfish over three time periods: June 7, June 16, and June 28. The election, where British voters chose to leave the EU, occurred on June 23. On PredictWise, which aggregates polls and betting markets, the probability of Brexit on June 7 was 25% and on June 16 was 37%.


Figure 1: Expectations of London Stock exchange two weeks before and one week after Brexit

The first figure shows the stock market expectations of the British, broken down by their referendum preference (voters who were unsure before the election are grouped with non-voters after the election). The three groups all start with similar expectations, but these diverge as the probability of Brexit increases. On June 7, all three are about evenly split on whether the market will go up (left pane) or go down (right pane), with the remain voters being a little more concerned. But, by June 16, with Brexit about 12 percentage points more likely to occur, the remain voters became much more likely to think the market would go down, with leave and unsure voters looking very similar. After the election, both remain voters and non-voters are very concerned about the market going down, while leave voters are more bullish than ever. The concern of the remain voters that the market will go down is much larger than the belief of the leave voters that the market will go up.


Figure 2: Expectation of family income two weeks before and one week after Brexit

Brits’ expectations of their family’s financial wellbeing show similar patterns to those for the stock market, but more muted. This is important in that Brits of all persuasions are convinced that the economic turmoil at a national level will not affect that as harshly on a personal level.


Figure 3: Cross-tabs of voting before and after Brexit

In our June 28 poll we also asked how respondents would vote in a hypothetical second referendum, where we saw small but meaningful post-election changes in attitude towards remain. Remain voters are steadfast: 94% would vote to remain again. Leave voters are a little wobblier: 5% would now vote to remain while 11% are unsure. The biggest moves were among the British who did not vote, who are now much more likely to support remain than leave. As a result, remain is up 8 points in our hypothetical second referendum.


Figure 4: June 7 and June 16 ask about the upcoming June 23 referendum, while June 28 asks about a hypothetical referendum at some future date.

We are confident our polling is generally accurate. First, our final pre-election poll had leave up 2 points, consistent with other polls in the final week and with the eventual result. Second, our estimate of financial expectations is subject to less measurement error than all of the other polls in the field. We ask about expectations first, then talk about the Brexit at the end. This avoids priming our respondents that this is a political, rather than financial, poll.

Follow the latest predictions in real-time on PredictWise. This article was written jointly with my graduate student Sam Corbett-Davies of Stanford.