Why did the Brexit polls get it so wrong?
The Brexit vote was a huge deal, both politically and economically. Tons of polls have been telling us for weeks that’s it’d be a close contest, but since the murder of Jo Cox’s, they had mostly been pointing one way: namely, to a Remain win.
To be clear, lots of people said it was too close to call, but the bulk of yesterday’s evidence said that Remain would win by 52% to 48%, with a margin of error of around 2%. The actual results were the opposite, Remain lost by 48% to 52%.
Stock markets can also embed beliefs, and in this case they definitely seemed to think Britain would vote to remain in the EU. For that matter, there were plenty of betting markets that allowed people to bet directly on the vote, and as of yesterday the odds were steeply in favor of Remain. Even the early exit polls pointed to Remain.
So, why did all the polls get it so wrong? I have no more information that anyone else, but I have some purely unsubstantiated, backwards-looking guesses:
- Older people are much more likely to vote, and they also tended to vote Leave.
- People who voted to Leave cared more about the issue.
- People lie in polls, and given that the Leave campaign was being accused of racism, it’s maybe easier to lie towards Remain than the other way around. Also could be a reason that more “undecided” voters were secretly planning to vote Leave but didn’t want to say it out loud.
- People might have actually put money in the betting markets, including the financial markets, that have nothing to do with their belief of the outcome but rather represents a hedge for another position.
- As for the exit polls, they are easier to take in cities, where there are a lot of people, but where there also tend to be more “Remain” voters.
What do you think? Here’s some demographic info from the Guardian that may or may not help: