There's more than a touch of grim irony in the fact that, in the midst of a series of blogposts alleging that SNP members have been systematically fleeced and hoodwinked by their party leadership, Stuart Campbell should yet again trot out a graph claiming that Yes support has remained absolutely static on 47% since 2015 - a claim he knows to be a downright lie. Whatever the rights and wrongs of the allegations against the SNP, they'd actually have a perfectly reasonable point if they said to Campbell that he needs to put his own house in order and stop repeatedly deceiving his readers before he can have any credibility in throwing stones himself.
Campbell's graph has been discredited so many times, by so many different people, and from so many different angles, that it's practically been beaten to a pulp by this point. It's been explained that Campbell has just hand-picked one single poll from each year, rather than averaging Yes support in all polls from each year. It's been explained that if he had averaged all polls, he'd have found Yes support had never been on 47% in any year and had in fact varied wildly from 45% in 2017 to 53% in 2020. It's been explained that if he had looked at the Scottish Social Attitudes Survey, which he has sometimes prayed in aid in other dodgy graphs, he's have found an even wider variation in support for independence from 39% in 2016 to 52% in 2021. It's been explained that the individual polls he's cherry-picked are not even directly comparable with each other, because they come from four different polling companies that use different methodologies.
Is there any other remaining way in which the graph can be debunked? Actually, there is. Campbell's single poll from each year has always been taken from April, implying that he thinks Yes support in April carries more significance than in other months (for some unspecified reason). So, as it actually is April right now, I decided to find out whether an average of all polls conducted in April of each year would bear out Campbell's claim that April-flavoured Yes support has flatlined at exactly 47% since 2015. Spoiler alert: no of course it doesn't. There have in fact only been two years in which the April average has come out at 47%, and one of those years is this year, which isn't even covered by Campbell's graph. There has been variation from 45% in 2017 to 49% in both 2021 and 2022.
Average Yes vote in conventional independence polls conducted (either in whole or in part) in April of each calendar year:
April 2015: 48.2%
April 2016: 47.0%
April 2017: 45.3%
April 2018: (No polls conducted)
April 2019: 48.4%
April 2020: (No polls conducted)
April 2021: 49.2%
April 2022: 49.0%
April 2023: 47.0%