There is little doubt that the Office for National Statistics has problems.
As the UK’s official statistical agency, it is supposed to operate free from political pressure, providing objective numbers about all aspects of the economy, demography and social change.
In most developed countries, there would be no problem with such a set-up. But in adversarial, target-driven Britain, the ONS is regularly dragged into controversy.
The Commons public administration select committee has recently published documents relating to the spat between Whitehall and the UK Statistics Authority — ONS’s independent protector — about the release of knife crime statistics by Number 10. The PASC helpfully requested the flow of e-mails about the subject between the Department of Health (which collected the figures), the ONS and Downing Street.
The first thing to note about the e-mails is the good light they shed on DoH civil servants. Faced with Number 10’s ‘adamant’ demand to publish the figures, the department officials made strenuous efforts to check if this would be OK, against a very tight deadline. Andy Sutherland, a senior statistician at the department, appears to have had near-perfect pitch in deciding what — and what not — to do.
Every effort was made to stop the publication of numbers that were only provisional and had not been pre-announced.
Number 10 was seen as ‘likely to publish the data irrespective of the concerns raised’. The national statistician was informed, which will have triggered the UK Statistics Authority’s decision to criticise the government.
Kevin Brennan, the Cabinet Office minister responsible for statistics, stated in a letter to PASC chair Tony Wright that the contested figures were ‘published under HO [Home Office] authority’, although the names of those giving permission were withheld. Home Office ministers ‘were not aware that there were outstanding concerns around the use of NHS data until after the factsheet [containing the knife crime statistics] had been published’.
No-one in a senior position in the Home Office, particularly ministers, knew what was going on. It is hard to see how the ‘do’s and don’ts’ summary will stop Number 10 overriding the junior officials who evidently make decisions there. The problem is, the government feels itself under such massive pressure to achieve its policy objectives that any scrap of evidence, however pitifully tiny, will be used.
The ONS is trapped in the middle of such a world. Not only do ministers want to cherry-pick good statistics as soon as possible, they also want to obscure the implications of any bad ones.
Do we imagine Number 10 would have been quite as keen to override the DoH’s statistician if the knife crime numbers had gone up? If the numbers suggested things were getting worse, they would not have been rushed into the public domain. Instead, they would have been carefully explained in a doubtless-delayed publication. Welcome to Orwell’s Britain.
Official statisticians also have to put up with ministers attacking their publications. John Reid, as health secretary, launched a full-throttle assault on ONS figures that showed NHS productivity was falling. Within the past fortnight, immigration minister Phil Woolas has hit out at the ONS’s decision to publish figures about the number of UK residents born abroad, describing it as ‘at best naive and at worst sinister’.
‘Sinister’ is an extraordinary word for a minister to use about the activities of the official statistics agency.
Woolas might have had a reasonable point to make. It is implausible to expect the ONS to get everything right. But to use such intemperate language about the statisticians will have only one effect — to reduce still further confidence in both politicians and official statisticians.
Trust in government is low. One of the ways we judge the performance of Whitehall is by using the output of the ONS. Both Labour and Conservative governments have put pressure on statistics by the use of performance indicators, targets and so on. By dragging official data into the contested centre of political discourse, it is inevitable that the statistics themselves will be challenged.
If ever there were an institution that needed support from the independent and the free-minded it is the ONS. As the complexity of measuring the way bank bail-outs, Treasury banking facilities and government support for industry becomes more apparent, the ONS will find itself under even more pressure.
Those who believe in good, politically neutral statistics should rally to support Karen Dunnell, the national statistician. The ONS is an important element in our democratic system.
Tony Travers is the director of the Greater London Group at the London School of Economics
