Polls didn't show a consistent Labour lead

20 September 2016
What was claimed

Labour was winning in the polls over the last year.

Our verdict

This is incorrect. Three isolated polls since the last general election have shown Labour ahead. Each time that has happened, most polls showed the Conservatives ahead. Individual polls aren’t enough evidence to see if one party is leading.

“We've been working, over the last year, to unite the party, and we were winning electorally and in the polls.”

John McDonnell, 15 September 2016

“85 out of 89 polls prior to the so-called coup, we were behind. In the other four, we were neck and neck. So he’s just not telling the truth about the polls.”

Alastair Campbell, 15 September 2016

“At no point this year have the polls ever shown a consistent Labour lead” according to polling expert Anthony Wells from YouGov.

That’s the problem with the claim: you need more than a poll or two to know that you’re winning in the polls, whatever party you’re from.

Polls are uncertain and have a margin of error, which means if one party is apparently only slightly ahead of another, you can’t tell it’s actually in the lead from a single poll. You need a series of polls to show a consistent lead in order to be surer.

Only three have shown Labour ahead of the Conservatives at certain points in the last year, according to Anthony Well’s list of polls. One shortly after the referendum showed the two parties were neck and neck. Each time this has happened, most polls around the same time showed the Conservatives ahead.

Labour did win about half of the local council seats up for election earlier this year, a similar proportion to the previous comparable elections in 2012. It lost a small number of seats overall, though the Conservatives lost more.

Update 20 September 2016

We updated the article to make it clearer, including the claim and conclusion

Full Fact fights bad information

Bad information ruins lives. It promotes hate, damages people’s health, and hurts democracy. You deserve better.