Sturgeon skews her stats (again)

Sturgeon skews her stats (again)
Fraser Bremner - Pool/Getty Images
Text settings

The statistical shenanigans of the SNP have been highlighted by Mr S before but it's always worth highlighting when the nationalist Holyrood government gets it wrong (again). At First Minister's Questions yesterday, Nicola Sturgeon told colleagues that England's infection rate is 20 per cent higher than that of Scotland, according to ONS figures. A surprising figure, given that the ONS estimated infection rate was 1 in 20 for both nations yesterday. The SNP leader said:

In terms of the ONS figures this week, infection levels in England right now are over 20 per cent higher than in Scotland. I don't think it's a competition but if Douglas Ross wants to make these comparisons, then they are the comparisons, there is the data.

But as journalist Chris Musson noted, this Sturgeon soundbite is just one of a number of statistical sleights-of-hand regularly used by the Scottish government. For the ONS infection survey figures that Sturgeon seems to have used to calculate an 'over 20 per cent' difference between Scotland and England are below. The last column says that the infection rate is both 1 in 20 for Scotland and England, with the figures rounded, presumably due to margins of error and the testing disparity between England and the other UK nations. 

But in the 1st column, the estimated average percent infected is 5.47 per cent in England and 4.49 per cent in Scotland – a difference of 0.98 percentage points. That then can be made to look like 'over 20 per cent' if you calculate the percentage difference between 4.49 and 5.47 which is 21.8 per cent: the figure to which Sturgeon refers. Describing the difference the way the First Minister did – but not giving the actual numbers – gives no perspective. 

Yet it is, of course, useful politically, as a small percentage change can seem significant when dealing with small numbers. For example a 0.98 percentage gap between 4.49 per cent and 5.47 per cent is 21.8 per cent while between 20 per cent and 20.98 per cent it's 4.9 per cent.

As Musson concluded, if Covid rates are 'not a competition' as the First Minister says, why is she using a stat that gives the public no perspective on the actual numbers, and arguably presents the difference in infections between Scotland and England in a misleading way? Mr S has his suspicions. 

It's by no means the first time that SNP ministers have performed such qualitative gymnastics: as Will Tanner of the Onward think-tank pointed out, data divergence is one of the hallmarks of nationalist rule:

Since taking power, the SNP has gradually, but deliberately, changed the way different public services collect data, measure outcomes and are held to account. The intention appears to be political: to create differentiation, evade scrutiny, and avoid intervention from Westminster. But it is the Scottish people who are suffering the consequences—with worse public services and an unaccountable state.

Such statistical sleights of hand have come to be synonymous with Sturgeon's government — hardly surprising from the party of the 'Curriculum for Excellence.'

Written bySteerpike

Steerpike is The Spectator's gossip columnist, serving up the latest tittle tattle from Westminster and beyond. Email tips to or message @MrSteerpike