Need to Know | 17th December 2014

ICM put Labour ahead 5: how does that compare to other polls?

ICM have, slightly surprisingly, put Labour ahead by five in their monthly poll for the Guardian […]

By

ICM have, slightly surprisingly, put Labour ahead by five in their monthly poll for the Guardian today.

This isn’t the first time Labour have been handed a 5-point lead since the race clearly narrowed in October. After Ukip defeated the Tories in the Rochester by-election, three pollsters – Populus, Ashcroft, and YouGov – handed Labour a 4 or 5 point lead.

At the time those leads looked unlikely to last. If you broke down the poll by party ID – as you can do for the past five years of polls using May2015‘s new tool: The Drilldown – you could see Labour’s sudden lead was about a Tory-to-Ukip swing, rather than a sudden Labour spike. The story is the same today: this lead is about a Tory dip, not a Labour rise.

As in late November, this lead is about a Tory lead, not a Labour spike.

The Autumn Statement, now a fortnight old, seems the most likely cause. George Osborne had a difficult pitch to make – it was unlikely to win him support and has possibly cost the party a point or two.

But this is only one poll, which means it has a margin of error of 3 per cent. While ICM are one of the industry’s most well-respected pollsters, looking at all recent polls offers a better understanding of where we are.

The graph averages the most recent three and five polls for Populus and YouGov respectively. Populus’ trio put Labour up two, as they often have since the autumn, while YouGov say we are in a tie: both parties are in the 32s.

The Tory vote has hovered between 30-33 per cent for almost all of 2014. It seems likely they will recover to those “heights”, where they spent most of the early 2000s. (After today’s poll, May2015‘s Poll of Polls has Labour on 33.9 per cent and the Tories on 31.6 per cent.)

The more interesting number in today’s poll was ICM’s figure for the Lib Dems. They put them on 14 per cent, which is far above other pollsters. There is a systematic difference between ICM and others: they gave the Lib Dems 11, 11 and 10 per cent in their last three monthly surveys.

This is likely to do with the way they ask their voting intention question. It is more constituency-focused, and, as we have often noted, Ashcroft’s seat-by-seat polls seem to be picking up a significant incumbency effect for the Lib Dems (and not the Tories).