It is now widely accepted that the ground war – what happens in individual constituencies – matters in British elections. In Scotland, SNP membership passed 100,000 in March, and is now said to be over 110,000. This makes the SNP the third largest party in the UK, but this rather underplays its size.
For a party fighting just 9% of seats, it is, proportionally, the largest political party in Great Britain by a long way. It averages more than 1500 members per constituency, members we are constantly being told are very fired up and energetic. It has, in other words, the sort of grassroots army that any other political party would kill for. Is there evidence of this making a difference on the ground in Scotland?
The first thing to note is that there has been an obvious explosion of grassroots campaigning in recent weeks in Scotland. Take the polls released by Lord Ashcroft in 16 Scottish constituencies in February 2015.
Asked about the contact they had had from political parties, the majority response in every seat Ashcroft polled in February was ‘none’. This ranged from 58% in Inverness, Nairn, Badenoch and Strathspey to 88% in Airdrie and Shotts. Things had improved slightly in the eight seats polled a month or so later in March, but still the majority response was none in six of the eight seats, and the lowest figure for no contact was 43%.
There has been an obvious explosion of grassroots campaigning.
The most recent polls, just released for eight constituencies, show that those reporting no contact from any of the parties has dropped to below 20% in every seat, and was as low as 5% in East Dunbartonshire. Of the eight latest seats to be polled, five had been polled before, allowing us to track the transformation in campaign activity across the last few months in more detail.
In East Renfrewshire, between March and April, the figure for contact from none of the parties has dropped from 56% to 9%; in Ross, Skye & Lochaber, it dropped from 46% to 10%; and in Dumfriesshire, Clydesdale & Tweeddale, it fell from 54% to 17%.
The change is even more obvious with the two seats first polled in February. In Glasgow South West, in February, the figure for contact from none of the parties was 86%, it is now 19%. In Paisley & Renfrewshire South it was 75%, it is now 14%. The individual figures for contact from each of the five parties which Ashcroft asks about – Con, Lab, LD, SNP, UKIP – had gone up in every case in all five seats.
In the February and March polls, of the (relatively minimal) contact that had taken place, the SNP were clearly having the better of it. Of the 16 seats polled in January, for example, the SNP had better contact rates than the other parties in all but three (and in two of those, the difference was at most two percentage points). Of the eight seats polled in March, the SNP were out contacting all other parties in the five Lab-SNP contests, albeit slightly behind in the other three (two of which were LD held, one Conservative-held).
In the most recent polling, however, the SNP are behind in voter contact in all eight constituencies polled. The difference is often not great, but it is consistent. In the three Lab-SNP contests polled in April, the gap ranges from four to 23 percentage points. In the four LD-SNP contests, the SNP trail by 5 to 11 points on contact, and in Dumfriesshire, Clydesdale and Tweeddale (a Con-SNP fight), the SNP trail both Labour and the Conservatives (by six and seven percentage points respectively).
In the most recent polling, the SNP are behind in voter contact in all eight constituencies polled.
The five Scottish seats which we can compare across time show a slightly more complex picture, but they were a slightly atypical bunch, in that the SNP were behind in their contact rates in three of them when first polled (despite being ahead in most of the other Scottish seats at that time). But even of these, the SNP’s relative position to their main challenger has weakened in terms of voter contact in three of the five.
One retort to this might be to say that it doesn’t seem to be mattering all that much. After all, in the most recent set of eight seats polled, and in which the SNP trailed in contact in all eight, the SNP were on track to win seven, often involving massive swings to them from their opponents. In the five seats polled previously, this involved their position improving since the last set of polling. Noting that they are behind on contact yet ahead on intention to vote is perhaps akin to noting that a football team has won fewer corners yet scored more goals.
Labour’s official reaction to the most recent Scottish polls was to say, bluntly, that there was no gloss that could be put on them. This certainly isn’t an attempt to do so. But it is still a puzzle. Why are they behind on contact? With a grassroots army like the one the SNP currently enjoy, they should be ahead in the ground war. So why aren’t they?
It is worth noting that the Ashcroft question about contact is a fairly broad one:
I would like to ask whether any of the main political parties have contacted you over the last few weeks – whether by delivering leaflets or newspapers, sending personally addressed letters, emailing, telephoning you at home or knocking on your door. Have you heard in any of these ways from the Conservatives, Labour, the Lib Dems, the SNP or UKIP?
