Whose Supporters are Most Likely to Vote?16 February 2015
It has long been recognised, both by academics studying elections and by parties fighting them, that a key part of electoral success is ensuring that a party’s supporters actually turn out on the day. Parties put a lot of effort into Getting Out The Vote, and they are right to do so: research about local campaigning suggests that much of the most effective on-the-ground work that parties can do is not so much about changing people’s minds as about identifying existing or potential support and then making sure that it turns out.
It is worthwhile, then, to keep a general eye on those polls (many, but not all of them) that enquire about how certain people are to vote and to see what if any patterns emerge from such data.
When we do this for GB-wide polls at present, the general picture tends to be distinctly discouraging for the Liberal Democrats. Not only have they lost the majority of their electoral support since 2010; but even those supporters that remain tend to score a lower percentage than those of other parties on certainty to vote. Among the other major parties we don’t tend to see great differences. One clear message from the polls is that anyone who is either expecting or hoping that UKIP support will melt away by election day may well be disappointed: many UKIP supporters appear highly motivated, and they generally score no lower – and sometimes even higher – than the other main GB-wide parties on likelihood to vote measures.
When we comparing across the different regions and nations of Britain, Scotland at present typically scores significantly higher in reported likelihood to vote. This is not, I should make clear, all about the SNP; rather this reflects the general political mobilisation and interest engendered by independence referendum. It may not persist into the longer-term but at present Scotland is simply a very politically-engaged and interested place. However, it is also clear that SNP supporters are currently no less motivated than those of other parties to turn out and vote.
What about Wales? Well, as I have had cause to celebrate on the blog previously, these days we do have fairly regular Welsh polling. In addition to Welsh Political Barometer polls that we at the Wales Governance Centre conducts with ITV-Wales and YouGov – and which we hope we may be able to make slightly more regular as the election approaches – there have been several polls conducted in the last year by ICM for the BBC. Moreover, not only do we have Welsh polls in greater quantity than before; we are also fortunate in that the two companies conducting them are two of the most highly-esteemed pollsters around, and that the polls are conducted using contrasting methods, with ICM polling by phone while YouGov work via the internet.
ICM typically ask about likelihood to vote in their political polls, not least because they use this information in weighting their vote intention figures. YouGov has typically placed less emphasis on likelihood to vote in their analysis; however, our recent Barometer polls have been asking about likelihood to vote in the general election and will continue to do so right through until the election itself.
The two companies ask about likelihood to vote in a very similar manner, requesting respondents to rate their certainty to vote on a scale running up to 10. Specifically, the questions are:
ICM: “Some people have said that they would not vote in a new general election to the Westminster parliament, while others have said that they would vote. I would like to know how certain it is that you would actually vote in a general election tomorrow”.
The scale then runs from 1-10, with 10 meaning ‘certain to vote’ and 1 meaning ‘certain not to vote’.
YouGov: “On a scale of 0 to 10, with 10 meaning definitely will vote and 0 meaning definitely will not vote, how likely are you to vote in the UK General Election next May?”
As ICM’s scale runs from 1-10 and YouGov’s from 0-10 they are not absolutely comparable. But they are close enough for most practical purposes. We might note that, given that YouGov’s scale runs right down to 0, then ceteris paribus we should expect lower average scores from YouGov than ICM. But things are not equal: in the two polls that these companies conducted in Wales in January, 53% of ICM’s whole sample said that they were 10/10 certain to vote, while 68% of YouGov’s sample put themselves at 10/10 on the YouGov scale. These differences are no fluke: they reflect the fact that internet samples are somewhat more likely to sample among those who tend to vote.
What about the relative certainty to vote of the supporters of the different parties in Wales? This table summarises the findings of the two most recent polls (the January ICM poll for BBC-Wales and the January Welsh Political Barometer poll):
|% 10/10||Average /10||% 10/10||Average /10|
In general there are not huge differences between the main parties, but a couple of points to note:
- One thing that is possibly emerging as a systematic difference between the two pollsters is the likelihood to vote of Plaid Cymru supporters. On the average out of 10 measure they have been clearly the lowest of the five main parties on both the last two ICM polls (the previous one having been conducted in September last year); neither of the last two YouGov Polls have found anything similar.
- The surprise in the latest ICM poll is that the Lib-Dems perform so strongly. This is not only contrary to the typical pattern in the GB-wide polls, but also contrary to the previous ICM poll in Wales, where the party scored worst out of the five on the 10/10 certainty measure. I suspect that this latest finding may simply be an outlier: given the low numbers of remaining Lib-Dem supporters in Wales (this finding depends on only 25 respondents) then it becomes easier to get blips like this.
Overall, the polling doesn’t suggest that any of the parties have established a clear and decisive advantage in terms of energising their potential support at present. All have some work to do – and in an election that could be very close, the fortunes of the parties could depend a great deal on how successfully they ensure that their potential vote actually ends up being manifest in the ballot box. So this is something I’ll be keeping a close eye on – and I’ll let you know if I find anything interesting.
Non-partisan thoughts on elections, voting and political representation from Roger Awan-Scully of Cardiff University.