‘But what about the SNP?’ That’s what I’ve been asked most, after writing a blog on the way the public see the ideological positions of the parties.
With ‘normal’ opinion polls it’s difficult to do sensible inter-country comparisons– because the number of respondents outside of England is usually not large enough, but the beauty of using the British Election Study – or rather, one of the many beauties – is its massive sample size. The N for England in the analysis that follows is 13k+; but even that for Scotland is c.4k and Wales c.2.3k. In other words, the N for Wales alone is larger than many normal polls of the whole of Great Britain.
The table (below) shows the mean average response to a series of questions asking respondents to place the parties on an eleven point scale, where 0 is left-wing and 10 is right-wing. The data are drawn from wave 3 of the BES, which was in the field in September and October this year, after the referendum result in Scotland. The table is broken down by each of the three countries that make up Great Britain. It also gives – in the top row of data – the average respondents’ self-placement.
Ideological placement, Sept/Oct 2014
As with the earlier analysis, it’s always worth remembering that not everyone can answer these sort of questions. The percentage saying don’t know ranges from 13% to 27%, depending on the question (self-placement is easier than placing the parties, and placing the minor parties is harder than the more established ones) and the country (Scotland had the lowest don’t knows for every question). Also remember that these are means: responses are distributed around these means, rather than seeing them as exact scores. The standard deviation for each of these questions is never less than 1.9, and in some cases reaches 2.6.
Here, for illustration, is the way the SNP are seen (by respondents in Scotland). As the graph shows, the mean response may be 3.9, but the distribution is much messier than that single figure indicates. Some see the SNP as as right-wing as the scale allows, other as left-wing as can be, and most are somewhere in between, clustering around 4.
A mean score of 3.9 puts the SNP very slightly to the left of how Labour are seen in Scotland on 4.1. Note also that Scots see themselves as more left-wing than respondents in England and Wales (even if they aren’t really) and that Scots see Labour as more right wing than people in either Wales or England do.
But the difference between the perceived positions of Labour and the SNP is not large. You can get a measure of this from this graph (above) which shows the relative position of Labour and the SNP in the minds of all Scottish respondents. Those to the right of zero on the x-axis see Labour as more right-wing than the SNP, those to the left see Labour as more left-wing. More people see the SNP as to the left of Labour (46%) than the other way around (36%), but the single most common response was to see them as identical (17%), and a full 40% see the two parties as at most one point apart.
The average Scot puts him or herself at 4.4 – that is, nearer to Labour (4.1) than to the SNP (3.9). The irony therefore is that the party which is seen as positioned closest to the voters in Scotland is the Labour party, the party currently in a deep existential crisis. Moreover, on the face of it, if it moved to the left – as many have argued it needs to in order to reconnect with the Scottish electorate – it would be moving away from the political centre of gravity, not towards it.
But these overall figures can be misleading, because there are some voters who see a bigger difference between the SNP and Labour: in particular, SNP and Labour voters. Labour voters in Scotland see themselves as left-wing (a mean of 3.4) and think they support a left-wing party (also 3.4), and they see the SNP as noticeably to their right (4.9). SNP voters see the world differently: they see themselves as almost equally left-wing (3.6) but they think they support a left-wing party (seeing the SNP at 3.8), and see Labour as even further to their right (5.3). The mean difference in all respondents’ views of Labour and the SNP was +0.3. But for Labour voters, it was -1.6. For SNP voters it was +1.6. They each see the world very differently.
Because the BES is a panel study, we can also see how people’s views have changed over time. These same questions were also asked in wave 2 of the study, which was in the field in May and June of this year. For most parties, there has been little or no change in their perceived position. (We wouldn’t expect much, to be fair, given that we are only talking about a four-month period). Most of the mean positions changed by 0.1 (or at most 0.2) on average, just random noise, and some changed not at all. The two biggest changes, however, were those of Scottish Labour and the SNP. The former became seen as more right-wing, with its mean position increasing from 3.6 to 4.0; the latter became seen as more left-wing (from 4.2 to 3.9). (The eagle-eyed may have noticed that the figures for wave 3 aren’t exactly the same as those cited above; that’s because we are reporting here only those respondents who completed both wave 2 and wave 3. But the differences are negligible). A shift of +0.4 or -0.3 may not sound like much, but they occurred in a short period of time, and were of sufficient magnitude to alter the parties’ relative positions: in May and June, the SNP were seen as (on average) slightly to the right of Labour; by September and October, they were seen as (on average) slightly to the left.
This is not, of course, to say that this is where the parties actually are. It’s where people think the parties are. Moreover, as just shown, perceptions of their positioning are seen through a partisan lens, in which what you think about your party helps determine what you think about your rival. Therefore some people might dismiss the data. (In which case, why have you read this far? Haven’t you got anything better to do?) But, right or wrong, perceptions are important. It’s where people think the parties are that will determine how they react to them, not where they actually are.
One last thing: the referendum clearly had a dramatic effect on the ‘don’t know’ responses for these questions. As noted above, in the more recent polling, Scottish respondents were the least likely to say don’t know to any of these questions. But in May and June, there was no substantive difference between the percentage saying ‘don’t know’ in Scotland compared to England or Wales; in the intervening months, as the referendum campaign reached its peak, the don’t knows in Scotland fell by as much as 10 percentage points, compared to much smaller drops in England and Wales.
As for how the Welsh see Plaid, that’s maybe for another day – although so far no one’s actually asked…