On Wednesday evening I spent an entertaining half-hour on local radio as part of Cambridge 105’s election panel, in the company of Chris Rand and Richard Taylor. With exquisite timing, earlier that day Lord Ashcroft released his latest batch of polling, which included a constituency poll for Cambridge. Here are the headline results:
On the radio I described Lord Ashcroft as “an eccentric billionaire who likes doing opinion polls”, and indeed this isn’t his first Cambridge constituency poll – his previous one was published in September last year. It looked like this:
If these polls are to be believed, then over the last few months Julian Huppert’s Lib Dem campaign has opened up a significant lead over his main rival, Labour’s Daniel Zeichner. Back in September they were running neck and neck, but now Julian is 9 points ahead. Indeed, this latest poll suggests that his support has actually increased since the 2010 election, which is a pretty extraordinary result given what has happened to the Lib Dems nationally in that time.
How much trust should we put in these polls? They’ve been performed by a reputable polling organisation following the British Polling Council rules (which means the full data tables are available), and they had a reasonably large sample size of 1000 people, which suggests a margin of error of about 3%. It is worth noting that both polls were conducted outside University term, and so presumably don’t include a number of students who will be in Cambridge on polling day. However, some students will be voting in their home constituencies, and they have a relatively low turnout in any case, so this is unlikely to be a decisive factor.
This poll will certainly be a boost to the Lib Dem campaign and a setback for Labour. One vivid illustration of this is in the effect on the Ladbrokes odds for the Cambridge seat. Here are the chances of winning the seat implied by the odds immediately before the poll was released:
Here’s how the odds shifted immediately after the release of the poll:
However, it’s worth emphasising that this is just one poll, and there is still a long way to go until May 7th, both in the”air war” (the national media campaign) and the “ground war” (feet on the streets). It’s certainly not over yet.
One particularly notable aspect of the poll was that people were asked about their general voting intention separately from their constituency voting intention. You can see the exact wording of the questions in the data tables, but essentially they were firstly asked who they support in the General Election, and secondly how they were planning to vote specifically in the Cambridge constituency. Neither question prompted with the names of the candidates. The headline result of the poll, shown in the first chart above, is the answer to the second, constituency question. On the first, general question, Labour actually came out ahead:
The difference between these numbers and the overall poll result in the first chart above is due to tactical voting in the context of the Cambridge constituency, and support for Julian Huppert personally, rather than as a Lib Dem. It’s known that incumbent MPs do get a support boost, particularly the first time they re-stand, but even so the result between these numbers and the overall poll result is pretty striking. For example, around 40% of the Green vote disappears between the general and constituency poll results.
Thanks to the data tables, we can look in more detail at the vote splits. This next graphic shows how each party’s supporters (from the general question) are planning to vote in Cambridge:
Here the rows represent general voting intentions, and the colours within each row represent how those people plan to vote in Cambridge. So while over 90% of Lib Dem supporters are planning to vote Lib Dem in Cambridge, only about 55% of Green supporters are planning to vote Green, with significant numbers switching to the Lib Dems, and some to Labour, either for tactical reasons or as a personal vote. The Lib Dems are also picking up significant support from Conservatives, and even a fair number of Labour supporters. Nobody except UKIP supporters is planning to vote for UKIP.
It’s uncertain whether Lord Ashcroft will favour us with another opinion poll before the end of the campaign, but if he does, it’ll be fascinating to see which direction the numbers are going. But I’ll conclude by repeating the mantra that Lord Ashcroft always applies to his opinion polls – this is a snapshot, not a prediction.