Skip to main content

The Ongoing Independence Referendum in Scotland: Implications for 7th May 2015

The British Election Study Team

By Jane Green and Chris Prosser

There is no doubting that the Independence Referendum has had a profound impact on vote intentions for May’s general election. British Election Study (BES) data reveals the depth of that impact, in particular the impact of the transfer of votes from the SNP to Labour. It is Labour’s hope that the losses it has seen to the SNP are temporary, and that those voters will come back to Labour in six weeks’ time. Our data suggests that while not impossible, that prospect is very unlikely.

Independence referendum voting has polarized support for Labour and the SNP. The depth of that polarization is seen in a stark comparison. Our most recent BES data gathered in March reveals that almost ninety per cent of Scottish Labour voters are ‘no’ voters and an equivalent ninety-one per cent of SNP voters are ‘yes’ voters. Table 1 reveals the percentage of Labour and SNP voters who voted ‘yes’ or ‘no’ in the independence referendum (using a preliminary sample of wave 4 data from the BES internet panel surveyed between 6- 13 March 2015).

Table 1: Independence Referendum Vote by 2015 General Election Vote Intention

Table 1 SNP

This polarization happened because of the movement of Yes voters away from Labour to the SNP and occurred between June and September 2014. ‘No’ voters show very little change in their general election vote intentions all the way from March 2014 through March 2015. These contrasts can be seen in the following two tables which present the general election vote intentions of ‘yes’ voters (Table 2) in the four waves of the BES Internet Panel, and the general election vote intentions of ‘no’ voters (Table 3).

Table 2: BES Wave 1-4 General Election Vote Intention among ‘Yes’ voters

Table 2 SNP

Table 3: BES Wave 1-4 General Election Vote Intention among ‘No’ voters

Table 3 SNP

There are two reasons to think that these new SNP voters won’t switch their support back to Labour.

If there were the possibility that the referendum effect was weakening among Yes voters, we would expect to find some movement back to Labour in 2015, as the prospect of the general election approaches. The parties have been in campaign-mode for some time. We don’t see a move back in net terms. In fact, the BES data in Table 2 (above) reveals a continuation of Labour vote losses among Yes voters to the SNP. The proportion of Yes voters intending to vote SNP in September 2014 (post referendum) was 70 per cent. That figure now stands at 79 per cent (March 2015). This trend is a worrying one for Labour in Scotland. Less than two months from the general election Yes voters are still moving to the SNP.

The second reason comes from understanding the certainty of people’s general election vote intentions. It is usually the case that people who switch support between parties report being less certain about their vote intentions. We find, for example, that people who have switched from the Conservatives to UKIP between 2010 and 2015 in UK-wide analyses are about 10% less certain in their vote intentions than people who voted Conservative in 2010 and who intend to vote Conservative again in 2015. That pattern holds across different patterns of UK party switching. However, Scottish voters who reported voting Labour in the 2010 general election but who now report a SNP vote intention in the 2015 general election are just as certain as those who have remained with Labour. This again is a worrying sign for Labour. It suggests that Labour-SNP vote switchers are no more moveable than SNP loyalists. Among all UK parties, the SNP’s vote base going into 2015 is the most certain on average of all.

BES respondents were given the following question if they provided a vote intention: “You said that you would be most likely to vote for [party] in a General Election. How certain are you that you would vote for this party?” The following table reports the percentage of respondents choosing points 6 or 7 on a scale where 1 = not at all certain and 7 = completely certain, and does so for respondents who remained with Labour in 2010 and 2015, with the SNP in 2010 and 2015, and who switched from Labour in 2010 to a SNP vote intention in 2015.[1]


[1] Note the N is too small to calculate a reliable figure for people who moved from the SNP in 2010 to a Labour vote intention in 2015.

Table 4: Proportion of loyal SNP and Labour voters and switchers certain of their vote choice

Table 4 SNP

BES data cannot tell us what will happen in the coming six weeks. Vote switching happens beneath the surface of average percentages and the campaign could easily throw up unexpected results. But what we can say is that while Labour may convince its previous voters to come back to the fold, we do know that the odds are certainly stacked against them.

  • Sure Scot

    All No voters –
    There is a Twitter trend gathering pace – #SNPout. It’s about organising tactical voting for the GE to keep the SNP out – based on party with most chance of stopping SNP. Time to put old party differences aside – this is not a normal GE – we need to keep SNP out to prevent their surge. 
    See the image attached to see your constituency.
    Website links to United Against Separation, SNPout and Scotland in Union can be found here

    Remember we need to put party differences aside – for this GE only. We need to hold our noses and vote for the candidate most likely to defeat the SNP in each constituency.

    • David

      How do you feel about the latest Daily Record poll? 51% Yes 49%No? Even if all the unionists voted for one party it still wouldn’t work

      • Sure Scot

        Hi David, I feel fine about that poll.
        Scottish polls are now disproportionately stacked with Nats and yes voters masquerading as “No” voters “who would now vote yes” and “labour” voters “who will now vote SNP”.
        Unfortunately this is another negative legacy from the referendum.
        See the attached excerpt from Survation study.

        • David

          Evidence? Or just an unsubstantiated claim to make you feel better?

          • Sure Scot

            Evidence? – I provided that in the attachment.

          • David

            Your source says that more people say that they voted(past tense) for the SNP than actually did. This leads to them being downweighted and actually produces a lower than accurate SNP score

          • Sure Scot

            Thats not the case or point of the statement. Scots that are being surveyed are giving false voting history.
            The fact is that this is being raised by some pollsters now since the referendum.
            If they are deliberately giving false information then this points towards the sample being unsafe. What other inaccurate information are they providing. Are they false/duplicate/multiple email accounts etc
            Strange that this phenomenon is only happening since the increase in SNP support isn’t it?

          • David

            I doubt that there is an organised campaign to falsify ALL opinion polls. Surely there is a difference with SNP voters feeling guilty about not voting and/or not voting SNP so they claim to have voted SNP in the past when in fact they voted Labour. Mistake and sabotage are totally different and I don’t think sabotage is really plausible here.