Statisticians are not the most colourful characters and they tend to play a fairly understated role in government.
Their days are spent staring at spreadsheets and their bread and butter are sampling methodologies and standard deviations. There is a reason no one ever made a high-octane political thriller about the UK Statistics Authority.
However, the Covid-19 pandemic has made the gathering and presentation of data, more than ever, a matter of life and death. Not only do ministers rely on forecasting models to make decisions about managing the virus, public confidence in health measures depends on clear, accurate and digestible information about risks and how best to mitigate them. The work of the Office for Statistics Regulation (OSR), the arm of the Statistics Authority that runs the rule over how politicians use (and misuse) data, has never been more vital.
Nicola Sturgeon has been reported to the independent body after claiming during First Minister’s Questions last week that Office for National Statistics figures show ‘infection levels in England are over 20 per cent higher than those in Scotland’. Willie Rennie, who made the referral, believes this may have been an inaccurate portrayal of the data.
Actually, he puts it in more robust terms. ‘I am concerned that these statistics may have been seriously twisted,’ he writes to the OSR. By his reckoning, the numbers point to a difference between the two countries of roughly one per cent. The OSR’s view will be awaited keenly.
Earlier this month, Sturgeon’s deputy John Swinney was reported for citing pre-Christmas infection rates to justify the Scottish Government’s post-Christmas restrictions. These are but the latest in a series of run-ins with the OSR. In June 2020, the Scottish Government made an unsupported assertion in a media release about the number of antibody tests Scotland had carried out.
This prompted the OSR to complain that, since Scotland was not making its testing data public, ’this figure cannot be verified’. The watchdog branded this ‘unacceptable for a figure of such importance used in a government news release’ and chided the Scottish Government: ‘[T]o ensure public confidence and equality of access, we urge you to publish the data quoted in the Scottish Government news release’. The impartial guardians of statistical integrity were, in effect, instructing the SNP to show its working.
The following month, Nicola Sturgeon rebuffed calls for a four-nations approach on travel quarantine rules because ‘we can’t allow ourselves to be dragged along in the wake of another government’s – to be quite frank about it – shambolic decision-making process’. She cited as evidence her government’s assessment that ‘the prevalence of the virus in Scotland, right now, is five times lower than it is in England’.
This time the OSR objected that ‘the sources used to underpin this claim have been difficult to identify’ and that the explanation given to them was ‘not clear’, with ‘different data sources being quoted to the media and to us’. A polite form of words but their implication was bracing. In the event, the oversight body found the hodgepodge of sources used to justify Sturgeon’s assertion did not ‘allow for a quantified and uncaveated comparison of the kind that was made’.
The OSR fired off a letter to the Scottish Government stating: ‘When unpublished figures are quoted in the public domain, we expect that this information is shared with the media and the public in a way that promotes transparency and clarity.’ That is as close to a stinging rebuke as statisticians get.
Then, in September 2020, the First Minister claimed in parliament that 40 per cent of Scottish care homes were allowing loved ones to visit. This was, it turned out, another highly dubious account of the facts. The OSR found it was ‘a very loose approximation based on incomplete data’ which ‘should have been more clearly reflected in the response’. How incomplete? Two of the biggest health boards (Greater Glasgow and Clyde and Lanarkshire) were among those unable to confirm how many care homes were letting in relatives. How unclear? As the OSR remarked: ‘It should not have been necessary to wait for the information to be published as part of an FOI response.’
June 2021 brought Humza Yousaf’s startling claim that ‘there was 10 young children – 0-9 years old – who were hospitalised because of COVID’ the previous week. Unsurprisingly, parents were alarmed, but the health secretary had misspoken. The OSR noted that ‘the figure used was not available publicly’ and ‘was inaccurately presented’. Embarrassingly, the watchdog underscored ‘the importance of ensuring ministers are appropriately briefed’.
That is not where the SNP’s many misfortunes with statistics ends. Recall that Yousaf, as Justice Secretary, told the health committee in June 2020 that officials were ‘carrying out spot checks’ and ‘contacting approximately 20 per cent of travellers’ to give them guidance on quarantine. An impressive feat when said checks wouldn’t begin until July.
He assured the same committee that Police Scotland had confirmed ‘compliance with the self-isolation requirements has been very high’. Police Scotland said its involvement in the matter had been ‘very limited’ because Public Health Scotland ‘would not pass on any information or details of apparent breaches of self-isolation’.
Recall, too, John Swinney’s surprise foray into memes when he shared on Twitter a graphic that claimed, without any evidence, that wearing masks and maintaining 6ft distance reduced the risk of transmission to zero. The Deputy First Minister — the minister charged with leading the Covid recovery — withdrew the graphic and admitted ‘the figures in it were not verified’.
The woman formerly in charge of the pandemic itself was not much better. As health secretary Jeane Freeman initially suggested only 300 patients had been discharged from hospitals to care homes prior to mandatory testing, only to later admit the figure was three times as many.
Some of these misstatements can be explained as just that: errors. Ministers are bombarded with information, under constant pressure and seldom far from a microphone. The potential for misspeaking is ever-present. The First Minister herself suggested tiredness as a reason for Jeane Freeman’s gaffe. Other failings can be explained by the sheer pace of a pandemic and the difficulties involved in gathering and publishing data. The OSR even acknowledges this.
That is not the problem. The problem is the unshakeable impression that the government’s priority in sharing data is not always public information and is too often public relations. The frequency with which these suspect statistics are deployed to demonstrate that Scotland is doing better than England, and therefore the SNP is doing better than the Tories, reeks of cynicism and partisanship and game-playing. It intimates to us that there is nothing, not even a pandemic, in which the SNP does not sniff political advantage and nothing, not even the immutable black-and-white of numbers, that will not be spun to advance the narrow agenda of this pressure group impersonating a government.
This is normally the point where a familiar howl shrieks forth from the it-wisnae-me chorus: ‘What about Westminster?’ What about it? Does the UK Government playing fast-and-loose with the facts justify the Scottish Government doing the same? Or is it somehow more ethical, more progressive when the public is gulled from St Andrew’s House rather than from Whitehall? Are the wicked Tories manipulating the data but the noble SNP spinning for Scotland? At some point Scottish politics is going to have to grow up and judge itself by a higher standard than the lesser angels of someone else’s nature.
At all times, but especially in a pandemic, statistics are not dry data points but the fundamental building blocks of collective understanding and public trust. Damage them in service of short-term political gain and you threaten to bring trust and understanding crumbling down. Every time government misuses statistics, the country gets a little more cynical. Every time a minister is caught slanting the numbers, the task of the anti-vaxxers gets a little easier and the reluctance of the vaccine-hesitant a little more entrenched. Statistics matter, and so does the truth.
Originally published in the Scottish Daily Mail on January 24, 2022.
We must also ask why they are allowed to get away with all this ,the media asks no questions !
Stephen. Great article thank you for that. You do yourself and the Real People of Scotland a disservice by referring to the SNPAssembly as a government. This is a self appointed title given or stolen by AlecSalmond and built on by Nicola Sturgeon and her Kitchen Cabinet.