This sets the bar low (one piece of literature from a party qualifies), it is a binary measure (and therefore does not measure intensity of contact), and it does not distinguish between types of contact (an intensive face-to-face chat counts the same as a pamphlet pushed through the door). It could, therefore, be that Labour’s contacts have mostly been by post – perhaps using targeted mail – whereas the SNP, with more bodies on the ground, have been contacting people on the doorstep. This, at least, was one suggestion made to me on twitter, when I first noticed this effect.
We can get at least a preliminary insight into this from the British Election Study. The most recent wave of BES data draws on interviews undertaken in March. The preliminary results from wave 4 of the BES include around 1500 respondents from Scotland, enough to get an indication into what is going on. Some 37% of Scottish respondents to the BES reported some contact from at least one of the political parties.
This is a slightly lower figure than most seats in the March Ashcroft survey (which averaged 44%), but the former is an all-Scotland figure, and will include those seats (SNP-held seats, for example) where there might be less campaigning. Of those people who had been contacted by the parties, respondents were most likely to have been contacted by Labour (58%) or the SNP (51%) (which wasn’t true in the Ashcroft polls, but again, this is an all-Scotland figure, not a sample of constituencies), with the Conservatives and Lib Dems (both on 20%) some way back, and with Greens (5%) and UKIP (2%) further behind still.
The BES asks about six types of contact, along with an ‘other’ option. The figure below shows the percentage of Scottish respondents reporting each type of contact from the two largest parties.
As the figure shows, despite being behind overall in contact in March, the SNP led on most forms of contact: home visits, street contact, email, (what the BES rather quaintly calls) SMS, and ‘other’. Most of these differences are not huge and they involve many of the least common forms of contact.
However, on both home visits and street contact – both of which are labour-intensive – there seems to be a clear advantage to the SNP, as there is on email contact (presumably building on databases from the independence referendum?). The SNP then are ahead in face-to-face contact, if behind overall. But in both cases this involves contact with less than 4% of the population.
On both home visits and street contact there seems to be a clear advantage to the SNP.
The SNP trailed Labour on both phone contact and contact by leaflet/letter. The latter was the most common form of contact, by a long way. The BES does not differentiate between the ability of a party to deliver leaflets to your door personally (some poor party member tramping up and down streets for free on behalf of a party he or she loves) and those that a party pays someone to deliver (some poor member of Royal Mail tramping up and down streets on behalf of a company he or she probably doesn’t love).
This is unfortunate, although it is perhaps understandable methodologically (will respondents know or remember how a leaflet got through their door?). I suspect – though cannot prove – that this is masking a difference in how the parties are campaigning, with SNP relying more on its members to deliver leaflets, Labour paying for the same service. Whether this difference actually matters, though, is more open to debate. If the public can’t tell how a leaflet arrives, why would it matter?
One simple way to measure campaign intensity is to sum these different types of contact, giving people a score from 0 to 7. There is some slight evidence here of SNP contact being more intensive. Of those contacted by the SNP, most (65%) had had just one type of contact. But 21% had two, 9% had three, 1% had four – and 1% (admittedly, only by rounding up) reported all seven types of contact.
This last group are either very lucky or unlucky depending on your point of view. For Labour, a full 80% reported just one type of contact, with another 15% two, 4% one, and 1% four. No one reported above four. Even this, of course, doesn’t allow us to measure the intensity of each type of contact – five leaflets as opposed to one, say – but it does at least help show that the SNP’s grassroots campaign is slightly more intensive than Labour’s.
The Labour grassroots campaign in Scotland therefore appears to be slightly broader than the SNPs, but slightly less intensive, with the SNP winning narrowly on the more face-to-face forms of contact. Even so, given the advantage the SNP have in terms of grassroots members, it is still puzzling that these differences are not larger – although, if the SNP continue to poll as they have been, I cannot imagine they will be losing too much sleep over it.
Given that this poll was taken in March, and Ashcroft’s polls show an explosion in campaign activity in Scotland in April, it will be interesting to revisit these data with the post-election wave of the BES (as well as any other Ashcroft polls), to see if things change.